Research On Bus Driving Behavior Characteristics And Safety Scheduling Optimization Method Based On Early Warning Data

Bus drivers are an important factor to ensure the safe operation of vehicles.How to dynamically identify and accurately control the working state of drivers is of great practical significance for reasonably arranging vehicle scheduling and ensuring the operation safety of urban public transport system.Taking the vehicle early warning data collected by Zhenjiang bus operation monitoring system as the object,this paper uses data mining technology to study the behavior state characteristics of drivers,and constructs a bus driver scheduling optimization model for safe operation,so as to provide technical support for public transport enterprises to improve operation safety.The main research contents and conclusions are as follows:Firstly,the early warning types of drivers are determined as high-frequency type,fatigue type,high-risk type and violation type.According to the early warning types,K-means clustering algorithm is adopted to divide bus drivers into three types:a(early warning high hairstyle),B(early warning low hairstyle)and C(early warning balanced type)The three dimensions of early warning are used to analyze the driving behavior characteristics of different types of drivers.The results show that class a drivers are older,have relatively low education,are prone to fatigue driving and have the lowest safety.Class B drivers are older,have the shortest working hours and have the highest safety except for high-risk early warning.Class C drivers are younger,have relatively high average speed and have medium safety;Secondly,the TOPSIS evaluation model of driver’s own attributes based on entropy weight method is constructed,the relationship between early warning frequency and working hours and weather conditions is statistically analyzed,and the quantitative characterization method is used to realize the operation safety evaluation of bus drivers.The total driver safety score is calculated by weighting the driver’s own attribute score,working period score and weather condition score,and then the driver scheduling optimization fitness function is constructed;Finally,aiming at the safe operation of public transport,a driver scheduling optimization model is established based on genetic algorithm,and an example is analyzed.The results show that the safety oriented scheduling optimization algorithm greatly improves the safety of individual drivers and the overall operation,and can improve the bus safety by 12.37%,which verifies the reliability of the optimization model.