摘要：Background: Traditional system fault diagnosis methods have low robustness. When it diagnoses system, the results are not accuracy. Materials and Methods: In this study, therefore, it proposes an improved discrete bat algorithm used for system fault diagnosis. The new algorithm includes three steps. Firstly, according to the actual meaning of system fault diagnosis, it adopts binary encoding to classify bat individuals. Secondly, it improves the fitness with a constraint equation. Thirdly, it applies a inertia coefficient into bat speed updating equation. Then it uses this new method to diagnose the system fault. Results: Finally, experiments show that the new algorithm can reduce the calculation difficulty, improve the diagnosis convergence speed and has higher diagnostic accuracy. Conclusion: The system fault diagnosis based on improved discrete bat algorithm is an effective method, which can effectively improve the accuracy of system fault diagnosis.
