Optimal Chiller For Operation Using A Teaching And Learning Optimization Algorithm

Air conditioning equipment is widely applicated with people’s life significantly improved.There is no denying that it provides people with more comfortable living conditions.However,the proportion of energy it consumes in the building increase sharply,which has accounted for40%～50%.Therefore,people attached great importance to energy saving in its operation,and various energy saving methods and technologies are also being studied and popularized.The paper aims to study the operation optimization algorithm of chiller running in parallel under partial load based on the goal of energy saving.The main research contents are as follows:(1)Selecting energy consumption model and optimization algorithm.The paper firstly elaborates the solution goal and steps of the optimization problem,then discuss the working principle of chiller and what determines operation energy consumption of chiller.Based on the existing research of chiller energy consumption model,confirming the model the paper used in the paper.What is more,we make analysis and performance testing for four common algorithms of solving optimization problems.Finally,teaching and learning based optimization algorithms(TLBO)are determined as the algorithm of this paper for further research.(2)Research on teaching and learning optimization algorithm.Firstly,on the premise of fully understanding the teaching and learning optimization algorithm,the origin deviation problem is studied,the reasons of the problem are explained and verification plan is designed to verify,using adaptive reference instead of origin reference as a solution.Secondly,Further research on the improvement of teaching and learning optimization algorithms,and constructing the teaching and learning optimization algorithm based on learning initiative(LITLBO),which improve convergence accuracy and convergence speed of the algorithm.Finally,in order to compare with the multi-class interactive(MCI)and random cross-self-learning strategy(CS)teaching and learning optimization algorithms,using six standard test functions to test their performance,and the results show that the teaching and learning optimization algorithm based on learning initiative has better performance.(3)Establishment and verification of building cooling load model.The research is the modeling process of building cooling load based on TRNSYS software,analyze the difference in the main parameters of buildings in the north and south.Then,select Jilin,Hangzhou and Guangzhou as representative cities.Obtaining the temperature data of the cities for ten years by Meteonorm software.The annual variation models of building hour cooling load in the abovementioned cities were established and compared with existing data to verify the correctness of the model parameter selection in this paper.It provides support for the subsequent research.(4)Example verification of LITLBO algorithm.After reading relevant literature,establish three cases based on real situation,respectively establish the hourly cooling load annual change model and operational energy consumption model of parallel water chiller.Using LITLBO algorithm compared with TLBO algorithm,particle swarm optimization(PSO),genetic algorithm(GA)and simulated annealing algorithm(SA).The results show that for the parallel operation of 3units,4 units,and 6 units,the LITLBO algorithm saves up to 3.8% energy,2.6% energy,and 6.9%energy respectively compared with other algorithms.This verifies that the LITLBO algorithm can achieve better effect for multi-unit parallel operation.