An adaptive inverse controller for nonliear discrete-time system is proposed in this paper. A compound neural network is constructed to identify the nonlinear system, which includes a linear part to approximate the no...An adaptive inverse controller for nonliear discrete-time system is proposed in this paper. A compound neural network is constructed to identify the nonlinear system, which includes a linear part to approximate the nonlinear system and a recurrent neural network to minimize the difference between the linear model and the real nonlinear system. Because the current control input is not included in the input vector of recurrent neural network (RNN), the inverse control law can be calculated directly. This scheme can be used in real-time nonlinear single-input single-output (SISO) and multi-input multi-output (MIMO) system control with less computation work. Simulation studies have shown that this scheme is simple and affects good control accuracy and robustness.展开更多
The results of an expert system of lanthanide intermetallic compounds using artificial neural networks and chemical bond parameter method were reported. Two pattern recognition neural models, one for prediction of the...The results of an expert system of lanthanide intermetallic compounds using artificial neural networks and chemical bond parameter method were reported. Two pattern recognition neural models, one for prediction of the occurrence of 1 : 1 lanthanide intermetallic compounds with CsClstructure and the other for prediction of congruent or incongruent melting types, were developed. Four regression neural models were also developed for prediction of melting point of these compounds. In order to get rid of overfitting, cross-vahdation method was used for the neural models. And satisfactory results were obtained in all of the neural models in this paper.展开更多
The bonding of solid steel plate to liquid aluminum was studied using rapidsolidification. The surface of solid steel plate was defatted, descaled, immersed (in K_2ZrF_6 fluxaqueous solution) and stoved. In order to d...The bonding of solid steel plate to liquid aluminum was studied using rapidsolidification. The surface of solid steel plate was defatted, descaled, immersed (in K_2ZrF_6 fluxaqueous solution) and stoved. In order to determine the thickness of Fe-Al compound layer at theinterface of steel-aluminum solid to liquid bonding under rapid solidification, the interface ofbonding plate was investigated by SEM (Scanning Electron Microscope) experiment. The relationshipbetween bonding parameters (such as preheat temperature of steel plate, temperature of aluminumliquid and bonding time) and thickness of Fe-Al compound layer at the interface was established byartificial neural networks (ANN) perfectly. The maximum of relative error between the output and thedesired output of the ANN is only 5.4%. From the bonding parameters for the largest interfacialshear strength of bonding plate (226℃ for preheat temperature of steel plate, 723℃ for temperatureof aluminum liquid and 15.8 s for bonding time), the reasonable thickness of Fe-Al compound layer10.8 μm was got.展开更多
With the artificial neural network（ANN） method combined with the multiple linear regression（MLR）,based on a series of quantum chemical descriptors and molecular connectivity indexes,quantitative structure-activity...With the artificial neural network（ANN） method combined with the multiple linear regression（MLR）,based on a series of quantum chemical descriptors and molecular connectivity indexes,quantitative structure-activity relationship（QSAR） models to predict the acute toxicity（-lgEC50） of substituted aromatic compounds to Photobacterium phosphoreum were established.Four molecular descriptors that appear in the MLR model,namely,the second order valence molecular connectivity index（2XV）,the energy of the highest occupied molecular orbital（EHOMO）,the logarithm of n-octyl alcohol/water partition coefficient（logKow） and the Connolly molecular area（MA）,were inputs of the ANN model.The root-mean-square error（RMSE） of the training and validation sets of the ANN model are 0.1359 and 0.2523,and the correlation coefficient（R） is 0.9810 and 0.8681,respectively.The leave-one-out（LOO） cross validated correlation coefficient（Q L2OO） of the MLR and ANN models is 0.6954 and 0.6708,respectively.The result showed that the two methods are complementary in the calculations.The regression method gave support to the neural network with physical explanation,and the neural network method gave a more accurate model for QSAR.In addition,some insights into the structural factors affecting the acute toxicity and toxicity mechanism of substituted aromatic compounds were discussed.展开更多
考虑模型不确定性和外部干扰等影响,基于径向基函数神经网络(radial basis function neural network,RBFNN)和改进的误差符号函数鲁棒积分(robust integral of signum error,RISE)技术,建立无人直升机(unmanned aerial helicopter,UAH)...考虑模型不确定性和外部干扰等影响,基于径向基函数神经网络(radial basis function neural network,RBFNN)和改进的误差符号函数鲁棒积分(robust integral of signum error,RISE)技术,建立无人直升机(unmanned aerial helicopter,UAH)轨迹跟踪控制设计方案。首先,建立包含模型不确定性和外部干扰的UAH非线性系统模型,利用跟踪误差作为RBFNN输入信号估计由模型不确定性和外部干扰组成的复合扰动。其次,将滤波信号及其变化率权重组合作为RISE输入信号设计控制器,从而降低控制设计方案对UAH动力学模型的依赖程度。进而,借助Lyapunov稳定性理论分析整合后闭环跟踪误差系统的稳定性,并给出控制参数的选取方法。最后,借助现有文献中UAH系统模型,仿真与比较结果均说明所提控制算法的有效性和优越性。展开更多
基金Supported by the National Natural Science Foundation of China (60575009, 60574036)
文摘An adaptive inverse controller for nonliear discrete-time system is proposed in this paper. A compound neural network is constructed to identify the nonlinear system, which includes a linear part to approximate the nonlinear system and a recurrent neural network to minimize the difference between the linear model and the real nonlinear system. Because the current control input is not included in the input vector of recurrent neural network (RNN), the inverse control law can be calculated directly. This scheme can be used in real-time nonlinear single-input single-output (SISO) and multi-input multi-output (MIMO) system control with less computation work. Simulation studies have shown that this scheme is simple and affects good control accuracy and robustness.
文摘The results of an expert system of lanthanide intermetallic compounds using artificial neural networks and chemical bond parameter method were reported. Two pattern recognition neural models, one for prediction of the occurrence of 1 : 1 lanthanide intermetallic compounds with CsClstructure and the other for prediction of congruent or incongruent melting types, were developed. Four regression neural models were also developed for prediction of melting point of these compounds. In order to get rid of overfitting, cross-vahdation method was used for the neural models. And satisfactory results were obtained in all of the neural models in this paper.
基金This project is financially supported by National Natural Science Foundation of China (No.50274047) and Advanced Technical Committee of China(No. 715-009-060)
文摘The bonding of solid steel plate to liquid aluminum was studied using rapidsolidification. The surface of solid steel plate was defatted, descaled, immersed (in K_2ZrF_6 fluxaqueous solution) and stoved. In order to determine the thickness of Fe-Al compound layer at theinterface of steel-aluminum solid to liquid bonding under rapid solidification, the interface ofbonding plate was investigated by SEM (Scanning Electron Microscope) experiment. The relationshipbetween bonding parameters (such as preheat temperature of steel plate, temperature of aluminumliquid and bonding time) and thickness of Fe-Al compound layer at the interface was established byartificial neural networks (ANN) perfectly. The maximum of relative error between the output and thedesired output of the ANN is only 5.4%. From the bonding parameters for the largest interfacialshear strength of bonding plate (226℃ for preheat temperature of steel plate, 723℃ for temperatureof aluminum liquid and 15.8 s for bonding time), the reasonable thickness of Fe-Al compound layer10.8 μm was got.
基金supported by the Natural Science Foundation of Fujian Province (D0710019)the Natural Science Foundation of Overseas Chinese Affairs Office of the State Council (06QZR09)
文摘With the artificial neural network（ANN） method combined with the multiple linear regression（MLR）,based on a series of quantum chemical descriptors and molecular connectivity indexes,quantitative structure-activity relationship（QSAR） models to predict the acute toxicity（-lgEC50） of substituted aromatic compounds to Photobacterium phosphoreum were established.Four molecular descriptors that appear in the MLR model,namely,the second order valence molecular connectivity index（2XV）,the energy of the highest occupied molecular orbital（EHOMO）,the logarithm of n-octyl alcohol/water partition coefficient（logKow） and the Connolly molecular area（MA）,were inputs of the ANN model.The root-mean-square error（RMSE） of the training and validation sets of the ANN model are 0.1359 and 0.2523,and the correlation coefficient（R） is 0.9810 and 0.8681,respectively.The leave-one-out（LOO） cross validated correlation coefficient（Q L2OO） of the MLR and ANN models is 0.6954 and 0.6708,respectively.The result showed that the two methods are complementary in the calculations.The regression method gave support to the neural network with physical explanation,and the neural network method gave a more accurate model for QSAR.In addition,some insights into the structural factors affecting the acute toxicity and toxicity mechanism of substituted aromatic compounds were discussed.
文摘考虑模型不确定性和外部干扰等影响,基于径向基函数神经网络(radial basis function neural network,RBFNN)和改进的误差符号函数鲁棒积分(robust integral of signum error,RISE)技术,建立无人直升机(unmanned aerial helicopter,UAH)轨迹跟踪控制设计方案。首先,建立包含模型不确定性和外部干扰的UAH非线性系统模型,利用跟踪误差作为RBFNN输入信号估计由模型不确定性和外部干扰组成的复合扰动。其次,将滤波信号及其变化率权重组合作为RISE输入信号设计控制器,从而降低控制设计方案对UAH动力学模型的依赖程度。进而,借助Lyapunov稳定性理论分析整合后闭环跟踪误差系统的稳定性,并给出控制参数的选取方法。最后,借助现有文献中UAH系统模型,仿真与比较结果均说明所提控制算法的有效性和优越性。
基金This work is supported by Ministry of Science and Technology of China(No.2016YFA0200602 and No.2018YFA0208603)the National Natural Science Foundation of China(No.21573204 and No.21421063)Anhui Initiative in Quantum Information Technologies,Fundamental Research Funds for the Central Universities,National Program for Support of Top-notch Young Professional,CAS Interdisciplinary Innovation Team.
