Research on Dynamic Force Formation in Aerospace Search and Rescue Missions Based on an Improved Contract Net Algorithm

2026-99-2310

10/30/2026

Authors
Abstract
Content
Aerospace search and rescue plays a crucial role in ensuring the safety of manned space missions. As China’s manned space program progresses toward normalization and higher-frequency, flight-like missions, the conventional static formation mode of search and rescue forces may no longer be adequate. To achieve the “high frequency, high parallelism, and multi-type” goals in future missions, we propose a more efficient contract net algorithm that incorporates hierarchical bidding and joint evaluation to dynamically organize and optimize distributed tasks. The force formation challenge is modeled as a distributed dynamic programming problem with strong spatiotemporal constraints, and a task-capability-context matching is developed to overcome the adaptability limitations of static formation modes in flight-type missions. To enhance the contract net, we introduce a Level 1 bidding algorithm based on efficacy objectives, which builds a “timeliness-loss-reliability” objective function using these three indicators. Then, we introduce a Level 2 bidding algorithm focused on equilibrium objectives by employing “rigid locking for high-priority tasks and dynamic optimization for secondary tasks” to alleviate resource load imbalance. Next, we implement a ”joint bidding” mechanism, which allows resources to form alliances and bid collaboratively. Simulation results demonstrate that the improved algorithm accelerates task formation and provides a systematic, scientific solution for aerospace search and recovery force formation.
Meta TagsDetails
DOI
https://doi.org/10.4271/2026-99-2310
Citation
Lv, N., Liu, R., Li, D., and Liu, X., "Research on Dynamic Force Formation in Aerospace Search and Rescue Missions Based on an Improved Contract Net Algorithm," 2026 2nd International Conference on Aerospace Engineering and Materials Technology, Xi'An, China, April 10, 2026, https://doi.org/10.4271/2026-99-2310.
Additional Details
Publisher
Published
Oct 30
Product Code
2026-99-2310
Content Type
Technical Paper
Language
English