The high cost and low efficiency of full-scale vehicle experiments and numerical simulations limit the efficient development of armored vehicle occupant protection systems.The floor-occupant-seat local simulation mode...The high cost and low efficiency of full-scale vehicle experiments and numerical simulations limit the efficient development of armored vehicle occupant protection systems.The floor-occupant-seat local simulation model provides an alternative solution for quickly evaluating the performance of occupant protection systems.However,the error and rationality of the loading of the thin-walled floor in the local model cannot be ignored.This study proposed an equivalent loading method for the local model,which includes two parts:the dimensionality reduction method for acceleration matrix and the joint optimization framework for equivalent node coordinates.In the dimensionality reduction method,the dimension of the acceleration matrix was reduced based on the improved kernel principal component analysis(KPCA),and a dynamic variable bandwidth was introduced to address the limitation of failing to effectively measure the similarity between acceleration data in conventional KPCA.In addition,a least squares problem with forced displacement constraints was constructed to solve the correction matrix,thereby achieving the scale restoration process of the principal component acceleration matrix.The joint optimization framework for coordinates consists of the error assessment of response time histories(EARTH)and Bayesian optimization.In this framework,the local loading error of the equivalent acceleration matrix is taken as the Bayesian optimization objective,which is quantified and scored by EARTH.The expected improvement acquisition function was used to select the new set of the equivalent acceleration node coordinates for the self-updating optimization of the observation dataset and Gaussian process surrogate model.We reduced the dimension of the acceleration matrix from 2256 to 7,while retaining 91%of the information features.The comprehensive error score of occupant's lower limb response in the local model increased from 58.5%to 80.4%.The proposed equivalent loading method provides a solution for the rapid and reliable development of occupant protection systems.展开更多
Considering chaotic time series multi-step prediction, multi-step direct prediction model based on partial least squares （PLS） is proposed in this article, where PLS, the method for predicting a set of dependent var...Considering chaotic time series multi-step prediction, multi-step direct prediction model based on partial least squares （PLS） is proposed in this article, where PLS, the method for predicting a set of dependent variables forming a large set of predictors, is used to model the dynamic evolution between the space points and the corresponding future points. The model can eliminate error accumulation with the common single-step local model algorithm~ and refrain from the high multi-collinearity problem in the reconstructed state space with the increase of embedding dimension. Simulation predictions are done on the Mackey-Glass chaotic time series with the model. The satisfying prediction accuracy is obtained and the model efficiency verified. In the experiments, the number of extracted components in PLS is set with cross-validation procedure.展开更多
Comparison of regular(diurnal,seasonal and solar cycle)variations of high-latitude,mid-latitude and low-latitude ionospheric characteristics has been provided on basis of local empirical models of the peak electron de...Comparison of regular(diurnal,seasonal and solar cycle)variations of high-latitude,mid-latitude and low-latitude ionospheric characteristics has been provided on basis of local empirical models of the peak electron density and the peak height.The local empirical models were derived from the hand-scaled ionogram data recorded by DPS-4 digisondes located at Norilsk(69°N,88°E),Irkutsk(52°N,104°E)and Hainan(19°N,109°E)for a 6-year period from December,2002 to December,2008.The technique used to build the local empirical model is described.The primary focus is diurnal-seasonal behavior under low solar activity and its change with increasing solar activity.Both common and specific features of the high-latitude(Norilsk),mid-latitude(Irkutsk)and low-latitude(Hainan)regular variations were revealed using their local empirical models.展开更多
基金supported by the National Natural Science Foundation of China(Grant Nos.52272437 and 52272370)the Postgraduate Research&Practice Innovation Program of Jiangsu Province(KYCX24_0635)。
文摘The high cost and low efficiency of full-scale vehicle experiments and numerical simulations limit the efficient development of armored vehicle occupant protection systems.The floor-occupant-seat local simulation model provides an alternative solution for quickly evaluating the performance of occupant protection systems.However,the error and rationality of the loading of the thin-walled floor in the local model cannot be ignored.This study proposed an equivalent loading method for the local model,which includes two parts:the dimensionality reduction method for acceleration matrix and the joint optimization framework for equivalent node coordinates.In the dimensionality reduction method,the dimension of the acceleration matrix was reduced based on the improved kernel principal component analysis(KPCA),and a dynamic variable bandwidth was introduced to address the limitation of failing to effectively measure the similarity between acceleration data in conventional KPCA.In addition,a least squares problem with forced displacement constraints was constructed to solve the correction matrix,thereby achieving the scale restoration process of the principal component acceleration matrix.The joint optimization framework for coordinates consists of the error assessment of response time histories(EARTH)and Bayesian optimization.In this framework,the local loading error of the equivalent acceleration matrix is taken as the Bayesian optimization objective,which is quantified and scored by EARTH.The expected improvement acquisition function was used to select the new set of the equivalent acceleration node coordinates for the self-updating optimization of the observation dataset and Gaussian process surrogate model.We reduced the dimension of the acceleration matrix from 2256 to 7,while retaining 91%of the information features.The comprehensive error score of occupant's lower limb response in the local model increased from 58.5%to 80.4%.The proposed equivalent loading method provides a solution for the rapid and reliable development of occupant protection systems.
文摘Considering chaotic time series multi-step prediction, multi-step direct prediction model based on partial least squares （PLS） is proposed in this article, where PLS, the method for predicting a set of dependent variables forming a large set of predictors, is used to model the dynamic evolution between the space points and the corresponding future points. The model can eliminate error accumulation with the common single-step local model algorithm~ and refrain from the high multi-collinearity problem in the reconstructed state space with the increase of embedding dimension. Simulation predictions are done on the Mackey-Glass chaotic time series with the model. The satisfying prediction accuracy is obtained and the model efficiency verified. In the experiments, the number of extracted components in PLS is set with cross-validation procedure.
基金Supported by Russian Foundation for Basic Research(13-05-91159-GFEN_a)Project 14.518.11.7065 and agreement N8388 of the Ministry of Education and Science of the Russian Federation+1 种基金the National Natural Science Foundation(41274146)the Specialized Research Fund for State Key Laboratory in China
文摘Comparison of regular(diurnal,seasonal and solar cycle)variations of high-latitude,mid-latitude and low-latitude ionospheric characteristics has been provided on basis of local empirical models of the peak electron density and the peak height.The local empirical models were derived from the hand-scaled ionogram data recorded by DPS-4 digisondes located at Norilsk(69°N,88°E),Irkutsk(52°N,104°E)and Hainan(19°N,109°E)for a 6-year period from December,2002 to December,2008.The technique used to build the local empirical model is described.The primary focus is diurnal-seasonal behavior under low solar activity and its change with increasing solar activity.Both common and specific features of the high-latitude(Norilsk),mid-latitude(Irkutsk)and low-latitude(Hainan)regular variations were revealed using their local empirical models.
