A Study On Optimizing Ventilation Module Of Metro System By Using Genetic Algorithm

With the citilized trend and the concentrated economic development, metro as one of the quick, comfortable and environment-protected transportation is being constructed in many big cities. To simulate the tunnel ventilation modules under different circumstances becomes an urgent requirement of the subway environment control system in the early period of the metro project (scheme design stage). Based the result of the simulation, the installation of the fans and their running modules are decided; the initial investments and subsequent operation expenses will decrease consequently. The possible running modules of n tunnel fans are 2n which make the process to find the optimal running module very difficult. By using the incomplete enumeration method, it is possible to find the satisfied but not optimal module which will definitely increase the initial investments and operation expenses. Therefore, the optimization of the running modules is very necessary. A new optimizing algorithm, genetic algorithm, is used in the paper to optimize the subway ventilation running module.Genetic algorithm is an effective optimizing method. It. does various genetic operations in the code space so as to find the whole optimal answer. The most advantage of the genetic algorithm is that it can get good result without difficulty from various non-linear, multi-module and multi-objective problems such as the subway ventilation modules.The difficult part of the application of genetic algorithm to real problems lies on the port part which includes the decision of the coding strategy, the map relationship of the problem, the definition of the adaptive value and the regulation of the genetic coefficients. The study deeply discussed the above problems and programmed the practical software to simulate and optimize two section stations and the tunnel section itself. With the simulation result, the genetic algorithm model, coding form, adaptive value and suitable geneticcoefficients of three running modules of the metro system are summarized. The genetic algorithm program can simulate and optimize different metro system and has general application prospect. It will be a powerful tool for the scheme design of the subway environment control system and the implications for the national metro construction are real and far-reaching.