Abstract
Proper emergency evacuation planning is a key to ensuring the safety and efficiency of resources allocation in disaster events. An efficient evacuation plan can save human lives and avoid other effects of disasters. To develop effective evacuation plans, this study proposed a multi-objective optimization model that assigns individuals to emergency shelters through safe evacuation routes during the available periods. The main objective of the proposed model is to minimize the total travel distance of individuals leaving evacuation zones to shelters, minimize the risk on evacuation routes and minimize the overload of shelters. The experimental results show that the Discrete Multi-Objective Cuckoo Search (DMOCS) has better and consistent performance as compared to the standard Multi-Objective Cuckoo Search (MOCS) in most cases in terms of execution time; however, the performance of MOCS is still within acceptable ranges. Metrics and measures such as hypervolume indicator, convergence evaluation and parameter tuning have been applied to evaluate the quality of Pareto front and the performance of the proposed algorithm. The results showed that the DMOCS has better performance than the standard MOCS.
Highlights
Many real-world problems, such as the evacuation planning in urban flooding, are modeled considering several factors such as evacuation time, shelter capacity, number of the individual at risk, and the distance between the risk zones and shelters
The Discrete Multi-Objective Cuckoo Search (DMOCS) algorithm demonstrated a good tradeoff between the solution quality and the computational time
According to Bish [12], this process of evacuation planning consists of three phases: (1) determination of the safe areas, (2) selecting the optimum path between risk zones and safe areas, and (3) joining risk zones associated with each safe area
Summary
Many real-world problems, such as the evacuation planning in urban flooding, are modeled considering several factors such as evacuation time, shelter capacity, number of the individual at risk, and the distance between the risk zones and shelters. Real-world problems may involve more than two conflicting objectives, and scalability tests of evolutionary algorithms have shown some problems related to convergence, diversity, and computation time with more than three objective functions In these cases, depending on the problem at hand, a multi-objective optimization approach might be required, and this is a relatively new research area [4], important advances have been made in recent years. To solve the problem of shelter location-allocation planning and the risk management on evacuation routes, we have identified the candidate shelters and their corresponding capacities, the origin points and their respective demands, and for each pair of shelter and demand points we have determined the five shortest paths and their corresponding risks After this data preparation, we can set up the parameters of the DMOCS whose output is the Pareto optimal set.
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
