Optimal design of distributed energy resource systems under large-scale uncertainties in energy demands based on decision-making theory

• Main Authors: Yang Yun, Li Da, Zhang Shi-Jie, Xiao Yun-Han
• Format: Article
• Language: English
• Published: VINCA Institute of Nuclear Sciences 2019-01-01
• Series: Thermal Science
• Subjects: distributed energy resource; mixed integer linear programming; optimal design; optimal operation; uncertain decision-making; uncertainty
• Online Access: http://www.doiserbia.nb.rs/img/doi/0354-9836/2019/0354-98361800199Y.pdf

This study focuses on the optimal design of distributed energy resource systems with consideration of large-scale uncertainty of energy demands based on decision- making theory. Five integrated modeling and optimization frameworks are developed through the combined use of mixed integer linear programming and uncertainty decision-making criteria (including optimistic criterion, pessimistic criterion, Hurwicz criterion, Laplace criterion, and minimax regret criterion). Superstructure-based mixed integer linear programming models are used for the optimal design and optimal operation of the system where the objective function is to minimize the annual cost. The uncertainty of energy demands is represented by assuming a set of possible scenarios. The proposed methods are applied to the planning of a distributed energy resource system for a hotel in city of Guangzhou, China and their validity and effectiveness are verified. Results show that each method has its specific feature. Optimistic method is risky and recommends a relative small-scale system, while pessimistic method is conservative presenting a relative large-scale system. Hurwicz method is with great subjectivity, making different decisions at different values of optimism coefficient. Both Laplace method and minimax regret method identify a moderate-scale system as the best alternative. Sensitivity analyses on the energy demand scenarios are conducted and results show that the five methods have high sensitivity to the choice of scenarios.