An improved fusion crossover genetic algorithm for a time-weighted maximal covering location problem for sensor siting under satellite-borne monitoring. (March 2020)

• Record Type: Journal Article
• Authors: Wang, Ke; Gong, Yue; Peng, Yuling; Hu, Chuli; Chen, Nengcheng
• Is Part Of: Computers & geosciences. Volume 136(2020)
• Publication Date: 2020-03
• Subjects: Genetic algorithm -- Partial coverage -- Spatio-temporal fusion crossover -- Time-weighted maximum coverage problem -- Precipitation station; Environmental policy -- Periodicals; 550.5
• DOI: 10.1016/j.cageo.2020.104406
• Languages: English
• ISSNs: 0098-3004

Abstract

Abstract: Traditional location problems usually focus on spatial and temporal impacts of facilities, but few studies have focused on sensor siting under satellite-borne monitoring in a space-ground integrated sensor network. Given the partial coverage and the requirement for continuous coverage in space and time, a time-weighted maximal covering location problem with partial coverage (TMCLP-PC) model is introduced. This model is solved using a modified genetic algorithm (GA)-based approach that utilizes the spatio-temporal characteristics of potential facility sites for faster convergence. The performance of the new GA is tested on a precipitation station siting problem in the Jinsha River Basin on the upper reaches of the Yangtze River in China. The results demonstrate that the proposed GA can significantly reduce the computation time compared with CPLEX, a commercial integer programming solver, and can outperform the greedy algorithm and the GAs with one-point, two-point, fusion, and uniform crossover operators. The applicability of the proposed method and exploration of the design in the new GA are also discussed. Highlights: The problem is to site stationary sensors under time-varying satellite monitoring. It is a time-weighted maximal coverage problem for which a new GA is proposed. The algorithm considers the importance of each site in crossover and mutation. The proposed method is tested on siting rain gauges in the Jinsha River Basin. The method can be extended to serve other spatio-temporal location problems.