Cooperative multi-agent reinforcement learning(MARL)is a key technology for enabling cooperation in complex multi-agent systems.It has achieved remarkable progress in areas such as gaming,autonomous driving,and multi-...Cooperative multi-agent reinforcement learning(MARL)is a key technology for enabling cooperation in complex multi-agent systems.It has achieved remarkable progress in areas such as gaming,autonomous driving,and multi-robot control.Empowering cooperative MARL with multi-task decision-making capabilities is expected to further broaden its application scope.In multi-task scenarios,cooperative MARL algorithms need to address 3 types of multi-task problems:reward-related multi-task,arising from different reward functions;multi-domain multi-task,caused by differences in state and action spaces,state transition functions;and scalability-related multi-task,resulting from the dynamic variation in the number of agents.Most existing studies focus on scalability-related multitask problems.However,with the increasing integration between large language models(LLMs)and multi-agent systems,a growing number of LLM-based multi-agent systems have emerged,enabling more complex multi-task cooperation.This paper provides a comprehensive review of the latest advances in this field.By combining multi-task reinforcement learning with cooperative MARL,we categorize and analyze the 3 major types of multi-task problems under multi-agent settings,offering more fine-grained classifications and summarizing key insights for each.In addition,we summarize commonly used benchmarks and discuss future directions of research in this area,which hold promise for further enhancing the multi-task cooperation capabilities of multi-agent systems and expanding their practical applications in the real world.展开更多
Deep neural networks(DNNs)are effective in solving both forward and inverse problems for nonlinear partial differential equations(PDEs).However,conventional DNNs are not effective in handling problems such as delay di...Deep neural networks(DNNs)are effective in solving both forward and inverse problems for nonlinear partial differential equations(PDEs).However,conventional DNNs are not effective in handling problems such as delay differential equations(DDEs)and delay integrodifferential equations(DIDEs)with constant delays,primarily due to their low regularity at delayinduced breaking points.In this paper,a DNN method that combines multi-task learning(MTL)which is proposed to solve both the forward and inverse problems of DIDEs.The core idea of this approach is to divide the original equation into multiple tasks based on the delay,using auxiliary outputs to represent the integral terms,followed by the use of MTL to seamlessly incorporate the properties at the breaking points into the loss function.Furthermore,given the increased training dificulty associated with multiple tasks and outputs,we employ a sequential training scheme to reduce training complexity and provide reference solutions for subsequent tasks.This approach significantly enhances the approximation accuracy of solving DIDEs with DNNs,as demonstrated by comparisons with traditional DNN methods.We validate the effectiveness of this method through several numerical experiments,test various parameter sharing structures in MTL and compare the testing results of these structures.Finally,this method is implemented to solve the inverse problem of nonlinear DIDE and the results show that the unknown parameters of DIDE can be discovered with sparse or noisy data.展开更多
This paper proposes that China,under the challenge of balancing its development and security,can aim for the Paris Agreement's goal to limit global warming to no more than 2℃by actively seeking carbonpeak and car...This paper proposes that China,under the challenge of balancing its development and security,can aim for the Paris Agreement's goal to limit global warming to no more than 2℃by actively seeking carbonpeak and carbon-neutrality pathways that align with China's national conditions,rather than following the idealized path toward the 1.5℃target by initially relying on extensive negative-emission technologies such as direct air carbon capture and storage(DACCS).This work suggests that pursuing a 1.5℃target is increasingly less feasible for China,as it would potentially incur 3–4 times the cost of pursuing the 2℃target.With China being likely to achieve a peak in its emissions around 2028,at about 12.8 billion tonnes of anthropogenic carbon dioxide(CO_(2)),and become carbon neutral,projected global warming levels may be less severe after the 2050s than previously estimated.This could reduce the risk potential of climate tipping points and extreme events,especially considering that the other two major carbon emitters in the world(Europe and North America)have already passed their carbon peaks.While natural carbon sinks will contribute to China's carbon neutrality efforts,they are not expected to be decisive in the transition stages.This research also addresses the growing focus on climate overshoot,tipping points,extreme events,loss and damage,and methane reductions in international climate cooperation,emphasizing the need to balance these issues with China's development,security,and fairness considerations.China's pursuit of carbon neutrality will have significant implications for global emissions scenarios,warming levels,and extreme event projections,as well as for climate change hotspots of international concern,such as climate tipping points,the climate crisis,and the notion that the world has moved from a warming to a boiling era.Possible research recommendations for global emissions scenarios based on China's 2℃target pathway are also summarized.展开更多
Predicting the health status of stroke patients at different stages of the disease is a critical clinical task.The onset and development of stroke are affected by an array of factors,encompassing genetic predispositio...Predicting the health status of stroke patients at different stages of the disease is a critical clinical task.The onset and development of stroke are affected by an array of factors,encompassing genetic predisposition,environmental exposure,unhealthy lifestyle habits,and existing medical conditions.Although existing machine learning-based methods for predicting stroke patients’health status have made significant progress,limitations remain in terms of prediction accuracy,model explainability,and system optimization.This paper proposes a multi-task learning approach based on Explainable Artificial Intelligence(XAI)for predicting the health status of stroke patients.First,we design a comprehensive multi-task learning framework that utilizes the task correlation of predicting various health status indicators in patients,enabling the parallel prediction of multiple health indicators.Second,we develop a multi-task Area Under Curve(AUC)optimization algorithm based on adaptive low-rank representation,which removes irrelevant information from the model structure to enhance the performance of multi-task AUC optimization.Additionally,the model’s explainability is analyzed through the stability analysis of SHAP values.Experimental results demonstrate that our approach outperforms comparison algorithms in key prognostic metrics F1 score and Efficiency.展开更多
Tropical cyclones(TCs)are one of the most serious types of natural disasters,and accurate TC activity predictions are key to disaster prevention and mitigation.Recently,TC track predictions have made significant progr...Tropical cyclones(TCs)are one of the most serious types of natural disasters,and accurate TC activity predictions are key to disaster prevention and mitigation.Recently,TC track predictions have made significant progress,but the ability to predict their intensity is obviously lagging behind.At present,research on TC intensity prediction takes atmospheric reanalysis data as the research object and mines the relationship between TC-related environmental factors and intensity through deep learning.However,reanalysis data are non-real-time in nature,which does not meet the requirements for operational forecasting applications.Therefore,a TC intensity prediction model named TC-Rolling is proposed,which can simultaneously extract the degree of symmetry for strong TC convective cloud and convection intensity,and fuse the deviation-angle variance with satellite images to construct the correlation between TC convection structure and intensity.For TCs'complex dynamic processes,a convolutional neural network(CNN)is used to learn their temporal and spatial features.For real-time intensity estimation,multi-task learning acts as an implicit time-series enhancement.The model is designed with a rolling strategy that aims to moderate the long-term dependent decay problem and improve accuracy for short-term intensity predictions.Since multiple tasks are correlated,the loss function of 12 h and 24 h are corrected.After testing on a sample of TCs in the Northwest Pacific,with a 4.48 kt root-mean-square error(RMSE)of 6 h intensity prediction,5.78 kt for 12 h,and 13.94 kt for 24 h,TC records from official agencies are used to assess the validity of TC-Rolling.展开更多
As the group-buying model shows significant progress in attracting new users,enhancing user engagement,and increasing platform profitability,providing personalized recommendations for group-buying users has emerged as...As the group-buying model shows significant progress in attracting new users,enhancing user engagement,and increasing platform profitability,providing personalized recommendations for group-buying users has emerged as a new challenge in the field of recommendation systems.This paper introduces a group-buying recommendation model based on multi-head attention mechanisms and multi-task learning,termed the Multi-head Attention Mechanisms and Multi-task Learning Group-Buying Recommendation(MAMGBR)model,specifically designed to optimize group-buying recommendations on e-commerce platforms.The core dataset of this study comes from the Chinese maternal and infant e-commerce platform“Beibei,”encompassing approximately 430,000 successful groupbuying actions and over 120,000 users.Themodel focuses on twomain tasks:recommending items for group organizers(Task Ⅰ)and recommending participants for a given group-buying event(Task Ⅱ).In model evaluation,MAMGBR achieves an MRR@10 of 0.7696 for Task I,marking a 20.23%improvement over baseline models.Furthermore,in Task II,where complex interaction patterns prevail,MAMGBR utilizes auxiliary loss functions to effectively model the multifaceted roles of users,items,and participants,leading to a 24.08%increase in MRR@100 under a 1:99 sample ratio.Experimental results show that compared to benchmark models,such as NGCF and EATNN,MAMGBR’s integration ofmulti-head attentionmechanisms,expert networks,and gating mechanisms enables more accurate modeling of user preferences and social associations within group-buying scenarios,significantly enhancing recommendation accuracy and platform group-buying success rates.展开更多
Aerosol optical depth(AOD)and fine particulate matter with a diameter of less than or equal to 2.5μm(PM_(2.5))play crucial roles in air quality,human health,and climate change.However,the complex correlation of AOD–...Aerosol optical depth(AOD)and fine particulate matter with a diameter of less than or equal to 2.5μm(PM_(2.5))play crucial roles in air quality,human health,and climate change.However,the complex correlation of AOD–PM_(2.5)and the limitations of existing algorithms pose a significant challenge in realizing the accurate joint retrieval of these two parameters at the same location.On this point,a multi-task learning(MTL)model,which enables the joint retrieval of PM_(2.5)concentration and AOD,is proposed and applied on the top-of-the-atmosphere reflectance data gathered by the Fengyun-4A Advanced Geosynchronous Radiation Imager(FY-4A AGRI),and compared to that of two single-task learning models—namely,Random Forest(RF)and Deep Neural Network(DNN).Specifically,MTL achieves a coefficient of determination(R^(2))of 0.88 and a root-mean-square error(RMSE)of 0.10 in AOD retrieval.In comparison to RF,the R^(2)increases by 0.04,the RMSE decreases by 0.02,and the percentage of retrieval results falling within the expected error range(Within-EE)rises by 5.55%.The R^(2)and RMSE of PM_(2.5)retrieval by MTL are 0.84 and 13.76μg m~(-3)respectively.Compared with RF,the R^(2)increases by 0.06,the RMSE decreases by 4.55μg m~(-3),and the Within-EE increases by 7.28%.Additionally,compared to DNN,MTL shows an increase of 0.01 in R^(2)and a decrease of 0.02 in RMSE in AOD retrieval,with a corresponding increase of 2.89%in Within-EE.For PM_(2.5)retrieval,MTL exhibits an increase of 0.05 in R^(2),a decrease of 1.76μg m~(-3)in RMSE,and an increase of 6.83%in Within-EE.The evaluation suggests that MTL is able to provide simultaneously improved AOD and PM_(2.5)retrievals,demonstrating a significant advantage in efficiently capturing the spatial distribution of PM_(2.5)concentration and AOD.展开更多
Accurate short-term forecast of offshore wind fields is still challenging for numerical weather prediction models.Based on three years of 48-hour forecast data from the European Centre for Medium-Range Weather Forecas...Accurate short-term forecast of offshore wind fields is still challenging for numerical weather prediction models.Based on three years of 48-hour forecast data from the European Centre for Medium-Range Weather Forecasts Integrated Forecasting System global model(ECMWF-IFS)over 14 offshore weather stations along the coast of Shandong Province,this study introduces a multi-task learning(MTL)model(TabNet-MTL),which significantly improves the forecast bias of near-surface wind direction and speed simultaneously.TabNet-MTL adopts the feature engineering method,utilizes mean square error as the loss function,and employs the 5-fold cross validation method to ensure the generalization ability of the trained model.It demonstrates superior skills in wind field correction across different forecast lead times over all stations compared to its single-task version(TabNet-STL)and three other popular single-task learning models(Random Forest,LightGBM,and XGBoost).Results show that it significantly reduces root mean square error of the ECMWF-IFS wind speed forecast from 2.20 to 1.25 m s−1,and increases the forecast accuracy of wind direction from 50%to 65%.As an explainable deep learning model,the weather stations and long-term temporal statistics of near-surface wind speed are identified as the most influential variables for TabNet-MTL in constructing its feature engineering.展开更多
The proposal of carbon neutrality target makes decarbonization and hydrogenation typical features of future energy development in China.With a wide range of application scenarios,hydrogen energy will experience rapid ...The proposal of carbon neutrality target makes decarbonization and hydrogenation typical features of future energy development in China.With a wide range of application scenarios,hydrogen energy will experience rapid growth in production and consumption.To formulate an effective hydrogen energy development strategy for the future of China,this study employs the departmental scenario analysis method to calculate and evaluate the future consumption of hydrogen energy in China’s heavy industry,transportation,electricity,and other related fields.Multidimensional technical parameters are selected and predicted accurately and reliably in combination with different development scenarios.The findings indicate that the period from 2030 to 2050 will enjoy rapid development of hydrogen energy,having an average annual growth rate of 2%to 4%.The technological progress and breakthroughs scenario has the greatest potential for hydrogen demand scale among the four development scenarios.Under this scenario,the total demand for hydrogen energy is expected to reach 446.37Mt in 2060.Thetransportation sector will be the sector with the greatest potential for hydrogen deployment growth from 2023 to 2060,which is expected to rise from 0.038Mt to about 163.18Mt,with the ambitious growth in the future.Additionally,hydrogen energy has a considerable development potential in the steel sector,and the trend of de-refueling coke by hydrogenation in this sector will be imperative for this energy-intensive industries.展开更多
Satellite communication technology has emerged as a key solution to address the challenges of data transmission in remote areas.By overcoming the limitations of traditional terrestrial communication networks,it enable...Satellite communication technology has emerged as a key solution to address the challenges of data transmission in remote areas.By overcoming the limitations of traditional terrestrial communication networks,it enables long-distance data transmission anytime and anywhere,ensuring the timely and accurate delivery of water level data,which is particularly crucial for fishway water level monitoring.To enhance the effectiveness of fishway water level monitoring,this study proposes a multi-task learning model,AS-SOMTF,designed for real-time and comprehensive prediction.The model integrates auxiliary sequences with primary input sequences to capture complex relationships and dependencies,thereby improving representational capacity.In addition,a novel timeseries embedding algorithm,AS-SOM,is introduced,which combines generative inference and pooling operations to optimize prediction efficiency for long sequences.This innovation not only ensures the timely transmission of water level data but also enhances the accuracy of real-time monitoring.Compared with traditional models such as Transformer and Long Short-Term Memory(LSTM)networks,the proposed model achieves improvements of 3.8%and 1.4%in prediction accuracy,respectively.These advancements provide more precise technical support for water level forecasting and resource management in the Diqing Tibetan Autonomous Prefecture of the Lancang River,contributing to ecosystem protection and improved operational safety.展开更多
With the popularization of microgrid construction and the connection of renewable energy sources to the power system,the problem of source and load uncertainty faced by the coordinated operation of multi-microgrid is ...With the popularization of microgrid construction and the connection of renewable energy sources to the power system,the problem of source and load uncertainty faced by the coordinated operation of multi-microgrid is becoming increasingly prominent,and the accuracy of typical scenario predictions is low.In order to improve the accuracy of scenario prediction under source and load uncertainty,this paper proposes a typical scenario identification model based on random forests and order parameters.Firstly,a method for ordinal parameter identification and quantification is provided for the coordinated operating mode of multi-microgrids,taking into account source-load uncertainty.Secondly,the dynamic change characteristics of the order parameters of the daily load curve,wind and solar curve,and load curve of typical scenarios are statistically analyzed to identify the key order parameters that have the most significant impact on the uncertainty of the load.Then,the order parameters and seasonal distribution are used as features to train a random forest classification model to achieve efficient scenario prediction.Finally,the simulation of actual data from a provincial distribution network shows that the proposed method can accurately classify typical scenarios with an accuracy rate of 92.7%.Additionally,sensitivity analysis is conducted to assess how changes in uncertainty levels affect the importance of each order parameter,allowing for adaptive uncertainty mitigation strategies.展开更多
Precision actuation is a foundational technology in high-end equipment domains,where stroke,velocity,and accuracy are critical for processing and/or detection quality,precision in spacecraft flight trajectories,and ac...Precision actuation is a foundational technology in high-end equipment domains,where stroke,velocity,and accuracy are critical for processing and/or detection quality,precision in spacecraft flight trajectories,and accuracy in weapon system strikes.Piezoelectric actuators(PEAs),known for their nanometer-level precision,flexible stroke,resistance to electromagnetic interference,and scalable structure,have been widely adopted across various fields.Therefore,this study focuses on extreme scenarios involving ultra-high precision(micrometer and beyond),minuscule scales,and highly complex operational conditions.It provides a comprehensive overview of the types,working principles,advantages,and disadvantages of PEAs,along with their potential applications in piezo-actuated smart mechatronic systems(PSMSs).To address the demands of extreme scenarios in high-end equipment fields,we have identified five representative application areas:positioning and alignment,biomedical device configuration,advanced manufacturing and processing,vibration mitigation,micro robot system.Each area is further divided into specific subcategories,where we explore the underlying relationships,mechanisms,representative schemes,and characteristics.Finally,we discuss the challenges and future development trends related to PEAs and PSMSs.This work aims to showcase the latest advancements in the application of PEAs and provide valuable guidance for researchers in this field.展开更多
Wind power generation is subjected to complex and variable meteorological conditions,resulting in intermittent and volatile power generation.Accurate wind power prediction plays a crucial role in enabling the power gr...Wind power generation is subjected to complex and variable meteorological conditions,resulting in intermittent and volatile power generation.Accurate wind power prediction plays a crucial role in enabling the power grid dispatching departments to rationally plan power transmission and energy storage operations.This enhances the efficiency of wind power integration into the grid.It allows grid operators to anticipate and mitigate the impact of wind power fluctuations,significantly improving the resilience of wind farms and the overall power grid.Furthermore,it assists wind farm operators in optimizing the management of power generation facilities and reducing maintenance costs.Despite these benefits,accurate wind power prediction especially in extreme scenarios remains a significant challenge.To address this issue,a novel wind power prediction model based on learning approach is proposed by integrating wavelet transform and Transformer.First,a conditional generative adversarial network(CGAN)generates dynamic extreme scenarios guided by physical constraints and expert rules to ensure realism and capture critical features of wind power fluctuations under extremeconditions.Next,thewavelet transformconvolutional layer is applied to enhance sensitivity to frequency domain characteristics,enabling effective feature extraction fromextreme scenarios for a deeper understanding of input data.The model then leverages the Transformer’s self-attention mechanism to capture global dependencies between features,strengthening its sequence modelling capabilities.Case analyses verify themodel’s superior performance in extreme scenario prediction by effectively capturing local fluctuation featureswhile maintaining a grasp of global trends.Compared to other models,it achieves R-squared(R^(2))as high as 0.95,and the mean absolute error(MAE)and rootmean square error(RMSE)are also significantly lower than those of othermodels,proving its high accuracy and effectiveness in managing complex wind power generation conditions.展开更多
With the rapid development of virtual reality(VR)and augmented reality(AR)technologies,their application potential in the field of education has become increasingly significant.For a long time,fire safety education in...With the rapid development of virtual reality(VR)and augmented reality(AR)technologies,their application potential in the field of education has become increasingly significant.For a long time,fire safety education in university laboratories has faced numerous challenges,and traditional teaching methods have been insufficiently effective,with high-risk scenarios difficult to realistically recreate.Especially in special scenarios involving hazardous chemicals,conventional training methods struggle to enable learners to achieve deep understanding and behavioral formation.This study systematically integrates immersive technology theory with safety education needs,providing a replicable technical solution for safety education in high-risk environments.Its modular design approach has reference value for expansion into other professional fields,offering practical evidence for innovation in safety education models in the digital age.展开更多
Toroidal torques,generated by the resonant magnetic perturbation(RMP)and acting on the plasma column,are numerically systematically investigated for an ITER baseline scenario.The neoclassical toroidal viscosity(NTV),i...Toroidal torques,generated by the resonant magnetic perturbation(RMP)and acting on the plasma column,are numerically systematically investigated for an ITER baseline scenario.The neoclassical toroidal viscosity(NTV),in particular the resonant portion,is found to provide the dominant contribution to the total toroidal torque under the slow plasma flow regime in ITER.While the electromagnetic torque always opposes the plasma flow,the toroidal torque associated with the Reynolds stress enhances the plasma flow independent of the flow direction.A peculiar double-peak structure for the net NTV torque is robustly computed for ITER,as the toroidal rotation frequency is scanned near the zero value.This structure is found to be ultimately due to a non-monotonic behavior of the wave-particle resonance integral(over the particle pitch angle)in the superbanana plateau NTV regime in ITER.These findings are qualitatively insensitive to variations of a range of factors including the wall resistivity,the plasma pedestal flow and the assumed frequency of the rotating RMP field.展开更多
Flood regulation service(FRS)stands as one of the key benefits that people get from the ecosystem.Under the influence of climate change and human activities,the relationship between supply and demand of FRS would incr...Flood regulation service(FRS)stands as one of the key benefits that people get from the ecosystem.Under the influence of climate change and human activities,the relationship between supply and demand of FRS would increasingly affect regional flood risk and sustainable development.However,there was currently a lack of systematic study on the future supply-demand relationship of FRS in the flood-vulnerable area undergoing rapidly development in China.This study integrated the Scenario Model Intercomparison Project(ScenarioMIP)with the Shared Socioeconomic Pathways(SSPs)datasets and climate model data to quantify the supply-demand ratio(SDR)of FRS in the Yangtze River Delta(YRD),China from 2020 to 2050.Trend analyses were conducted using linear regres-sion,Theil-Sen median estimation,and Hurst exponent analysis,while key drivers of SDR changes were identified and quantified through the Lindeman-Merenda-Gold(LMG)method between 2021 and 2050.Results show that the supply of FRS in the YRD was generally insufficient to meet the demand.The imbalanced subbasins covered 88.24%of the total study area,with 34.48%of this imbal-anced area concentrated in the Southeastern Basin in China.During 2021 and 2050,the imbalance of FRS supply-demand relationship would largely aggravate in the YRD,of which the aggravated area would account for 77.23%.Under different scenarios,the SDR for FRS would decrease significantly,with rates ranging from-5.45×10^(-4) to-2.06×10^(-4)(P<0.05).Especially,the decline rate of SDR in the YRD Basin(DeltaB)reached 2.92 times that the average of YRD.Human activities were the primary factors that exacerbated the imbalance in FRS supply-demand relationship,of which the relative contribution rate exceeds 75%.Particular attention should be direc-ted toward critical regions like the Southeast Basin in China(SEB)and DeltaB where substantial aggravation of supply-demand imbal-ances of FRS is projected.展开更多
A literature review on AI applications in the field of railway safety shows that the implemented approaches mainly concern the operational,maintenance,and feedback phases following railway incidents or accidents.These...A literature review on AI applications in the field of railway safety shows that the implemented approaches mainly concern the operational,maintenance,and feedback phases following railway incidents or accidents.These approaches exploit railway safety data once the transport system has received authorization for commissioning.However,railway standards and regulations require the development of a safety management system(SMS)from the specification and design phases of the railway system.This article proposes a new AI approach for analyzing and assessing safety from the specification and design phases of the railway system with a view to improving the development of the SMS.Unlike some learning methods,the proposed approach,which is dedicated in particular to safety assessment bodies,is based on semi-supervised learning carried out in close collaboration with safety experts who contributed to the development of a database of potential accident scenarios(learning example database)relating to the risk of rail collision.The proposed decision support is based on the use of an expert system whose knowledge base is automatically generated by inductive learning in the form of an association rule(rule base)and whose main objective is to suggest to the safety expert possible hazards not considered during the development of the SMS to complete the initial hazard register.展开更多
Accurate forecasting of blast furnace gas(BFG)production is an essential prerequisite for reasonable energy scheduling and management to reduce carbon emissions.Coupling forecasting between BFG generation and consumpt...Accurate forecasting of blast furnace gas(BFG)production is an essential prerequisite for reasonable energy scheduling and management to reduce carbon emissions.Coupling forecasting between BFG generation and consumption dynamics was taken as the research object.A multi-task learning(MTL)method for BFG forecasting was proposed,which integrated a coupling correlation coefficient(CCC)and an inverted transformer structure.The CCC method could enhance key information extraction by establishing relationships between multiple prediction targets and relevant factors,while MTL effectively captured the inherent correlations between BFG generation and consumption.Finally,a real-world case study was conducted to compare the proposed model with four benchmark models.Results indicated significant reductions in average mean absolute percentage error by 33.37%,achieving 1.92%,with a computational time of 76 s.The sensitivity analysis of hyperparameters such as learning rate,batch size,and units of the long short-term memory layer highlights the importance of hyperparameter tuning.展开更多
Emergency resources play a vital role in the emergency rescue process.The adequate and timely supply of emergency resources can effectively control the development of accidents and reduce accident losses.However,the c...Emergency resources play a vital role in the emergency rescue process.The adequate and timely supply of emergency resources can effectively control the development of accidents and reduce accident losses.However,the current emergency resource allocation of chemical enterprises lacks scientific analysis of accident scenarios,and the individual allocation method of enterprises increases the cost of emergency resource allocation.Given the above problems,this paper proposes a regional collaborative allocation method of emergency resources for enterprises within the chemical industry park(CIP)based on the worst credible accident scenario(WCAS).Firstly,the concept and analysis method of the WCAS is proposed.Then,based on the characteristics and consequences of the accident,the mapping relationship between accident scenarios and emergency resources is established.Finally,an optimization model for regional collaborative allocation of emergency resources is constructed to determine the amount of emergency resource allocation for each enterprise.Through the case study,the emergency resource allocation method based on the WCAS analysis can better meet the demands of accident emergency rescue.Simultaneously,the regional collaborative allocation optimization model can strengthen the cooperation ability among enterprises,greatly reducing the cost of emergency resource allocation for each enterprise.展开更多
基金The National Natural Science Foundation of China(62136008,62293541)The Beijing Natural Science Foundation(4232056)The Beijing Nova Program(20240484514).
文摘Cooperative multi-agent reinforcement learning(MARL)is a key technology for enabling cooperation in complex multi-agent systems.It has achieved remarkable progress in areas such as gaming,autonomous driving,and multi-robot control.Empowering cooperative MARL with multi-task decision-making capabilities is expected to further broaden its application scope.In multi-task scenarios,cooperative MARL algorithms need to address 3 types of multi-task problems:reward-related multi-task,arising from different reward functions;multi-domain multi-task,caused by differences in state and action spaces,state transition functions;and scalability-related multi-task,resulting from the dynamic variation in the number of agents.Most existing studies focus on scalability-related multitask problems.However,with the increasing integration between large language models(LLMs)and multi-agent systems,a growing number of LLM-based multi-agent systems have emerged,enabling more complex multi-task cooperation.This paper provides a comprehensive review of the latest advances in this field.By combining multi-task reinforcement learning with cooperative MARL,we categorize and analyze the 3 major types of multi-task problems under multi-agent settings,offering more fine-grained classifications and summarizing key insights for each.In addition,we summarize commonly used benchmarks and discuss future directions of research in this area,which hold promise for further enhancing the multi-task cooperation capabilities of multi-agent systems and expanding their practical applications in the real world.
文摘Deep neural networks(DNNs)are effective in solving both forward and inverse problems for nonlinear partial differential equations(PDEs).However,conventional DNNs are not effective in handling problems such as delay differential equations(DDEs)and delay integrodifferential equations(DIDEs)with constant delays,primarily due to their low regularity at delayinduced breaking points.In this paper,a DNN method that combines multi-task learning(MTL)which is proposed to solve both the forward and inverse problems of DIDEs.The core idea of this approach is to divide the original equation into multiple tasks based on the delay,using auxiliary outputs to represent the integral terms,followed by the use of MTL to seamlessly incorporate the properties at the breaking points into the loss function.Furthermore,given the increased training dificulty associated with multiple tasks and outputs,we employ a sequential training scheme to reduce training complexity and provide reference solutions for subsequent tasks.This approach significantly enhances the approximation accuracy of solving DIDEs with DNNs,as demonstrated by comparisons with traditional DNN methods.We validate the effectiveness of this method through several numerical experiments,test various parameter sharing structures in MTL and compare the testing results of these structures.Finally,this method is implemented to solve the inverse problem of nonlinear DIDE and the results show that the unknown parameters of DIDE can be discovered with sparse or noisy data.
基金supported by the top-level design of the National Natural Science Foundation of China(NSFC)Major Project“Realization of optimal carbon neutral pathway and coupling of multi-scale interaction patterns of natural-social systems in China”(42341202)the Basic Scientific Research Fund of the Chinese Academy of Meteorological Sciences(2021Z014)。
文摘This paper proposes that China,under the challenge of balancing its development and security,can aim for the Paris Agreement's goal to limit global warming to no more than 2℃by actively seeking carbonpeak and carbon-neutrality pathways that align with China's national conditions,rather than following the idealized path toward the 1.5℃target by initially relying on extensive negative-emission technologies such as direct air carbon capture and storage(DACCS).This work suggests that pursuing a 1.5℃target is increasingly less feasible for China,as it would potentially incur 3–4 times the cost of pursuing the 2℃target.With China being likely to achieve a peak in its emissions around 2028,at about 12.8 billion tonnes of anthropogenic carbon dioxide(CO_(2)),and become carbon neutral,projected global warming levels may be less severe after the 2050s than previously estimated.This could reduce the risk potential of climate tipping points and extreme events,especially considering that the other two major carbon emitters in the world(Europe and North America)have already passed their carbon peaks.While natural carbon sinks will contribute to China's carbon neutrality efforts,they are not expected to be decisive in the transition stages.This research also addresses the growing focus on climate overshoot,tipping points,extreme events,loss and damage,and methane reductions in international climate cooperation,emphasizing the need to balance these issues with China's development,security,and fairness considerations.China's pursuit of carbon neutrality will have significant implications for global emissions scenarios,warming levels,and extreme event projections,as well as for climate change hotspots of international concern,such as climate tipping points,the climate crisis,and the notion that the world has moved from a warming to a boiling era.Possible research recommendations for global emissions scenarios based on China's 2℃target pathway are also summarized.
基金funded by the Excellent Talent Training Funding Project in Dongcheng District,Beijing,with project number 2024-dchrcpyzz-9.
文摘Predicting the health status of stroke patients at different stages of the disease is a critical clinical task.The onset and development of stroke are affected by an array of factors,encompassing genetic predisposition,environmental exposure,unhealthy lifestyle habits,and existing medical conditions.Although existing machine learning-based methods for predicting stroke patients’health status have made significant progress,limitations remain in terms of prediction accuracy,model explainability,and system optimization.This paper proposes a multi-task learning approach based on Explainable Artificial Intelligence(XAI)for predicting the health status of stroke patients.First,we design a comprehensive multi-task learning framework that utilizes the task correlation of predicting various health status indicators in patients,enabling the parallel prediction of multiple health indicators.Second,we develop a multi-task Area Under Curve(AUC)optimization algorithm based on adaptive low-rank representation,which removes irrelevant information from the model structure to enhance the performance of multi-task AUC optimization.Additionally,the model’s explainability is analyzed through the stability analysis of SHAP values.Experimental results demonstrate that our approach outperforms comparison algorithms in key prognostic metrics F1 score and Efficiency.
基金jointly supported by the National Natural Science Foundation of China(Grant Nos.42075138 and 42375147)the Program on Key Basic Research Project of Jiangsu(Grant No.BE2023829)。
文摘Tropical cyclones(TCs)are one of the most serious types of natural disasters,and accurate TC activity predictions are key to disaster prevention and mitigation.Recently,TC track predictions have made significant progress,but the ability to predict their intensity is obviously lagging behind.At present,research on TC intensity prediction takes atmospheric reanalysis data as the research object and mines the relationship between TC-related environmental factors and intensity through deep learning.However,reanalysis data are non-real-time in nature,which does not meet the requirements for operational forecasting applications.Therefore,a TC intensity prediction model named TC-Rolling is proposed,which can simultaneously extract the degree of symmetry for strong TC convective cloud and convection intensity,and fuse the deviation-angle variance with satellite images to construct the correlation between TC convection structure and intensity.For TCs'complex dynamic processes,a convolutional neural network(CNN)is used to learn their temporal and spatial features.For real-time intensity estimation,multi-task learning acts as an implicit time-series enhancement.The model is designed with a rolling strategy that aims to moderate the long-term dependent decay problem and improve accuracy for short-term intensity predictions.Since multiple tasks are correlated,the loss function of 12 h and 24 h are corrected.After testing on a sample of TCs in the Northwest Pacific,with a 4.48 kt root-mean-square error(RMSE)of 6 h intensity prediction,5.78 kt for 12 h,and 13.94 kt for 24 h,TC records from official agencies are used to assess the validity of TC-Rolling.
基金supported by the Key Research and Development Program of Heilongjiang Province(No.2022ZX01A35).
文摘As the group-buying model shows significant progress in attracting new users,enhancing user engagement,and increasing platform profitability,providing personalized recommendations for group-buying users has emerged as a new challenge in the field of recommendation systems.This paper introduces a group-buying recommendation model based on multi-head attention mechanisms and multi-task learning,termed the Multi-head Attention Mechanisms and Multi-task Learning Group-Buying Recommendation(MAMGBR)model,specifically designed to optimize group-buying recommendations on e-commerce platforms.The core dataset of this study comes from the Chinese maternal and infant e-commerce platform“Beibei,”encompassing approximately 430,000 successful groupbuying actions and over 120,000 users.Themodel focuses on twomain tasks:recommending items for group organizers(Task Ⅰ)and recommending participants for a given group-buying event(Task Ⅱ).In model evaluation,MAMGBR achieves an MRR@10 of 0.7696 for Task I,marking a 20.23%improvement over baseline models.Furthermore,in Task II,where complex interaction patterns prevail,MAMGBR utilizes auxiliary loss functions to effectively model the multifaceted roles of users,items,and participants,leading to a 24.08%increase in MRR@100 under a 1:99 sample ratio.Experimental results show that compared to benchmark models,such as NGCF and EATNN,MAMGBR’s integration ofmulti-head attentionmechanisms,expert networks,and gating mechanisms enables more accurate modeling of user preferences and social associations within group-buying scenarios,significantly enhancing recommendation accuracy and platform group-buying success rates.
基金supported by the National Natural Science Foundation of China(Grant Nos.42030708,42375138,42030608,42105128,42075079)the Opening Foundation of Key Laboratory of Atmospheric Sounding,China Meteorological Administration(CMA),and the CMA Research Center on Meteorological Observation Engineering Technology(Grant No.U2021Z03),and the Opening Foundation of the Key Laboratory of Atmospheric Chemistry,CMA(Grant No.2022B02)。
文摘Aerosol optical depth(AOD)and fine particulate matter with a diameter of less than or equal to 2.5μm(PM_(2.5))play crucial roles in air quality,human health,and climate change.However,the complex correlation of AOD–PM_(2.5)and the limitations of existing algorithms pose a significant challenge in realizing the accurate joint retrieval of these two parameters at the same location.On this point,a multi-task learning(MTL)model,which enables the joint retrieval of PM_(2.5)concentration and AOD,is proposed and applied on the top-of-the-atmosphere reflectance data gathered by the Fengyun-4A Advanced Geosynchronous Radiation Imager(FY-4A AGRI),and compared to that of two single-task learning models—namely,Random Forest(RF)and Deep Neural Network(DNN).Specifically,MTL achieves a coefficient of determination(R^(2))of 0.88 and a root-mean-square error(RMSE)of 0.10 in AOD retrieval.In comparison to RF,the R^(2)increases by 0.04,the RMSE decreases by 0.02,and the percentage of retrieval results falling within the expected error range(Within-EE)rises by 5.55%.The R^(2)and RMSE of PM_(2.5)retrieval by MTL are 0.84 and 13.76μg m~(-3)respectively.Compared with RF,the R^(2)increases by 0.06,the RMSE decreases by 4.55μg m~(-3),and the Within-EE increases by 7.28%.Additionally,compared to DNN,MTL shows an increase of 0.01 in R^(2)and a decrease of 0.02 in RMSE in AOD retrieval,with a corresponding increase of 2.89%in Within-EE.For PM_(2.5)retrieval,MTL exhibits an increase of 0.05 in R^(2),a decrease of 1.76μg m~(-3)in RMSE,and an increase of 6.83%in Within-EE.The evaluation suggests that MTL is able to provide simultaneously improved AOD and PM_(2.5)retrievals,demonstrating a significant advantage in efficiently capturing the spatial distribution of PM_(2.5)concentration and AOD.
基金the National Key Research and Development Plan of China[Grant No.2023YFB3002400]the Shanghai 2021 Natural Science Foundation[Grant Nos.21ZR1420400 and 21ZR1419800]+1 种基金the Shanghai 2023 Natural Science Foundation[Grant No.23ZR1463000]the Shandong Provincial Meteorological Bureau Scientific Research Project[Grant No.2023SDBD05].
文摘Accurate short-term forecast of offshore wind fields is still challenging for numerical weather prediction models.Based on three years of 48-hour forecast data from the European Centre for Medium-Range Weather Forecasts Integrated Forecasting System global model(ECMWF-IFS)over 14 offshore weather stations along the coast of Shandong Province,this study introduces a multi-task learning(MTL)model(TabNet-MTL),which significantly improves the forecast bias of near-surface wind direction and speed simultaneously.TabNet-MTL adopts the feature engineering method,utilizes mean square error as the loss function,and employs the 5-fold cross validation method to ensure the generalization ability of the trained model.It demonstrates superior skills in wind field correction across different forecast lead times over all stations compared to its single-task version(TabNet-STL)and three other popular single-task learning models(Random Forest,LightGBM,and XGBoost).Results show that it significantly reduces root mean square error of the ECMWF-IFS wind speed forecast from 2.20 to 1.25 m s−1,and increases the forecast accuracy of wind direction from 50%to 65%.As an explainable deep learning model,the weather stations and long-term temporal statistics of near-surface wind speed are identified as the most influential variables for TabNet-MTL in constructing its feature engineering.
基金supported by the National Natural Science Foundation of China(No.71704178)Beijing Municipal Excellent Talents Foundation(No.2017000020124G133)Major consulting project of the Chinese Academy of Engineering(Nos.2023-JB-08,2022-PP-03).
文摘The proposal of carbon neutrality target makes decarbonization and hydrogenation typical features of future energy development in China.With a wide range of application scenarios,hydrogen energy will experience rapid growth in production and consumption.To formulate an effective hydrogen energy development strategy for the future of China,this study employs the departmental scenario analysis method to calculate and evaluate the future consumption of hydrogen energy in China’s heavy industry,transportation,electricity,and other related fields.Multidimensional technical parameters are selected and predicted accurately and reliably in combination with different development scenarios.The findings indicate that the period from 2030 to 2050 will enjoy rapid development of hydrogen energy,having an average annual growth rate of 2%to 4%.The technological progress and breakthroughs scenario has the greatest potential for hydrogen demand scale among the four development scenarios.Under this scenario,the total demand for hydrogen energy is expected to reach 446.37Mt in 2060.Thetransportation sector will be the sector with the greatest potential for hydrogen deployment growth from 2023 to 2060,which is expected to rise from 0.038Mt to about 163.18Mt,with the ambitious growth in the future.Additionally,hydrogen energy has a considerable development potential in the steel sector,and the trend of de-refueling coke by hydrogenation in this sector will be imperative for this energy-intensive industries.
基金supported in part by the National Natural Science Foundation of China under Grant 62371181in part by the Changzhou Science and Technology International Cooperation Program under Grant CZ20230029The Institute of Information&Communications Technology Planning&Evaluation(IITP)grant funded by the Korea government(MSIT)(No.RS-2024-00396797,Development of core technology for intelligent O-RAN security platform).
文摘Satellite communication technology has emerged as a key solution to address the challenges of data transmission in remote areas.By overcoming the limitations of traditional terrestrial communication networks,it enables long-distance data transmission anytime and anywhere,ensuring the timely and accurate delivery of water level data,which is particularly crucial for fishway water level monitoring.To enhance the effectiveness of fishway water level monitoring,this study proposes a multi-task learning model,AS-SOMTF,designed for real-time and comprehensive prediction.The model integrates auxiliary sequences with primary input sequences to capture complex relationships and dependencies,thereby improving representational capacity.In addition,a novel timeseries embedding algorithm,AS-SOM,is introduced,which combines generative inference and pooling operations to optimize prediction efficiency for long sequences.This innovation not only ensures the timely transmission of water level data but also enhances the accuracy of real-time monitoring.Compared with traditional models such as Transformer and Long Short-Term Memory(LSTM)networks,the proposed model achieves improvements of 3.8%and 1.4%in prediction accuracy,respectively.These advancements provide more precise technical support for water level forecasting and resource management in the Diqing Tibetan Autonomous Prefecture of the Lancang River,contributing to ecosystem protection and improved operational safety.
基金supported by Science and Technology Project Managed by the State Grid Jiangsu Electric Power Co.,Ltd.(No.J2024163).
文摘With the popularization of microgrid construction and the connection of renewable energy sources to the power system,the problem of source and load uncertainty faced by the coordinated operation of multi-microgrid is becoming increasingly prominent,and the accuracy of typical scenario predictions is low.In order to improve the accuracy of scenario prediction under source and load uncertainty,this paper proposes a typical scenario identification model based on random forests and order parameters.Firstly,a method for ordinal parameter identification and quantification is provided for the coordinated operating mode of multi-microgrids,taking into account source-load uncertainty.Secondly,the dynamic change characteristics of the order parameters of the daily load curve,wind and solar curve,and load curve of typical scenarios are statistically analyzed to identify the key order parameters that have the most significant impact on the uncertainty of the load.Then,the order parameters and seasonal distribution are used as features to train a random forest classification model to achieve efficient scenario prediction.Finally,the simulation of actual data from a provincial distribution network shows that the proposed method can accurately classify typical scenarios with an accuracy rate of 92.7%.Additionally,sensitivity analysis is conducted to assess how changes in uncertainty levels affect the importance of each order parameter,allowing for adaptive uncertainty mitigation strategies.
基金financially supported by the National Key R&D Program of China(Grant No.2022YFC2204203)the National Natural Science Foundation of China(Grant No.52305107)。
文摘Precision actuation is a foundational technology in high-end equipment domains,where stroke,velocity,and accuracy are critical for processing and/or detection quality,precision in spacecraft flight trajectories,and accuracy in weapon system strikes.Piezoelectric actuators(PEAs),known for their nanometer-level precision,flexible stroke,resistance to electromagnetic interference,and scalable structure,have been widely adopted across various fields.Therefore,this study focuses on extreme scenarios involving ultra-high precision(micrometer and beyond),minuscule scales,and highly complex operational conditions.It provides a comprehensive overview of the types,working principles,advantages,and disadvantages of PEAs,along with their potential applications in piezo-actuated smart mechatronic systems(PSMSs).To address the demands of extreme scenarios in high-end equipment fields,we have identified five representative application areas:positioning and alignment,biomedical device configuration,advanced manufacturing and processing,vibration mitigation,micro robot system.Each area is further divided into specific subcategories,where we explore the underlying relationships,mechanisms,representative schemes,and characteristics.Finally,we discuss the challenges and future development trends related to PEAs and PSMSs.This work aims to showcase the latest advancements in the application of PEAs and provide valuable guidance for researchers in this field.
基金funded by the Science and Technology Project of State Grid Corporation of China under Grant No.5108-202218280A-2-299-XG.
文摘Wind power generation is subjected to complex and variable meteorological conditions,resulting in intermittent and volatile power generation.Accurate wind power prediction plays a crucial role in enabling the power grid dispatching departments to rationally plan power transmission and energy storage operations.This enhances the efficiency of wind power integration into the grid.It allows grid operators to anticipate and mitigate the impact of wind power fluctuations,significantly improving the resilience of wind farms and the overall power grid.Furthermore,it assists wind farm operators in optimizing the management of power generation facilities and reducing maintenance costs.Despite these benefits,accurate wind power prediction especially in extreme scenarios remains a significant challenge.To address this issue,a novel wind power prediction model based on learning approach is proposed by integrating wavelet transform and Transformer.First,a conditional generative adversarial network(CGAN)generates dynamic extreme scenarios guided by physical constraints and expert rules to ensure realism and capture critical features of wind power fluctuations under extremeconditions.Next,thewavelet transformconvolutional layer is applied to enhance sensitivity to frequency domain characteristics,enabling effective feature extraction fromextreme scenarios for a deeper understanding of input data.The model then leverages the Transformer’s self-attention mechanism to capture global dependencies between features,strengthening its sequence modelling capabilities.Case analyses verify themodel’s superior performance in extreme scenario prediction by effectively capturing local fluctuation featureswhile maintaining a grasp of global trends.Compared to other models,it achieves R-squared(R^(2))as high as 0.95,and the mean absolute error(MAE)and rootmean square error(RMSE)are also significantly lower than those of othermodels,proving its high accuracy and effectiveness in managing complex wind power generation conditions.
文摘With the rapid development of virtual reality(VR)and augmented reality(AR)technologies,their application potential in the field of education has become increasingly significant.For a long time,fire safety education in university laboratories has faced numerous challenges,and traditional teaching methods have been insufficiently effective,with high-risk scenarios difficult to realistically recreate.Especially in special scenarios involving hazardous chemicals,conventional training methods struggle to enable learners to achieve deep understanding and behavioral formation.This study systematically integrates immersive technology theory with safety education needs,providing a replicable technical solution for safety education in high-risk environments.Its modular design approach has reference value for expansion into other professional fields,offering practical evidence for innovation in safety education models in the digital age.
基金funded by National Natural Science Foundation of China(NSFC)(Nos.12075053,11505021 and 11975068)by National Key R&D Program of China(No.2022YFE 03060002)+1 种基金by Fundamental Research Funds for the Central Universities(No.2232024G-10)supported by the U.S.DoE Office of Science(No.DE-FG02–95ER54309)。
文摘Toroidal torques,generated by the resonant magnetic perturbation(RMP)and acting on the plasma column,are numerically systematically investigated for an ITER baseline scenario.The neoclassical toroidal viscosity(NTV),in particular the resonant portion,is found to provide the dominant contribution to the total toroidal torque under the slow plasma flow regime in ITER.While the electromagnetic torque always opposes the plasma flow,the toroidal torque associated with the Reynolds stress enhances the plasma flow independent of the flow direction.A peculiar double-peak structure for the net NTV torque is robustly computed for ITER,as the toroidal rotation frequency is scanned near the zero value.This structure is found to be ultimately due to a non-monotonic behavior of the wave-particle resonance integral(over the particle pitch angle)in the superbanana plateau NTV regime in ITER.These findings are qualitatively insensitive to variations of a range of factors including the wall resistivity,the plasma pedestal flow and the assumed frequency of the rotating RMP field.
基金Under the auspices of National Natural Science Foundation of China(No.42101251)。
文摘Flood regulation service(FRS)stands as one of the key benefits that people get from the ecosystem.Under the influence of climate change and human activities,the relationship between supply and demand of FRS would increasingly affect regional flood risk and sustainable development.However,there was currently a lack of systematic study on the future supply-demand relationship of FRS in the flood-vulnerable area undergoing rapidly development in China.This study integrated the Scenario Model Intercomparison Project(ScenarioMIP)with the Shared Socioeconomic Pathways(SSPs)datasets and climate model data to quantify the supply-demand ratio(SDR)of FRS in the Yangtze River Delta(YRD),China from 2020 to 2050.Trend analyses were conducted using linear regres-sion,Theil-Sen median estimation,and Hurst exponent analysis,while key drivers of SDR changes were identified and quantified through the Lindeman-Merenda-Gold(LMG)method between 2021 and 2050.Results show that the supply of FRS in the YRD was generally insufficient to meet the demand.The imbalanced subbasins covered 88.24%of the total study area,with 34.48%of this imbal-anced area concentrated in the Southeastern Basin in China.During 2021 and 2050,the imbalance of FRS supply-demand relationship would largely aggravate in the YRD,of which the aggravated area would account for 77.23%.Under different scenarios,the SDR for FRS would decrease significantly,with rates ranging from-5.45×10^(-4) to-2.06×10^(-4)(P<0.05).Especially,the decline rate of SDR in the YRD Basin(DeltaB)reached 2.92 times that the average of YRD.Human activities were the primary factors that exacerbated the imbalance in FRS supply-demand relationship,of which the relative contribution rate exceeds 75%.Particular attention should be direc-ted toward critical regions like the Southeast Basin in China(SEB)and DeltaB where substantial aggravation of supply-demand imbal-ances of FRS is projected.
文摘A literature review on AI applications in the field of railway safety shows that the implemented approaches mainly concern the operational,maintenance,and feedback phases following railway incidents or accidents.These approaches exploit railway safety data once the transport system has received authorization for commissioning.However,railway standards and regulations require the development of a safety management system(SMS)from the specification and design phases of the railway system.This article proposes a new AI approach for analyzing and assessing safety from the specification and design phases of the railway system with a view to improving the development of the SMS.Unlike some learning methods,the proposed approach,which is dedicated in particular to safety assessment bodies,is based on semi-supervised learning carried out in close collaboration with safety experts who contributed to the development of a database of potential accident scenarios(learning example database)relating to the risk of rail collision.The proposed decision support is based on the use of an expert system whose knowledge base is automatically generated by inductive learning in the form of an association rule(rule base)and whose main objective is to suggest to the safety expert possible hazards not considered during the development of the SMS to complete the initial hazard register.
基金supported by the National Natural Science Foundation of China(No.52474435)China Baowu Low Carbon Metallurgy Innovation Foundation(BWLCF202307).
文摘Accurate forecasting of blast furnace gas(BFG)production is an essential prerequisite for reasonable energy scheduling and management to reduce carbon emissions.Coupling forecasting between BFG generation and consumption dynamics was taken as the research object.A multi-task learning(MTL)method for BFG forecasting was proposed,which integrated a coupling correlation coefficient(CCC)and an inverted transformer structure.The CCC method could enhance key information extraction by establishing relationships between multiple prediction targets and relevant factors,while MTL effectively captured the inherent correlations between BFG generation and consumption.Finally,a real-world case study was conducted to compare the proposed model with four benchmark models.Results indicated significant reductions in average mean absolute percentage error by 33.37%,achieving 1.92%,with a computational time of 76 s.The sensitivity analysis of hyperparameters such as learning rate,batch size,and units of the long short-term memory layer highlights the importance of hyperparameter tuning.
基金support provided by the Qingdao Science and Technology Benefits People Demonstration and Guidance Project(21-1-4-sf-4-nsh).
文摘Emergency resources play a vital role in the emergency rescue process.The adequate and timely supply of emergency resources can effectively control the development of accidents and reduce accident losses.However,the current emergency resource allocation of chemical enterprises lacks scientific analysis of accident scenarios,and the individual allocation method of enterprises increases the cost of emergency resource allocation.Given the above problems,this paper proposes a regional collaborative allocation method of emergency resources for enterprises within the chemical industry park(CIP)based on the worst credible accident scenario(WCAS).Firstly,the concept and analysis method of the WCAS is proposed.Then,based on the characteristics and consequences of the accident,the mapping relationship between accident scenarios and emergency resources is established.Finally,an optimization model for regional collaborative allocation of emergency resources is constructed to determine the amount of emergency resource allocation for each enterprise.Through the case study,the emergency resource allocation method based on the WCAS analysis can better meet the demands of accident emergency rescue.Simultaneously,the regional collaborative allocation optimization model can strengthen the cooperation ability among enterprises,greatly reducing the cost of emergency resource allocation for each enterprise.
