Hybrid multiple attribute group decision-making for power system restoration. Issue 19 (1st November 2015)

• Record Type: Journal Article
• Authors: Sun, Pengbo; Liu, Yutian; Qiu, Xizhao; Wang, Liang
• Is Part Of: Expert systems with applications. Volume 42:Issue 19(2015)
• Page Start: 6795
• Page End: 6805
• Publication Date: 2015-11-01
• Subjects: Backbone-network reconfiguration -- Hybrid multiple attribute decision-making -- Power system restoration -- Sensitivity analysis; Expert systems (Computer science) -- Periodicals; Systèmes experts (Informatique) -- Périodiques; Electronic journals; 006.33
• DOI: 10.1016/j.eswa.2015.05.001
• Languages: English
• ISSNs: 0957-4174

Abstract

Highlights: Hybrid MADM method for backbone-network reconfiguration in restoration. Mathematical programming for combining different forms of preference relations. Extend VIKOR for crisp data, fuzzy numbers, interval numbers and linguistic terms. Sensitivity analysis on weights of attributes for guiding the decision-making. Abstract: Due to deregulated power industry, distributed power generation, aging infrastructure and many other factors, the modern society is exposed to higher blackout risks. Decision-making optimization is indispensable for ensuring fast and secure power supply restoration to end users. As an important stage of power system restoration, backbone-network reconfiguration is necessary to re-establish the skeleton network and restore loads. Backbone-network reconfiguration after a large-scale outage is influenced by many factors about system safety and restoration speed. In order to evaluate candidate restoration schemes, multiple types of attributes including crisp data, fuzzy numbers, interval numbers and linguistic terms are employed. An extended VIKOR method is proposed to provide compromise solutions considering hybrid attributes. The method can reflect the vagueness and uncertainties in practical restoration problems, and avoid too much fuzzification. Different forms of preference relations and maximum deviation model are integrated by the minimum relative entropy to determine combined weights of attributes. Sensitivity analysis on the weights provides efficient guidelines for decision makers. Finally, an actual power system demonstrates the basic features of the developed method. It is more reasonable and creditable to consider multiple types of information and unsatisfactory attributes in decision-making.