Research On Optimal Dispatching Strategy Of Power System Considering High-temperature Weather

As the "forerunner" of economic development,electric power is an important pillar industry for China to win the "battle against poverty".Economy is the primary goal of the modern power grid,and accurate short-term load forecasting results can make units obtain the optimal combination mode and improve the effectiveness of economic dispatching.However,in recent years,during the peak summer period,the country has been affected by high-temperature weather on a large scale.The surge in electricity load has led to a gap in the power balance.And residents’ power outages and industrial power rationing have occurred from time to time,which have seriously affected the people’s production and life.In order to improve the economic benefits of the power system,a short-term load forecasting model based on load decomposition under high-temperature weather is established in this paper.Based on the accurate short-term load forecasting results,an economic dispatching model for optimal combination of units is established with the goal of minimizing the total operating cost of the system.The specific work contents are as follows:(1)An Isolation Forest(IF)algorithm is proposed to detect outliers in electricity consumption data.The method has high processing speed and high precision.The improved K-means clustering algorithm combining CHI and DBI indexes is adopted to perform clustering analysis on the electricity consumption data and obtain diffe rent subgroups of electricity consumption.(2)After decomposing the high-temperature load in the summer hot weather,the basic load gray prediction model based on the fractional operator and the high temperature load prediction model based on nonlinear r egression are constructed for the base load and the high temperature load respectively.Compare the results based on the load decomposition model with the total load forecast results.The results show that the precision of the combined prediction model bas ed on fractional operator and nonlinear regression is higher than that of total load prediction.At the same time,based on the above load clustering results and the above load decomposition method,the industry refined short-term load forecasting model based on the load decomposition is established to further improve the forecasting accuracy of short-term load forecasting.(3)The combined STLF results of the grey base load prediction model based on fractional order operator and the high-temperature load prediction model based on LSTM are used as the boundary conditions of the system.After considering the uncertainty of the load,the objective function is to minim ize the total operating cost of the system,that is,to minimize the sum of the generation cos t function,the start-stop cost function and the power loss cost function.An economic dispatch considering the uncertainty of load forecasting(ED-ULF)model was established with the constraints of power supply and demand balance,generating unit output,re serve capacity,generating unit climbing and line transmission as constraints.