Faulty components detection of smart meters based on transfer function

With the rapid growth of smart energy meters and the transformation of management models, fault detection has become a key technology that affects rotation decisions and benefits. The existing fault detection relies heavily on manual work, resulting in low detection efficiency. This paper proposes a faulty component identification method based on the transfer function, which detect the fault characteristics of the electric smart meter without opening the smart meters, and then identify typical faulty components through a novel classification algorithm based on improved two-branch convolutional neural network (CNN) model, so as to avoid the dependence on human experience and improve the detection efficiency. Experimental results demonstrate that the method can accurately identify smart meter faults, offering an innovative solution for the recycling of retired meters.