基于改进果蝇神经网络的数控工作台PID控制

Speed Control of CNC Worktable Based on Improved Fruit Fly Optimization Algorithm and Neural Networks PID

传统数控机床工作台速度控制多采用PID控制,传统PID控制存在响应慢、超调大、动态性差、抗干扰能力差等问题。提出了一种变步长果蝇神经网络PID控制方法,将变步长果蝇算法与神经网络联合使用。在线寻找控制器最优控制参数,实现电机速度的智能化控制。通过Simulink仿真,对数控工作台速度控制内环转速跟踪性能进行分析。实验结果证明,改进果蝇神经网络的PID控制比传统PID控制方案响应更快,抗干扰性能和鲁棒性能更好。

The speed control of traditional CNC table regularly use PID control, however the traditional PID control has many problems, such as response slowly, over adjustment, poolly dynamic performance and poolly anti-interference ability. On the basis of previous PID control,a strategy of speed control will be designed which uses variable step size fruit fly optimization algorithm and neural networks. The best control parameters will be searched online throught combine variable step size fruit fly optimization algorithm and neural networks, thus the speed of motor will be intelligent controlled. By Simulink simulation, the rotate speed tracking performance of CNC worktable has been analyzed. The results of simulation proved that the improved fruit fly optimization algorithm and neural networks control compared with traditional PID control has a rapidly response and great dynamic tracking performance and robustness.