Study Of The Intelligent Dispatching Algorithm Based On Artificial Neural Network

Enter since twenty-first Century, China rapid economic development, the city changes a process to also make a spurt of progress, people’s living standard enhances increasingly, which makes the vehicle in the city also increases quickly, city road problems become more and more serious, people can get around the difficult problems more obvious. Each local government realized the need to solve problems, road, public transportation is the core, developed public traffic can relieve a city public pressure on the road. China has been more emphasis on public transport, but the public transportation in China is relatively backward, there exists a gap with the developed countries, can not meet the needs of the people. So, to solve these problems of the urban traffic, we must first solve the urban public traffic occurring situation, optimization of urban public transportation.Bus dispatching is reasonable, efficient, stable, is the fundamental solution to the problem of urban public traffic, public transport is the basic. Bus dispatching reasonably efficient, can bus is fully utilized and improve the efficiency of utilization of the bus, and maximize the passengers and the interests of the company, to solve the contradiction produced between bus companies and passengers, thereby improving the benefit maximization and the bus company, but also solve the people can get around the difficult problems, make people travel more convenient. The use of private cars less rate, make the urban traffic pressure eased, the road becomes less crowded.This paper introduces the background and the status quo of the development of intelligent transportation, selecting Changchun City Road 304 bus line as the research object, field investigation on the line, and after a long time of sample collection, public transport departure time scheduling fully understand the study, master grid principle. The purpose of this paper is to improve the efficiency in the use of the bus through the intelligent bus time scheduling, to reduce the average waiting time of passengers, so as to maximize the benefits of public transportation company. According to the past successful and efficient departure experience, to the departure time for the data set, through the realization of the training of the neural network programming on the platform of.NET, to assist the public transit dispatcher for the reasonable and efficient bus departure time scheduling.The classification problem of the bus departure time is a non-linear, according to this characteristic, this paper use the knowledge of the principle of artificial neural network, to solve this kind of problem. Methods: the main steps of various information departure time scheduling bus need to code, and code attributes as input training data, testing for a variety of artificial neural network method, find out a suitable way, and by the method of training, combined with previous data, the trained model is applied in practice, will the results and the actual results are analyzed and compared, the conclusion and the past successful departure time.