Aiming at the problem of poor model accuracy caused by poor stability and strong randomness of wind power output. A short-term prediction model of wind power based on quadratic decomposition error compensation is proposed. Firstly, BiLSTM prediction model is established to predict wind power and output prediction error. Secondly, an Improved Dung Beetle Optimization Algorithm (IDBO), which uses chaotic mapping to initialize the population, introduces golden sine strategy to update the location of rolling dung beetles, and adds dynamic adaptive weight coefficient to update the location of thieving dung beetles, is used to optimize the parameters of the prediction model, prevent the network from falling into the local optimal solution, and adaptively search for the optimal parameter combination. Then, the decomposition-refactoring-decomposition strategy is adopted to conduct the first decomposition by using CEEMDAN, and sample entropy and K-means (SE, K-means) are introduced to reconstruct the sequence by frequency, and the high-frequency error sequence is decomposed into error sequences of different frequency bands by variational mode decomposition (VMD). Improve the prediction efficiency and accuracy of subsequent models. Finally, the input error compensation model of each component is used to predict and the Attention mechanism is introduced to learn the feature relationship of different time steps and give different weight values to enhance the attention to key information. Through the measured data of a wind farm in Xinjiang, the prediction accuracy of the proposed model is proved to be high and has significant advantages.