The Method of Traffic State Identification Based on BP Neural Network

According to the complex characteristic of traffic flow,the method of traffic state identification is proposed based on BP neural network.Taking the highway tunnel as an example,the paper realizes the analog simulation through the Matlab software programming,and eventually determined its reflected traffic state through comparing neural network's output result with the threshold value.The simulation result indicates that this algorithm can be used in the real-time traffic state identification,and has very high precision and good convergence.