Optimising driving strategies for long heavy haul trains

Rail operators worldwide are increasing the length and weight of Heavy Haul trains to increase network capacity and reduce operating costs. But with these changes to the train weight and length there is a need to reconsider the driving strategy. It is essential that these new trains are driven safely, quickly and efficiently. Determining this point is difficult as there are many possible driving strategies to consider. Currently new driving strategies are determined by using computer simulation, computer tools, expert drivers and experimental field tests. While these current methods provide good driving strategies they might not determine the optimal driving strategy. The aim of this research is to create a computational method that can produce an optimal driving strategy for a given train, track and operational requirements. The computational method being developed has the advantage as it can create and test many different driving strategies and use the results to incrementally create better strategies. The nature of the computational method also means that different train configurations or track modifications can be tested relatively easily. Initial results have shown that a method using genetic algorithms can determine optimal driving strategies based on energy. The research is continuing to incorporate longitudinal dynamics into the optimisation method.