【作者】 宋丽芳；

【导师】 丁宝；

【作者基本信息】 哈尔滨工业大学 ， 电气工程， 2010， 硕士

【摘要】 随着通讯事业的飞速发展,越来越多的通讯机房投入到了建设和使用中。通讯设备不同于一般的工业设备,它需要一个苛刻的工作环境,对所处环境的温度、湿度、洁净度有着明确的要求。通讯机房内的空调机组设备就是用于维护这样一个环境的必要设备。有数据显示一个通讯基站的主要电能消耗在通讯机房内置空调上,提高这些空调机组的效率是降低通讯机房电能消耗量的主要途径。同时,由于近年来国家有关节能减排政策的提出,以及公众环保意识的加强,大能耗设备的节能方面倍受关注。因此,开展对通讯机房专用空调的节能控制研究,对工业产业、机械产业、通讯产业等涉及到节能控制思想的领域有重要的理论意义和实际价值。本文对应用在通讯机房的空调领域变频节能控制策略进行分析,并将模糊神经网络的控制应用到空调变频节能技术中。研究了基于模糊神经网络控制算法的具体实施过程,包括模糊化方法和神经网络作用方式。借助Matlab/Simulink的控制系统仿真功能,建立基于热力学理想房间模型的模糊神经网络仿真模型,对所设计的空调控制器算法的响应特性,抗干扰性,鲁棒性进行仿真研究,同时与传统的控制方式相比较,在理论上验证基于模糊神经网络的空调控制器的优越性。对通讯机房专用空调控制器进行变频节能实验研究,设计了基于Microchip公司的dsPIC30F4013为核心的硬件电路,包括微控制器dsPIC及其外围电路、温度检测和转换电路、隔离输出电路、操作和显示电路以及电源电路。采用本文控制策略,实现了软件程序设计、包括主程序、中断子程序、A/D采样和温度测量子程序、温度求解子程序、系统控制子程序、除湿子程序的设计。完成了基于模糊神经网络控制算法的通讯机房专用空调控制器的样机制造。借助于天津移动的通讯机组设备,进行长达一个月的实地测试,将样机投入到实际的应用中。实际运行表明,将模糊神经网络控制策略用于通讯机房专用空调的变频节能系统,该控制策略是合理与有效的。更多还原

【Abstract】 With the rapid development of communication industry, more and more communication rooms are put into construction and using process. Communication equipment is unique with other industry equipments, while it needs a severe working environment. It has clear requirements on the working temperature, humidity and cleanness degree. Air conditions are the equipment for serving this kind of environment. Data shows that the main electricity consumption was introduced by the air condition. To improve the efficiency of air condition is a way to cut the electricity consumption of the communication room. At the mean time, with the governmental policies on saving energy and reducing emissions and public awareness on environmental protection, more and more focuses on energy saving for huge equipments come up. Research on the energy saving control for communication room will contribute production industry, mechanical industry and communication industry for both theoretical and practical values.This paper analyzed the existing control theory on air condition area. Put fuzzy neural network control thought on air condition energy saving area. This paper gave a practical application method for fuzzy neural network algorithm, including fuzzy algorithm and neural network algorithm.This paper used the simulation function for control system provided by Matlab/Simulink to construct a fuzzy neural network simulation model based on thermo dynamical ideal model, and then tested the response characteristics, interference characteristic and robustness, and at last compared with the traditional control methods. The conclusion was that the control method basing on fuzzy neural network was better than tradition control method. This paper realized the equipment for communication room special used air condition, designed the hardware circuit based on Microchip Company’s dsPIC30F4013, including dsPIC and its peripheral expansion circuit, temperature testing and transferring circuit, isolation and output circuit, display and operation circuit and power supply circuit, designed software program, including main program, interrupt sub program, A/D sampling sub program, temperature testing sub program, temperature changing rate calculating sub program, system control selection sub program, dehumidification sub program. At last, a sample equipment for communication room’s air condition basing on fuzzy neural network control algorithm was produced.This paper did a practical test in the communication room of Tianjing Mobile, put the sample machine into real life, and then approved the energy saving advantage in practical situation after the theoretical discussion.This paper analyzed and designed a fuzzy neural network algorithm for the energy saving control equipment for communication room special used air condition. This paper approved the energy saving advantage from both theoretical and practical aspects.更多还原