Study on Model Reference Adaptive Control of ATO Systems

Improving ATO's control accuracy is the major factor to achieve driverless operation over a whole journey.In this paper the dynamic performance of the train braking system was analyzed,and then the state-equation based model reference adaptive control system was built.For the control algorithm,the asymptotic stability was proved theoretically,and the inherent defect to cause local controller oscillation was pointed out.By introducing an appropriate auxiliary system,the augmented-error based adaptive control system was built,which was made able to overcome the defect of the former control algorithm and to possess a more rigorous theoretical structure.Both theoretical analysis and numerical simulation reveal that the proposed algorithm can effectively compensate the uncertainties existing in train operation and achieve accurate tracking performance.