Machining precision is of great significance for ensuring production quality and improving production efficiency. It has always been a research hotspot in the field of machinery. However, existing researches are difficult to adapt to the changes of actual machining state and cannot realize real-time judgment of machining precision. This paper takes gear hobbing precision as the research object, and conducts research from the perspective of actual production. First, a classification prediction model of gear hobbing precision based on capsule neural network is established. The model takes vibration signals as input to evaluate gear hobbing precision in real time. The data validation results show that the prediction accuracy of the model can reach 99.5. Then, the correlation between process parameters and gear hobbing precision is evaluated. Moreover, an iterative adjustment strategy of process parameters is developed to address the unqualified gear hobbing precision.
展开▼
