Assessing Tolerance-Based Robust Short-Term Load Forecasting in Buildings

Abstract

Short-term load forecasting (STLF) in buildings differs from its broader counterpart in that the load to be predicted does not seem to be stationary, seasonal and regular but, on the contrary, it may be subject to sudden changes and variations on its consumption behaviour. Classical STLF methods do not react fast enough to these perturbations (i.e., they are not robust) and the literature on building STLF has not yet explored this area. Hereby, we evaluate a well-known post-processing method (Learning Window Reinitialization) applied to two broadly-used STLF algorithms (Autoregressive Model and Support Vector Machines) in buildings to check their adaptability and robustness. We have tested the proposed method with real-world data and our results state that this methodology is especially suited for buildings with non-regular consumption profiles, as classical STLF methods are enough to model regular-profiled ones.

A review on artificial intelligence based load demand forecasting techniques for smart grid and buildings

“…However, in order to achieve the higher forecast results, there is need to accommodate all factors affecting on load demand as forecast model inputs such as; historical load and respective weather data. In this modern era of technology, an accurate load forecast plays a vital role to implement the concept of smart grids and smart buildings [4].…”

“…The result of proposed intelligent technique is reduction of energy losses by limiting the number of switching operations. 4 ANN and SVM based STLF model for smart grid [25] Artificial neural network and Support vector machine based short term forecast model is designed for multiple loads. The objective to this research is to forecast the demand response in smart grids.…”

More Buildings Make More Generalizable Models—Benchmarking Prediction Methods on Open Electrical Meter Data

“…Prediction is often made both on the short-term (hours or days ahead) or long-term (weeks, months or years ahead). Short-term prediction is generally used for real-time HVAC control and efficiency of upcoming hours [2,3], scheduling and management of power stations and demand response schemes [4][5][6] and the analysis of residential metering and sub-metering [7], in addition to many other applications. Long-term prediction is used for the evaluation of energy conservation measures through a baseline model generation [8] and capacity expansion and planning.…”

On the influence of surrounding load demand to improve primary substation STLF

“…Therefore, the operators have a huge interest in improving the forecast accuracy (see [2] for example), and this same interest explains the remarkable effort performed by the scientific community to push the borders of research on STLF further. The focus from the prediction ranges from country loads (as in [3], [4] and [5]), to buildings (see [6] for a detailed state of the art account in this field).…”