Study On The Area Traffic Control Evaluation Method Based On Fuzzy Neural Network

Since the 1990s, Europe together with some other countries proposed the intelligent transport systems (ITS). As an important component of the intelligent transport systems, the advanced traffic signal control system (ATCS) is not only a important traffic information collection systems, but also an implementation subsystems to regulate traffic flow distribution. In parallel with the use of traffic engineering, psychology, applied mathematics, automation and information network technologies, as well as systems engineering and other subjects of the theory, ATCS has already developed into a application system.Traffic Signal Control method has already developed from fixed timing signal control to sensor, adaptive signal control after several decades. Lots of excellent urban traffic signal control systems has been developed and put in to use, such as: TRANSYT, SCATS, SCOOT. Development and application of China's signal control system started in a late period ever since it was listed in the "July 5" key scientific and technological project. Although a number of systems has been put into market in 1990s, development and applications of the those products are not so good while comparing with the contemporaneous foreign ones. Based on the actual conditions of urban traffic signal control system developed in China, this paper carried out a in-depth research on the regional urban traffic signal control.In order to enable the effectiveness of area traffic control, this paper presented regional traffic control evaluation method which based on fuzzy neural network. When controlling a road network, it is necessary to feedback the effectiveness of area control, this action has great significance. So we should evaluate the effectiveness of area traffic control.Chapter 1: IntroductionThis chapter mainly introduced the significance of this paper and summed up the status quo at home and abroad. First introduced the problem with the effectively easing of the traffic congestion problems in such a rapid growth of private cars and high density of urban population. How to enhance the security of urban traffic? It has become the most serious problem in the development of China's urban cities. With the adopt of adaptive system, some of the traffic congestion problem can be solved. However a single point or line control of signals can only take some independent and some certain sub area intersections of traffic into account. Control of the area can seek the optimal scheme overall based on the local optimization, so that it is indispensable to research on the evaluate the area traffic control. Finally, the paper presents the argument of the main contents and chapters.Chapter 2: Establish of Area Traffic Control Indicators SystemThis chapter mainly selected the area traffic control indicators and established the indicators system. According to the characteristics of scientific, independent, practical, comprehensive, universal principles, the paper selected the regional average delay, regional average stop frequency, regional average queue length and regional average capacity as the area traffic control indicators, then used the AHP to determine weights of intersections and used the object-arrangment-expandedness method to establish indicators system.Chapter 3: Analysis of Comprehensive Evaluation MethodAccording to the characteristics of fuzzy comprehensive evaluation and neural network, this chapter summarized the characteristics of the ANN-FCE method. While introducing the fuzzy comprehensive evaluation methods, the paper analyzes their concept , advantages and disadvantages of steps in brief at the beginning. Then this paper have a brief introduction of BP neural network, and summed up the characteristics of ANN-FCEChapter 4: Study on Fuzzy Comprehensive Evaluation Method Based on Neural NetworkIn order to eliminate the influences of different indicators units, the paper introduced the method to measure the indicators units, and established the fuzzy comprehensive evaluation model which is based on the neural network. In the end of this part, VISSIM is used to verify the evaluation results.Chapter 5: Summary and OutlookThis chapter summarized the whole research process, and then put forward to the deducts need to be improved and the further research work.