A multi-criteria decision analysis-based approach for dispatch of electric microgrids. (June 2017)

• Record Type: Journal Article
• Authors: Panwar, Mayank; Suryanarayanan, Siddharth; Hovsapian, Rob
• Is Part Of: International journal of electrical power & energy systems. Volume 88(2017)
• Page Start: 99
• Page End: 107
• Publication Date: 2017-06
• Subjects: Discrete compromise programming -- Dispatch -- Multi-criteria decision analysis -- Microgrid -- Multi-objective optimization -- Peak load reduction; Electrical engineering -- Periodicals; Electric power systems -- Periodicals; Électrotechnique -- Périodiques; Réseaux électriques (Énergie) -- Périodiques; Electric power systems; Electrical engineering; Periodicals; 621.3
• DOI: 10.1016/j.ijepes.2016.12.018
• Languages: English
• ISSNs: 0142-0615

Abstract

Highlights: Discrete compromise programming used to rank alternatives based on preference of objectives. DCP is used to select non-dominated solutions without generating a Pareto front. Operating cost, emissions, and peak load reduction are considered as the objectives. Uncertainty in renewables considered by scenario-based sampling of distributions. Abstract: This paper presents a decision support system (DSS) for multi-objective dispatch of an electric microgrid considering cost of operation, peak load reduction, and emissions. Discrete compromise programming (DCP) is used as the multi-criteria decision analysis (MCDA) technique for providing the decision support to the distribution system operator (DSO) by ranking various alternatives based on preference of the objectives. The focus of this paper is the application of DCP-based MCDA to select feasible dispatch solutions closest to the preference of the DSO when multiple objectives are considered in the microgrid dispatch. This technique obtains non-dominated solutions in case of conflicting objectives without generating a Pareto front, and hence avoiding prohibitive computational cost. DCP can be used for MCDA in both exact and metaheuristic dispatch algorithms. Uncertainty in renewable energy forecasting and load demand is included through a scenario-based approach by sampling empirical distributions. The dispatch algorithm considers a two hour look-ahead time horizon. Simulations are performed on a notional electric microgrid with diesel generator, solar photovoltaic, and energy storage. The microgrid is based on the IEEE 13-node test feeder and the results are presented.