|
Abstract
|
PURPOSE The aim of this study was to determine the number of shelters, specify some optimal paths among building blocks towards shelters, and assign population to shelters.
DESIGN/METHODOLOGY/APPROACH Imperialist competition algorithm (ICA) and particle swarm optimization (PSO) were used to optimize the objectives of this study.
FINDINGS The optimal value for PSO objective function was with the number of function evaluations (NFE) of 5300 and the optimal value of ICA objective function was with NFE of 1062. Repetition test for both algorithms showed that imperialist competition algorithm enjoys better stability and constancy and higher speed of convergence compared to particle swarm algorithm. This has been also shown in larger environments. 92% of the existing populations have access to shelters at a distance of less than600 meters. This means that evacuation from the building blocks to shelters takes less than 8 minutes. The average distance from a block (for example, a residential complex) to an optimal shelter is approximately273meters. The greatest risk of route and shelter has been 239 and 121, respectively. Research limitations/implications To address these goals, four following objective functions were considered: a) minimization of the distance for getting all the people to shelters b) the lowest total risk of the discharge path c) minimization of the total time required to transfer people to shelters or hospitals if necessary, and d) the lowest total risk in shelters. Social implications Over the recent decades, the frequency of so-called 'natural' disasters has increased significantly worldwide and resulted in escalating human and economic losses. Among them, the earthquake is one of the major concerns of the various stakeholders related to urban planning.
ORIGINALITY/VALUE In addition, the maximum time of discharge from the helter to the hospital has been 17 minutes, which means the presence of good access to selected shelters.
Language: en |