Short-Term Prediction Of Traffic Flow Parameters Based On License Plate Recognition Data

With the development of the society, the perfecting of motor industry and theimprovement of the level of people life, the amount of sedans is becoming greaterand greater Whether urban managers or travelers want to find a feasible solution to this problem, intelligent transportation system is the most important means to solve the problem of traffic congestion. Short term prediction of traffic flow parameters is the basis of traffic control and guidance module in intelligent transportation system. With the development of modern science and technology, the accuracy of license plate capture and recognition is significantly improved, and the data acquisition technology based on the traffic flow parameters of vehicle license plate recognition is produced. Short term prediction of traffic flow parameters based on vehicle license plate recognition data is a new choice.First of all, this paper analyzes the increasingly serious traffic congestion problems and the advantages and disadvantages of existing traffic information collection technology, the method of short time prediction of traffic flow parameters by using the data of vehicle license plate recognition is presented, and the feasibility of the method is pointed out by analyzing its technical characteristics.Secondly, introduces the process of license plate recognition data acquisition. And edit distance algorithm in fuzzy matching is used to match the upstream and downstream of the vehicle license plate. And the process of the traffic flow data still exist abnormal data through statistical calculation, similarity based on traffic flow is compensated by using historical data of week. The travel speed of the abnormal data are eliminated by using variation coefficient method. Then, the grey forecasting model is introduced to the data after the pretreatment. This paper introduces some operators of the grey prediction model, and makes a detailed exposition of the modeling steps of the grey prediction model. And on the basis of analyzing the background value of grey forecasting model, an improved method based on background value is proposed. Finally, select the Chengdu high-speed beltway outer ring K66+850, K67+985 two monitoring points in the license plate recognition data, with 5min interval traffic flow parameters for short-term forecasting, and compared with the actual test data, proved the accuracy of short-term prediction of traffic flow parameters based on license plate recognition data.