An Adaptive Nonlinear Predictor with Orthogonal Escalator Structure for Short-Term Load Forecasting

An adaptive Hammerstein model with an orthogonal escalator structure as well as a lattice structure for joint processes is developed for short-term load forecasting from one hour to several hours in the future. The method uses a Hammerstein nonlinear time-varying functional relationship between load and temperature. Parameters in both linear and nonlinear parts of the predictor are updated systematically using a scalar orthogonalization procedure. Matrix operations are avoided, thereby allowing better model tracking ability, numerical properties, and performance. Prediction results using actual load-temperature data demonstrate that this algorithm performs better than the commonly used matrix-oriented recursive least-squares algorithm for one-hour-ahead forecasts.