LightGBM and GPR coupled wind speed probability prediction method

The invention discloses a wind speed probability prediction method which combines and couples a light gradient hoisting machine (Light GBM) and Gaussian process regression (GPR), and the method effectively overcomes the defect that only a wind speed determinacy prediction result can be obtained and the wind speed uncertainty cannot be quantified when the Light GBM is independently used, so that the prediction accuracy of the wind speed probability prediction is improved, and the prediction accuracy of the wind speed probability prediction method is improved. Meanwhile, the method also solves the problem that a high-precision wind speed prediction result cannot be obtained by independently using a GPR model, so that the LightGBM-GPR model can give full play to the respective advantages of the LightGBM model and the GPR model, the wind speed process with nonlinear, volatility and uncertainty characteristics can be accurately simulated, and the method not only can provide the high-precision wind speed prediction