An optimization method of low orbit constellation satellite fault diagnosis based on improved artificial bee colony algorithm

In this paper, we propose the Dynamic-weight Levy Artificial Bee Colony (DLABC) algorithm for satellite fault diagnosis to address key limitations in processing high-dimensional telemetry data and complex fault modes. This algorithm includes key innovations: a dynamic weighting mechanism that balances global exploration and local exploitation, a Levy flight-inspired search operator for escaping local optima, and a hybrid feature selection framework that combines mutual information pre-screening with ABC optimization. The experimental results show that compared with the traditional swarm intelligence method, the proposed method has excellent performance in feature dimension reduction while ensuring the accuracy of diagnosis. At the same time, the algorithm has strong robustness and fast convergence under noise conditions.