International Journal of Innovative Technology and Exploring Engineering · 2019 · 13 citations · 0 references
ComputingData ScienceDemand ForecastingForecastingTime Series AnalysisEnergy Demand ForecastingLoad ForecastingTime Series AnalyticsTime Series ForecastingTime Series ModelingTime Series PredictionAccuracyArima ModelsData-driven Energy ModelingDynamic Forecasting ModelsDynamic ModelsElectricity Consumption ForecastingEnergy Use PredictionNonstationary Time Series Forecasting
Healthcare buildings have an immense demand for electricity, because of which they exhibit distinctive ability for forecasting of electricity consumption. In this work, ability for forecasting electricity consumption of a large healthcare building was researched. Both the techniques change non-stationary data into stationary data to make an effective and simple data representation and removing of noise subspaces. The comparison of experimental results is done among the SARIMA and ARIMA models. Analysis of the results concludes that performance of SARIMA is better when compared to ARIMA model. The analysis of data from 11 years in the hospital demonstrates that these dynamic models are sufficiently adaptable to forecast the electricity consumption at required accuracy levels.