Real-time forecasting and visualization toolkit for multi-seasonal time series. (July 2018)

• Record Type: Journal Article
• Authors: Chen, Jinduan; Boccelli, Dominic L.
• Is Part Of: Environmental modelling & software. Volume 105(2018)
• Page Start: 244
• Page End: 256
• Publication Date: 2018-07
• Subjects: Seasonality -- Autocorrelation -- Time series -- Forecasting -- Visualization; Environmental monitoring -- Computer programs -- Periodicals; Ecology -- Computer simulation -- Periodicals; Digital computer simulation -- Periodicals; Computer software -- Periodicals; Environmental Monitoring -- Periodicals; Computer Simulation -- Periodicals; Environnement -- Surveillance -- Logiciels -- Périodiques; Écologie -- Simulation, Méthodes de -- Périodiques; Simulation par ordinateur -- Périodiques; Logiciels -- Périodiques; Computer software; Digital computer simulation; Ecology -- Computer simulation; Environmental monitoring -- Computer programs; Periodicals; Electronic journals; 363.70015118
• DOI: 10.1016/j.envsoft.2018.03.034
• Languages: English
• ISSNs: 1364-8152

Abstract

Abstract: Many environmental data sets are driven by multiple superimposed periods, yet most time series analysis software packages only support single-seasonality. The objective of this research was to develop a software toolkit utilizing multi-seasonal Autoregressive Integrated (msARI) models. A toolkit in MATLAB was developed for msARI-based identification, estimation, forecasting, and visualization. In the toolkit, an adaptive forecasting routine uses a continual event loop for real-time data acquisition and parameter re-estimation. A statistical quality control algorithm monitors model performance and re-estimates parameters when necessary. A set of visualization tools provide animated graphical representations of forecasts, prediction intervals and key performance metrics. The toolkit was applied to three case studies: electricity demands, water demands, and sewer flows. The analysis of the results demonstrated that the explicit modeling of multi-seasonality improved model predictions. Therefore, the msARI software presents a promising tool for modeling and predicting real-time data series. Highlights: A software package for mulit-seasonal time series data analysis is developed. The time series package includes identification, estimation, forecasting, and visualization. The multi-seasonal functionality has been demonstrated with three environmental data sets.