神经元网络控制器在热网中的应用研究

Study and Application of Neural Network Controller in the Heat Supply Network

城市供热管网运行质调节和量调节之间的耦合制约了质量并调的控制效果。针对这个问题,提出采用PID神经元网络解耦控制器对质量并调供热系统进行解耦控制。深入分析了PID神经元网络的优缺点和适用范围,采用PSO学习算法优化PID神经元网络初始权值;并对PSO优化后的PID神经元网络控制器进行了仿真验证。分析表明,PSO优化的PID神经元网络控制器可在供热管网质量并调控制中实现解耦。

In urban heat supply network operation, the coupling between qualitative （ temperature ） regulation and quantitative （ flow ） regulation constrains the control effect for collaborative quality and quantity control. Aiming at this problem, the decoupling control based on PID neural network de-coupler for such control system is proposed. Depth analysis on the advantages, disadvantages, and scope of application for PID neural network is conducted;the initial weights of PID neural network are optimized by adopting PSO learning algorithm;and the PID neural network controller optimized by PSO is simulated and verified. The analysis indicates that the PID neural network controller optimized by PSO can be used in heat supply network for decoupling and implement collaborative control.