Identify of bus incident reason based on Fuzzy cluster analysis

摘要

In the course of the operation of bus system, once the bus incident occurs, bus dispatch center needs to develop real-time adjustment measures. Identify the courses of the incident is the basis for development online dispatching measure, the paper researches the measures of bus incident auto-detection and identification the courses led to incident. According to bus location and speed information collecting from GPS satellite positioning system, compare the actual running situation and driving program, study the method of bus incident auto-detection. According to bus stations passenger data and sections occupancy data when bus incident occurring, using Fuzzy cluster analysis method to study the measure of identifying the courses led incident. According to bus operation data comes from a bus line in shanghai, calculate the similarity of validation sample and history typical sample to identify the courses led to bus incident, and compare the simulation results with actual situation. The comparison shows the fuzzy cluster analysis is an effective method to apply into identifying the courses led bus incident.