# A Rolling Bearing Fault Diagnosis Method Based on Wild Horse Optimizer-Enhanced VMD and Improved GoogLeNet

## TL;DR

This paper introduces a new method for diagnosing rolling bearing faults using optimized signal decomposition and an improved neural network, achieving high accuracy even in noisy conditions.

## Contribution

The novel contribution is combining an improved wild horse optimizer with enhanced VMD and a modified GoogLeNet for effective fault diagnosis in rolling bearings.

## Key findings

- The proposed method achieved 99.17% accuracy across four fault types in rolling bearings.
- It maintained over 95.80% accuracy under noisy conditions, demonstrating robustness.
- The modified GoogLeNet with TReLU and cascaded kernels improved adaptability and convergence.

## Abstract

To address the challenges of weak fault features and strong non-stationarity in early-stage vibration signals, this study proposes a novel fault diagnosis method combining enhanced variational mode decomposition (VMD) with a structurally improved GoogLeNet. Specifically, an improved wild horse optimizer (IWHO) with tent chaotic mapping is employed to automatically optimize critical VMD parameters, including the number of modes K and the penalty factor α, enabling precise decomposition of non-stationary signals to extract weak fault features. The vibration signal is decomposed, and the top five intrinsic mode functions (IMFs) are selected based on the kurtosis criterion. Time–frequency features are then extracted from these IMFs and input into a modified GoogLeNet classifier. The GoogLeNet structure is improved by replacing standard n × n convolution kernels with cascaded 1 × n and n × 1 kernels, and by substituting the ReLU activation function with a parameterized TReLU function to enhance adaptability and convergence. Experimental results on two public rolling bearing datasets demonstrate that the proposed method effectively handles non-stationary signals, achieving 99.17% accuracy across four fault types and maintaining over 95.80% accuracy under noisy conditions.