Currently, simultaneously ensuring the machining accuracy and efficiency of thin-walled structures especially high performance parts still remains a challenge. Existing compensating methods are mainly focusing on 3-ai...Currently, simultaneously ensuring the machining accuracy and efficiency of thin-walled structures especially high performance parts still remains a challenge. Existing compensating methods are mainly focusing on 3-aixs machining, which sometimes only take one given point as the compensative point at each given cutter location. This paper presents a redesigned surface based machining strategy for peripheral milling of thin-walled parts. Based on an improved cutting force/heat model and finite element method（FEM） simulation environment, a deflection error prediction model, which takes sequence of cutter contact lines as compensation targets, is established. And an iterative algorithm is presented to determine feasible cutter axis positions. The final redesigned surface is subsequently generated by skinning all discrete cutter axis vectors after compensating by using the proposed algorithm. The proposed machining strategy incorporates the thermo-mechanical coupled effect in deflection prediction, and is also validated with flank milling experiment by using five-axis machine tool. At the same time, the deformation error is detected by using three-coordinate measuring machine. Error prediction values and experimental results indicate that they have a good consistency and the proposed approach is able to significantly reduce the dimension error under the same machining conditions compared with conventional methods. The proposed machining strategy has potential in high-efficiency precision machining of thin-walled parts.展开更多
针对现有时间序列在线预测方法存在对数据特性变化感知与预测及时性不足的问题,创新设计了一种基于信息感知权重与误差预测的时间序列在线预测方法。该方法利用信息感知权重替换代价函数中遗忘因子λ0参量;通过建立输入数据与预测误差...针对现有时间序列在线预测方法存在对数据特性变化感知与预测及时性不足的问题,创新设计了一种基于信息感知权重与误差预测的时间序列在线预测方法。该方法利用信息感知权重替换代价函数中遗忘因子λ0参量;通过建立输入数据与预测误差的映射关系进行误差预测,采用加权误差补偿系数实现误差补偿。通过改变隐含层节点数方法进行多次单步预测实验,实验结果从预测精度和泛化性等多方面验证了设计方法优异的单步预测能力。其中,Sinc、Mackey-Glass和Solar Energy 3个数据选取点的单步预测方差分别达到1.56×10-13、2.29×10-7与1.43。根据实际失效情况分别设定失效电压为5.8与5.6 V,并针对封装降压电源模块加速寿命实验实测数据进行多步预测。五步与十步预测结果显示设计方法均有效预测电源失效。实验结果全面说明设计方法在预测数据特性发生变化情况时,能够稳定、精准且有效地完成在线单步与多步预测。展开更多
Due to the complexity of economic system and the interactive effects between all kinds of economic variables and foreign trade, it is not easy to predict foreign trade volume. However, the difficulty in predicting for...Due to the complexity of economic system and the interactive effects between all kinds of economic variables and foreign trade, it is not easy to predict foreign trade volume. However, the difficulty in predicting foreign trade volume is usually attributed to the limitation of many conventional forecasting models. To improve the prediction performance, the study proposes a novel kernel-based ensemble learning approach hybridizing econometric models and artificial intelligence (AI) models to predict China's foreign trade volume. In the proposed approach, an important econometric model, the co-integration-based error correction vector auto-regression (EC-VAR) model is first used to capture the impacts of all kinds of economic variables on Chinese foreign trade from a multivariate linear anal- ysis perspective. Then an artificial neural network (ANN) based EC-VAR model is used to capture the nonlinear effects of economic variables on foreign trade from the nonlinear viewpoint. Subsequently, for incorporating the effects of irregular events on foreign trade, the text mining and expert's judgmental adjustments are also integrated into the nonlinear ANN-based EC-VAR model. Finally, all kinds of economic variables, the outputs of linear and nonlinear EC-VAR models and judgmental adjustment model are used as input variables of a typical kernel-based support vector regression (SVR) for en- semble prediction purpose. For illustration, the proposed kernel-based ensemble learning methodology hybridizing econometric techniques and AI methods is applied to China's foreign trade volume predic- tion problem. Experimental results reveal that the hybrid econometric-AI ensemble learning approach can significantly improve the prediction performance over other linear and nonlinear models listed in this study.展开更多
This paper proposes a novel method to predict the spur gear pair’s static transmission error based on the accuracy grade,in which manufacturing errors(MEs),assembly errors(AEs),tooth deflections(TDs)and profile modif...This paper proposes a novel method to predict the spur gear pair’s static transmission error based on the accuracy grade,in which manufacturing errors(MEs),assembly errors(AEs),tooth deflections(TDs)and profile modifications(PMs)are considered.For the prediction,a discrete gear model for generating the error tooth profile based on the ISO accuracy grade is presented.Then,the gear model and a tooth deflection model for calculating the tooth compliance on gear meshing are coupled with the transmission error model to make the prediction by checking the interference status between gear and pinion.The prediction method is validated by comparison with the experimental results from the literature,and a set of cases are simulated to study the effects of MEs,AEs,TDs and PMs on the static transmission error.In addition,the time-varying backlash caused by both MEs and AEs,and the contact ratio under load conditions are also investigated.The results show that the novel method can effectively predict the range of the static transmission error under different accuracy grades.The prediction results can provide references for the selection of gear design parameters and the optimization of transmission performance in the design stage of gear systems.展开更多
In order to solve the problem of difficult modeling and identification caused by time-variable parameters,multiple inputs and outputs and unstable open loop,a subsystem model-based close-loop grey-box identification m...In order to solve the problem of difficult modeling and identification caused by time-variable parameters,multiple inputs and outputs and unstable open loop,a subsystem model-based close-loop grey-box identification method was put forward when consider the main coupling effects of hydraulic Stewart platform.Firstly,the whole system is divided into three TITO(Two Input Two Output) subsystems according to the characteristics of the pseudo-mass matrix,hence transfer function matrix model of the subsystem can also be found.Secondly,since the Stewart platform is unstable,the close-loop transfer model of the subsystem is derived under the proportional controllers.The inverse M serial is adopted as the identification signal to get the experimental data.All parameters of the subsystem are determined in close-loop indirect identification by PEM(Prediction Error Method).Finally,a case study validates the correctness and effectiveness of the subsystem model-based close-loop grey-box identification method for hydraulic Stewart platform.展开更多
Human error,an important factor,may lead to serious results in various operational fields.The human factor plays a critical role in the risks and hazards of the maritime industry.A ship can achieve safe navigation whe...Human error,an important factor,may lead to serious results in various operational fields.The human factor plays a critical role in the risks and hazards of the maritime industry.A ship can achieve safe navigation when all operations in the engine room are conducted vigilantly.This paper presents a systematic evaluation of 20 failures in auxiliary systems of marine diesel engines that may be caused by human error.The Cognitive Reliability Error Analysis Method(CREAM)is used to determine the potentiality of human errors in the failures implied thanks to the answers of experts.Using this method,the probabilities of human error on failures were evaluated and the critical ones were emphasized.The measures to be taken for these results will make significant contributions not only to the seafarers but also to the ship owners.展开更多
基金supported by Key Program of National Natural Science Foundation of China (Grant No. 50835001) General Program of National Natural Science Foundation of China (Grant No. 50775023)Program for New Century Excellent Talents of Ministry of Education of China (Grant No. NCET-08-081)
文摘Currently, simultaneously ensuring the machining accuracy and efficiency of thin-walled structures especially high performance parts still remains a challenge. Existing compensating methods are mainly focusing on 3-aixs machining, which sometimes only take one given point as the compensative point at each given cutter location. This paper presents a redesigned surface based machining strategy for peripheral milling of thin-walled parts. Based on an improved cutting force/heat model and finite element method（FEM） simulation environment, a deflection error prediction model, which takes sequence of cutter contact lines as compensation targets, is established. And an iterative algorithm is presented to determine feasible cutter axis positions. The final redesigned surface is subsequently generated by skinning all discrete cutter axis vectors after compensating by using the proposed algorithm. The proposed machining strategy incorporates the thermo-mechanical coupled effect in deflection prediction, and is also validated with flank milling experiment by using five-axis machine tool. At the same time, the deformation error is detected by using three-coordinate measuring machine. Error prediction values and experimental results indicate that they have a good consistency and the proposed approach is able to significantly reduce the dimension error under the same machining conditions compared with conventional methods. The proposed machining strategy has potential in high-efficiency precision machining of thin-walled parts.
文摘针对现有时间序列在线预测方法存在对数据特性变化感知与预测及时性不足的问题,创新设计了一种基于信息感知权重与误差预测的时间序列在线预测方法。该方法利用信息感知权重替换代价函数中遗忘因子λ0参量;通过建立输入数据与预测误差的映射关系进行误差预测,采用加权误差补偿系数实现误差补偿。通过改变隐含层节点数方法进行多次单步预测实验,实验结果从预测精度和泛化性等多方面验证了设计方法优异的单步预测能力。其中,Sinc、Mackey-Glass和Solar Energy 3个数据选取点的单步预测方差分别达到1.56×10-13、2.29×10-7与1.43。根据实际失效情况分别设定失效电压为5.8与5.6 V,并针对封装降压电源模块加速寿命实验实测数据进行多步预测。五步与十步预测结果显示设计方法均有效预测电源失效。实验结果全面说明设计方法在预测数据特性发生变化情况时,能够稳定、精准且有效地完成在线单步与多步预测。
基金the National Natural Science Foundation of China under Grant Nos.70601029 and 70221001the Knowledge Innovation Program of the Chinese Academy of Sciences under Grant Nos.3547600,3046540,and 3047540the Strategy Research Grant of City University of Hong Kong under Grant No.7001806
文摘Due to the complexity of economic system and the interactive effects between all kinds of economic variables and foreign trade, it is not easy to predict foreign trade volume. However, the difficulty in predicting foreign trade volume is usually attributed to the limitation of many conventional forecasting models. To improve the prediction performance, the study proposes a novel kernel-based ensemble learning approach hybridizing econometric models and artificial intelligence (AI) models to predict China's foreign trade volume. In the proposed approach, an important econometric model, the co-integration-based error correction vector auto-regression (EC-VAR) model is first used to capture the impacts of all kinds of economic variables on Chinese foreign trade from a multivariate linear anal- ysis perspective. Then an artificial neural network (ANN) based EC-VAR model is used to capture the nonlinear effects of economic variables on foreign trade from the nonlinear viewpoint. Subsequently, for incorporating the effects of irregular events on foreign trade, the text mining and expert's judgmental adjustments are also integrated into the nonlinear ANN-based EC-VAR model. Finally, all kinds of economic variables, the outputs of linear and nonlinear EC-VAR models and judgmental adjustment model are used as input variables of a typical kernel-based support vector regression (SVR) for en- semble prediction purpose. For illustration, the proposed kernel-based ensemble learning methodology hybridizing econometric techniques and AI methods is applied to China's foreign trade volume predic- tion problem. Experimental results reveal that the hybrid econometric-AI ensemble learning approach can significantly improve the prediction performance over other linear and nonlinear models listed in this study.
基金Project(51675061)supported by the National Natural Science Foundation of China。
文摘This paper proposes a novel method to predict the spur gear pair’s static transmission error based on the accuracy grade,in which manufacturing errors(MEs),assembly errors(AEs),tooth deflections(TDs)and profile modifications(PMs)are considered.For the prediction,a discrete gear model for generating the error tooth profile based on the ISO accuracy grade is presented.Then,the gear model and a tooth deflection model for calculating the tooth compliance on gear meshing are coupled with the transmission error model to make the prediction by checking the interference status between gear and pinion.The prediction method is validated by comparison with the experimental results from the literature,and a set of cases are simulated to study the effects of MEs,AEs,TDs and PMs on the static transmission error.In addition,the time-varying backlash caused by both MEs and AEs,and the contact ratio under load conditions are also investigated.The results show that the novel method can effectively predict the range of the static transmission error under different accuracy grades.The prediction results can provide references for the selection of gear design parameters and the optimization of transmission performance in the design stage of gear systems.
文摘In order to solve the problem of difficult modeling and identification caused by time-variable parameters,multiple inputs and outputs and unstable open loop,a subsystem model-based close-loop grey-box identification method was put forward when consider the main coupling effects of hydraulic Stewart platform.Firstly,the whole system is divided into three TITO(Two Input Two Output) subsystems according to the characteristics of the pseudo-mass matrix,hence transfer function matrix model of the subsystem can also be found.Secondly,since the Stewart platform is unstable,the close-loop transfer model of the subsystem is derived under the proportional controllers.The inverse M serial is adopted as the identification signal to get the experimental data.All parameters of the subsystem are determined in close-loop indirect identification by PEM(Prediction Error Method).Finally,a case study validates the correctness and effectiveness of the subsystem model-based close-loop grey-box identification method for hydraulic Stewart platform.
文摘Human error,an important factor,may lead to serious results in various operational fields.The human factor plays a critical role in the risks and hazards of the maritime industry.A ship can achieve safe navigation when all operations in the engine room are conducted vigilantly.This paper presents a systematic evaluation of 20 failures in auxiliary systems of marine diesel engines that may be caused by human error.The Cognitive Reliability Error Analysis Method(CREAM)is used to determine the potentiality of human errors in the failures implied thanks to the answers of experts.Using this method,the probabilities of human error on failures were evaluated and the critical ones were emphasized.The measures to be taken for these results will make significant contributions not only to the seafarers but also to the ship owners.
