An Intelligent Screening Algorithm for Mining Key Dangerous Sources of Urban Ground Transport

With increasing bus capacity, operational intensity, etc., urban public transport emergencies are more and more characterized by heavy loads with high frequency. To build a collaborative public transport emergency command system (CPTECS) based on existing systems and datasets, bus emergency scenes and categories of sources of danger are defined. Emergency cases in Beijing are selected for analysis, designing new means of encoding and expanding the decision attributes of the rough set model. Cellular genetic algorithm (CGA) is used to screen key hazards highly correlated to existing information systems. By comparing with genetic algorithm (GA), it is found that CGA can better solve attribute reduction problems of multi-decision attribute rough set in stability, convergence quality, and algorithm efficiency. Based on the meteorological hazards screened out, a CPTECS is designed, enriching research in such territories. Research findings provide quantitative support for the design of CPTECS, and have certain practical significance.