High-frequency partial discharge type identification method and system based on gradient descent method

The invention discloses a high-frequency partial discharge type identification method and system based on a gradient descent method, and the method comprises the steps: reading data of a determined partial discharge type, solving all coefficients of a fitting polynomial based on the gradient descent method, and building a database of feature data and discharge types; reading data of partial discharge types of the tested equipment, calculating a polynomial-based total discharge amplitude, and identifying and classifying the partial discharge types of errors; defining a new type of partial discharge, solving each coefficient of a polynomial based on self-learning, and updating a new discharge type database; the system comprises a database module, an identification and classification module and an updating module. According to the invention, online monitoring of the current operation state of the power equipment is realized, early warning and prediction are realized in advance according to faults possibly caused