摘要
焊接电源输出是一个多变量、非线性、参数时变的被控对象,传统的PID控制参数固定不变,难以根据实际情况对参数进行自适应调整,最终输出的焊接电流效果较差。针对以上情况,提出了一种基于RBF(Radial Basis Function)神经网络的PID的焊接电流智能控制方法。建立了3层神经网络模型,由RBF神经网络在线辨识得到了梯度信息,最后由梯度信息对PID中的三个参数进行在线调整,从而提高了焊接电源信号的精确控制。最后,对文中所提方法进行了仿真验证,结果表明所设计控制方法具有响应速度快、超调量小、收敛速度快、稳态精度高等优点。
Welding power output is a multi variable,nonlinear,time-varying parameters of the controlled object,the traditional PID control parameters are unvaried,and difficult to be adaptive adjusted according to the actual situation of parameters,the effect of the welding current output is poor. In view of the above situation,an intelligent control method of the welding current were proposed based on PID RBF( Radial Basis Function) neural network. A three-layer RBF neural network model was founded to obtain the gradient information by on-line identification,and finally PID three parameters were adjusted on-line according to the gradient information,thus the control precision of the welding power supply signal was improved. The simulation results show that the proposed control method has fast response,small overshoot,fast convergence and steady precision.
出处
《焊接》
北大核心
2016年第10期55-58,72,共4页
Welding & Joining
关键词
焊接电源
传统PID控制
RBF神经网络
仿真
welding power source
traditional PID control
RBF neural network
simulation
