摘要
近几年来,随着国家经济的不断发展,国家电力的负担越来越大。随着国家倡导建设节约型社会,蓄冷空调在我国也得到了越来越广泛的应用。
如何有效的利用蓄冷空调,确定第二天所需要的蓄冰量呢?这时,人们就非常希望能够预测第二天蓄冷空调的负荷。如果能够精确的预测建筑物的动态冷负荷,不仅可以更准确的帮助设计人员完成设计,更可以使设备控制人员根据预测的冷负荷确定系统的控制时间、方式以及尽早发现系统故障。这一点在蓄冷空调系统中尤为重要,因为每天晚上蓄冰量的多少将直接影响第二天整个系统的工作效率,并且是提高蓄冷空调系统经济效益的重要途径之一。所以,这里要解决的主要问题就是进行“精确预测”。
该预测系统采用的是人工神经网络(ANN)的BP算法。主要实现两大块功能。一是对未来一天外界环境参数及空调负荷的预测;二是根据当前时刻及外界参数计算负荷。该预测系统主要以BP算法为核心,通过利用BP算法对以往的历史数据的不断学习训练,而不断去修改算法中的中间值(达到预定的精度范围),以达到精确预测和计算的目的。从而解决“精确预测”的问题来提高蓄冷空调的经济效益。
关键字:蓄冷空调;人工神经网络;BP算法;预测
Abstract
In recent years, The national power more and more burdener with the constant development of the national economy. The Holding Cold Air-Conditioning have increasingly broad applications in our country by the advocacy of national saving society.
How to effective use of The Holding Cold Air-Conditioning, to ensure how much ice is store on the next day? At that time, it was very much like to be able to predict the next day the load of Holding Cold Air-Conditioning. If we can accurately predict the dynamics of the cold load building, not only can help the designers more accurate completion of the design, but also to equipment control personnel determined on the basis of the projected cold load time control systems, methods and early detection system failures. This is particularly important in The Holding Cold Air-Conditioning system, because every night how much ice is store will directly affect the efficiency of the entire system on the second day, and air-conditioning systems is one of the important way