For the diagnostics and health management of lithium-ion batteries,numerous models have been developed to understand their degradation characteristics.These models typically fall into two categories:data-driven models...For the diagnostics and health management of lithium-ion batteries,numerous models have been developed to understand their degradation characteristics.These models typically fall into two categories:data-driven models and physical models,each offering unique advantages but also facing limitations.Physics-informed neural networks(PINNs)provide a robust framework to integrate data-driven models with physical principles,ensuring consistency with underlying physics while enabling generalization across diverse operational conditions.This study introduces a PINN-based approach to reconstruct open circuit voltage(OCV)curves and estimate key ageing parameters at both the cell and electrode levels.These parameters include available capacity,electrode capacities,and lithium inventory capacity.The proposed method integrates OCV reconstruction models as functional components into convolutional neural networks(CNNs)and is validated using a public dataset.The results reveal that the estimated ageing parameters closely align with those obtained through offline OCV tests,with errors in reconstructed OCV curves remaining within 15 mV.This demonstrates the ability of the method to deliver fast and accurate degradation diagnostics at the electrode level,advancing the potential for precise and efficient battery health management.展开更多
The Internet of Things(IoT)ecosystem faces growing security challenges because it is projected to have 76.88 billion devices by 2025 and $1.4 trillion market value by 2027,operating in distributed networks with resour...The Internet of Things(IoT)ecosystem faces growing security challenges because it is projected to have 76.88 billion devices by 2025 and $1.4 trillion market value by 2027,operating in distributed networks with resource limitations and diverse system architectures.The current conventional intrusion detection systems(IDS)face scalability problems and trust-related issues,but blockchain-based solutions face limitations because of their low transaction throughput(Bitcoin:7 TPS(Transactions Per Second),Ethereum:15-30 TPS)and high latency.The research introduces MBID(Multi-Tier Blockchain Intrusion Detection)as a groundbreaking Multi-Tier Blockchain Intrusion Detection System with AI-Enhanced Detection,which solves the problems in huge IoT networks.The MBID system uses a four-tier architecture that includes device,edge,fog,and cloud layers with blockchain implementations and Physics-Informed Neural Networks(PINNs)for edge-based anomaly detection and a dual consensus mechanism that uses Honesty-based Distributed Proof-of-Authority(HDPoA)and Delegated Proof of Stake(DPoS).The system achieves scalability and efficiency through the combination of dynamic sharding and Interplanetary File System(IPFS)integration.Experimental evaluations demonstrate exceptional performance,achieving a detection accuracy of 99.84%,an ultra-low false positive rate of 0.01% with a False Negative Rate of 0.15%,and a near-instantaneous edge detection latency of 0.40 ms.The system demonstrated an aggregate throughput of 214.57 TPS in a 3-shard configuration,providing a clear,evidence-based path for horizontally scaling to support overmillions of devices with exceeding throughput.The proposed architecture represents a significant advancement in blockchain-based security for IoT networks,effectively balancing the trade-offs between scalability,security,and decentralization.展开更多
The packaging quality of coaxial laser diodes(CLDs)plays a pivotal role in determining their optical performance and long-term reliability.As the core packaging process,high-precision laser welding requires precise co...The packaging quality of coaxial laser diodes(CLDs)plays a pivotal role in determining their optical performance and long-term reliability.As the core packaging process,high-precision laser welding requires precise control of process parameters to suppress optical power loss.However,the complex nonlinear relationship between welding parameters and optical power loss renders traditional trial-and-error methods inefficient and imprecise.To address this challenge,a physics-informed(PI)and data-driven collaboration approach for welding parameter optimization is proposed.First,thermal-fluid-solid coupling finite element method(FEM)was employed to quantify the sensitivity of welding parameters to physical characteristics,including residual stress.This analysis facilitated the identification of critical factors contributing to optical power loss.Subsequently,a Gaussian process regression(GPR)model incorporating finite element simulation prior knowledge was constructed based on the selected features.By introducing physics-informed kernel(PIK)functions,stress distribution patterns were embedded into the prediction model,achieving high-precision optical power loss prediction.Finally,a Bayesian optimization(BO)algorithm with an adaptive sampling strategy was implemented for efficient parameter space exploration.Experimental results demonstrate that the proposedmethod effectively establishes explicit physical correlations between welding parameters and optical power loss.The optimized welding parameters reduced optical power loss by 34.1%,providing theoretical guidance and technical support for reliable CLD packaging.展开更多
Energy efficiency stands as an essential factor when implementing deep reinforcement learning(DRL)policies for robotic control systems.Standard algorithms,including Deep Deterministic Policy Gradient(DDPG),primarily o...Energy efficiency stands as an essential factor when implementing deep reinforcement learning(DRL)policies for robotic control systems.Standard algorithms,including Deep Deterministic Policy Gradient(DDPG),primarily optimize task rewards but at the cost of excessively high energy consumption,making them impractical for real-world robotic systems.To address this limitation,we propose Physics-Informed DDPG(PI-DDPG),which integrates physics-based energy penalties to develop energy-efficient yet high-performing control policies.The proposed method introduces adaptive physics-informed constraints through a dynamic weighting factor(λ),enabling policies that balance reward maximization with energy savings.Our motivation is to overcome the impracticality of rewardonly optimization by designing controllers that achieve competitive performance while substantially reducing energy consumption.PI-DDPG was evaluated in nine MuJoCo continuous control environments,where it demonstrated significant improvements in energy efficiency without compromising stability or performance.Experimental results confirm that PI-DDPG substantially reduces energy consumption compared to standard DDPG,while maintaining competitive task performance.For instance,energy costs decreased from 5542.98 to 3119.02 in HalfCheetah-v4 and from1909.13 to 1586.75 in Ant-v4,with stable performance in Hopper-v4(205.95 vs.130.82)and InvertedPendulum-v4(322.97 vs.311.29).Although DDPG sometimes yields higher rewards,such as in HalfCheetah-v4(5695.37 vs.4894.59),it requires significantly greater energy expenditure.These results highlight PI-DDPG as a promising energy-conscious alternative for robotic control.展开更多
Physics-informed neural networks(PINNs)have shown remarkable prospects in solving the forward and inverse problems involving partial differential equations(PDEs).The method embeds PDEs into the neural network by calcu...Physics-informed neural networks(PINNs)have shown remarkable prospects in solving the forward and inverse problems involving partial differential equations(PDEs).The method embeds PDEs into the neural network by calculating the PDE loss at a set of collocation points,providing advantages such as meshfree and more convenient adaptive sampling.However,when solving PDEs using nonuniform collocation points,PINNs still face challenge regarding inefficient convergence of PDE residuals or even failure.In this work,we first analyze the ill-conditioning of the PDE loss in PINNs under nonuniform collocation points.To address the issue,we define volume weighting residual and propose volume weighting physics-informed neural networks(VW-PINNs).Through weighting the PDE residuals by the volume that the collocation points occupy within the computational domain,we embed explicitly the distribution characteristics of collocation points in the loss evaluation.The fast and sufficient convergence of the PDE residuals for the problems involving nonuniform collocation points is guaranteed.Considering the meshfree characteristics of VW-PINNs,we also develop a volume approximation algorithm based on kernel density estimation to calculate the volume of the collocation points.We validate the universality of VW-PINNs by solving the forward problems involving flow over a circular cylinder and flow over the NACA0012 airfoil under different inflow conditions,where conventional PINNs fail.By solving the Burgers’equation,we verify that VW-PINNs can enhance the efficiency of existing the adaptive sampling method in solving the forward problem by three times,and can reduce the relative L 2 error of conventional PINNs in solving the inverse problem by more than one order of magnitude.展开更多
Enforcing initial and boundary conditions(I/BCs)poses challenges in physics-informed neural networks(PINNs).Several PINN studies have gained significant achievements in developing techniques for imposing BCs in static...Enforcing initial and boundary conditions(I/BCs)poses challenges in physics-informed neural networks(PINNs).Several PINN studies have gained significant achievements in developing techniques for imposing BCs in static problems;however,the simultaneous enforcement of I/BCs in dynamic problems remains challenging.To overcome this limitation,a novel approach called decoupled physics-informed neural network(d PINN)is proposed in this work.The d PINN operates based on the core idea of converting a partial differential equation(PDE)to a system of ordinary differential equations(ODEs)via the space-time decoupled formulation.To this end,the latent solution is expressed in the form of a linear combination of approximation functions and coefficients,where approximation functions are admissible and coefficients are unknowns of time that must be solved.Subsequently,the system of ODEs is obtained by implementing the weighted-residual form of the original PDE over the spatial domain.A multi-network structure is used to parameterize the set of coefficient functions,and the loss function of d PINN is established based on minimizing the residuals of the gained ODEs.In this scheme,the decoupled formulation leads to the independent handling of I/BCs.Accordingly,the BCs are automatically satisfied based on suitable selections of admissible functions.Meanwhile,the original ICs are replaced by the Galerkin form of the ICs concerning unknown coefficients,and the neural network(NN)outputs are modified to satisfy the gained ICs.Several benchmark problems involving different types of PDEs and I/BCs are used to demonstrate the superior performance of d PINN compared with regular PINN in terms of solution accuracy and computational cost.展开更多
Dynamic wake field information is vital for the optimized design and control of wind farms.Combined with sparse measurement data from light detection and ranging(LiDAR),the physics-informed neural network(PINN)framewo...Dynamic wake field information is vital for the optimized design and control of wind farms.Combined with sparse measurement data from light detection and ranging(LiDAR),the physics-informed neural network(PINN)frameworks have recently been employed for forecasting freestream wind and wake fields.However,these PINN frameworks face challenges of low prediction accuracy and long training times.Therefore,this paper constructed a PINN framework for dynamic wake field prediction by integrating two accuracy improvement strategies and a step-by-step training time saving strategy.The results showed that the different performance improvement routes significantly improved the overall performance of the PINN.The accuracy and efficiency of the PINN with spatiotemporal improvement strategies were validated via LiDAR-measured data from a wind farm in Shandong province,China.This paper sheds light on load reduction,efficiency improvement,intelligent operation and maintenance of wind farms.展开更多
Numerical simulation plays an important role in the dynamic analysis of multibody system.With the rapid development of computer science,the numerical solution technology has been further developed.Recently,data-driven...Numerical simulation plays an important role in the dynamic analysis of multibody system.With the rapid development of computer science,the numerical solution technology has been further developed.Recently,data-driven method has become a very popular computing method.However,due to lack of necessary mechanism information of the traditional pure data-driven methods based on neural network,its numerical accuracy cannot be guaranteed for strong nonlinear system.Therefore,this work proposes a mechanism-data hybrid-driven strategy for solving nonlinear multibody system based on physics-informed neural network to overcome the limitation of traditional data-driven methods.The strategy proposed in this paper introduces scaling coefficients to introduce the dynamic model of multibody system into neural network,ensuring that the training results of neural network conform to the mechanics principle of the system,thereby ensuring the good reliability of the data-driven method.Finally,the stability,generalization ability and numerical accuracy of the proposed method are discussed and analyzed using three typical multibody systems,and the constrained default situations can be controlled within the range of 10^(-2)-10^(-4).展开更多
This research introduces a spectrum-based physics-informed neural network(SP-PINN)model to significantly improve the accuracy of calculation of two-phase flow parameters,surpassing existing methods that have limitatio...This research introduces a spectrum-based physics-informed neural network(SP-PINN)model to significantly improve the accuracy of calculation of two-phase flow parameters,surpassing existing methods that have limitations in global and continuous data sampling.SP-PINNs address the challenges of traditional methods in terms of continuous sampling by integrating the spectral analysis and pressure correction into the Navier-Stokes(N-S)equations,enhancing the predictive accuracy especially in critical regions like gas-phase boundaries and velocity peaks.The novel introduction of a pressure-correction module within SP-PINNs mitigates prediction errors,achieving a substantial reduction to 1‰compared with the conventional physics-informed neural network(PINN)approaches.Experimental applications validate the model’s ability to accurately and rapidly predict flow parameters with different sampling time intervals,with the computation time of predicting unsampled data less than 0.01 s.Such advancements signify a 100-fold improvement over traditional DNS calculations,underscoring the model’s potential in the real-time calculation and analysis of multiphase flow dynamics.展开更多
In this paper,we propose a symmetric difference data enhancement physics-informed neural network(SDE-PINN)to study soliton solutions for discrete nonlinear lattice equations(NLEs).By considering known and unknown symm...In this paper,we propose a symmetric difference data enhancement physics-informed neural network(SDE-PINN)to study soliton solutions for discrete nonlinear lattice equations(NLEs).By considering known and unknown symmetric points,numerical simulations are conducted to one-soliton and two-soliton solutions of a discrete KdV equation,as well as a one-soliton solution of a discrete Toda lattice equation.Compared with the existing discrete deep learning approach,the numerical results reveal that within the specified spatiotemporal domain,the prediction accuracy by SDE-PINN is excellent regardless of the interior or extrapolation prediction,with a significant reduction in training time.The proposed data enhancement technique and symmetric structure development provides a new perspective for the deep learning approach to solve discrete NLEs.The newly proposed SDE-PINN can also be applied to solve continuous nonlinear equations and other discrete NLEs numerically.展开更多
With the rapid development of deep learning neural networks,new solutions have emerged for addressing fluid flow problems in porous media.Combining data-driven approaches with physical constraints has become a hot res...With the rapid development of deep learning neural networks,new solutions have emerged for addressing fluid flow problems in porous media.Combining data-driven approaches with physical constraints has become a hot research direction,with physics-informed neural networks(PINNs) being the most popular hybrid model.PINNs have gained widespread attention in subsurface fluid flow simulations due to their low computational resource requirements,fast training speeds,strong generalization capabilities,and broad applicability.Despite success in homogeneous settings,standard PINNs face challenges in accurately calculating flux between irregular Eulerian cells with disparate properties and capturing global field influences on local cells.This limits their suitability for heterogeneous reservoirs and the irregular Eulerian grids frequently used in reservoir.To address these challenges,this study proposes a physics-informed graph neural network(PIGNN) model.The PIGNN model treats the entire field as a whole,integrating information from neighboring grids and physical laws into the solution for the target grid,thereby improving the accuracy of solving partial differential equations in heterogeneous and Eulerian irregular grids.The optimized model was applied to pressure field prediction in a spatially heterogeneous reservoir,achieving an average L_(2) error and R_(2) score of 6.710×10^(-4)and 0.998,respectively,which confirms the effectiveness of model.Compared to the conventional PINN model,the average L_(2) error was reduced by 76.93%,the average R_(2) score increased by 3.56%.Moreover,evaluating robustness,training the PIGNN model using only 54% and 76% of the original data yielded average relative L_(2) error reductions of 58.63% and 56.22%,respectively,compared to the PINN model.These results confirm the superior performance of this approach compared to PINN.展开更多
Shape prediction of deformable linear objects(DLO)plays critical roles in robotics,medical devices,aerospace,and manufacturing,especially in manipulating objects such as cables,wires,and fibers.Due to the inherent fle...Shape prediction of deformable linear objects(DLO)plays critical roles in robotics,medical devices,aerospace,and manufacturing,especially in manipulating objects such as cables,wires,and fibers.Due to the inherent flexibility of DLO and their complex deformation behaviors,such as bending and torsion,it is challenging to predict their dynamic characteristics accurately.Although the traditional physical modeling method can simulate the complex deformation behavior of DLO,the calculation cost is high and it is difficult to meet the demand of real-time prediction.In addition,the scarcity of data resources also limits the prediction accuracy of existing models.To solve these problems,a method of fiber shape prediction based on a physical information graph neural network(PIGNN)is proposed in this paper.This method cleverly combines the powerful expressive power of graph neural networks with the strict constraints of physical laws.Specifically,we learn the initial deformation model of the fiber through graph neural networks(GNN)to provide a good initial estimate for the model,which helps alleviate the problem of data resource scarcity.During the training process,we incorporate the physical prior knowledge of the dynamic deformation of the fiber optics into the loss function as a constraint,which is then fed back to the network model.This ensures that the shape of the fiber optics gradually approaches the true target shape,effectively solving the complex nonlinear behavior prediction problem of deformable linear objects.Experimental results demonstrate that,compared to traditional methods,the proposed method significantly reduces execution time and prediction error when handling the complex deformations of deformable fibers.This showcases its potential application value and superiority in fiber manipulation.展开更多
Physics-informed neural networks(PINNs)are promising to replace conventional mesh-based partial tial differen-equation(PDE)solvers by offering more accurate and flexible PDE solutions.However,PINNs are hampered by the...Physics-informed neural networks(PINNs)are promising to replace conventional mesh-based partial tial differen-equation(PDE)solvers by offering more accurate and flexible PDE solutions.However,PINNs are hampered by the relatively slow convergence and the need to perform additional,potentially expensive training for new PDE parameters.To solve this limitation,we introduce LatentPINN,a framework that utilizes latent representations of the PDE parameters as additional(to the coordinates)inputs into PINNs and allows for training over the distribution of these parameters.Motivated by the recent progress on generative models,we promote using latent diffusion models to learn compressed latent representations of the distribution of PDE parameters as they act as input parameters for NN functional solutions.We use a two-stage training scheme in which,in the first stage,we learn the latent representations for the distribution of PDE parameters.In the second stage,we train a physics-informed neural network over inputs given by randomly drawn samples from the coordinate space within the solution domain and samples from the learned latent representation of the PDE parameters.Considering their importance in capturing evolving interfaces and fronts in various fields,we test the approach on a class of level set equations given,for example,by the nonlinear Eikonal equation.We share results corresponding to three Eikonal parameters(velocity models)sets.The proposed method performs well on new phase velocity models without the need for any additional training.展开更多
Accurately estimating the battery state of health(SOH)is essential for ensuring the safe and reliable operation of battery systems of electric vehicles.However,due to the complex and variable operating conditions enco...Accurately estimating the battery state of health(SOH)is essential for ensuring the safe and reliable operation of battery systems of electric vehicles.However,due to the complex and variable operating conditions encountered in practical applications,achieving precise and physics-informed SOH estimation remains challenging.To address these problems,this paper develops a lightweight two-stage physicsinformed neural network(TSPINN)method for SOH estimation of lithium-ion batteries with different chemistries.Specifically,this paper utilizes firstly relaxation voltage data obtained after a full charge to determine the aging-related parameters of physical equivalent circuit model(ECM).Additionally,incremental capacity(IC)feature is extracted by analyzing peak values of the IC curve during the charging phase,which thereby constitutes the first stage of the proposed TSPINN,termed as physics-informed data augmentation for SOH estimation.Additionally,the physical information can be further embedded by incorporating feature knowledge related to mechanisms into the loss function,and ultimately,the second stage of the proposed TSPINN is developed,which is named the physics-informed loss function.The effectiveness of the TSPINN method was confirmed through the experimental data for LiNi_(0.86)Co_(0.11)Al_(0.03)O_(2)(NCA)and LiNi_(0.83)Co_(0.11)Mn_(0.07)O_(2)(NCM)battery materials under different temperature conditions.The final experimental results indicate that the TSPINN method achieved SOH estimation with a mean absolute error(MAE)of 0.675%,showing improvements of approximately 29.3%,60.3%,and 8.1% compared to methods using only ECM,IC,and integrated features,respectively.The results validate the effectiveness and adaptability of TSPINN,establishing it as a reliable solution for advanced battery management systems.展开更多
Despite the significant progress over the last 50 years in simulating flow problems using numerical discretization of the Navier–Stokes equations(NSE),we still cannot incorporate seamlessly noisy data into existing a...Despite the significant progress over the last 50 years in simulating flow problems using numerical discretization of the Navier–Stokes equations(NSE),we still cannot incorporate seamlessly noisy data into existing algorithms,mesh-generation is complex,and we cannot tackle high-dimensional problems governed by parametrized NSE.Moreover,solving inverse flow problems is often prohibitively expensive and requires complex and expensive formulations and new computer codes.Here,we review flow physics-informed learning,integrating seamlessly data and mathematical models,and implement them using physics-informed neural networks(PINNs).We demonstrate the effectiveness of PINNs for inverse problems related to three-dimensional wake flows,supersonic flows,and biomedical flows.展开更多
Physics-informed deep learning has drawn tremendous interest in recent years to solve computational physics problems,whose basic concept is to embed physical laws to constrain/inform neural networks,with the need of l...Physics-informed deep learning has drawn tremendous interest in recent years to solve computational physics problems,whose basic concept is to embed physical laws to constrain/inform neural networks,with the need of less data for training a reliable model.This can be achieved by incorporating the residual of physics equations into the loss function.Through minimizing the loss function,the network could approximate the solution.In this paper,we propose a mixed-variable scheme of physics-informed neural network(PINN)for fluid dynamics and apply it to simulate steady and transient laminar flows at low Reynolds numbers.A parametric study indicates that the mixed-variable scheme can improve the PINN trainability and the solution accuracy.The predicted velocity and pressure fields by the proposed PINN approach are also compared with the reference numerical solutions.Simulation results demonstrate great potential of the proposed PINN for fluid flow simulation with a high accuracy.展开更多
Accurate insight into the heat generation rate(HGR) of lithium-ion batteries(LIBs) is one of key issues for battery management systems to formulate thermal safety warning strategies in advance.For this reason,this pap...Accurate insight into the heat generation rate(HGR) of lithium-ion batteries(LIBs) is one of key issues for battery management systems to formulate thermal safety warning strategies in advance.For this reason,this paper proposes a novel physics-informed neural network(PINN) approach for HGR estimation of LIBs under various driving conditions.Specifically,a single particle model with thermodynamics(SPMT) is first constructed for extracting the critical physical knowledge related with battery HGR.Subsequently,the surface concentrations of positive and negative electrodes in battery SPMT model are integrated into the bidirectional long short-term memory(BiLSTM) networks as physical information.And combined with other feature variables,a novel PINN approach to achieve HGR estimation of LIBs with higher accuracy is constituted.Additionally,some critical hyperparameters of BiLSTM used in PINN approach are determined through Bayesian optimization algorithm(BOA) and the results of BOA-based BiLSTM are compared with other traditional BiLSTM/LSTM networks.Eventually,combined with the HGR data generated from the validated virtual battery,it is proved that the proposed approach can well predict the battery HGR under the dynamic stress test(DST) and worldwide light vehicles test procedure(WLTP),the mean absolute error under DST is 0.542 kW/m^(3),and the root mean square error under WLTP is1.428 kW/m^(3)at 25℃.Lastly,the investigation results of this paper also show a new perspective in the application of the PINN approach in battery HGR estimation.展开更多
We consider solving the forward and inverse partial differential equations(PDEs)which have sharp solutions with physics-informed neural networks(PINNs)in this work.In particular,to better capture the sharpness of the ...We consider solving the forward and inverse partial differential equations(PDEs)which have sharp solutions with physics-informed neural networks(PINNs)in this work.In particular,to better capture the sharpness of the solution,we propose the adaptive sampling methods(ASMs)based on the residual and the gradient of the solution.We first present a residual only-based ASM denoted by ASMⅠ.In this approach,we first train the neural network using a small number of residual points and divide the computational domain into a certain number of sub-domains,then we add new residual points in the sub-domain which has the largest mean absolute value of the residual,and those points which have the largest absolute values of the residual in this sub-domain as new residual points.We further develop a second type of ASM(denoted by ASMⅡ)based on both the residual and the gradient of the solution due to the fact that only the residual may not be able to efficiently capture the sharpness of the solution.The procedure of ASMⅡis almost the same as that of ASMⅠ,and we add new residual points which have not only large residuals but also large gradients.To demonstrate the effectiveness of the present methods,we use both ASMⅠand ASMⅡto solve a number of PDEs,including the Burger equation,the compressible Euler equation,the Poisson equation over an Lshape domain as well as the high-dimensional Poisson equation.It has been shown from the numerical results that the sharp solutions can be well approximated by using either ASMⅠor ASMⅡ,and both methods deliver much more accurate solutions than the original PINNs with the same number of residual points.Moreover,the ASMⅡalgorithm has better performance in terms of accuracy,efficiency,and stability compared with the ASMⅠalgorithm.This means that the gradient of the solution improves the stability and efficiency of the adaptive sampling procedure as well as the accuracy of the solution.Furthermore,we also employ the similar adaptive sampling technique for the data points of boundary conditions(BCs)if the sharpness of the solution is near the boundary.The result of the L-shape Poisson problem indicates that the present method can significantly improve the efficiency,stability,and accuracy.展开更多
Neural networks with physical governing equations as constraints have recently created a new trend in machine learning research.In this context,a review of related research is first presented and discussed.The potenti...Neural networks with physical governing equations as constraints have recently created a new trend in machine learning research.In this context,a review of related research is first presented and discussed.The potential offered by such physics-informed deep learning models for computations in geomechanics is demonstrated by application to one-dimensional(1D)consolidation.The governing equation for 1D problems is applied as a constraint in the deep learning model.The deep learning model relies on automatic differentiation for applying the governing equation as a constraint,based on the mathematical approximations established by the neural network.The total loss is measured as a combination of the training loss(based on analytical and model predicted solutions)and the constraint loss(a requirement to satisfy the governing equation).Two classes of problems are considered:forward and inverse problems.The forward problems demonstrate the performance of a physically constrained neural network model in predicting solutions for 1D consolidation problems.Inverse problems show prediction of the coefficient of consolidation.Terzaghi’s problem,with varying boundary conditions,is used as a numerical example and the deep learning model shows a remarkable performance in both the forward and inverse problems.While the application demonstrated here is a simple 1D consolidation problem,such a deep learning model integrated with a physical law has significant implications for use in,such as,faster realtime numerical prediction for digital twins,numerical model reproducibility and constitutive model parameter optimization.展开更多
Accurate prediction of formation pore pressure is essential to predict fluid flow and manage hydrocarbon production in petroleum engineering.Recent deep learning technique has been receiving more interest due to the g...Accurate prediction of formation pore pressure is essential to predict fluid flow and manage hydrocarbon production in petroleum engineering.Recent deep learning technique has been receiving more interest due to the great potential to deal with pore pressure prediction.However,most of the traditional deep learning models are less efficient to address generalization problems.To fill this technical gap,in this work,we developed a new adaptive physics-informed deep learning model with high generalization capability to predict pore pressure values directly from seismic data.Specifically,the new model,named CGP-NN,consists of a novel parametric features extraction approach(1DCPP),a stacked multilayer gated recurrent model(multilayer GRU),and an adaptive physics-informed loss function.Through machine training,the developed model can automatically select the optimal physical model to constrain the results for each pore pressure prediction.The CGP-NN model has the best generalization when the physicsrelated metricλ=0.5.A hybrid approach combining Eaton and Bowers methods is also proposed to build machine-learnable labels for solving the problem of few labels.To validate the developed model and methodology,a case study on a complex reservoir in Tarim Basin was further performed to demonstrate the high accuracy on the pore pressure prediction of new wells along with the strong generalization ability.The adaptive physics-informed deep learning approach presented here has potential application in the prediction of pore pressures coupled with multiple genesis mechanisms using seismic data.展开更多
基金supported by the Beijing Natural Science Foundation(Grant No.L223013)。
文摘For the diagnostics and health management of lithium-ion batteries,numerous models have been developed to understand their degradation characteristics.These models typically fall into two categories:data-driven models and physical models,each offering unique advantages but also facing limitations.Physics-informed neural networks(PINNs)provide a robust framework to integrate data-driven models with physical principles,ensuring consistency with underlying physics while enabling generalization across diverse operational conditions.This study introduces a PINN-based approach to reconstruct open circuit voltage(OCV)curves and estimate key ageing parameters at both the cell and electrode levels.These parameters include available capacity,electrode capacities,and lithium inventory capacity.The proposed method integrates OCV reconstruction models as functional components into convolutional neural networks(CNNs)and is validated using a public dataset.The results reveal that the estimated ageing parameters closely align with those obtained through offline OCV tests,with errors in reconstructed OCV curves remaining within 15 mV.This demonstrates the ability of the method to deliver fast and accurate degradation diagnostics at the electrode level,advancing the potential for precise and efficient battery health management.
基金supported in part by Multimedia University under the Research Fellow Grant MMUI/250008in part by Telekom Research&Development Sdn Bhd underGrantRDTC/241149Princess Nourah bint Abdulrahman University Researchers Supporting Project number(PNURSP2025R140),Princess Nourah bint Abdulrahman University,Riyadh,Saudi Arabia.
文摘The Internet of Things(IoT)ecosystem faces growing security challenges because it is projected to have 76.88 billion devices by 2025 and $1.4 trillion market value by 2027,operating in distributed networks with resource limitations and diverse system architectures.The current conventional intrusion detection systems(IDS)face scalability problems and trust-related issues,but blockchain-based solutions face limitations because of their low transaction throughput(Bitcoin:7 TPS(Transactions Per Second),Ethereum:15-30 TPS)and high latency.The research introduces MBID(Multi-Tier Blockchain Intrusion Detection)as a groundbreaking Multi-Tier Blockchain Intrusion Detection System with AI-Enhanced Detection,which solves the problems in huge IoT networks.The MBID system uses a four-tier architecture that includes device,edge,fog,and cloud layers with blockchain implementations and Physics-Informed Neural Networks(PINNs)for edge-based anomaly detection and a dual consensus mechanism that uses Honesty-based Distributed Proof-of-Authority(HDPoA)and Delegated Proof of Stake(DPoS).The system achieves scalability and efficiency through the combination of dynamic sharding and Interplanetary File System(IPFS)integration.Experimental evaluations demonstrate exceptional performance,achieving a detection accuracy of 99.84%,an ultra-low false positive rate of 0.01% with a False Negative Rate of 0.15%,and a near-instantaneous edge detection latency of 0.40 ms.The system demonstrated an aggregate throughput of 214.57 TPS in a 3-shard configuration,providing a clear,evidence-based path for horizontally scaling to support overmillions of devices with exceeding throughput.The proposed architecture represents a significant advancement in blockchain-based security for IoT networks,effectively balancing the trade-offs between scalability,security,and decentralization.
基金funded by the National Key R&D Program of China,Grant No.2024YFF0504904.
文摘The packaging quality of coaxial laser diodes(CLDs)plays a pivotal role in determining their optical performance and long-term reliability.As the core packaging process,high-precision laser welding requires precise control of process parameters to suppress optical power loss.However,the complex nonlinear relationship between welding parameters and optical power loss renders traditional trial-and-error methods inefficient and imprecise.To address this challenge,a physics-informed(PI)and data-driven collaboration approach for welding parameter optimization is proposed.First,thermal-fluid-solid coupling finite element method(FEM)was employed to quantify the sensitivity of welding parameters to physical characteristics,including residual stress.This analysis facilitated the identification of critical factors contributing to optical power loss.Subsequently,a Gaussian process regression(GPR)model incorporating finite element simulation prior knowledge was constructed based on the selected features.By introducing physics-informed kernel(PIK)functions,stress distribution patterns were embedded into the prediction model,achieving high-precision optical power loss prediction.Finally,a Bayesian optimization(BO)algorithm with an adaptive sampling strategy was implemented for efficient parameter space exploration.Experimental results demonstrate that the proposedmethod effectively establishes explicit physical correlations between welding parameters and optical power loss.The optimized welding parameters reduced optical power loss by 34.1%,providing theoretical guidance and technical support for reliable CLD packaging.
文摘Energy efficiency stands as an essential factor when implementing deep reinforcement learning(DRL)policies for robotic control systems.Standard algorithms,including Deep Deterministic Policy Gradient(DDPG),primarily optimize task rewards but at the cost of excessively high energy consumption,making them impractical for real-world robotic systems.To address this limitation,we propose Physics-Informed DDPG(PI-DDPG),which integrates physics-based energy penalties to develop energy-efficient yet high-performing control policies.The proposed method introduces adaptive physics-informed constraints through a dynamic weighting factor(λ),enabling policies that balance reward maximization with energy savings.Our motivation is to overcome the impracticality of rewardonly optimization by designing controllers that achieve competitive performance while substantially reducing energy consumption.PI-DDPG was evaluated in nine MuJoCo continuous control environments,where it demonstrated significant improvements in energy efficiency without compromising stability or performance.Experimental results confirm that PI-DDPG substantially reduces energy consumption compared to standard DDPG,while maintaining competitive task performance.For instance,energy costs decreased from 5542.98 to 3119.02 in HalfCheetah-v4 and from1909.13 to 1586.75 in Ant-v4,with stable performance in Hopper-v4(205.95 vs.130.82)and InvertedPendulum-v4(322.97 vs.311.29).Although DDPG sometimes yields higher rewards,such as in HalfCheetah-v4(5695.37 vs.4894.59),it requires significantly greater energy expenditure.These results highlight PI-DDPG as a promising energy-conscious alternative for robotic control.
基金supported by the National Natural Science Foundation of China(Grant No.92152301)the National Key Research and Development Program of China(Grant No.2022YFB4300200)the Shaanxi Provincial Key Research and Development Program(Grant No.2023-ZDLGY-27).
文摘Physics-informed neural networks(PINNs)have shown remarkable prospects in solving the forward and inverse problems involving partial differential equations(PDEs).The method embeds PDEs into the neural network by calculating the PDE loss at a set of collocation points,providing advantages such as meshfree and more convenient adaptive sampling.However,when solving PDEs using nonuniform collocation points,PINNs still face challenge regarding inefficient convergence of PDE residuals or even failure.In this work,we first analyze the ill-conditioning of the PDE loss in PINNs under nonuniform collocation points.To address the issue,we define volume weighting residual and propose volume weighting physics-informed neural networks(VW-PINNs).Through weighting the PDE residuals by the volume that the collocation points occupy within the computational domain,we embed explicitly the distribution characteristics of collocation points in the loss evaluation.The fast and sufficient convergence of the PDE residuals for the problems involving nonuniform collocation points is guaranteed.Considering the meshfree characteristics of VW-PINNs,we also develop a volume approximation algorithm based on kernel density estimation to calculate the volume of the collocation points.We validate the universality of VW-PINNs by solving the forward problems involving flow over a circular cylinder and flow over the NACA0012 airfoil under different inflow conditions,where conventional PINNs fail.By solving the Burgers’equation,we verify that VW-PINNs can enhance the efficiency of existing the adaptive sampling method in solving the forward problem by three times,and can reduce the relative L 2 error of conventional PINNs in solving the inverse problem by more than one order of magnitude.
基金Project supported by the Basic Science Research Program through the National Research Foundation(NRF)of Korea funded by the Ministry of Science and ICT(No.RS-2024-00337001)。
文摘Enforcing initial and boundary conditions(I/BCs)poses challenges in physics-informed neural networks(PINNs).Several PINN studies have gained significant achievements in developing techniques for imposing BCs in static problems;however,the simultaneous enforcement of I/BCs in dynamic problems remains challenging.To overcome this limitation,a novel approach called decoupled physics-informed neural network(d PINN)is proposed in this work.The d PINN operates based on the core idea of converting a partial differential equation(PDE)to a system of ordinary differential equations(ODEs)via the space-time decoupled formulation.To this end,the latent solution is expressed in the form of a linear combination of approximation functions and coefficients,where approximation functions are admissible and coefficients are unknowns of time that must be solved.Subsequently,the system of ODEs is obtained by implementing the weighted-residual form of the original PDE over the spatial domain.A multi-network structure is used to parameterize the set of coefficient functions,and the loss function of d PINN is established based on minimizing the residuals of the gained ODEs.In this scheme,the decoupled formulation leads to the independent handling of I/BCs.Accordingly,the BCs are automatically satisfied based on suitable selections of admissible functions.Meanwhile,the original ICs are replaced by the Galerkin form of the ICs concerning unknown coefficients,and the neural network(NN)outputs are modified to satisfy the gained ICs.Several benchmark problems involving different types of PDEs and I/BCs are used to demonstrate the superior performance of d PINN compared with regular PINN in terms of solution accuracy and computational cost.
基金supported by the National Natural Science Foundation of China(Grant Nos.12072105,11932006,and 52308498)the Natural Science Foundation of Jiangsu Province,China(Grant No.BK20220976).
文摘Dynamic wake field information is vital for the optimized design and control of wind farms.Combined with sparse measurement data from light detection and ranging(LiDAR),the physics-informed neural network(PINN)frameworks have recently been employed for forecasting freestream wind and wake fields.However,these PINN frameworks face challenges of low prediction accuracy and long training times.Therefore,this paper constructed a PINN framework for dynamic wake field prediction by integrating two accuracy improvement strategies and a step-by-step training time saving strategy.The results showed that the different performance improvement routes significantly improved the overall performance of the PINN.The accuracy and efficiency of the PINN with spatiotemporal improvement strategies were validated via LiDAR-measured data from a wind farm in Shandong province,China.This paper sheds light on load reduction,efficiency improvement,intelligent operation and maintenance of wind farms.
基金supported by the National Natural Science Foundation of China(Grant No.U2241263)the fellowship of China Postdoctoral Science Foundation(Grant No.2024M750310).
文摘Numerical simulation plays an important role in the dynamic analysis of multibody system.With the rapid development of computer science,the numerical solution technology has been further developed.Recently,data-driven method has become a very popular computing method.However,due to lack of necessary mechanism information of the traditional pure data-driven methods based on neural network,its numerical accuracy cannot be guaranteed for strong nonlinear system.Therefore,this work proposes a mechanism-data hybrid-driven strategy for solving nonlinear multibody system based on physics-informed neural network to overcome the limitation of traditional data-driven methods.The strategy proposed in this paper introduces scaling coefficients to introduce the dynamic model of multibody system into neural network,ensuring that the training results of neural network conform to the mechanics principle of the system,thereby ensuring the good reliability of the data-driven method.Finally,the stability,generalization ability and numerical accuracy of the proposed method are discussed and analyzed using three typical multibody systems,and the constrained default situations can be controlled within the range of 10^(-2)-10^(-4).
基金Supported by the National Natural Science Foundation of China(No.62304022)。
文摘This research introduces a spectrum-based physics-informed neural network(SP-PINN)model to significantly improve the accuracy of calculation of two-phase flow parameters,surpassing existing methods that have limitations in global and continuous data sampling.SP-PINNs address the challenges of traditional methods in terms of continuous sampling by integrating the spectral analysis and pressure correction into the Navier-Stokes(N-S)equations,enhancing the predictive accuracy especially in critical regions like gas-phase boundaries and velocity peaks.The novel introduction of a pressure-correction module within SP-PINNs mitigates prediction errors,achieving a substantial reduction to 1‰compared with the conventional physics-informed neural network(PINN)approaches.Experimental applications validate the model’s ability to accurately and rapidly predict flow parameters with different sampling time intervals,with the computation time of predicting unsampled data less than 0.01 s.Such advancements signify a 100-fold improvement over traditional DNS calculations,underscoring the model’s potential in the real-time calculation and analysis of multiphase flow dynamics.
基金supported by the National Natural Science Foundation of China(Grant No.12071042)the Beijing Natural Science Foundation(Grant No.1202004)Promoting the Development of University Classification-Student Innovation and Entrepreneurship Training Programme(Grant No.5112410857)。
文摘In this paper,we propose a symmetric difference data enhancement physics-informed neural network(SDE-PINN)to study soliton solutions for discrete nonlinear lattice equations(NLEs).By considering known and unknown symmetric points,numerical simulations are conducted to one-soliton and two-soliton solutions of a discrete KdV equation,as well as a one-soliton solution of a discrete Toda lattice equation.Compared with the existing discrete deep learning approach,the numerical results reveal that within the specified spatiotemporal domain,the prediction accuracy by SDE-PINN is excellent regardless of the interior or extrapolation prediction,with a significant reduction in training time.The proposed data enhancement technique and symmetric structure development provides a new perspective for the deep learning approach to solve discrete NLEs.The newly proposed SDE-PINN can also be applied to solve continuous nonlinear equations and other discrete NLEs numerically.
基金supported by the National Natural Science Foundation of China (No. 52274048)Beijing Natural Science Foundation (No. 3222037)。
文摘With the rapid development of deep learning neural networks,new solutions have emerged for addressing fluid flow problems in porous media.Combining data-driven approaches with physical constraints has become a hot research direction,with physics-informed neural networks(PINNs) being the most popular hybrid model.PINNs have gained widespread attention in subsurface fluid flow simulations due to their low computational resource requirements,fast training speeds,strong generalization capabilities,and broad applicability.Despite success in homogeneous settings,standard PINNs face challenges in accurately calculating flux between irregular Eulerian cells with disparate properties and capturing global field influences on local cells.This limits their suitability for heterogeneous reservoirs and the irregular Eulerian grids frequently used in reservoir.To address these challenges,this study proposes a physics-informed graph neural network(PIGNN) model.The PIGNN model treats the entire field as a whole,integrating information from neighboring grids and physical laws into the solution for the target grid,thereby improving the accuracy of solving partial differential equations in heterogeneous and Eulerian irregular grids.The optimized model was applied to pressure field prediction in a spatially heterogeneous reservoir,achieving an average L_(2) error and R_(2) score of 6.710×10^(-4)and 0.998,respectively,which confirms the effectiveness of model.Compared to the conventional PINN model,the average L_(2) error was reduced by 76.93%,the average R_(2) score increased by 3.56%.Moreover,evaluating robustness,training the PIGNN model using only 54% and 76% of the original data yielded average relative L_(2) error reductions of 58.63% and 56.22%,respectively,compared to the PINN model.These results confirm the superior performance of this approach compared to PINN.
基金Supported by the Fundamental Research Funds for the Central Universities(Grant Nos.2232024Y-01,LZB2023001)DHU Distinguished Young Professor Program+1 种基金National Natural Science Foundation of China(Grant No.52275478)AI-Enhanced Research Program of Shanghai Municipal Education Commission(Grant No.SMEC-AI-DHUY-05)。
文摘Shape prediction of deformable linear objects(DLO)plays critical roles in robotics,medical devices,aerospace,and manufacturing,especially in manipulating objects such as cables,wires,and fibers.Due to the inherent flexibility of DLO and their complex deformation behaviors,such as bending and torsion,it is challenging to predict their dynamic characteristics accurately.Although the traditional physical modeling method can simulate the complex deformation behavior of DLO,the calculation cost is high and it is difficult to meet the demand of real-time prediction.In addition,the scarcity of data resources also limits the prediction accuracy of existing models.To solve these problems,a method of fiber shape prediction based on a physical information graph neural network(PIGNN)is proposed in this paper.This method cleverly combines the powerful expressive power of graph neural networks with the strict constraints of physical laws.Specifically,we learn the initial deformation model of the fiber through graph neural networks(GNN)to provide a good initial estimate for the model,which helps alleviate the problem of data resource scarcity.During the training process,we incorporate the physical prior knowledge of the dynamic deformation of the fiber optics into the loss function as a constraint,which is then fed back to the network model.This ensures that the shape of the fiber optics gradually approaches the true target shape,effectively solving the complex nonlinear behavior prediction problem of deformable linear objects.Experimental results demonstrate that,compared to traditional methods,the proposed method significantly reduces execution time and prediction error when handling the complex deformations of deformable fibers.This showcases its potential application value and superiority in fiber manipulation.
基金King Abdullah University of Science and Technol-ogy(KAUST)for supporting this research and the Seismic Wave Anal-ysis group for the supportive and encouraging environment.
文摘Physics-informed neural networks(PINNs)are promising to replace conventional mesh-based partial tial differen-equation(PDE)solvers by offering more accurate and flexible PDE solutions.However,PINNs are hampered by the relatively slow convergence and the need to perform additional,potentially expensive training for new PDE parameters.To solve this limitation,we introduce LatentPINN,a framework that utilizes latent representations of the PDE parameters as additional(to the coordinates)inputs into PINNs and allows for training over the distribution of these parameters.Motivated by the recent progress on generative models,we promote using latent diffusion models to learn compressed latent representations of the distribution of PDE parameters as they act as input parameters for NN functional solutions.We use a two-stage training scheme in which,in the first stage,we learn the latent representations for the distribution of PDE parameters.In the second stage,we train a physics-informed neural network over inputs given by randomly drawn samples from the coordinate space within the solution domain and samples from the learned latent representation of the PDE parameters.Considering their importance in capturing evolving interfaces and fronts in various fields,we test the approach on a class of level set equations given,for example,by the nonlinear Eikonal equation.We share results corresponding to three Eikonal parameters(velocity models)sets.The proposed method performs well on new phase velocity models without the need for any additional training.
基金supported by the Scientific Research and Innovation Team Program of Sichuan University of Science and Engineering(No.SUSE652B005)Anhui Province Applied Peak Discipline Mechanical Engineering(No.XK-XJGF004)。
文摘Accurately estimating the battery state of health(SOH)is essential for ensuring the safe and reliable operation of battery systems of electric vehicles.However,due to the complex and variable operating conditions encountered in practical applications,achieving precise and physics-informed SOH estimation remains challenging.To address these problems,this paper develops a lightweight two-stage physicsinformed neural network(TSPINN)method for SOH estimation of lithium-ion batteries with different chemistries.Specifically,this paper utilizes firstly relaxation voltage data obtained after a full charge to determine the aging-related parameters of physical equivalent circuit model(ECM).Additionally,incremental capacity(IC)feature is extracted by analyzing peak values of the IC curve during the charging phase,which thereby constitutes the first stage of the proposed TSPINN,termed as physics-informed data augmentation for SOH estimation.Additionally,the physical information can be further embedded by incorporating feature knowledge related to mechanisms into the loss function,and ultimately,the second stage of the proposed TSPINN is developed,which is named the physics-informed loss function.The effectiveness of the TSPINN method was confirmed through the experimental data for LiNi_(0.86)Co_(0.11)Al_(0.03)O_(2)(NCA)and LiNi_(0.83)Co_(0.11)Mn_(0.07)O_(2)(NCM)battery materials under different temperature conditions.The final experimental results indicate that the TSPINN method achieved SOH estimation with a mean absolute error(MAE)of 0.675%,showing improvements of approximately 29.3%,60.3%,and 8.1% compared to methods using only ECM,IC,and integrated features,respectively.The results validate the effectiveness and adaptability of TSPINN,establishing it as a reliable solution for advanced battery management systems.
基金The research of the second author(ZM)was sup-539 ported by the National Natural Science Foundation of China(Grant 54012171404)The last author(GEK)would like to acknowledge support 541 by the Alexander von Humboldt fellowship.
文摘Despite the significant progress over the last 50 years in simulating flow problems using numerical discretization of the Navier–Stokes equations(NSE),we still cannot incorporate seamlessly noisy data into existing algorithms,mesh-generation is complex,and we cannot tackle high-dimensional problems governed by parametrized NSE.Moreover,solving inverse flow problems is often prohibitively expensive and requires complex and expensive formulations and new computer codes.Here,we review flow physics-informed learning,integrating seamlessly data and mathematical models,and implement them using physics-informed neural networks(PINNs).We demonstrate the effectiveness of PINNs for inverse problems related to three-dimensional wake flows,supersonic flows,and biomedical flows.
文摘Physics-informed deep learning has drawn tremendous interest in recent years to solve computational physics problems,whose basic concept is to embed physical laws to constrain/inform neural networks,with the need of less data for training a reliable model.This can be achieved by incorporating the residual of physics equations into the loss function.Through minimizing the loss function,the network could approximate the solution.In this paper,we propose a mixed-variable scheme of physics-informed neural network(PINN)for fluid dynamics and apply it to simulate steady and transient laminar flows at low Reynolds numbers.A parametric study indicates that the mixed-variable scheme can improve the PINN trainability and the solution accuracy.The predicted velocity and pressure fields by the proposed PINN approach are also compared with the reference numerical solutions.Simulation results demonstrate great potential of the proposed PINN for fluid flow simulation with a high accuracy.
基金funded by the Artificial Intelligence Technology Project of Xi’an Science and Technology Bureau in China(No.21RGZN0014)。
文摘Accurate insight into the heat generation rate(HGR) of lithium-ion batteries(LIBs) is one of key issues for battery management systems to formulate thermal safety warning strategies in advance.For this reason,this paper proposes a novel physics-informed neural network(PINN) approach for HGR estimation of LIBs under various driving conditions.Specifically,a single particle model with thermodynamics(SPMT) is first constructed for extracting the critical physical knowledge related with battery HGR.Subsequently,the surface concentrations of positive and negative electrodes in battery SPMT model are integrated into the bidirectional long short-term memory(BiLSTM) networks as physical information.And combined with other feature variables,a novel PINN approach to achieve HGR estimation of LIBs with higher accuracy is constituted.Additionally,some critical hyperparameters of BiLSTM used in PINN approach are determined through Bayesian optimization algorithm(BOA) and the results of BOA-based BiLSTM are compared with other traditional BiLSTM/LSTM networks.Eventually,combined with the HGR data generated from the validated virtual battery,it is proved that the proposed approach can well predict the battery HGR under the dynamic stress test(DST) and worldwide light vehicles test procedure(WLTP),the mean absolute error under DST is 0.542 kW/m^(3),and the root mean square error under WLTP is1.428 kW/m^(3)at 25℃.Lastly,the investigation results of this paper also show a new perspective in the application of the PINN approach in battery HGR estimation.
基金Project supported by the National Key R&D Program of China(No.2022YFA1004504)the National Natural Science Foundation of China(Nos.12171404 and 12201229)the Fundamental Research Funds for Central Universities of China(No.20720210037)。
文摘We consider solving the forward and inverse partial differential equations(PDEs)which have sharp solutions with physics-informed neural networks(PINNs)in this work.In particular,to better capture the sharpness of the solution,we propose the adaptive sampling methods(ASMs)based on the residual and the gradient of the solution.We first present a residual only-based ASM denoted by ASMⅠ.In this approach,we first train the neural network using a small number of residual points and divide the computational domain into a certain number of sub-domains,then we add new residual points in the sub-domain which has the largest mean absolute value of the residual,and those points which have the largest absolute values of the residual in this sub-domain as new residual points.We further develop a second type of ASM(denoted by ASMⅡ)based on both the residual and the gradient of the solution due to the fact that only the residual may not be able to efficiently capture the sharpness of the solution.The procedure of ASMⅡis almost the same as that of ASMⅠ,and we add new residual points which have not only large residuals but also large gradients.To demonstrate the effectiveness of the present methods,we use both ASMⅠand ASMⅡto solve a number of PDEs,including the Burger equation,the compressible Euler equation,the Poisson equation over an Lshape domain as well as the high-dimensional Poisson equation.It has been shown from the numerical results that the sharp solutions can be well approximated by using either ASMⅠor ASMⅡ,and both methods deliver much more accurate solutions than the original PINNs with the same number of residual points.Moreover,the ASMⅡalgorithm has better performance in terms of accuracy,efficiency,and stability compared with the ASMⅠalgorithm.This means that the gradient of the solution improves the stability and efficiency of the adaptive sampling procedure as well as the accuracy of the solution.Furthermore,we also employ the similar adaptive sampling technique for the data points of boundary conditions(BCs)if the sharpness of the solution is near the boundary.The result of the L-shape Poisson problem indicates that the present method can significantly improve the efficiency,stability,and accuracy.
基金The research is supported by internal funding from SINTEF through a strategic project focusing on Machine Learning and Digitalization in the infrastructure sector.
文摘Neural networks with physical governing equations as constraints have recently created a new trend in machine learning research.In this context,a review of related research is first presented and discussed.The potential offered by such physics-informed deep learning models for computations in geomechanics is demonstrated by application to one-dimensional(1D)consolidation.The governing equation for 1D problems is applied as a constraint in the deep learning model.The deep learning model relies on automatic differentiation for applying the governing equation as a constraint,based on the mathematical approximations established by the neural network.The total loss is measured as a combination of the training loss(based on analytical and model predicted solutions)and the constraint loss(a requirement to satisfy the governing equation).Two classes of problems are considered:forward and inverse problems.The forward problems demonstrate the performance of a physically constrained neural network model in predicting solutions for 1D consolidation problems.Inverse problems show prediction of the coefficient of consolidation.Terzaghi’s problem,with varying boundary conditions,is used as a numerical example and the deep learning model shows a remarkable performance in both the forward and inverse problems.While the application demonstrated here is a simple 1D consolidation problem,such a deep learning model integrated with a physical law has significant implications for use in,such as,faster realtime numerical prediction for digital twins,numerical model reproducibility and constitutive model parameter optimization.
基金funded by the National Natural Science Foundation of China(General Program:No.52074314,No.U19B6003-05)National Key Research and Development Program of China(2019YFA0708303-05)。
文摘Accurate prediction of formation pore pressure is essential to predict fluid flow and manage hydrocarbon production in petroleum engineering.Recent deep learning technique has been receiving more interest due to the great potential to deal with pore pressure prediction.However,most of the traditional deep learning models are less efficient to address generalization problems.To fill this technical gap,in this work,we developed a new adaptive physics-informed deep learning model with high generalization capability to predict pore pressure values directly from seismic data.Specifically,the new model,named CGP-NN,consists of a novel parametric features extraction approach(1DCPP),a stacked multilayer gated recurrent model(multilayer GRU),and an adaptive physics-informed loss function.Through machine training,the developed model can automatically select the optimal physical model to constrain the results for each pore pressure prediction.The CGP-NN model has the best generalization when the physicsrelated metricλ=0.5.A hybrid approach combining Eaton and Bowers methods is also proposed to build machine-learnable labels for solving the problem of few labels.To validate the developed model and methodology,a case study on a complex reservoir in Tarim Basin was further performed to demonstrate the high accuracy on the pore pressure prediction of new wells along with the strong generalization ability.The adaptive physics-informed deep learning approach presented here has potential application in the prediction of pore pressures coupled with multiple genesis mechanisms using seismic data.
