Fault Diagnosis of Wind Turbine Rolling Bearings Based on DCS-EEMD-SSA

Addressing the challenges of non-stationarity, nonlinearity, and noise interference in vibration signals of wind turbine rolling bearings, this paper proposes a fault diagnosis method combining differentiated creative search (DCS), ensemble empirical mode decomposition (EEMD), and singular spectrum analysis (SSA)—termed as DCS-EEMD-SSA. Initially, the DCS algorithm adaptively selects parameters for EEMD to decompose the fault signals. The decomposed signals are then filtered and reconstructed based on criteria such as variance contribution ratio, correlation coefficients, and permutation entropy. Subsequently, DCS adaptively selects parameters for SSA to further decompose the reconstructed signals into multiple subsequences. By analyzing the w-correlation graphs, signals of the same cycle are merged. The merged signals undergo envelope spectrum analysis, based on the highest variance contribution ratio, to diagnose faults in the wind turbine rolling bearings. The effectiveness of the proposed method is demonstrated through analysis of a publicly available rolling bearing dataset from Case Western Reserve University, showing its capability in accurately diagnosing faults in wind turbine rolling bearings.