| Thanks to the low-cost, renewable and clean shallow geothermal resources, air conditioners with ground source pumps are developing rapidly with the advantage of high energy utilization rate. However, there is still some room for improvement from the perspective of saving energy in the operation process of most air conditioners with ground source heat pumps. The objective of this paper is to realize the optimal control over air conditioners with ground source heat pumps and reduce energy consumption while ensuring the comfort of people so as to improve energy efficiency.In this paper, a TRNSYS-based simulation system of air conditioners with ground source heat pumps is established on the basis of the air conditioning system of an office building. The simulation system consists of the energy consumption model of the office building and the input-and-output models of heat pump units, water-circulating pumps, heat-exchange coils, fans and VAV terminals. The working state of the fans and the VAV terminals are controlled by PID controllers.The input and output parameters of the energy consumption prediction model and the temperature prediction model of the simulation system are determined by analyzing the operation data of the simulation system, the energy consumption mechanism model of the heat pump units and the measurable and controllable parameters of the control system of air conditioners. The input parameters comprise the measurable and controllable parameters of the system. The energy consumption prediction model and the temperature prediction model of the simulation system are established through collecting the opeation data of the simulation system working in different default parameters and using a multilayer perceptron neural network. Based on the established prediction models, an optimal control model is established with the ultimate optimal object being the sum of temperature control errors and the total energy consumption of the system. The control model is validated by the simulation system’s operation data which are processed by the PSO optimization algorithm. The on-line intelligent optimization control over the simulation system is realized through the joint simulation of MATLAB and TRNSYS. Compared with the operation results with constant flows and constant pressure in the same temperature, the simulation system can save energy by nearly 11.12%.In order to realize the optimization control over energy consumption of the air conditioning system of the office buliding, an energy consumption optimization control system for air conditioners with ground source heat pumps which can realize the above-mentioned prediction models and optimization control models is designed while taking the current situation of the control system of the air conditioners of the office building into consideration. |