改进的PID神经网络在黑液液位控制中的应用

Study on the Control of Black Liquor Level Based on Improved PID Neural Network

针对黑液液位控制系统的非线性、大惯性及时变性的特点,对PID神经网络(PIDNN)在黑液液位控制中的可行性进行了深入研究与探讨。通过Matlab仿真,比较了常规PID算法、基于BP的神经网络算法和PIDNN算法对黑液液位的控制效果,验证了PIDNN算法的良好自适应性和鲁棒性。

Because the level control system of black liquor has characteristics of nonlinear, large delay and time-varying, this paper discussed the feasibility of using PID neural network （PIDNN） in black liquor level control. The simulation results in Matlab environment showed that the PIDNN had the good adaptability and strong robustness compared with conventional PID and PID control based on BP neural network.