During unexpected operational disruptions, addressing how to develop targeted guidance information for affected passengers in urban rail transit systems to ensure their safety and maintain network operations is a pressing issue. This thesis focuses on passengers present in the network during disruptions, categorizing them and disseminating guided paths through station broadcasts. This thesis primarily minimizes the total travel time of passengers by formulating an optimized passenger flow guidance information release model based on station broadcasts, solved using a genetic algorithm to determine the optimal paths release strategy between affected origin-destination pairs during disruptions. A case thesis is conducted on the regional rail network of Guangzhou. The induced path release strategy calculated by the model results in an average passenger travel time of 21.31 min. The release strategy with the maximum target value generated during the optimization process yielded an average travel time of 24.79 min, an increase of 16.34%. In contrast, the strategy of universally releasing the shortest path information to passengers resulted in an average travel time of 23.25 min, an increase of 9.11%. The computational results indicate that the model effectively generates reasonable solutions that significantly reduce passenger travel costs, providing theoretical support for management departments in organizing passenger flow during disruptions.