In view of the uncertainty and randomness of wind power and the existing wind power point prediction can not reflect its uncertainty information, a short-term wind power interval prediction model based on local characteristic-scale decomposition (LCD) -sample entropy (SE) and improved whale optimization algorithm (IWOA) -kernel extreme learning machine (KELM) was proposed. LCD decomposition is used to reduce the non-stationarity of the original wind power sequence. By measuring the sample entropy of each ISC component, the new sequence is reconstructed to reduce the influence of excessive components on the prediction accuracy, and then the interval of each new sequence is established. The prediction model is finally superimposed on the prediction results of each new sequence to obtain the final prediction result. The improved WOA algorithm is used to optimize the parameters of the kernel extreme learning machine. The experimental simulations show that the proposed model can obtain good interval prediction results, which has certain practical significance and application value.