Research on fault diagnosis method for milling robot electric spindle based on generative adversarial networks

Milling robots operating under prolonged high-load conditions are susceptible to failures, which, if not detected promptly, can lead to severe consequences. These consequences manifest in two primary ways. Firstly, such failures pose a significant risk to worker safety. Secondly, they impair the surface quality and assembly precision of machined components, resulting in production interruptions, delays, and downtime, which ultimately undermine profitability and disrupt delivery schedules. To address this challenge, this study proposes a bearing fault diagnosis method based on a Conditional Generative Adversarial Network (CGAN). The approach integrates label information into the generator network to achieve data augmentation. Additionally, a Stacked Convolutional Autoencoder (S-CAE) is employed to extract features from both the generated and original samples, enabling effective fault diagnosis. The proposed method is validated using fault data collected from the bearings of electric spindles in milling robots, demonstrating high accuracy and confirming its feasibility.