Abstract: Aiming at the local freight trains transship system for the local freight flow in railway terminal, the simultaneous placing-in and taking-out wagons with multiple engines on branch-shaped sidings is presented. The problem is formulated as a bi-objective programming model. The first term in objective function focuses on balancing the workload of each engine. The second term in objective function aims at minimizing shunting cost which including the running cost of engines and idling cost of wagons. Some constraints indicating the time interval of placing-in and taking-out wagons and traction number of engine are considered into the model. And then a hybrid procedure combining comprehensive correlation indicator and asynchronous iteration heuristic for solving the model is given and is abbreviated as HP-CCI&AIH. HP-CCI&AIH includes two stages: clustering approach with comprehensive correlation indicator for shunting engine allocation, and asynchronous iteration procedure for placing-in and taking-out routing. In first stage, engine allocation strategy based on comprehensive correlation indicator is proposed. According to the idea of clustering division, the comprehensive correlation indicator abbreviated as CCI is proposed to determine the optimum number of engines, divide the work area and assign the tasks to each engine. In second stage, shunting strategy based on asynchronous iteration heuristic is provided. On the basis of the technique of finding the optimal solution using iterative procedure, an asynchronous iteration heuristic is provided and is abbreviated as AIH. Here the cycle body and updating procedure of solution are designed. Further the crossover and mutation procedure of genetic algorithm is introduced to optimize the cycle body. And then the artificial fish swarm behavior is also introduced to achieve the secondary optimization procedure. Finally, the experimental scenarios are designed to test the proposed algorithm, and the proposed algorithm is compared with some other algorithms by some different sized cases. The results show that the proposed algorithm is effective and superior.

Key words: branch-shaped siding; multiple engines; local transship trains; placing-in and taking-out wagons