感知器修正的大林算法在配煤系统中应用

Application of Dahlin Algorithm Modified by Perceptron in Coal Blending System

针对配煤控制系统大滞后、非线性、时变性的特点,设计了一种基于感知器修正的大林算法控制器,采用神经网络感知器对大林算法的时间常数进行调节,通过控制器输出与期望出煤量的偏差值反向传播再度修正算法时间常数,并采用模糊控制对感知器学习率进行调整,将该控制器运用在成庄选煤厂配煤环节。结果表明:系统最大超调量从原有70%降低至4%,稳态误差从原有的9%降低至3%,有效解决了配煤系统中模型参数不稳定带来的曲线波动问题。

Aiming at the characteristics of large delay,non-linearity and time-varying of coal blending system,a Dahlin algorithm controller based on perceptron correction is designed.Adjusting the time constant of the Dahlin algorithm using a neural network perceptron,the time constant is again corrected by the deviation of the controller output value and the expected coal output using the backward error propagation algorithm,and use fuzzy control to adjust the learning rate of the neural network.The results show that the maximum overshoot of the system is reduced from 70%to 4%,the steady-state error is reduced from the original 9%to 3%.Effectively solved the problem of curve fluctuation caused by unstable model parameters in coal blending system.