Long-gap streamer discharge development image prediction method under lightning impulse

The invention discloses a long-gap streamer discharge development image prediction method under lightning impulse, and the method comprises the steps: collecting a streamer discharge image, discharge voltage and time data, which are shot by an ICCD, through a streamer discharge test observation system, converting the image into a discharge light intensity matrix, determining the relation between the light intensity matrix and the discharge voltage as well as the relation between the light intensity matrix and the discharge time, and carrying out the prediction of the long-gap streamer discharge development image. Establishing a multi-output regression model by adopting an XGBoost algorithm, selecting a loss function to evaluate the optimal model and parameters by adopting an n-fold cross validation method, and inputting instantaneous voltage and corresponding time which change along with time into the trained model according to a voltage discharge sequence to obtain a series of predicted light intensity matri