Fault diagnosis model training method, power grid fault diagnosis method and storage medium

The invention provides a fault diagnosis model training method, a power grid fault diagnosis method and a storage medium, and relates to the technical field of power grid fault diagnosis. The model training method comprises the following steps: inputting original data of various working conditions into a preprocessing layer to obtain a standard data set; inputting the standard data set into an input layer to obtain fused feature data; the fusion features sequentially pass through a convolution layer, a batch normalization layer and a pooling layer to obtain output data; inputting the output data into the input layer again, and obtaining the convolution data and the component vector of the output data under the current cycle; inputting the component vector into an output layer for iteration to obtain a diagnosis result, and when the component vector reaches a preset maximum number of iterations and the accuracy of the diagnosis result is greater than or equal to a preset accurate threshold, determining convolu