Data-Driven Load Forecasting Method for 10 kV Distribution Lines

The 10 kV distribution line load prediction method suffers from the problem of large absolute errors in the prediction results, and a data-driven 10 kV distribution line load prediction method is designed. The actual values of demand coefficients in the region are derived from historical data, the power characteristics of 10 kV distribution lines are obtained, the set of upstream load points at each of the two end nodes of the contact line is obtained, the load transfer threshold is set, the percentage of heavy-duty distribution substations is calculated, and the data-driven load prediction model is constructed. Experimental results: The mean absolute errors of the 10 kV distribution line load prediction method designed this time and the other two 10 kV distribution line load prediction methods are: 6.896%, 10.461% and 11.224% respectively, indicating that the designed 10 kV distribution line load prediction method works better when combined with data-driven technology.