变风量空调末端室温滞后Elman网络预测控制方法
Predication control for temperature lag using Elman neural network in variable air volume air conditioning system
摘要：
对比分析了压力相关型和压力无关型变风量空调末端室温控制原理，阐述了压力相关型末端室温预测控制的研究意义。描述了时滞系统周期性预测控制基本原理，提出了基于Elman网络多步预测模型的室温预测控制方法。试验研究结果显示：在满足末端负荷需求的前提下，提出的基于压力相关型末端的室温预测控制方法改变了变风量末端的调节方式，有利于提高变风量空调系统室温控制回路的稳定性。
Abstract：
Compares and analyses the control principles of pressure-dependent and pressure-independent variable air volume air conditioning terminals, and elaborates the research significance of room temperature prediction control based on the pressure-dependent terminal. Describes the basic principle of periodicity predication control for time-delay system, and presents a prediction control method for room temperature based on Elman neural network multiple-step prediction model. The experimental results indicate that with the satisfied terminal load demand, the proposed prediction control method based on the pressure-dependent terminal could change the control mode of the VAV terminal unit, which is beneficial to improving the stability of room temperature control loop of the VAV air conditioning system.
Keywords:variable air volume air conditioning, pressure-dependent terminal, room temperature prediction control, time-delay system, Elman neural network