MAINTENANCE SCHEDULING OF ROLLING STOCK USING A GENETIC ALGORITHM

We have developed a Genetic algorithm (GA) for the optimisation of mai ntenance overhaul scheduling of rolling stock (trains) at the Hong Kon g Mass Transit Railway Corporation (MTRC). The problem is one of combi natorial optimisation. Genetic algorithms (GAs) belong to the class of heuristic optimisation techniques that utilise randomisation as well as directed smart search to seek the global optima. The workshop at MT RC does have difficulties in establishing good schedules for the overh aul maintenance of the rolling stock. Currently, an experienced schedu ler at MTRC performs this task manually. In this paper, we study the p roblem in a scientific manner and propose ways in which the task can b e automated with the help of an algorithm embedded in a computer progr am. The algorithm enables the scheduler to establish the annual mainte nance schedule of the trains in an efficient manner, the objective bei ng to satisfy the maintenance requirements of various units of the tra ins as closely as possible to their due dates since there is a cost as sociated with undertaking the maintenance tasks either 'too early' or 'too late'. The genetic algorithm developed is found to be very effect ive for solving this intractable problem. Computational results indica te that the genetic algorithm consistently provides significantly bett er schedules than those established manually at MTRC. More over, we pr ovide evidence that the algorithm delivers close to optimal solutions for randomly generated problems with known optimal solutions. We also propose a local search method to reconfigure the trains in order to im prove the schedule and to balance the work load of the overhaul mainte nance section of the workshop throughout the planning horizon. We demo nstrate that the reconfiguration of trains improves the schedule and r educes cost significantly.