基于神经网络PID电子负载控制系统设计

Design of Electronic Load Control System Based on Neural Network PID

以电子负载为研究对象,针对其控制系统复杂的非线性和时变性,常规的PID控制方法存在着精度低,自适应能力不强等缺点,以定点32位DSPTMS320F2812作为控制芯片,结合神经网络和常规PID设计了一种神经网络PID控制器。仿真结果表明,通过神经网络PID控制器,能够在线调整PID控制参数,使系统具有更小的超调量和更短的调节时间,而且系统的精度和稳定性也得到大大的提高。神经网络PID控制器是一种动态特性和静态特性良好的控制器,在电子负载控制实验中可以获得满意的控制效果。

Electronic load as the research object,in view of its control system of complex non-linear and dynamicity,conventional PID control methods exist low accuracy,adaptive capability is not strong and other shortcomings.To fixed-point 32-bit DSPTMS320F2812 as control chip,combined with neural network and the conventional PID designed a kind of neural network PID controller.The result of simulation indicates that neural network PID controller can adjust PID controlling parameters on line,then the system could have smaller overshoot and shorter settling time,and the precision and stability of the system has also been greatly enhanced.Such controller is a controller with good dynamic characteristics and static characteristics,a satisfactory control effect is obtained in the experiment of the electronic load system.