In order to further simplify the data processing process and improve the accuracy of production process identification, a process identification method based on non-invasive load decomposition was proposed. Firstly, each process was defined as a kind of electrical equipment. Then, according to the relevant theories of non-invasive load decomposition, bidirectional long short-term memory network and temporal convolution network were selected to construct the load decomposition model, and the corresponding power and total power data of each electrical equipment were selected to construct the data set for training and testing the model. Finally, the corresponding process data was obtained by relevant processing of the load decomposition results of the test set. The results show that the process identification model constructed by the load decomposition method based on the temporal convolution network has high recognition accuracy, and the process identification accuracy for the test set is 98.83%.