Short-term load forecasting considering improved cumulative effect of hourly temperature

Abstract

As the proportion of air-conditioning load in the total power load increases year by year, meteorological factors such as temperature and humidity have more significant influences on power load fluctuations. Power load fluctuations often lag behind changes in meteorological factors, which increases the difficulty of short-term load forecasting. This paper proposes an equivalent temperature model based on the heat island effect, the temperature and humidity effect, and the improved hourly air temperature cumulative effect. This paper improves the population initial value generation method of the traditional genetic algorithm to improve the algorithm's efficiency in solving the optimal temperature change parameter of the hourly temperature's cumulative effect. The equivalent temperature is then used as the input feature of the prediction model, and the Elman neural network is used for short-term load prediction. Finally, the PJM hourly load data and weather station data in the United States Washington area are used for example analysis. The results show that the improved genetic algorithm can effectively solve the optimal temperature change correction parameters, and the generated equivalent temperature has a higher correlation with the power load; compared with the prediction model considering the actual temperature, the model proposed has a higher prediction accuracy.