Parameter Correction-based Adhesion Characteristic Curve Identification for Train Wheelset on Fault Condition

A parameter correction-based adhesion characteristic curve identification strategy is proposed for train wheelset on fault condition. Firstly, the parameters of the fault adhesion characteristic curve(FACC) are initially identified by the nonlinear least square(NLLS) method. Then, a trust interval particle filter(TIPF)-based parameter correction method is proposed. The TIPF iteratively corrects the initially identified parameters of the FACC, and then evaluates the corrected parameters by recollecting adhesion data. Finally, the proposed strategy is verified on the simulink platform. The test results show that the proposed strategy can achieve fast and accurate FACC identification.