Optimal control strategies for multi-chiller system based on probability density distribution of cooling load ratio

Abstract

Probability density distribution profile of cooling load ratio is obtained by Parzen window density estimation method, which is used to describe the building cooling load variation. Based on the analysis of this profile, the optimal sequence control strategies (AI, AII and AIII) are presented. In order to optimize the COP of total chillers, the optimal load allocation control strategy B is presented. The optimal load allocation control strategy BI is presented to improve the strategy B for simplifying the process of the optimal calculation and maintaining stability of the system operating. Simulation results show that these optimal control strategies presented can save more power consumption of chillers operating than the original sequence control strategy. Comparing the strategy B, the strategy BI can avoid unnecessary switch on/off actions of chillers and decrease computational costs.