Research On Energy Saving Of Hospital Air-Conditioning System Based On System Energy Consumption Forecasting And Operation Optimization

The energy consumption of central air conditioning system is the main part of building energy consumption.It is of great significance to excavate its energy saving potential for the central air conditioning system of large hospitals with ’ full space and full time ’ operation mode.Optimizing the operation strategy based on data mining to realize the energy-saving operation of central air conditioning system is an important research direction of building energy conservation.Although there are many researches on energy saving of hospital air conditioning system at home and abroad,it is often not considered that the secondary pump variable flow system has the characteristics of sub-regional cooling,but directly predicts the cooling capacity of the whole system,and then optimizes the operation of cooling water system or chilled water system alone.This paper proposes an energy-saving method for the hospital air-conditioning system based on system energy consumption prediction and operation optimization.The method preprocesses the historical operation data and weather data of the air conditioning system,determines the prediction order of input variables and target variables by correlation analysis,and uses five prediction models to predict the cooling load of the central air conditioning system,and obtains the total cooling load of the system at the next time.The energy consumption model of main equipment of air conditioning system is obtained by regression modeling.The energy consumption model of chilled water system and cooling water system are obtained by coupling different energy consumption models.Finally,the energy consumption model of air conditioning system is obtained.The predicted total cooling load of the air conditioning system is taken as the goal of operation optimization,and the particle swarm optimization algorithm is used to optimize the operation of the air conditioning system.In this paper,three multi-objective regression methods are used to predict the regional cooling load for the central air conditioning system with multiple cooling zones.The three multi-objective regression methods are the combination prediction of regression chain and support vector machine,the combination prediction of regression chain and multi-layer perceptron,and the bidirectional long-term and short-term memory neural network with multi-objective output.Then,the total cooling load of the system is obtained by summing the cooling load of all cooling zones.Two single objective regression methods are used to predict the cooling load of the system directly.The single objective regression method is support vector machine and bidirectional long-short term memory neural network.The results show that the prediction accuracy of the multi-objective regression prediction model is better than that of the single-objective regression model,and the regional cooling load prediction method is more suitable for the central air conditioning system.In the multi-objective regression prediction model,the determination coefficient of the regression chain and support vector machine combination prediction model is 0.9078,which is higher than the other two multi-objective regression models,and the prediction speed is 6.97 s.Therefore,the multi-objective regression model combined with regression chain and support vector machine is more suitable for cooling load forecasting of the system,and has higher prediction accuracy and faster prediction speed.Overall,its prediction effect is best in all models.The operation data of the main equipment of the hospital air-conditioning system are sorted out and the regression modeling is carried out respectively.The energy consumption models of two water chilling units are established,and the accuracy is over 96 %.The energy consumption regression models of primary cooling pump,cooling pump and cooling tower are established.The accuracy of the energy consumption regression model of cooling pump is more than 90 %,and the accuracy of the energy consumption regression model of cooling tower is more than 98 %.In addition,the flow-frequency regression model of cooling pump and the heat release regression model of cooling tower are established to prepare for the system operation optimization.The energy consumption model of water chilling unit in air conditioning system is obtained by combining two energy consumption models of water chilling unit.The predicted total cooling load of air conditioning system is used as input,and the optimal operation scheme of water chilling unit is obtained by particle swarm optimization.Five working conditions are selected from the operation data of the chiller to optimize.After comparison,it is found that when the temperature difference is not changed,the energy saving effect of the five working conditions is general,with an average energy saving of 7.5 %.When the temperature difference is minimum,the energy saving effect is the highest,with a maximum energy saving of 17 %.The greater the cooling load required by the system,the greater the energy saving potential of the chiller.The energy consumption model of air conditioning cooling water system is obtained by combining the energy consumption model of cooling pump and cooling tower.The particle swarm optimization algorithm is used for operation optimization in five working conditions.The comparison shows that the energy saving efficiency of cooling water system is between 5 % and 10 % after optimization in five working conditions.The energy consumption model of system chilled water system is obtained by combining the energy consumption model of water chilling unit and chilled primary pump,and coupled with the energy consumption model of cooling water system to obtain the energy consumption model of air conditioning system.On the premise of satisfying the best operation scheme of the chiller,the five working conditions of the system are optimized.The results show that the energy saving effect decreases due to the increase of energy consumption of the primary pump and cooling water system.The average energy saving rate of the five working conditions can reach 8.67 %,and the maximum energy saving rate can reach 11.8 %.