基于神经网络的分数阶PID控制器在磨矿分级系统中的应用研究

磨矿分级过程机理复杂,存在惯性大、滞后时间长、参数时变、非线性等影响因素,难以建立精确实用的数学模型,随之带来的控制效果也不理想,针对此提出基于BP神经网络的分数阶PID控制的系统模型。仿真结果表明,该控制模型具有收敛速度快、无超调量、鲁棒性强等特点,并能提高磨机的效率和选矿厂的经济效益。

Based on Neural Network Fractional Order PID Controller and Application in Milling-Classification System

Milling-classification operation system is complexed,there is inertia,long lag time,timevara- tion parameters,nonlinear factors and so on,it is difficult to establish accurate mathematical modelling, the control effect is not satisfactory,based on BP neural network fractional order PID control of system model is present.The simulation results show that the control model has fast convergence,no overshoot, robustness,and can improve the efficiency and economic benefits.