Aiming at the terminal defense problem of aircraft,this paper proposes a method to simultaneously achieve terminal defense and seize the dominant position.The method employs aλ-return based reinforcement learning alg...Aiming at the terminal defense problem of aircraft,this paper proposes a method to simultaneously achieve terminal defense and seize the dominant position.The method employs aλ-return based reinforcement learning algorithm,which can be applied to the flight assistance decision-making system to improve the pilot’s survivability.First,we model the environment to simulate the interaction between air-to-air missiles and aircraft.Subsequently,we propose aλ-return based approach to improve the deep Q learning network(DQN),deep advantageous actor criticism(A2C),and proximity policy optimization(PPO)algorithms used to train manoeuvre strategies.The method employs an action space containing nine manoeuvres and defines the off-target distance at the end of the scene as a sparse reward for algorithm training.Simulation results show that the convergence speed of the three improved algorithms is significantly improved when using theλ-return method.Moreover,the effect of the fetch value on the convergence speed is verified by ablation experiments.In order to solve the illegal behavior problem in the training process,we also design a backtracking-based illegal behavior masking mechanism,which improves the data generation efficiency of the environment model and promotes effective algorithm training.展开更多
Target maneuver trajectory prediction plays an important role in air combat situation awareness and threat assessment.To solve the problem of low prediction accuracy of the traditional prediction method and model,a ta...Target maneuver trajectory prediction plays an important role in air combat situation awareness and threat assessment.To solve the problem of low prediction accuracy of the traditional prediction method and model,a target maneuver trajectory prediction model based on phase space reconstruction-radial basis function(PSR-RBF)neural network is established by combining the characteristics of trajectory with time continuity.In order to further improve the prediction performance of the model,the rival penalized competitive learning(RPCL)algorithm is introduced to determine the structure of RBF,the Levenberg-Marquardt(LM)and the hybrid algorithm of the improved particle swarm optimization(IPSO)algorithm and the k-means are introduced to optimize the parameter of RBF,and a PSR-RBF neural network is constructed.An independent method of 3D coordinates of the target maneuver trajectory is proposed,and the target manuver trajectory sample data is constructed by using the training data selected in the air combat maneuver instrument(ACMI),and the maneuver trajectory prediction model based on the PSR-RBF neural network is established.In order to verify the precision and real-time performance of the trajectory prediction model,the simulation experiment of target maneuver trajectory is performed.The results show that the prediction performance of the independent method is better,and the accuracy of the PSR-RBF prediction model proposed is better.The prediction confirms the effectiveness and applicability of the proposed method and model.展开更多
Full-scale measurements are regarded as the most reliable method to evaluate wind effects on large buildings and structures. Some selected results are presented in this paper from the full-scale measurement of wind ef...Full-scale measurements are regarded as the most reliable method to evaluate wind effects on large buildings and structures. Some selected results are presented in this paper from the full-scale measurement of wind effects on a long-span steel roof structure during the passage of Typhoon Fanapi. Some fi eld data, including wind speed and direction, acceleration responses, etc., were continuously and simultaneously recorded during the passage of the typhoon. Comprehensive analysis of the measured data is conducted to evaluate the typhoon-generated wind characteristics and its effects on a long-span steel roof. The fi rst four natural frequencies and their vibration mode shapes of the Guangzhou International Sports Arena(GISA) roof are evaluated by the stochastic subspace identifi cation(SSI) method and comparisons with those from fi nite element(FE) analysis are made. Meanwhile, damping ratios of the roof are also identifi ed by the SSI method and compared with those identifi ed by the random decrement method; the amplitude-dependent damping behaviors are also discussed. The fullscale measurement results are further compared with the corresponding wind tunnel test results to evaluate its reliability. The results obtained from this study are valuable for academic and professional engineers involved in the design of large-span roof structures.展开更多
Accidents in engineered systems are usually generated by complex socio-technical factors.It is beneficial to investigate the increasing complexity and coupling of these factors from the perspective of system safety.Ba...Accidents in engineered systems are usually generated by complex socio-technical factors.It is beneficial to investigate the increasing complexity and coupling of these factors from the perspective of system safety.Based on system and control theories,System-Theoretic Accident Model and Processes(STAMP)is a widely recognized approach for accident analysis.In this paper,we propose a STAMP-Game model to analyze accidents in oil and gas storage and transportation systems.Stakeholders in accident analysis by STAMP can be regarded as players of a game.Game theory can,thus,be adopted in accident analysis to depict the competition and cooperation between stakeholders.Subsequently,we established a game model to study the strategies of both supervisory and supervised entities.The obtained results demonstrate that the proposed game model allows for identifying the effectiveness deficiency of the supervisory entity,and the safety and protection altitudes of the supervised entity.The STAMP-Game model can generate quantitative parameters for supporting the behavior and strategy selections of the supervisory and supervised entities.The quantitative data obtained can be used to guide the safety improvement,to reduce the costs of safety regulation violation and accident risk.展开更多
Target maneuver trajectory prediction is an important prerequisite for air combat situation awareness and maneuver decision-making.However,how to use a large amount of trajectory data generated by air combat confronta...Target maneuver trajectory prediction is an important prerequisite for air combat situation awareness and maneuver decision-making.However,how to use a large amount of trajectory data generated by air combat confrontation training to achieve real-time and accurate prediction of target maneuver trajectory is an urgent problem to be solved.To solve this problem,in this paper,a hybrid algorithm based on transfer learning,online learning,ensemble learning,regularization technology,target maneuvering segmentation point recognition algorithm,and Volterra series,abbreviated as AERTrOS-Volterra is proposed.Firstly,the model makes full use of a large number of trajectory sample data generated by air combat confrontation training,and constructs a Tr-Volterra algorithm framework suitable for air combat target maneuver trajectory prediction,which realizes the extraction of effective information from the historical trajectory data.Secondly,in order to improve the real-time online prediction accuracy and robustness of the prediction model in complex electromagnetic environments,on the basis of the TrVolterra algorithm framework,a robust regularized online Sequential Volterra prediction model is proposed by integrating online learning method,regularization technology and inverse weighting calculation method based on the priori error.Finally,inspired by the preferable performance of models ensemble,ensemble learning scheme is also incorporated into our proposed algorithm,which adaptively updates the ensemble prediction model according to the performance of the model on real-time samples and the recognition results of target maneuvering segmentation points,including the adaptation of model weights;adaptation of parameters;and dynamic inclusion and removal of models.Compared with many existing time series prediction methods,the newly proposed target maneuver trajectory prediction algorithm can fully mine the prior knowledge contained in the historical data to assist the current prediction.The rationality and effectiveness of the proposed algorithm are verified by simulation on three sets of chaotic time series data sets and a set of real target maneuver trajectory data sets.展开更多
In recent years,with the rapid development of e-commerce,people need to classify the wide variety and a large number of clothing images appearing on e-commerce platforms.In order to solve the problems of long time con...In recent years,with the rapid development of e-commerce,people need to classify the wide variety and a large number of clothing images appearing on e-commerce platforms.In order to solve the problems of long time consumption and unsatisfactory classification accuracy arising from the classification of a large number of clothing images,researchers have begun to exploit deep learning techniques instead of traditional learning methods.The paper explores the use of convolutional neural networks(CNNs)for feature learning to enhance global feature information interactions by adding an improved hybrid attention mechanism(HAM)that fully utilizes feature weights in three dimensions:channel,height,and width.Moreover,the improved pooling layer not only captures local feature information,but also fuses global and local information to improve the misclassification problem that occurs between similar categories.Experiments on the Fashion-MNIST and DeepFashion datasets show that the proposed method significantly improves the accuracy of clothing classification(93.62%and 67.9%)compared with residual network(ResNet)and convolutional block attention module(CBAM).展开更多
The strong static magnetic fields(SMFs)have caused great health concerns on the potential occupational and public exposure in recent year.However,the evidence in support of the cellular effects induced by the strong S...The strong static magnetic fields(SMFs)have caused great health concerns on the potential occupational and public exposure in recent year.However,the evidence in support of the cellular effects induced by the strong SMFs is inconsistent and the mechanisms are largely unknown.The present study focused on the effects of the strong SMFs on ATP contents in the two cell lines:human-hamster hybrid(AL)cells and mitochondria deficient human-hamster hybrid(ρ0 AL)cells.展开更多
文摘Aiming at the terminal defense problem of aircraft,this paper proposes a method to simultaneously achieve terminal defense and seize the dominant position.The method employs aλ-return based reinforcement learning algorithm,which can be applied to the flight assistance decision-making system to improve the pilot’s survivability.First,we model the environment to simulate the interaction between air-to-air missiles and aircraft.Subsequently,we propose aλ-return based approach to improve the deep Q learning network(DQN),deep advantageous actor criticism(A2C),and proximity policy optimization(PPO)algorithms used to train manoeuvre strategies.The method employs an action space containing nine manoeuvres and defines the off-target distance at the end of the scene as a sparse reward for algorithm training.Simulation results show that the convergence speed of the three improved algorithms is significantly improved when using theλ-return method.Moreover,the effect of the fetch value on the convergence speed is verified by ablation experiments.In order to solve the illegal behavior problem in the training process,we also design a backtracking-based illegal behavior masking mechanism,which improves the data generation efficiency of the environment model and promotes effective algorithm training.
文摘Target maneuver trajectory prediction plays an important role in air combat situation awareness and threat assessment.To solve the problem of low prediction accuracy of the traditional prediction method and model,a target maneuver trajectory prediction model based on phase space reconstruction-radial basis function(PSR-RBF)neural network is established by combining the characteristics of trajectory with time continuity.In order to further improve the prediction performance of the model,the rival penalized competitive learning(RPCL)algorithm is introduced to determine the structure of RBF,the Levenberg-Marquardt(LM)and the hybrid algorithm of the improved particle swarm optimization(IPSO)algorithm and the k-means are introduced to optimize the parameter of RBF,and a PSR-RBF neural network is constructed.An independent method of 3D coordinates of the target maneuver trajectory is proposed,and the target manuver trajectory sample data is constructed by using the training data selected in the air combat maneuver instrument(ACMI),and the maneuver trajectory prediction model based on the PSR-RBF neural network is established.In order to verify the precision and real-time performance of the trajectory prediction model,the simulation experiment of target maneuver trajectory is performed.The results show that the prediction performance of the independent method is better,and the accuracy of the PSR-RBF prediction model proposed is better.The prediction confirms the effectiveness and applicability of the proposed method and model.
基金National Natural Science Foundation of China under Grant Nos.51222801 and 51378134Yangcheng Scholarship in Guangzhou Municipal Universities under Project No.12A004Sthe Research Funding for Ph.D Programme in Higher Education Universities under Project No.20124410110005
文摘Full-scale measurements are regarded as the most reliable method to evaluate wind effects on large buildings and structures. Some selected results are presented in this paper from the full-scale measurement of wind effects on a long-span steel roof structure during the passage of Typhoon Fanapi. Some fi eld data, including wind speed and direction, acceleration responses, etc., were continuously and simultaneously recorded during the passage of the typhoon. Comprehensive analysis of the measured data is conducted to evaluate the typhoon-generated wind characteristics and its effects on a long-span steel roof. The fi rst four natural frequencies and their vibration mode shapes of the Guangzhou International Sports Arena(GISA) roof are evaluated by the stochastic subspace identifi cation(SSI) method and comparisons with those from fi nite element(FE) analysis are made. Meanwhile, damping ratios of the roof are also identifi ed by the SSI method and compared with those identifi ed by the random decrement method; the amplitude-dependent damping behaviors are also discussed. The fullscale measurement results are further compared with the corresponding wind tunnel test results to evaluate its reliability. The results obtained from this study are valuable for academic and professional engineers involved in the design of large-span roof structures.
基金supported by the National Natural Science Foundation of China(Grant No.52004030)the R&D Program of Beijing Municipal Education Commission(Grant No.KM202310016003)the Exchange Program of High-end Foreign Experts of Ministry of Science and Technology,China(Grant No.G2022178013L)。
文摘Accidents in engineered systems are usually generated by complex socio-technical factors.It is beneficial to investigate the increasing complexity and coupling of these factors from the perspective of system safety.Based on system and control theories,System-Theoretic Accident Model and Processes(STAMP)is a widely recognized approach for accident analysis.In this paper,we propose a STAMP-Game model to analyze accidents in oil and gas storage and transportation systems.Stakeholders in accident analysis by STAMP can be regarded as players of a game.Game theory can,thus,be adopted in accident analysis to depict the competition and cooperation between stakeholders.Subsequently,we established a game model to study the strategies of both supervisory and supervised entities.The obtained results demonstrate that the proposed game model allows for identifying the effectiveness deficiency of the supervisory entity,and the safety and protection altitudes of the supervised entity.The STAMP-Game model can generate quantitative parameters for supporting the behavior and strategy selections of the supervisory and supervised entities.The quantitative data obtained can be used to guide the safety improvement,to reduce the costs of safety regulation violation and accident risk.
基金the support of the Fundamental Research Funds for the Air Force Engineering University under Grant No.XZJK2019040。
文摘Target maneuver trajectory prediction is an important prerequisite for air combat situation awareness and maneuver decision-making.However,how to use a large amount of trajectory data generated by air combat confrontation training to achieve real-time and accurate prediction of target maneuver trajectory is an urgent problem to be solved.To solve this problem,in this paper,a hybrid algorithm based on transfer learning,online learning,ensemble learning,regularization technology,target maneuvering segmentation point recognition algorithm,and Volterra series,abbreviated as AERTrOS-Volterra is proposed.Firstly,the model makes full use of a large number of trajectory sample data generated by air combat confrontation training,and constructs a Tr-Volterra algorithm framework suitable for air combat target maneuver trajectory prediction,which realizes the extraction of effective information from the historical trajectory data.Secondly,in order to improve the real-time online prediction accuracy and robustness of the prediction model in complex electromagnetic environments,on the basis of the TrVolterra algorithm framework,a robust regularized online Sequential Volterra prediction model is proposed by integrating online learning method,regularization technology and inverse weighting calculation method based on the priori error.Finally,inspired by the preferable performance of models ensemble,ensemble learning scheme is also incorporated into our proposed algorithm,which adaptively updates the ensemble prediction model according to the performance of the model on real-time samples and the recognition results of target maneuvering segmentation points,including the adaptation of model weights;adaptation of parameters;and dynamic inclusion and removal of models.Compared with many existing time series prediction methods,the newly proposed target maneuver trajectory prediction algorithm can fully mine the prior knowledge contained in the historical data to assist the current prediction.The rationality and effectiveness of the proposed algorithm are verified by simulation on three sets of chaotic time series data sets and a set of real target maneuver trajectory data sets.
文摘In recent years,with the rapid development of e-commerce,people need to classify the wide variety and a large number of clothing images appearing on e-commerce platforms.In order to solve the problems of long time consumption and unsatisfactory classification accuracy arising from the classification of a large number of clothing images,researchers have begun to exploit deep learning techniques instead of traditional learning methods.The paper explores the use of convolutional neural networks(CNNs)for feature learning to enhance global feature information interactions by adding an improved hybrid attention mechanism(HAM)that fully utilizes feature weights in three dimensions:channel,height,and width.Moreover,the improved pooling layer not only captures local feature information,but also fuses global and local information to improve the misclassification problem that occurs between similar categories.Experiments on the Fashion-MNIST and DeepFashion datasets show that the proposed method significantly improves the accuracy of clothing classification(93.62%and 67.9%)compared with residual network(ResNet)and convolutional block attention module(CBAM).
基金supported by National Nature Science Foundation of China under Grant nos.30570442Hundred Talents Program of The Chinese Academy of SciencesKey Talent Program of Hefei institutes of Physical Science,CAS
文摘The strong static magnetic fields(SMFs)have caused great health concerns on the potential occupational and public exposure in recent year.However,the evidence in support of the cellular effects induced by the strong SMFs is inconsistent and the mechanisms are largely unknown.The present study focused on the effects of the strong SMFs on ATP contents in the two cell lines:human-hamster hybrid(AL)cells and mitochondria deficient human-hamster hybrid(ρ0 AL)cells.
