The Research On Energy-saving Of The Central Air Conditioning System Based On The Hierarchical Steady-state Optimization

When the air conditioning system is designed,it can satisfy the highest need of cooling load and heating load.However,the air conditioning system often runs with the part load and it must be adjusted consistently according to the change of weather condition outdoors and the heat and humidity indoors.Therefore,the global optimal control can make the air conditioning system run steadily to meet the environment change indoors and outdoors and make it satisfy the thermal comfort requirements indoors at the minimum cost of energy.The energy-saving of central air conditioning with the global optimization method involves in the detail compose,the energy consumption model of equipment,control system and optimal methods.This study focuses on the on energy-saving of the central air conditioning system based on the hierarchical steady-state optimization.Firstly,the experimental system of central air conditioning is introduced,which includes the whole hardware and software platform of the VAV central air conditioning and the function of each subsystem.At the same time,the energy models of all energy consumption equipments are built based on the measured data with fitting.In order to understand the characters of energy consumption equipments and be helpful for the energy-saving control,the adjusting equipments are analyzed based on the measured data.Secondly,in order to ensure that the central air conditioning has the good dynamic performance before it runs steadily.The math models of each subsystem are built by system identification.The control methods are designed according to the characters of different subsystem.The control methods consists of PID,single PID and MFAC(model free adaptive control).The MFAC is applied firstly in the subsystem with large time delay and the decoupling between different subsystems in central air conditioning.The effect of MFAC to the system with uncertainty is analyzed.The optimal parameters of MFAC are found out based on the simplex on the premise of the absence of good methods that can compute the control parameters.Thirdly,in order to enhance the load forecasting accuracy of central air conditioning,an air-conditioning load forecasting based on dynamical combined residual error correction is posed based on the combination forecasting aiming at multiple single forecasting methods with fixed number.Fourthly,according to the characters of central air-conditioning and large scale system theory,the decomposition method based on the system construction is choosed to decompose the central air-conditioning,and the IBMF(Interaction Balance Method with Feedbacak)choosed as the coordination method.The decentralized identification method is choosed to build the steady-state models of subsystems of the central air-conditioning,and the steady-state models got.Finally,the steady-state models of subsystem are built with decentralized identification method,and the global steady-state optimization model of central air conditioning is built.In order to solve the problem of dimension calamity in the process of optimization of central air conditioning,the IBMF is used to optimize the steady-state with hierarchical method aiming at the summer condition.The result shows that the global scheme can solve the problems of global control and optimization,which is valuable for application.