Prediction to Temperature Field of Heating Efficiency Detecting Cavity for Wood Heating Floor based on BP Network

Abstract

Heat storage efficiency of wood floor made by various tree species is different. Author’s research group has developed the equipment for testing the thermal storage efficiency. The number of temperature sensors is few and it’s difficult to construct a complete and continuous temperature field in cavity of this equipment. In this paper, BP neural network is used to predict and analyze the temperature field in the closed cavity in time and space dimensions based on the temperature data of the finite points obtained from the test. In the time dimension, the coordinate and the temperatures of 3 nodes are taken as inputs and the temperature of the fourth node is taken as the output. The results are: MRE= 0.2694%, MAE= 5.9162%,MSE= 0.4225%, $R^{2}$=0.9988. In the space dimension, the temperatures of 150 measured points are used as the model training and testing sample which can chose from geometric view. The results are: MRE=.3641, MAE =4.7817, MSE=0.5216, $R^{2}$ =0.9985. It can be concluded from above: that the BP neural network used in this paper can effectively obtain the continuous and complete temperature field inside the testing cavity of the wood floor, which provides a new theoretical support for the analysis and calculation of the thermal storage efficiency of the wood floor.