Locomotive Brake Control Method Based on T-S Fuzzy Modeling Predictive Control

To achieve braking control of locomotive brake control system (LBCS) accurately and steadily under high nonlinearity various time delay condition, a locomotive brake control method based on T-S fuzzy modeling predictive control (MPC) is proposed. Firstly, the paper uses fuzzy clustering method (FCM) to initial parameters, and uses back-propagation algorithm to rectify rectified its premise parameters by learning off-line. The consequent parameters of the fuzzy rules are self-learning online by recursive least square method with the forgetting factor. By introducing the conception of quality satisfying degree to rectify forgetting factor, the paper can precision and construct T-S modeling.Secondly, the paper uses MPC to control LBCS. The fuzzy genetic algorithm (FGA) of importing excellent subpopulation migrating strategy is used as rolling optimization (RO) method to reduce the influence of the model parameters, the noise coupling and the random interference of this system. It can improve control capability and accelerate quicken the speed of convergence. At last, Simulation and practical application in new generation locomotive brake system show that this improved method is effective.