Multi-objective Optimization of Emergency Evacuation Using Improved Genetic Algorithm

孟永昌,杨赛霓,史培军
DOI: https://doi.org/10.13203/j.whugis20120584
2014-01-01
Abstract:Based on the actual demands of emergency evacuation,this paper establishes a multi-objective optimization model which takes the flow of each path as a control variable.Maximum flow,minimum cost,and maximum reliability are considered as objectives to integrate the timeliness,economy and security of emergency evacuation.An improved Pareto multiple objective genetic algorithm is proposed,to encode the control variables directly.It introduces a fitness function based on the degree of Pareto domination and self-adaption punishment,and designs a selection operator based on tournament and niche technology.The algorithm provides a practical tool to solve the problem with complex constraints and multiple objectives.Finally,a real world road network is used for simulation and analyses,which validates the effectiveness and applicability of the proposed methodology.
What problem does this paper attempt to address?