Permanent magnet synchronous motor(PMSM),known for their compact size and high-power density,is widely used in fields such as electric vehicles and servo drives.However,traditional PID control methods for PMSM cannot ...Permanent magnet synchronous motor(PMSM),known for their compact size and high-power density,is widely used in fields such as electric vehicles and servo drives.However,traditional PID control methods for PMSM cannot dynamically adjust parameters according to varying operating conditions.To address this issue,this paper proposes a PID control method based on a radial basis function(RBF)neural network,which adaptively tunes the PID controller parameters.First,an offline RBF neural network with optimal structural parameters is trained using the current and speed data of the PMSM,and then employed to construct the RBF-PID controller.During online training,the Jacobian information calculated via the RBF neural network is used to adaptively adjust the PID parameters.Simultaneously,the structural parameters of the RBF network are updated in reverse based on the error between the predicted and reference speed values.Finally,numerical simulations and experiments in the context of electric vehicle drive control show that the maximum speed errors of the SMC controller and the RBF-PID controller are 1.97 km/h and 0.17 km/h,respectively.Moreover,the RBF-PID controller outperforms both the SMC and traditional PID controllers in handling sudden speed changes.展开更多
文摘Permanent magnet synchronous motor(PMSM),known for their compact size and high-power density,is widely used in fields such as electric vehicles and servo drives.However,traditional PID control methods for PMSM cannot dynamically adjust parameters according to varying operating conditions.To address this issue,this paper proposes a PID control method based on a radial basis function(RBF)neural network,which adaptively tunes the PID controller parameters.First,an offline RBF neural network with optimal structural parameters is trained using the current and speed data of the PMSM,and then employed to construct the RBF-PID controller.During online training,the Jacobian information calculated via the RBF neural network is used to adaptively adjust the PID parameters.Simultaneously,the structural parameters of the RBF network are updated in reverse based on the error between the predicted and reference speed values.Finally,numerical simulations and experiments in the context of electric vehicle drive control show that the maximum speed errors of the SMC controller and the RBF-PID controller are 1.97 km/h and 0.17 km/h,respectively.Moreover,the RBF-PID controller outperforms both the SMC and traditional PID controllers in handling sudden speed changes.
文摘针对工业生产中加热炉物料出口温度的非线性、时变性、大滞后性等特点,无法建立精确的数学模型,并且为提高该系统控制的可靠性和安全性,达到精确控制。提出一种基于RBF(Radial Basis Function)神经网络的PID串级控制器,即先用建立的三层RBF神经网络在线辨识得到梯度信息,再用梯度信息在线整定PID控制的三个参数,最后将整定的PID控制物料出口温度-炉膛温度串级系统的主回路。仿真结果表明,RBF-PID串级控制较传统P I D串级控制有较强的鲁棒性,提高了控制品质,获得了更好的控制效果。
