Condition-Based Operation and Maintenance Strategy for Load-Sharing Systems Based on Wiener Process

Abstract

As a distinctive redundant form in various practical applications, load-sharing systems consist of stochastically dependent units bearing system load altogether. Conventional load-sharing systems usually operate under an equal load allocation policy, and the system load is evenly distributed among all working units. However, this static policy neglects the individual dynamic and heterogenous characteristics during unit degradation processes, and leads to nonnegligible individual differences between unit reliability and lifetime distributions. Faced with this problem, this article proposes a novel condition-based operation and maintenance strategy for two-unit load-sharing systems. Each unit undergoes nonmonotonic continuous degradation following the Wiener process, and the system reliability is evaluated by considering a possible two-phase degradation process of the surviving unit once one unit fails. At each periodic inspection time, the system load is dynamically allocated by minimizing the Jensen–Shannon divergence between unit remaining useful lifetime distributions. Furthermore, a condition-based maintenance model is established according to semi-renewal process characteristics, along with specific theoretical analysis for the stationary distribution of system states. Compared with traditional operation and maintenance strategies, the effectiveness of the proposed strategy is validated through numerical experiments, and a practical case study of a two-cell lithium-ion battery pack illustrates robust economic benefit in dynamically adjusting the battery cell loads.