Objectives  This paper concerns about test set optimization and test sequence optimization for system-level testability analysis.

  Methods  The Discrete Bat Algorithm(DBA)is used to acquire the optimal complete test set and test sequence for each type of fault based on a multi-signal flow graph model. The convergence speed of the algorithm is accelerated while the optimization accuracy is guaranteed. Taking warship combat systems as the research object, such main indexes as fault isolation rate, average test cost and average expected test cost are compared with the existing optimization schemes.

  Results  The simulation results show that the proposed method offers better optimization performance compared with other algorithms, among which the fault isolation rate is increased by 5.5% and the average test cost and expected test cost are decreased by 6.35% and 17.01% respectively.

  Conclusions  The feasibility and advantages of the DBA applied to system-level testability analysis are verified, providing effective guidance for system testability engineering in the design phase.