基于变流量解耦的制冷机房节能优化研究  

Research on Energy Saving Optimization of Chilled Plant Based on Variable Flow Decoupling

作　　者：许成炎 闫秀英[1] XU Chengyan;YAN Xiuying(School of Building Services Sciences and Engineering,Xi'an University of Architecture and Technology,Xi'an 710055,Shaanxi,China)

机构地区：[1]西安建筑科技大学建筑设备科学与工程学院,陕西西安710055

出　　处：《制冷技术》2022年第4期47-53,共7页Chinese Journal of Refrigeration Technology

基　　金：陕西省低能耗建筑节能创新示范项目(No.2017ZDXM-GY-025)。

摘　　要：针对制冷机房节能降耗问题,本文运用广义回归神经网络(General Regression Neural Network,GRNN)对冷却水进水温度进行建模,从而对冷水机组的多元多项式(Multivariate Polynomial,MP)模型进行改进。结合工程案例,将变流量解耦控制策略植入Trnsys仿真平台,确定冷负荷工况下,在该平台上进行四因素二次回归正交试验,得到冷冻水流量、冷却水流量、冷冻水供水温度和冷却水进水温度的参数优化值。结果表明,采用正交回归试验法对系统参数进行优化后的制冷机房总功率仿真结果较实际运行能耗下降了26.44 kW,其中水泵能耗增加3.44 kW,冷水机组能耗降低29.88 kW,节能率达到了12.23%。Aiming at the problem of energy saving and consumption reduction in the chilled plant,use general regression neural network(GRNN)to model the cooling water inlet temperature,thereby improving the multivariate molynomial(MP)model of the chiller.Combined with the engineering case,the variable flow decoupling control strategy was implanted into the Trnsys simulation platform.Under the cooling load condition,the four-factor quadratic regression orthogonal test was performed on the platform to obtain the parameter optimization values of chilled water flow,cooling water flow,chilled water supply temperature and cooling water inlet temperature.The results show that the total power simulation result of the chilled plant after optimizing the system parameters by the orthogonal regression test method is 26.44 kW lower than the actual operating energy consumption.Among them,the energy consumption of water pump increased by 3.44 k W,and the energy consumption of chiller decreased by 29.88 kW,the energy saving rate reaches 12.23%.

关 键 词：制冷机房 广义回归神经网络 TRNSYS 变流量解耦 能耗 

分 类 号：TB657[一般工业技术—制冷工程] TP391.9[自动化与计算机技术—计算机应用技术]