In order to solve the problems of road traffic safety macroscopic evaluation and traffic safety level classification in China, an index system which consists of 14 relative indexes for road traffic safety macroscopic evaluation was established taking into account the synthetic influences of human, vehicle, road, and socio-economic developmental levels on traffic safety. Then, a new PCA (principal component analysis)-CA (cluster analysis) method was put forward based on the math theory of PCA and CA. Firstly, the PCA method was used to reduce the dimensions of the multi-index data, and then the Irts (index of road traffic safety) was constructed for the order of road traffic safety situation of samples being evaluated. Secondly, the K-means clustering method was applied to safety level classifications taking Irts as variables. Last, 2006 road traffic accident statistical data of certain provinces in China was selected as examples, and the analysis results show that there are differences in the order of road traffic safety situation by comparative study on the PCA-CA method, projection pursuit (PP) method, and fuzzy comprehensive evaluation (PCE) method.
展开▼