Short-term load forecasting with chaos time series analysis

Abstract

This paper presents a new approach to short-term load forecasting in power systems. The proposed method makes use of chaos time series analysis that is based on deterministic chaos to capture characteristics of complicated load behaviour. Deterministic chaos allows us to reconstruct a time series and determine the number of input variables. This paper describes chaos time series analysis of daily power system peak loads. The nonlinear mapping of deterministic chaos is identified by the multi-layer perceptron of an artificial neural network. The proposed approach is demonstrated in an example.