Research on performance of passive heat supply tower based on the back propagation neural network

In this study, the performance parameters of passive heat supply tower (PHST) were analyzed. The optimal daily water-air ratio of PHST was obtained using non dominated sorting genetic algorithm (NSGA-II), BP neural network algorithm and decision score (DMS). The energy saving potential and energy efficiency ratio range of 48.8-50.8 were obtained for PHST operation with the air parameters corresponding to the optimal water-to-air ratio. The study shows that PHST has superior energy-saving characteristics when used to supplement heat for underground soil. The BP neural network model can accurately calculate and predict the variation of PHST performance parameters. Higher energy efficiency of PHST system can be obtained by effectively controlling the water-air ratio. This study provides a new scheme for the operation of the supplementary heat system and a new idea for the efficient operation of the heat exchange equipment.