神经网络自适应控制技术在空调系统中的应用  

The Use of Neural Networks Adaptive Control in Air-conditioning System

作　　者：李界家[1] 盛庆飞[1] 马斌[1] 阚洪亮[1] 

机构地区：[1]沈阳建筑大学信息与控制工程学院,辽宁沈阳110168

出　　处：《沈阳建筑大学学报（自然科学版）》2007年第3期522-524,共3页Journal of Shenyang Jianzhu University：Natural Science

基　　金：建设部基金项目(03-2-117)

摘　　要：目的采用神经网络自适应控制方法,来解决中央空调系统大惯性、大时滞问题,提高控制效果.方法利用两个BP神经网络NNC和NNI,分别做控制器和辨识器,神经网络辨识器通过采集相关的输出量,监控系统的工作状态,并通过实时调整神经网络控制器参数.结果实现中央空调智能控制,仿真结果验证有良好的动态性能和稳态性能.结论神经网络自适应控制器优于常规控制器.调节速度快,超调小,具有良好的控制性能.In this paper, neural networks adaptive control is used to solve the problem of great inertia and great delay in air-conditioning system and to improve the Control effect. An algorithm is proposed based on neural networks adaptive control. Two BP neural networks, NNC and NNI, working respectively as controller and identification, can solve the problem of the temperature control in the terminal installment of variable amount of wind air-conditioning system by means of mutually erroneous transmission, gathering correlative output, monitoring the work state of the system and regulating the real-time parameter of neural networks controller. Neural networks identification can make the air conditioning working state good, realizing intelligent control in air-conditioning system. The simulative result indicates that the system has good dynamic performance and stable performance. We can conclude that neural network adaptive control is better than conventional controller, with the characteristics of quick adjustable speed, small overshoot and good control performance.

关 键 词：自适应 神经网络 温度控制 控制器 

分 类 号：TP273[自动化与计算机技术—检测技术与自动化装置]