Publication: Heuristic algorithms for surveyor standby location planning with multiple plans
Issued Date
2020-01-01
Resource Type
ISSN
19984464
Other identifier(s)
2-s2.0-85099782118
Rights
Mahidol University
Rights Holder(s)
SCOPUS
Bibliographic Citation
International Journal of Circuits, Systems and Signal Processing. Vol.14, (2020), 1154-1161
Suggested Citation
Rawee Suwandechochai, Wasin Padungwech Heuristic algorithms for surveyor standby location planning with multiple plans. International Journal of Circuits, Systems and Signal Processing. Vol.14, (2020), 1154-1161. doi:10.46300/9106.2020.14.143 Retrieved from: https://repository.li.mahidol.ac.th/handle/20.500.14594/60919
Research Projects
Organizational Units
Authors
Journal Issue
Thesis
Title
Heuristic algorithms for surveyor standby location planning with multiple plans
Author(s)
Other Contributor(s)
Abstract
© 2020, North Atlantic University Union NAUN. All rights reserved. This paper concerns a location problem which arises from an auto insurance company which aims to meet customers’ satisfaction by sending their surveyors to accident locations as promptly as possible. The goal of the problem is to find a suitable set of locations at which the surveyors stand by and to determine coverage area for each standby location by using heuristic algorithms. One challenge of this problem is that the number and locations of accidents can vary from time to time and may not be evenly distributed over a given time horizon. In this paper, a computational study is conducted to make a comparison between standby location planning strategies that involve one and two plans. For the strategies with two plans, several strategies for deciding when to switch between the plans are investigated. Experimental studies suggest that using multiple standby location plan throughout the time horizon can improve the efficiency of surveyors in terms of their total distance. Moreover, compared with usual rules of splitting dates such as weekdays versus weekends, the total distance can be reduced by switching between the location plans according to classification of dates by k-means clustering based on historical data of accident frequencies on each day.