Research on Optimization Technology for Sequential Diagnostic Strategy Based on Improved Quasi-Depth First Search Algorithm.

Xiaofeng Lv,Deyun Zhou,Fuqiang Li
DOI: https://doi.org/10.1145/3603781.3603902
2023-01-01
Abstract:Aiming at the problem that sequential diagnosis strategy optimization techniques for complex avionics systems are currently difficult to obtain optimal solutions quickly and effectively, this paper proposes an improved Quasi-Depth First Search (QDFS) algorithm based on the Rollout algorithm to optimize the design of sequential diagnosis strategies. The algorithm determines a test in two steps. Firstly, the Rollout algorithm is used to construct a temporary diagnostic tree with the selected test as the vertex, and then the "evaluation function" of the improved QDFS algorithm is used to calculate and compare the constructed temporary diagnostic tree to obtain the optimal diagnostic tree. The first test of the diagnostic tree is the current optimal test. Through a case study of sequential diagnosis strategy optimization, the results show that the improved algorithm can obtain a global optimal solution and generate an optimal diagnosis tree compared to the improved QDFS algorithm.
What problem does this paper attempt to address?