An Efficient Configuration for Energy Hub to Peak Reduction Considering Demand Response Using Metaheuristic Automatic Data Clustering

Background and Objectives: Different energy demand calls the need for utilizing Energy Hub Systems (EHS), but the economic dispatch issue has become complicated due to uncertainty in demand. So, scenario generation and reduction techniques are used to considering the uncertainty of the EH demand. Dependent on the amount of fuel used, each system has various generation costs. Configuration selection stands as a challenging dilemma in the EHS designing besides economic problems. In this paper, the optimal EHS operation along with configuration issue is tackled. Methods: To do so, two EHS types are investigated to evaluate the configuration effect besides energy prices simultaneously change. Typically, the effect of the Demand Response (DR) feature is rarely considered in EHSs management which is considered in this paper. Also, Metaheuristic Automatic Data Clustering (MADC) is used to reduce the decision-making problem dimension instead of using human decision-makers in the subject of cluster center numbers and considering uncertainty. The Shannon s Entropy and the TOPSIS methods are also used in the decision-making. The study is carried out in MATLAB© and GAMS©. Results: In addition to minimizing the computational burden, the proposed EHS not only serves an enhancement in benefit by reducing the cost but also provides a semi-flat load curve in peak period by employing the Emergency Demand Response Program (EDRP) and Time of Use (TOU). Conclusion: The results show that significant computational burden reduction is possible in the field of demand data by using the automatic clustering method without human interference. In addition to the proposed configuration s results betterment, the approach demonstrated EH s configuration effect could consider as important as other features in the presence of DRPs for reaching the desires of EHs customers which is rarely considered. Also, Shannon s Entropy and the TOPSIS methods integration could select the best DRP scenario without human interference. The results of this study are encouraging and warrant further analysis and researches.