The short-term load forecasting using the kernel recursive least-squares algorithm

This paper presents a new approach for short-term load forecasting problem based on the kernel recursive least-square algorithm (KRLS). The kernel recursive least-square algorithm is an online real-time kernel-based algorithm and also capable of efficiently solving in recursive manner nonlinear least-square predictive problems. In this paper we consider the loads as a time series, through training the KRLS, we give the one-step ahead load forecasting. The test result of short term load forecasting series shows that the precision of load forecasting is greatly improved by means of the new method.