Subway Environmental Control Heat Load Calculation Based On Artificial Neural Networks

Subway has a history of 130 years and becomes the symbol and the primary transportation of the modern cities. The fast construction of subways in China need a simple and time-saving method to estimate the environmental control heat load in the early stage of scheme design of subway project, so that the scale of subway environmental control and investment can be predicted. The calculation of subway environmental control heat load is very complicated, especially for the system without platform screen door whose environmental control heat load is closely related to the run of the train. The heat given off by the train has a great influence on the station cold load. Comparing with other cold loads, this heat load is difficult to calculate. The study involves the method for the calculation of subway station heat load in the system without platform screen door according the real need of engineering design institutes.In the early stage of subway design, because of many unknown variables and the complicated character of subway station heat load calculation, it is hard to use a simplified mathematical model to get the precise result. The neural network method is presented to calculate the subway station heat load for the first time. Based on the comprehensive analysis of the composition of the station heat load of the system without platform screen door and the influential variables in the process of heat load calculation, a simple and fast computing method is presented and a practically useful software is programmed. The results of the simulated calculation of two domestic subways verify the correctness and feasibility of the neural network method and the software. The calculating method of the heat load presented by the paper is a new idea to compute the environmental control heat load in the early scheme design stage of subway, the software will be a powerful tool to predict to the scale of the subway environmental control and the amount of investment. The application of the method has an important real implication.