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
Nowadays,there has been an increasing focus on integrated flight propulsion control and the inlet-exhaust design for the aero-propulsion system.Traditional component-level models are inadequate due to installed perfor...Nowadays,there has been an increasing focus on integrated flight propulsion control and the inlet-exhaust design for the aero-propulsion system.Traditional component-level models are inadequate due to installed performance deviations and mismatches between the real engine and the model,failing to meet the accuracy requirements of supersonic conditions.This paper establishes a quasi-one-dimensional model for the inlet-exhaust system and conducts experimental calibration.Additionally,a mechanism-data fusion adaptive modeling scheme using an Extreme Learning Machine based on the Salp Swarm Algorithm(SSA-ELM)is proposed.The study reveals the inlet model’s efficacy in reflecting installed performance,flow matching,and mitigating pressure distortion,while the nozzle model accurately predicts flow coefficients and thrust coefficients,and identifies various operational states.The model’s output closely aligns with typical experimental parameters.By combining offline optimization and online adaptive correction,the mechanismdata fusion adaptive model substantially reduces output errors during regular flights and varying levels of degradation,and effectively handles gradual degradation within a single flight cycle.Notably,the mechanism-data fusion adaptive model holistically addresses total pressure errors within the inlet-exhaust system and normal shock location correction.This approach significantly curbs performance deviations in supersonic conditions.For example,at Ma=2.0,the system error impressively drops from 34.17%to merely 6.54%,while errors for other flight conditions consistently stay below the 2.95%threshold.These findings underscore the clear superiority of the proposed method.展开更多
The real-time prediction of head warping and lower buckling during the production process of rough-rolled slabs has long been a persistent technical problem at the production site.An online real-time prediction system...The real-time prediction of head warping and lower buckling during the production process of rough-rolled slabs has long been a persistent technical problem at the production site.An online real-time prediction system was proposed for head warping and lower buckling of rough-rolled slab based on mechanism-data dual drive.The modified Johnson–Cook constitutive model was derived and established to provide parameter support for the finite element simulation.Visual detection technology was employed to provide data.Industrial applications showed that the prediction accuracy of the prediction system described is as follows:prediction error≤±3 mm,type prediction rate≥98%.Moreover,the head warping and lower buckling of slab in the production site have been significantly improved,and the shape quality of slab has been increased by 3 times compared with that before adjustment,which meet site production requirements.展开更多
Digital twin technology used to realize the interactive mapping between digital model and physical entity in virtual space plays a crucial role in promoting the transformation of battery management to digitalization a...Digital twin technology used to realize the interactive mapping between digital model and physical entity in virtual space plays a crucial role in promoting the transformation of battery management to digitalization and intelligence.The key to achieving a digital twin is developing a virtual model that can accurately reflect the physical object.However,the intricate time-varying and nonlinear reaction characteristics within the battery often pose challenges in simulating complex operational conditions and maintaining high accuracy throughout the full life cycle.Incremental learning algorithms are suitable for online streaming data processing and can adapt to concept drift of the data stream by receiving new data online without retraining the entire model from scratch.This paper employs a simplified electrochemical model of the battery and integrates an Aggregated Mondrian Forest with incremental learning capabilities to construct a hybrid mechanism-data battery model.The developed hybrid model can acquire battery data and train in real time during battery operation to realize co-evolution with the physical battery to ensure the voltage prediction accuracy during the full life cycle of the battery.展开更多
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co-supported by the National Natural Science Foundation of China(Nos.61890921,61890924)the National Science and Technology Major Project,China(No.J2019-1-0019-0018).
文摘Nowadays,there has been an increasing focus on integrated flight propulsion control and the inlet-exhaust design for the aero-propulsion system.Traditional component-level models are inadequate due to installed performance deviations and mismatches between the real engine and the model,failing to meet the accuracy requirements of supersonic conditions.This paper establishes a quasi-one-dimensional model for the inlet-exhaust system and conducts experimental calibration.Additionally,a mechanism-data fusion adaptive modeling scheme using an Extreme Learning Machine based on the Salp Swarm Algorithm(SSA-ELM)is proposed.The study reveals the inlet model’s efficacy in reflecting installed performance,flow matching,and mitigating pressure distortion,while the nozzle model accurately predicts flow coefficients and thrust coefficients,and identifies various operational states.The model’s output closely aligns with typical experimental parameters.By combining offline optimization and online adaptive correction,the mechanismdata fusion adaptive model substantially reduces output errors during regular flights and varying levels of degradation,and effectively handles gradual degradation within a single flight cycle.Notably,the mechanism-data fusion adaptive model holistically addresses total pressure errors within the inlet-exhaust system and normal shock location correction.This approach significantly curbs performance deviations in supersonic conditions.For example,at Ma=2.0,the system error impressively drops from 34.17%to merely 6.54%,while errors for other flight conditions consistently stay below the 2.95%threshold.These findings underscore the clear superiority of the proposed method.
基金supported by Regional Joint Funds of the National Natural Science Foundation of China(U20A20289)Innovative Talents International Cooperative Training Project of China Scholarship Council(CXXM20240010)+2 种基金The General Program of National Natural Science Foundation of China(52075471)Innovation Capacity Enhancement Program of Hebei Province(24461901D)National Natural Science Foundation of China(Grant No.52475409).
文摘The real-time prediction of head warping and lower buckling during the production process of rough-rolled slabs has long been a persistent technical problem at the production site.An online real-time prediction system was proposed for head warping and lower buckling of rough-rolled slab based on mechanism-data dual drive.The modified Johnson–Cook constitutive model was derived and established to provide parameter support for the finite element simulation.Visual detection technology was employed to provide data.Industrial applications showed that the prediction accuracy of the prediction system described is as follows:prediction error≤±3 mm,type prediction rate≥98%.Moreover,the head warping and lower buckling of slab in the production site have been significantly improved,and the shape quality of slab has been increased by 3 times compared with that before adjustment,which meet site production requirements.
基金the Shandong Province National Natural Science Foundation of China(No.ZR2023QE036)the Natural Science Foundation of Jiangsu Province(No.BK20210600)for their financial support.
文摘Digital twin technology used to realize the interactive mapping between digital model and physical entity in virtual space plays a crucial role in promoting the transformation of battery management to digitalization and intelligence.The key to achieving a digital twin is developing a virtual model that can accurately reflect the physical object.However,the intricate time-varying and nonlinear reaction characteristics within the battery often pose challenges in simulating complex operational conditions and maintaining high accuracy throughout the full life cycle.Incremental learning algorithms are suitable for online streaming data processing and can adapt to concept drift of the data stream by receiving new data online without retraining the entire model from scratch.This paper employs a simplified electrochemical model of the battery and integrates an Aggregated Mondrian Forest with incremental learning capabilities to construct a hybrid mechanism-data battery model.The developed hybrid model can acquire battery data and train in real time during battery operation to realize co-evolution with the physical battery to ensure the voltage prediction accuracy during the full life cycle of the battery.
