Research on fault diagnosis method for milling robot based on spatiotemporal feature fusion

Traditional CNC milling machines, due to their limited degrees of freedom, are unable to meet the demands of complex machining tasks. In contrast, milling robots are capable of multi-dimensional milling. However, the electric spindle, a critical component of the milling robot, is prone to faults under sustained high-load conditions. Additionally, environmental noise in real-world conditions interferes with signal feature extraction, further impacting the diagnostic accuracy of the model. To address this issue, this paper proposes a model that combines Convolutional Neural Networks (CNN) with Bidirectional Long Short-Term Memory Networks (BiLSTM). In this model, CNN is responsible for extracting spatial features from the data, while BiLSTM focuses on processing time-series data to capture bidirectional long-term dependencies. Collecting fault data from the electric spindle bearings of milling robots and comparing it with other models demonstrates that the CNN-BiLSTM model achieves a diagnostic accuracy of up to 95.5%.