A Gaussian Type-2 Fuzzy Programming Approach for Multicrowd Congestion-Relieved Evacuation Planning

Large-scale emergencies occur frequently around the world, causing serious casualties. Emergency evacuation is one of the top priorities after a disaster. Due to the uncertain and complex evacuation processes, evacuation plans obtained in a deterministic context may not meet the requirements of practical engineering applications. This research considers a multi-crowd congestion-relieved evacuation problem in Gaussian Type-2 fuzzy environments. A path-based one destination network flow (P-ODNF) model is developed for the problem formulation. To relieve the traffic congestion, a multi-point diversion evacuation strategy is employed in the model's construction. The uncertainties associated with the evacuation process are expressed as Gaussian Type-2 fuzzy variables. A critical value-based defuzzification technique is adopted to handle the Type-2 fuzziness. Based on the credibility measure, the uncertain P-ODNF model is transformed into its deterministic counterpart. An efficient adaptive chaos particle swarm optimization algorithm (A-CPSO) is designed for model solving. Several numerical experiments are performed to demonstrate the proposed methodology. A sensitivity analysis is performed to illustrate the logical correctness of the proposed defuzzification process. The computational results show that A-CPSO is competitive in both its effectiveness and efficiency. The proposed methodology can coordinate multiple simultaneous evacuation processes to relieve congestion and improves the overall evacuation efficiency.

Automated summary

This research addresses the problem of multi-crowd congestion-relieved evacuation in Gaussian Type-2 fuzzy environments. It proposes a path-based one destination network flow model with a multi-point diversion evacuation strategy to alleviate traffic congestion. Uncertainties in the evacuation process are handled using Gaussian Type-2 fuzzy variables and a critical value-based defuzzification technique. An efficient adaptive chaos particle swarm optimization algorithm is designed for model solving. The proposed methodology improves overall evacuation efficiency by coordinating multiple simultaneous evacuation processes.