基于递归函数链模糊神经网络的永磁直线同步电动机位置控制

针对高精度永磁直线同步电动机(permanent magnet linear synchronous motor,PMLSM)存在参数变化、负载扰动、摩擦力等不确定性因素而影响电机伺服性能的问题,提出递归函数链模糊神经网络控制(RFLFNN)保证系统的伺服性能。首先在磁场定向控制下建立PMLSM伺服系统动态数学模型。其次,将函数链神经网络(FLNN)和递归模糊神经网络(RFNN)相结合设计RFLFNN控制策略,利用FLNN实现神经网络的函数扩展,提高系统的非线性逼近能力并对系统参数进行辨识; RFNN采用反向传播算法实时更新并调整神经网络的参数值,对系统中存在的不确定性因素进行估计以抑制不确定性因素对系统的影响。最后,通过系统实验证明所提方法的有效性,实验结果表明,与RFNN相比,该方法极大地改善了PMLSM伺服系统的位置跟踪性能。

Position Control of PMLSM Based on Recurrent Functional Link Fuzzy Neural Network

In order to solve the problem that the servo performance of high precision permanent magnet linear synchronous motor (PMLSM) is affected by uncertainties such as parameter changes, load disturbances, frictions and so on, a recurrent functional link fuzzy neural network control (RFLFNN) was proposed to ensure the servo performance of the system. Firstly, the dynamic mathematical model of PMLSM servo system was established under the field-oriented control. Secondly, the RFLFNN control strategy was designed by combining the functional link neural network (FLNN) and the recurrent fuzzy neural network (RFNN). FLNN was used to realize function expansion of the neural network, which could improve the non-linear approximation ability of the system and identify the system parameters. RFNN could update and adjust the parameters of the neural network in real time by using the back propagation algorithm, thus it could estimate the uncertainties in the system and suppress the influence of uncertainties on the system. Finally, the effectiveness of the proposed method was proved by system experiments. The experimental results show that the proposed method not only improves the position tracking performance of PMLSM servo system, but also improves the robust performance of the system.