Surface Roughness Prediction Method of Milling Based on NCA-SAE
- Resource Type
- Conference
- Authors
- Wang, Jingshu; Chen, Tao; Zhou, Wu; Ma, Jinghua; Chu, Xingrong
- Source
- 2023 5th International Conference on Intelligent Control, Measurement and Signal Processing (ICMSP) Intelligent Control, Measurement and Signal Processing (ICMSP), 2023 5th International Conference on. :1029-1032 May, 2023
- Subject
- Computing and Processing
Robotics and Control Systems
Signal Processing and Analysis
Adaptation models
Redundancy
Milling
Predictive models
Signal processing
Surface roughness
Rough surfaces
surface roughness prediction
feature fusion
neighbourhood component analysis
sparse autoencoders
- Language
Roughness is a critical parameter to characterize the quality of surface processing. To improve the prediction accuracy, a novel NCA-SAE feature fusion method is proposed for surface roughness prediction in milling process. Based on the manufacturing signals collected in milling process, NCA is applied for feature selection to screen out the features that are more sensitive to the surface roughness, while the feature fusion to eliminate redundancy and complementarity between features in time and space is carried out by SAE. The key parameters of NCA-SAE method are optimized by particle swarm optimization (PSO). The milling experiment is conducted to verify the effectiveness of the improved method. The results show that the obtained roughness model based on NCA-SAE obtains the best prediction effect.