Multitarget Assignment Under Uncertain Information Through Decision Support Systems

• Published in: IEEE transactions on industrial informatics 2024-08, Vol.20 (8), p.10636-10646
• Main Authors: Jia, Qianlei, Xiao, Jiaping, Feroskhan, Mir
• Format: Article
• Language: English
• Subjects: Decision support system; Decision support systems; Fire fighting; Fire prevention; Fuzzy sets; Heuristic algorithms; Informatics; Linguistics; multitarget assignment; Optimization models; Polar coordinates; Probability distribution; Transforms; Uncertainty; Unmanned aerial vehicles; unmanned aerial vehicles (UAVs); Z-numbers
• ISSN: 1551-3203; 1941-0050
• DOI: 10.1109/TII.2024.3397392

Unmanned aerial vehicles (UAVs) play an important role in advancing fire prevention technology. However, existing research usually assumes that the firefighting equipment carried by UAVs, such as fire extinguishing bombs and water, can significantly satisfy actual requirements. However, in many practical scenarios, the firefighting resources available on UAVs are limited, necessitating an assessment and prioritization of affected areas. This article proposes a group decision-making framework for uncertain environments, employing Z-numbers and q-rung orthopair fuzzy sets to address this challenge. Specifically, a polar coordinate system is employed to transform the parameters in Z-numbers into the corresponding membership and nonmembership. Meanwhile, the potential probability distribution of Z-numbers is calculated based on an optimization model. To combine multiple sets of uncertain information into a final overall assessment, we propose an aggregation operator and provide strict proof using mathematical induction. Furthermore, a new weight calculation method is introduced to determine the weights of multiple pieces of information based on the potential probability distribution of Z-numbers and an improved golden rule representative value. Based on the sigmoid function and generalized knowledge measure, a novel score function is defined to rank different Z-information. In addition, a distance measure between Z-information is proposed and rigorously proved. Finally, the proposed method is validated through practical flight tests. The results and comparative analysis illustrate that our proposed method effectively mitigates other existing approaches' computational and informational loss limitations.