Abstract：The technological core of the automatic train operation（ATO）for urban trains lies in the highprecision control of traction motors，which requires online identification and correction of the motor closed-loop parameters. In order to satisfy the high-precision requirement that the ATO operation and stopping accuracy of

urban trains must be controlled within ±15 cm，a model reference adaptive system（MRAS）online identification

method was proposed for optimizing the rotor chain observer，carrying out identification parameter matching，and

combined with an improved positional ATO algorithm to verify the practical application effect of control

optimization. Through simulation analysis and experimental verification，this method can accurately identify the

motor parameters，and the deviation from the actual value is less than 0.1，which improves the control performance of the train traction motor，and further optimizes the ATO control of the train within the range of train stopping accuracy can be controlled ±10 cm.

      Key words：urban rail transit；automatic train operation（ATO）；parameter identification；rotor magnetic chain