Research Of Road Traffic Accident Combined Forecasting Method Based On Grey Correlation Degree

Road accidents not only bring huge economic losses, but threaten people’s lives seriously. China as the world’s largest developing country has a higher rate of death compared with other countries. Therefore, to analyze the mechanism and influencing factors of road accidents, and to predict the index of road traffic accidents are helpful for taking traffic preventive measures, preventing and reducing road accidents and improving road traffic safety management level.This paper begins with the mechanism of road accidents, combines the characteristics of road traffic system:complexity, nonlinearity and grey. Firstly, to introduce the influencing factors of road traffic accidents from people, vehicles, roads and environment, based on this, using the grey correlation analysis to identify the main influencing factors. Secondly, we construct the combined forecasting models of IOWA operator based on grey correlation degree and its improved model. Which are described in detail as follow:(1)Factors Affecting Traffic AccidentsSelecting the influencing factors of road traffic accidents as a factor sequence and the predictor of road traffic accidents as a behavior sequence, statistical data of road traffic accidents in China and the regional condition as examples, we using the grey correlation analysis to solve the gray correlation matrix, and then to analyze each factor’s influence on road traffic accidents. The results show that the motor vehicle driver’s violations are the important factors affecting road traffic accidents on the whole, however, the influencing factors affecting traffic accidents have different relevancy in different regions.(2)Combined Forecasting Models on the Basis of Grey Correlation DegreeBased on the grey correlation degree and combined forecasting model of IOWA operator, this paper established a combined forecasting model of IOWA operator based on grey correlation degree and its improved model. These two models combine grey correlation with IOWA operator, overcome some limitations of using single error index as forecasting precision and better reflect the complex, non-linear and grey features of road traffic system, and provide a new idea to study combined forecasting at the same time. The model is converted to Quadratic programming problem with constraints, the MATLAB optimization toolbox is used to solve the optimal weighting coefficients of combined forecasting models. (3)Examples of Road Traffic Accidents ForecastTake China’s road traffic accident statistics for example, the models of the fourth chapter were used to predict. By comparing the single forecasting models and the combined forecasting model of IOWA operator, results shows that the established models of this paper have a low error, higher precision and accuracy, which provide a reference for accurating prediction of road traffic accidents.