基于误差反馈学习的永磁同步电机变负载有限时间速度控制

针对传统永磁同步电机调速系统面对变负载和大范围调速时,P、I参数需要频繁调整且速度跟踪不理想的问题,提出了一种基于误差反馈学习结构的永磁同步电机有限时间速度控制方法。在对永磁同步电机运动方程分析的基础上,使用非线性PI和径向基神经网络建立了速度环控制器模型。前者保证控制系统收敛和稳定,其输出作为神经网络的误差学习参数;后者基于终端滑模理论设计参数调整律,加快神经网络的参数收敛速度,使得神经网络的输出逐渐取代非线性PI成为控制系统的主要控制器。利用李雅普诺夫稳定判据分析了控制器的收敛性,并在永磁同步电机调速系统上进行了试验。研究结果表明,基于误差反馈学习结构的有限时间控制策略能够减小系统静态误差和抖振,具有一定的抗干扰能力。

Finite time speed control of permanent magnet synchronous motor with variable load based on feedback-error learning

Aiming at improving the speed tracking performance and reducing the effect caused by mismatching parameters of proportion integral （PI）controllers when wide range of speed and variable load were considered in traditional permanent magnet synchronous motor （PMSM） speed control system, a finite time speed control approach was proposed based on feedback error learning structure. After the analysis of the motion equation of PMSM, the speed controller was established by combining with a nonlinear PI （NPI） controller in parallel with a radial basis function neural network （RBFNN）controller. The former was provided to guarantee system convergence and stability, and its output was used to update the NN parameters. The adaption laws of NN parameters were designed based on terminal sliding mode （TSM）principle to accelerate the parameters convergence speed. Then, output of the RBFNN controller can gradually replace that of the nonlinear PI controller. The stypticity of the controller was analyzed based on Lyapunov, and it was simulated on PMSM speed control system. The results indicate that the proposed finite time speed control approach which based on feedback error learning structure can reduce the static system error and chattering, and have strong anti-disturbance ability.