Weighing sensor is the key component of dynamic truck scale

once the sensor is failed in work

it will cause serious impact on dynamic weighing system. In order to diagnose fault of weighing sensor accurately

a fault diagnosis method is proposed based on wavelet transform and deep belief network（ DBN）. The method can not only judge whether the sensor signal is abnormal signal

but also can ensure output accuracy of dynamic weighing system by replacing value of fault by predicted value of DBN model. Simulation results show that this method can identify and estimate fault sensor and effectively improve precision of dynamic weighing system.