Research On Control And Energy-Saving Optimization Technology Of Large-Scale Building Cold Source System

With the rapid development of China’s economy,the number of large buildings is increasing,and energy consumption remains high.The central air conditioning cold source system accounts for more than 50% of the building energy consumption.At the beginning of the design,the rated cooling capacity of the cold source system generally has a surplus of more than 20%,which causes it to be under partial load for a long time.Considering the stability and convenience of the system control,the frequency conversion function of the device is not enabled,the system is in the maximum refrigeration output state as a whole,and there is a large space for energy saving.In the existing energy-saving strategies,most of them set up cold source system models to optimize the operating parameters of the system and match the cold demand in real time.However,the modeling of related equipment often does not fully reflect the field operation conditions,and the calculation accuracy of the cold demand is not high.At the same time,the optimization is limited to a single device or a local system.In view of the above problems,this paper designs an energy-saving optimization platform architecture based on the energy-saving field of large-scale cold source system as a specific application scenario,and conducts research on cooling load calibration and prediction,system modeling,optimization algorithm,etc.to achieve dynamic matching between cold demand side and cold supply side.The main research contents are as follows:(1)A cooling load calibration and prediction scheme based on multi-feature fusion data calibration and long-term memory network(LSTM)is designed to solve the problem of large error and large lag in cooling load measurement on the cold demand side.A characteristic engineering study was carried out on the traditional cold conformance measurements,and the calibration of cooling load is realized by using multiple feature fusion,which solved the problem that isolated points and measurement noise affect the accuracy of the original data.The main influencing factors of building cooling load are analyzed theoretically,and an LSTM model suitable for predicting cooling load is established,which overcomes the common problem that the lag of cooling load measurement results in the deviation of control effect.(2)In order to solve the problem that it’s difficult to predict the working condition and energy consumption of the cold source system,this paper establishes a simulation model of the cold source system which combines the mechanism characteristics with data-driven.The mechanism characteristics of the cold source equipment are studied in detail,and used as the overall framework of the model.The data-driven method is used to modify the parts of the device characteristic curve parameters that are offset with time to achieve the self-adaptive modification of the device operating curve.At the same time,the accuracy of the model is verified by comparing the energy consumption of the cold source,the working condition data and the simulation data.(3)In order to ensure the long-term stable operation of the cold source system at high energy efficiency,a hybrid intelligent optimization algorithm(PSOCC-NM)combining the particle swarm optimization algorithm with shrinkage factor and the downhill simplex method is presented,which achieves fast convergence and good optimization results,and solves the problem that the complex coupling relationship of control variables in the cold source system affects the optimization efficiency.In this paper,the dynamic matching problem between the cold supply side and the cold demand side is transformed into the system energy consumption minimization problem,the energy consumption coupling relationship of the whole system is analyzed to determine the optimization variables,and the global optimal control strategy for the cold source system based on PSOCC-NM is designed.In this paper,the above energy-saving optimization platform is deployed and put into use in a large office building in Hangzhou,Zhejiang Province,to verify that the architecture can fully meet the design requirements.Among them,the accuracy of cooling load prediction model is more than 98%.The accuracy of energy consumption of cold source simulation model is more than 95%,and the variance coefficient of mean square deviation of important system parameters is less than 15%.Data from the energy-saving optimization platform deployed on September 8,2022,show that the cold source system saves 14% energy,and the system energy efficiency ratio(EER)is4.48,which increases by more than 20% compared with the platform deployed before.