A Bayesian Inference Based Operator Skill Evaluation Method for Reliable and Efficient Maintenance of Railway Traction Power Supply Systems

Quantifying the skills of the maintenance operators of railway traction power supply systems (TPSSs) is crucial for developing scientific maintenance strategies, thus improving the maintenance efficiency and achieving reliable operations. In this context, this paper proposes a Bayesian inference based skill evaluation method for TPSS maintenance operators which leverage both expert knowledge and field observations. Firstly, the evaluation index set is constructed by comprehensive analysis of the competency requirements for different maintenance tasks. Then, a multi‐criteria evaluation method integrating expert knowledge and a fuzzy comprehensive evaluation method are presented to calculate the initial skill levels of maintenance operators. Subsequently, the operator skills are re‐evaluated using observed performance during actual maintenance operations, and the resulting observed skill levels are used to update the initial skill level by employing the Bayesian inference method. Finally, a real maintenance scenario is introduced for the case study, which illustrates the skill evaluation process of the operator performing a maintenance task in the TPSS catenary system. In addition, a sensitivity analysis is carried out to investigate the influence of key factors such as index weights, errors in expert scoring and field observations, etc., on the evaluation results of operator skills, which further validates the effectiveness and robustness of the proposed method.