Online Condition Monitoring Method for the Cooling System of Urban Rail Train

The cooling system plays an important role in the safe operation of traction converters for urban rail trains. While in long-term operation, the cooling system would suffer from various failures, reducing the heat dissipation performance and even causing damage to converter. Thus, it is necessary to monitor the health status of cooling system. This paper proposes to use the temperature response curve of negative temperature coefficient (NTC) thermistor built in IGBT module as indicator to monitor the health status of cooling system of traction converters. Back Propagation Neural Network (BPNN) is employed to predict the reference temperature response curve of healthy state under different working conditions and One Way Distance (OWD) algorithm is applied to achieve the difference comparison between the measured actual curve and the predicted one. This method does not require thermal balance state, which is suitable for the design of cooling system of urban rail train. Moreover, it enables the concurrent monitoring of multiple components in the cooling system without additional measurement circuit. Experimental tests are utilized to verify this approach.