Variable Speed Gear Fault Identification Based on ATVCF and IAOA-SDAE

To enhance the gear fault identification accuracy under variable speed operating conditions, a variable speed gear fault identification method was proposed based on adaptive time-varying comb filtering (ATVCF) and improved arithmetic optimization algorithm (IAOA) for optimizing stack denoising auto encoder (SDAE). Aiming at the noise reduction of vibration signals of variable speed gear, the ATVCF was used to preprocess the data so as to filter out noise components and retain the effective signals. Aiming at the shortcomings of AOA algorithm in global search and local development, cosine regulation factor was introduced for improving the mathematical optimizer acceleration (MOA) function so as to increase the global search ability and local development full optimization ability. Random opposition-based learning (ROBL) was introduced to increase the population diversity of the algorithm, enhancing search ability. In addition, the parameters of IAOA-SDAE model were optimized to ensure fault identification accuracy and stability of the model. The analysis of the vibration test data of the variable speed gears verifies the effectiveness and superiority of the proposed method in the intelligent identification of the variable speed gear fault.