A multi-dimensional method of nodal load forecasting in power grid

It is very important for intelligent dispatch and control of the future smart grid to grasp an arbitrary nodal load pattern in the general great grid to realize the load forecast of an arbitrary nodal load. With this in mind, a multi-dimensional nodal load forecast system and the corresponding method for effective prediction of an arbitrary nodal load of the grid are proposed in light of the weakness of isolated nodal load forecast and based on the correlation analysis. This system includes topological partitioning of the grid energy flow according to layers and regions, the basic forecasting unit composed of each layer's total amount of load and its nodal loads, and combined forecast for any one node. The forecasting method is based on techniques including the least square support vector machine, Kalman filtering and weighted recursive least squares, respectively. A case study shows that the multi-dimensional nodal load forecasting model helps to improve the forecasting accuracy and has prospects for application. © 2012 State Grid Electric Power Research Institute Press.