Client exception identification method, system, equipment and medium

The invention discloses a client abnormity identification method, system and device and a medium, and the method comprises the steps: obtaining a plurality of current data of at least one client terminal in a first preset time period, and determining a current unbalance rate according to the current extreme value of the current data in the first preset time period. And extracting a plurality of historical current data in a second preset time period from the client terminal of which the current imbalance rate meets the first preset abnormal standard, and determining a corresponding current variation according to the historical current data and the current extreme value. And comparing the current variable quantity with a second preset abnormal standard, and determining the client terminal with the special transformer current abnormity according to a comparison result. Through the first preset abnormal standard, the current unbalance rate, the current variation and the second preset abnormal standard, the curren