Risk Assessment and Safety Barrier Importance Measure for Multi-State Electromechanical Systems Considering Competing Failure Mechanisms

Electromechanical systems comprise multiple interconnected components that are susceptible to progressive degradation and competing failure modes driven by various underlying physical mechanisms during prolonged operation. Most existing risk assessment studies for such systems focus on binary-state assumptions and single failure mode scenarios, overlooking the importance of multi-state degradation characterization and the competing interactions among these modes. This paper proposes a risk assessment and safety barrier importance evaluation method for multi-state electromechanical systems that quantifies system risk and barrier importance by incorporating competing failure modes and multi-state degradation processes. A multi-state Bow-tie modeling framework based on semi-Markov process theory is first established to capture the evolutionary behavior of system components under dynamically varying degradation conditions through state transition mechanisms. A competing failure semi-Markov model (CF-SMM) is then constructed by integrating these multiple failure modes, thereby overcoming the limitations of conventional exponential distribution assumptions and single failure mode assumptions. Subsequently, a safety barrier importance measure (SBIM) is developed by fusing risk sensitivity, structural relevance, and state degradation degree to maximize barrier discrimination capability. Finally, the proposed method is validated through a case study of a five-axis machining center as a representative electromechanical system. The results demonstrate that the monotonicity index of the SBIM method reaches 1.00 at t=8760  h, representing improvements of 8.7 % and 13.6 % over the Birnbaum importance and Fussell-Vesely importance measures, respectively. The proposed method enhances both risk assessment accuracy and barrier importance discrimination for electromechanical systems, providing methodological support for maintenance resource optimization and operational safety assurance in dynamic degradation environments.

Keywords: Electromechanical systems, competing failures, semi-Markov processes, bow-tie models, safety-barrier importance.