Evaluation of city road running condition based on GRBF neural network

To alleviate the urban traffic congestion, this paper proposed an evaluation method of city road operation and gave the specific implementation technology. The city road running condition was measured by means of the traffic congestion analysis, and an evaluation index system of road congestion degree was presented. Then the model of computing road traffic congestion degree was set up through the optimization algorithm of generalized radial basis function neural network. For a typical local road network of Xi'an, the traffic congestion degrees of each sample road in the road network was computed based on MATLAB 2010, and the specific location and occurrence time of congestion were obtained. Then the traffic condition of the total road network was evaluated by using the weighted average method and one optimal travel path of the road network was gotten. The results show that the east of South Second Ring road is the most congested in the road network at the same time interval and the traffic congestion index arrives 0.9744; for the sample section on Cuihua road, it is congested at the evening peak 16:00~18:00 and the traffic congestion degree arrives 0.9118, which are consistent with the actual situation. Quantitative value of urban traffic congestion can be acquired by the method presented in this paper, and the evaluation of traffic congestion in certain area can be completed in this way. ©, 2015, Editorial Department of Journal of Chang'an University (Natural Science Edition). All right reserved.