Multi-source information fusion based fault diagnosis for complex electromechanical equipment considering replacement parts

The research on fault diagnosis based on multi-source information fusion technology aims to comprehensively integrate the diagnostic information of complex mechanical and electrical equipment, providing a scientific and precise decision-making basis for decision-makers. However, in diagnostic practice, issues such as the impact of component replacement, rule combination explosion, and information redundancy have become research difficulties. To address these challenges, this paper innovatively combines equipment mechanisms with expert knowledge to build an optimized model that considers the impact of component replacement based on the traditional Belief Rule Base (BRB-h). Meanwhile, under the framework of traditional independent component analysis, this paper proposes an Independent Component Analysis (ICA) method that considers Expert knowledge (ICA-E). Furthermore, to quantify the impact of component replacement on equipment performance, this paper delves into the transparency and traceability of replacement impact factors and conducts a sensitivity analysis. Through empirical case studies, the advancement and practicability of this new method in the field of fault diagnosis are verified.