Risk assessment of ship components under component-function interdependency using evidential weighted influence nonlinear gauge system

In complex shipboard systems, degradation of safety-critical components can affect multiple technical functions through shared dependencies. This study develops an evidential weighted influence nonlinear gauge system (WINGS) framework for risk assessment of interdependent ship components. Fifteen physical components and five ship-level safety functions are represented as nodes in a coupled component-function network. Five experts provide criterion-level basic belief assignments (BBAs) for intrinsic risk and directed influence. Professional-background information and inter-expert BBA similarity are combined to derive evidence-source reliability coefficients, which are used to discount the assessments before fusion by Dempster's rule. The fused BBAs are transformed into expected scores, with intrinsic-risk scores placed on the diagonal of the WINGS matrix and directed-influence scores placed on the off-diagonal entries. Total involvement, net-influence role, relative closeness, residual intrinsic-risk uncertainty, and model-based risk-control classes are then obtained. In the cargo-ship case, the main switchboard has the highest calculated priority, followed by the diesel generator, emergency generator, fuel oil supply system, and cooling water pump. Electric power, fire and emergency, and navigation and communication functions occupy the most involved net-receiving positions. The proposed method provides a novel and effective approach for ship component risk assessment considering component-function interdependency.