论文部分内容阅读
针对火电厂锅炉水质调节过程的大时滞时变特性,常规控制算法控制效果不好的问题,本文提出了基于BP神经网络的Smith-PID鲁棒自适应控制算法,利用BP神经网络的任意非线性表达能力和很强的自学习能力,在线自学习整定PID参数,被控对象不需要精确辩识,控制器参数跟踪被控对象自适应调整,克服了常规PID算法不适用于大时滞过程控制和常规Smith预估补偿控制对模型不确定性敏感的缺陷。MATLAB仿真表明,本文控制算法的静态特性、动态品质良好,鲁棒性强。
In view of the large time-delay characteristics of boiler water quality regulation in thermal power plants and the poor control effect of conventional control algorithms, this paper proposes a Smith-PID robust adaptive control algorithm based on BP neural network. Linear expression ability and strong self-learning ability, online self-learning tuning PID parameters, the controlled object does not need to be precisely identified, the controller parameters follow the controlled object adaptive adjustment, to overcome the conventional PID algorithm is not suitable for large delay process Control and Conventional Smith Predictor Compensation Control Defects Sensitive to Model Uncertainty. MATLAB simulation shows that the static characteristics of this control algorithm, dynamic quality is good and robust.