制冷剂自然循环能量回收装置的智能仿真  

Intelligent simulation on refrigerant natural circulation energy recovery installation

作　　者：孙丽颖[1] 李铁磊 

机构地区：[1]哈尔滨工程大学航天与建筑工程学院,黑龙江哈尔滨150001

出　　处：《哈尔滨工程大学学报》2013年第8期984-989,共6页Journal of Harbin Engineering University

基　　金：黑龙江省自然科学基金面上资助项目(E201139);中央高校基本科研业务费专项资金资助项目(HEUCF130210)

摘　　要：为分析制冷剂自然循环能量回收装置的运行特性,建立了装置中冷凝器、蒸发器等部件的数学模型.为了提高仿真精度,将神经网络模型中的多层前传网模型引入到部件模型中,用神经网络产生的修正系数进一步校正部件的分布参数模型.在部件模型的基础上建立装置的系统仿真模型,通过对模型的求解,分析室内外温差等因素对能量回收装置运行特性的影响规律.将理论计算结果与实验数据进行对比,验证了能量回收装置仿真数学模型的可靠性,同时说明人工神经网络模型与分布参数模型的结合为提高装置性能预测的精度提供了有效途径.In order to analyze the operating characteristics of energy recovery installation in refrigerant natural circu-lation , mathematical models are built for the device components such as condensers and evaporators. To improve the simulation accuracy, multilayer forward propagation network model in the neural network is applied to the device mathematical model, and the distributed parameter model of components is further corrected by correction factors got from artificial neural networks. The working rules of influencing factors such as the temperature difference be-tween indoor and outdoor air on the energy recovery installation's operation characteristics are analyzed through in-stallation system simulation model, which is set up on the basis of component models. Comparing the theoretical calculation results with experimental data, it verified the reliability of simulation mathematical models of the energy recovery installation. Meanwhile, it shows that combination of the artificial neural network model with the distribu-ted parameter model provides an effective way to improve the accuracy of the installation performance prediction.

关 键 词：制冷剂 自然循环 能量回收 神经网络 智能仿真 

分 类 号：TU831.36[建筑科学—供热、供燃气、通风及空调工程]