A multi-objective improved novel discrete particle swarm optimization for emergency resource center location problem. (May 2022)

• Record Type: Journal Article
• Authors: Peng, Dajiang; Ye, Chunming; Wan, Mengran
• Is Part Of: Engineering applications of artificial intelligence. Volume 111(2022)
• Publication Date: 2022-05
• Subjects: Emergency management -- Panic perception -- Multi-objective metaheuristic -- Solution memory -- Random crossover; Engineering -- Data processing -- Periodicals; Artificial intelligence -- Periodicals; Expert systems (Computer science) -- Periodicals; Ingénierie -- Informatique -- Périodiques; Intelligence artificielle -- Périodiques; Systèmes experts (Informatique) -- Périodiques; Artificial intelligence; Engineering -- Data processing; Expert systems (Computer science); Periodicals; 620.00285
• DOI: 10.1016/j.engappai.2022.104725
• Languages: English
• ISSNs: 0952-1976

Abstract

Abstract: The location of emergency resource centers has always been a great challenge in emergency management, which directly influences the recovery speed of disaster areas and the credibility of the government. Although most studies propose mathematical models based on economic indicators and solve them with various algorithms, few of them consider the panic perceived by the victims, and several algorithms are designed from the characteristics of the problem. To address these issues, we develop an emergency resource center location model (ERCLM) with objectives of the panic perception and the total weighted distance, and then design a multi-objective improved novel discrete particle swarm optimization (MOINDPSO) for it. The main idea of the proposed algorithm is three-fold: 1) The representation of the solution is modified and the search operator is improved by a new mechanism named solution memory. 2) The solutions are evaluated by the idea of fuzzy correlation entropy analysis (FCEA), which is efficient in the selection of better solutions. 3) An external archive is used to store the nondominated solutions, and a mechanism for choosing leaders from the external archive along with a mechanism to enhance the stability for the obtained solutions is put forward. Eventually, numeric experiments of comparison algorithms and ablation experiments of two designed mechanisms are carried out on 26 data sets to show the effectiveness and the universality of the proposed algorithm. Highlights: A multi-objective emergency resource center location model (ERCLM) is proposed. An improved novel discrete particle swarm optimization (INDPSO) is presented. Several new mechanisms are designed to propose a multi-objective INDPSO (MOINDPSO). Numeric experiments are carried out to validate the effectiveness of MOINDPSO.