An Evolutionary Strategy for the Parametrization of a Linear Predictive Model Generator in Time Series Modelling

摘要

This paper describes the application of an Evolutionary Strategy of the (μ,λ) class to the problem of automatic parametrization of a linear predictive model generator, which is used for short time forecasting in noisy time series with changing dynamics and variable time spacing. Results with real time series of pressure and volume of flow in a pipeline are presented.