Cooling load forecasting for public buildings

Abstract

In order to realize optimization control strategies for central air conditioning systems, an air conditioning load forecasting model for public buildings is developed with BP neural networks. Bayesian normalization method is used to train the neural network model for improving its generalization abilities. The neural network model constructed in this thesis is simple in structure, and does not need to know the internal mechanism between the load and its influencing factors, and has fewer input parameters, avoiding cumbersome calculations. This project utilizes the real-time data samples of the water storage air-conditioning system from June to August for network training and network verification. The validity and accuracy of the model model are verified by comparing the load prediction value with the actual load value, which can be used for central air conditioning. Research on system control strategy and predictive control of water storage air conditioning systems. It can be used for the research of control strategy of central air conditioning system and the research of predictive control of water storage air conditioning system.