作者: Grzegorz Dudek
DOI: 10.1109/TEVC.2016.2586049
关键词:
摘要: In this paper, a new forecasting model based on artificial immune system (AIS) is proposed. The used for short-term electrical load as an example of time series with multiple seasonal cycles. Artificial learns to recognize antigens (AGs) representing two fragments the series: 1) fragment preceding forecast (input vector) and 2) forecasted (output vector). Antibodies recognition units AGs by selected features input vectors learn output vectors. test procedure, AG only vector recognized some antibodies (ABs). Its reconstructed from activated ABs. unique feature proposed AIS embedded property local selection. Each AB in clonal selection process its optimal subset (a paratope) improve prediction abilities. simulation studies was tested real power data compared other AIS-based models well neural networks, autoregressive integrated moving average, exponential smoothing. obtained results confirm good performance model.