某于功率反馈的涡轴发动机神经网络PID控制研究

通过对功率平衡关系进行分析，提出了利用功率反馈设计智能神经网络PIO控制器的方法。基于BP(Back Propagation)神经网络，将功率信号作为神经网络的输入层信号，并改进了网络权值的学习规则。通过在线整定PID参数，控制器能够根据功率误差信号的变化实时调整控制参数，从而使系统自主寻找到功率平衡点，具有良好的稳态和动态响应特性。仿真结果表明：该方法可以使涡轴发动机在全包线范围内具有理想的控制性能。

Neural Networks PID Control Study of Turboshaft Engine Based on Power Feedback

An intelligent neural networks PID controller based on power feedback was proposed by analyzing power balance. Based on BP neural network, the power signal was taken as the input layer signal, and the network weight learning rules was improved. According to on-line setting of PID parameters, thecontroller parameters was adjusted by the changes of power error signal, to find the power balance point independently and obatain excellent steady and dynamic response characteristics of system. The simulation results show that the turboshaft engine has satisfied control performance over the whole envelope.