Research On Intelligent Algorithms In Expert System Of Public Transit

Nowadays, the major cities in the country give priority to the development of urban public transit. The Expert System of Public Transit (ESPT) has been an important measure to improve the information and modernization of urban public transit system. ESPT will play the important role in vehicle scheduling and networks optimization.The application of Ant Colony Algorithm and Genetic Algorithm in the field of artificial intelligence research is hot these years. As a kind of intelligent algorithm, they resolve a lot of practical problems, such as engineering and computer simulation.Ant Colony Algorithm, simulating the habits of ants, is a simulated evolutionary algorithm which shows many promising characters and has been applied successfully to applications of data mining. Genetic Algorithm is a kind of randomized search algorithms based on natural chosen and natural inheritance mechanism. Its main characteristic is the colony search strategy and information interchange between individuals in each colony and the search does not depend on grads information.Vehicle scheduling is part of ESPT. In the case of limited investment, it can effectively make it convenient to take an urban public vehicle, which is widely facing in most of cities at the present time, and has important practical signification. On the basis of previous studies, this paper explores the genetic algorithm intelligent algorithm to solve the scheduling problem in the application.As part of ESPT, the optimization public transit networks will essentially make the allocation of resources for urban public transport reasonable. An optimal model is important to public transit networks optimization. The dissertation studies optimal models of one public transit line and integer networks, discusses their restrict condition each other on the base of analyzing optimal aim and effective factors of public transit networks optimization. To compare with several optimal methods in common use, it is excellent to apply to ant colony algorithm to optimize public transit networks.This paper respectively introduces the applications of Ant Colony Algorithm and Genetic Algorithm in ESPT, and addresses the principle and characteristics of algorithms. The application of intelligent algorithms in data mining is still at an early stage. This paper, based on the results of previous studies, gives the research of the intelligent algorithms in the application of intelligent systems, including the rules design, rules changes, pheromone updates, convergence test, the sample data revised, and so on.Based on the experimental data, the performance of the algorithm was tested. Results show that the algorithm can make ESPT trusted. By continuing to optimize the rules, optimization parameter settings and will surely achieve more perfect results. However, the intelligent algorithms are complex, more optimization rules and parameters can make more perfect results. It can be predicted that the ESPT will play an important role in the urban public transports.