武器站伺服系统的模糊神经自适应PI控制

Weapon Station Servo System Based on Fuzzy Neural Adaptive PI Control

针对武器站伺服控制中存在的转动惯量和负载力矩变化大、冲击力矩强等特点,提出了一种模糊神经自适应PI控制策略,运用于自动武器站转速控制。该策略采用双输入单输出的结构,根据模糊神经原理,将转速误差及误差积分作为模糊神经PI控制器的输入量,经Sugeno模型的神经网络处理之后得到的输出值。由PI值推导出的初值能够使系统快速响应,提高了自适应能力。仿真及实验结果表明该控制策略具有更好的稳定性和动态跟踪精度。

To overcome wide variations in loads and parameters and large disturbed moment,an adaptive fuzzy neural network PI controller is designed and applied in rotating velocity control of auto-weapon station.The controller has the structure of dual inputs and single output,which both the error and integration of the error become the inputs of the proposed controller,and output can be obtained by Sugeno model fuzzy neural network.The response and adaptive ability of drive system are improved by initial value derived by PI values.Simulation and experimental results show that the control strategy has better stability and dynamic tracking accuracy.