Optimization of trains dynamic wagon-flow allocation based on Improved Genetic Algorithm

Currently, most of the wagon-flow allocation optimization processes are based on constant break-up time and marshalling time. Consequently, it is difficult to implement in practice. According to the field trains operation data, the operation time is fitted by linear regression with the number of operation hooks. Taking the total dwell time of trains at stations as the objective function. Comprehensively considering trains break-up time, marshalling time and continuous wagon-flow, a dynamic wagon-flow allocation optimization model is established. The simulation results show that using the improved genetic algorithm can reduce the total dwell time of trains by 37%. Estimating the break-up time and marshalling time can reduce the total dwell time at stations by 25%.