Energy Efficiency Monitoring And Diagnosis Of Central Air Conditioning Chilled Water System In Large Public Buildings

The energy consumption of air-conditioning systems in large-scale public buildings accounts for a huge proportion of energy consumption.Therefore,studying the operational energy efficiency of air-conditioning systems in combination with the needs of actual projects plays an important role in reducing the energy consumption of public buildings.The operating energy efficiency ratio is not only closely related to the control mode of air conditioning system,but also affected by various operating parameters.Aiming at the characteristics of high data dimension and large data volume of central air-conditioning unit equipment,a method for predicting the energy efficiency ratio of central air-conditioning cold source system operation based on the extreme learning machine（ELM）artificial intelligence algorithm is proposed.On the premise of accurate prediction of the energy efficiency ratio of the central air-conditioning cold source system operation,would be better realize the optimization research on the operation control of the central air-conditioning system.According to the operating characteristics of large public buildings,the concept of energy efficiency ratio of cold source system operation is proposed.Based to the composition of the central air-conditioning system of a large public building in Liaoning,the historical operating parameter data of the equipment operation monitoring system is collected to obtain the operating energy efficiency ratio of the cold source system,mathematical models calculate the energy efficiency ratio of each period in the historical operation process,establish the historical operation database of the central air-conditioning system.Selecting 12 kinds of central air conditioning cold source systems were selected as input（host current,refrigeration pump current,cooling pump current,chilled water outlet temperature,chilled water return temperature,chilled water outlet temperature,chilled water return temperature,chilled water outlet Pressure,chilled water return pressure,cooling water outlet pressure,cooling water return pressure,make-up pressure）from the historical operation database of central air conditioning system,the energy efficiency ratio of the cold source system is used as the output,and the data used for building model training and verification is obtained through data preprocessing and data cleaning set.A mathematical model of energy efficiency ratio prediction based on the BP neural network algorithm and the ELM algorithm is established,and the two prediction models are machine-learned and debugged through the organized training data set,and a BP-based algorithm is constructed.Neural network and ELM energy efficiency ratio prediction is modeled.By predicting the test data set and comparing the predicted results with the real results,the error values of different prediction models are obtained.The BP neural network model and different ELM model predictions are compared,and the final comparison result is that the excitation function adopts the sine function,and the number of hidden layer nodes is set to 50with best performance.Use the established ELM prediction model to predict the real-time operating energy efficiency ratio of cold source system,and compare the historical operating data of the central air-conditioning system according to the performance coefficient of cold source system under the standard condition of partial load operation state.The COP_smin of central air-conditioning cold source system,when the real-time predicted energy efficiency ratio is lower than the alarm value,the system was diagnosed as running inefficiently.At the same time,the predicted values of COP of chillers,ECR-a of refrigerating water pump and ECR-a of cooling water pump are compared with their limits to infer the areas where problems may occur analyzing the possible causes of data abnormalities in time and give suggested measures.