Introduction  This study focuses on the prediction of summer daily maximum power load in Wuhan, Huangshi, and Yichang of the Hubei section in the Yangtze River Economic Belt based on climatic forecast model temperature data.

  Method  By analyzing the daily maximum power load data from 2008 to 2019, along with meteorological elements such as average temperature, maximum temperature, minimum temperature, and regional climate model (RegCM4) forecast data, the characteristics of meteorologically sensitive power load in the three regions were analyzed. Regression analysis and a group-particle optimized back-propagation (BP) neural network algorithm were used to quantitatively predict the future (from 2020 to 2096) daily maximum power load.

  Result  The results indicate that there is a significant correlation between summer average temperature and meteorologically sensitive load. The predicted values of the summer daily maximum power load in Wuhan and Yichang show a steady increase similar to the past decade, with the growth rate of regression prediction slightly higher than that of neural network prediction. The growth rate in Yichang is higher than that in Wuhan, exceeding 40% at its peak. The expected values of the daily maximum power load in Huangshi show distinctly different prediction results compared to the other two locations.

  Conclusion  Predicting the variation patterns of summer daily maximum power load in medium-to-large cities along the Yangtze River Economic Belt is beneficial for planning the required additional grid capacity in the future.