Optimal condition-based maintenance strategy for a three-state system under dependent competing failures

Industrial systems function in dynamic settings characterized by changing conditions. This manuscript develops a reliability model and a condition-based maintenance approach for a single-component system with three states (normal, defective, and failed). The system is prone to failures resulting from both internal degradation and external shocks, and these shocks can cause abrupt rises in degradation. Internal degradation is simulated through a Wiener process that features elevated drift and diffusion coefficients when the system is in the defective state. External shocks follow a non-homogeneous Poisson process and are classified as safe, damaging, and fatal shocks. A preventive degradation threshold is defined according to the state of the system, and particle swarm optimization is utilized to determine the inspection interval and the optimal preventive threshold. Finally, the proposed maintenance model is verified using a numerical example.