Predicting power load is a traditional problem in the power grid.Effective peak load forecasting can economically and reasonably arrange the start and stop of generator units

maintain the safety and stability of the power system

under the premise of ensuring the normal production and life of society

reduce the cost of electricity generation

and effectively improve social and economic benefits.The characteristics of power load forecasting have been studied in detail

We proposed a DAAGA-BP neural network prediction model and conducted in-depth comparative analysis of multiple prediction algorithms.Through simulation calculations

comparing the fitted data with the actual load values

the predicted value curve is closer to the actual value curve.It has been proven that the model has adaptability and is generally feasible and effective when applied to power grid load forecasting analysis.