人工气候室照度神经网络控制系统的研究

人工气候室光照系统具有复杂的非线性，不确定性，滞后性，常规控制难以达到较好的效果。针对这一问题，采用神经网络的PID控制，可实现对系统的在线辨识和控制；为克服局部最小问题。采用遗传算法时神经网络权阈值进行优化。仿真结果表明，该系统控制效果好，稳态精度高，响应较快，动态性能好，超调量较小，完全满足控制要求。

Neural Network Control System of Luminance in the Artificial Climate Room Research

illumination system in the artificial climate room has the complex non-linearity , the uncertainty and the lag, so that the conventional control with difficulty can receive the better control effect. Adopting the neural network PID illumination control can recognize the system on- line and can control system. In order to overcome partially the smallest question, adopting the heredity algorithm optimizes the power and threshold value of the nerve network. The simulation result indicates, this control effect of the system is good, the response is quicker, the dynamic performance is good, the over modulation is smaller, the precision of stable state is high. It completely satisfies the control request.