摘要
针对永磁同步直线电机的初级磁链近似为常数这一特点,在d-q轴下建立了直线电机的数学模型。直线电机具有非线性、耦合性、负载扰动、时变不确定性等特点。常规PID控制虽然结构简单、输出稳定、易实现,但在高速高精度应用场合却不能达到理想的控制效果。提出了一种基于RBF神经网络与传统PID控制相结合的新策略,形成RBF神经网络整定PID控制,在一定程度上改进了PID控制的性能。仿真结果表明,RBF神经网络PID控制具有更好的动态响应性和更加稳定的跟踪性能。
According to the characteristic of the magnet flux of permanent magnet linear synchronous motors ( PMLSM ) is a constant, the mathematical mode of PMLSM can be obtained by analyzing it's d-q model. Linear motors are characteristic with non-linearity, coupling, load disturbance, time-varying uncertainty, hard to establish precise mathematical model. Although the conventional PID control has simple structure, stable output and easy to realize, it can not achieve ideal control effect in high speed, high precision applications. A RBF neural network PID controller was designed by combined RBF neural network with conventional PID control. To some extent, it could improve the performance of traditional PID controller. Simulation results showed that the RBF neural network PID controller was superior to the conventional PID controller in dynamic stability performance and speed tracking power.
出处
《电机与控制应用》
北大核心
2012年第6期29-32,共4页
Electric machines & control application
基金
上海市教委科研创新项目(11cxy64)
上海市教委重点学科电力电子与电力传动(J51901)