Research And Application Of Expressway Toll Data Split And Verification Algorithm

Expressway networking toll system has a lot of advantages, such as high efficient, energy saving, environment friendly and so on. It effectively relieves the traffic congestion, environmental pollution and other issues, providing a strong guarantee for the construction of three strategies and new urbanization. However, there are some shortcomings in terms of networking toll split. Unfair and unreasonable networking toll split problem will be more prominent. On the one hand, the fuzzy split of networking toll inevitably leads to disputes, but the owner of highway sections lacks of credible evidence to find and resolve disputes that may result in huge economic losses, and needs to establish a system to check the networking toll split. On the other hand, the hardware-based verification method is high cost, low efficiency, low accuracy, poor stability, and difficult to promote. Therefore, the study of verification method that is low-cost, high-accuracy, high stability, and easy to promote has important practical significance and economic value.According to the defect of the hardware-based verification method the verification program based on data mining technology is first proposed in the paper, it sets up networking toll splitting and checking model by analyzing historical networking toll data, using incomplete data of the highway sections to predict the data can’t be obtained, and compares the data of settlement center to check the networking toll split. Firstly, half a year’s historical networking toll data from Guangzhou North Second Ring Expressway section and settlement center is collected, integrated and pre-processed, and the data warehouse in the theme of networking toll splitting and checking is established. Then variables are filtered out by Lasso algorithm and stepwise regression algorithm, traffic and tolls were accurately predicted based on multivariate linear regression algorithm, and the model is evaluated. Finally, the model is verified and applied, and its application effect is analyzed.The results show that the combination of Lasso algorithm and stepwise regression algorithm can well select significant variables and solve the multi-correlation problems among the variables, and this algorithm is high-accuracy, high stability, and easy to promote.