Multi-Task Roughness and Tool Wear Prediction From Machining Vibration
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current prediction methods for part surface roughness and tool wear can only predict one condition at a time, limiting their ability to learn complex nonlinear mappings and requiring separate models for each, which hampers simultaneous prediction of both.
Innovation Solution
A multi-task learning approach using tri-axial acceleration sensors to collect vibration signals, adding Gaussian white noise, extracting and normalizing features, and constructing an improved deep belief network (DBN) to predict both part surface roughness and tool wear condition in real-time.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional intelligent algorithms and machine learning algorithms with shallow structures are used, then the network can learn the mapping relationship between cutting parameters and surface roughness/tool wear, but the ability to learn complex nonlinear mapping relationships is limited
Solution Approach 1:
The patent transitions from shallow network structures to deep network structures, adding more layers to the network architecture. This dimensional change in network depth enables the model to learn complex nonlinear mapping relationships between cutting parameters and surface roughness/tool wear more effectively, resolving the limitation of shallow networks while maintaining manageable complexity through systematic design.
Solution Approach 2:
The patent combines multiple algorithmic components (genetic algorithm, convolutional neural network, and other processing modules) to create a hybrid intelligent algorithm system. This composite approach integrates the strengths of different algorithms to achieve high prediction accuracy for complex nonlinear relationships without requiring excessively complex individual network structures.
2Reliability
If separate prediction models are established for surface roughness and tool wear, then each condition can be predicted individually, but the workload and cost of model establishment increase
Solution Approach 1:
The patent merges the prediction of surface roughness and tool wear into a single unified intelligent algorithm system. The model processes cutting parameters and dynamically predicts both surface roughness and tool wear conditions simultaneously, eliminating the need to establish and maintain separate prediction models for each condition, thus reducing workload and establishment time while maintaining prediction reliability.
3Measurement precision
If separate prediction models are established for surface roughness and tool wear, then each prediction can be optimized independently, but the cost of model establishment increases
Solution Approach 1:
The patent designs a universal intelligent algorithm system that performs multiple prediction functions (surface roughness and tool wear) within a single model framework. This multi-functional approach allows the system to maintain high prediction precision for both conditions while avoiding the need to develop, train, and deploy multiple separate specialized models, thereby reducing overall system complexity and establishment cost.
Data Source
AI summary
A prediction method of part surface roughness and tool wear based on multi-task learning belong to the file of machining technology. Firstly, the vibration signals in the machining process are collected; next, the part surface roughness and tool wear are measured, and the measured results are corresponding to the vibration signals respectively; secondly, the samples are expanded, the features are extracted and normalized; then, a multi-task prediction model based on deep belief networks (DBN) is constructed, and the part surface roughness and tool wear are taken as the output of the model, and the features are extracted as the input to establish the multi-task DBN prediction model; finally, the vibration signals are input into the multi-task prediction model to predict the surface roughness and tool wear.