Assessing the Impacts of Real-Time Price Prediction Quality on Demand Response Management for Sustainable Smart Manufacturing

Abstract

The emerging smart manufacturing technologies pave the way for flexible and autonomous monitoring and control of complex manufacturing systems, which facilitate the implementation of real-time price (RTP) based demand response management towards sustainability. The demand response management requires scheduling of smart manufacturing systems in advance, and thus the quality of RTP predictions directly impacts the performance of demand response. Although several prediction evaluation metrics are currently available, they are designed to show the similarities between prediction and actual RTP, which are not necessarily related to demand response performance. Therefore, in this study, the daily energy cost reductions obtained by solving a demand response management problem are adopted as an indicator of demand response performance. Six commonly used evaluation metrics are examined, and their correlations with energy cost reductions are investigated. In addition, a new metric called k-peak distance considering the characteristics of the demand response problem is proposed and compared with the other six metrics. The case studies show that the proposed metric has two to four times higher correlation with energy cost reductions and only about half of the standard error compared to other metrics. The results indicate that the proposed metric can better represent the prediction quality in the demand response problem.