基于Trnsys平台的中央空调迭代学习控制研究

Research on Iterative Learning Control of Central Air Conditioning Based on Trnsys Platform

中央空调在生活中应用越来越广泛,但其能耗占据大型建筑总能耗的50%以上,同时全球能源危机向各国敲响了警钟,因此中央空调系统的节能改造刻不容缓;中央空调系统复杂,为了建立精确的模型以研究控制算法应用于中央空调末端风机控制的效果,采用了Trnsys(瞬时系统模拟程序)仿真平台进行建模,通过Trnsys平台搭建冷负荷建筑和中央空调系统模型;为克服Trnsys软件控制模块少、简单、不便修改等缺点,采用了Trnsys与Matlab(矩阵实验室)联合仿真系统,通过Matlab编程实现各种控制算法;基于中央空调周期性工作的特性,提出将迭代学习控制应用于其末端控制;迭代学习控制算法计算简单、适应性强,易于实现,对于重复动作的系统有很好的控制效果;对迭代学习控制的学习率、稳定性、收敛性等进行研究,并通过仿真检验其控制效果;在迭代学习控制作用下,经过一定次数的学习,被控变量能实现对目标曲线的无差跟踪,达到比较理想的控制效果。

Central air conditioning is more and more applied in our lives.The energy consumption of central air conditioning accounts for more than 50%as which of large buildings.The global energy crisis sets off the alarm bells,thus energy conservation of central air conditioning system should be implemented without delay.The system of central air conditioning is complex.In order to provide an accurate model to study the control effect of the iterative learning control algorithm on the terminals of central air conditioner,it builds the model with the Trnsys platform.In order to overcome the weakness of the Trnsys control modules and realize various control algorithms,the joint simulation of Trnsys and Matlab is adopted.The iterative learning control can be applied to central air conditioning system that works periodically.The iterative learning control algorithm is simple,adaptable,easy to implement and takes good effect on repetitive working systems as well.The characteristics of iterative learning control such as rules,stability and convergence are studied and its performance is shown through simulation.It’s shown that the controlled variable can track without steady-state error after learning for several times.