Network EMU scheduling based on expert system

According to the complex situations of EMU operation on network, such as more lines, more operation lines, more EMUs, more EMU depots as well as more repair systems and cycles, we propose a network EMU scheduling diagram, which is clearer to display the information of EMU scheduling in network circumstance. Then a bi-objective programming optimization model for EMU operation is established with the regular running mileage and time of maintenance operation package, the repair capacity of EMU depot (EMU running shed) and the minimum connecting time as its constraints, with the minimum number of EMUs as its optimal target and with the least repair times as its suboptimal target. The experience for making EMU scheduling is added to expert system. The Max-Min Ant System (MMAS) directed by Expert System is proposed to solve the model. Based on 432 train operation lines, we solve an EMU scheduling of a network composed of Beijing-Shanghai, Shanghai-Hangzhou, Nanjing-Hangzhou, Beijing-Tianjin, Shanghai-Nanjing and Jinan-Qingdao lines. Results show that the solution has reduced the number of EMUs to a certain extent compared to the result by using ant colony optimization alone.