Research on the Safe Driving Behavior of Transport Vehicles Based on Big Data and Micro Trips

In today's transportation field, the widespread application of vehicle Internet technology has enabled transportation vehicles to accumulate a large amount of driving behavior data. However, existing methods for evaluating driving behavior have many limitations. This study aims to integrate multi-source data such as vehicle trajectory data, meteorological conditions, and road conditions to construct a comprehensive and accurate comprehensive evaluation model of transportation vehicle driving behavior. Through data preprocessing of the DriVeBahn dataset, including denoising, map matching, and micro-trip division, an evaluation index system for driving safety was determined, and mining algorithms for various unsafe driving behaviors were designed. At the same time, a comprehensive evaluation model was established by combining Principal Component Analysis (PCA) (retaining 95% variance) and Analytic Hierarchy Process (AHP). The research results show that this comprehensive evaluation model is more scientific and objective. Most vehicles have a comprehensive score above 60 points, and vehicles with low scores have more unsafe driving behaviors. This study provides scientific and effective decision-making support for the safety management of transportation enterprises, helps to reduce the incidence of traffic accidents, and improves transportation efficiency and safety.