Power Battery Fault Diagnosis of Electric Vehicles Based on Modified Shannon Entropy in Real Scenarios

Fast and accurate fault diagnosis of electric vehicle power battery systems is important to ensure the safe and reliable operation of vehicles. For a long time, power battery fault detection methods have been widely studied and a rich literature library has been formed, in which the interval probability-based Shannon entropy method has been applied in many literatures. However, when we used real-world vehicle data from the cloud platform to validate and analyze the model, a large number of false alarm single cells are found in the diagnostic results, based on this, we further extended our research for the traditional Shannon entropy method. First, we analyze the abnormal voltage fluctuation fault and the fault diagnosis principle of this method. Then, the misdiagnosis mechanism of the method is explored in the context of two typical vehicle driving conditions. Finally, a solution to mitigate false alarms is proposed and its effectiveness is verified based on real-world vehicle data.