CNN Contact Spot Distribution Estimation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing numerical analysis methods for estimating contact spot distribution between rough contact surfaces are computationally intensive and time-consuming, making them inefficient for accurately predicting contact characteristics such as friction, electrical resistance, and wear.
Innovation Solution
A deep learning model based on a convolutional neural network (CNN) is used to quickly estimate contact spot distribution and pressure distribution from contact surface images, incorporating pressure distribution data and surface characteristics to improve accuracy and reduce calculation time.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a numerical analysis method based on contact dynamics is used to estimate contact spot distribution, then measurement precision is improved, but loss of time increases
Solution Approach 1:
The patent applies preliminary action by pre-training a deep learning model offline using numerical analysis results. The model learns the mapping between surface topography and contact distribution in advance, so that during actual use, contact spot distribution can be estimated quickly from surface images without performing time-consuming numerical calculations again.
Solution Approach 2:
The patent uses copying by creating a deep learning model that replicates the behavior of the numerical analysis method. Instead of directly running complex numerical simulations, the trained neural network model copies the estimation capability, providing similar accuracy with much faster computation speed for real-time or repeated assessments.
2Manufacturing precision
If a numerical analysis method based on contact dynamics is used to estimate contact spot distribution, then manufacturing precision is improved, but productivity decreases
Solution Approach 1:
The deep learning model is pre-trained offline using comprehensive numerical analysis data, capturing the complex relationships between surface characteristics and contact distribution. This preliminary training enables the model to provide accurate manufacturing precision estimates rapidly during production or inspection phases without repeating full numerical simulations.
Solution Approach 2:
The patent substitutes the mechanical numerical calculation system with an intelligent deep learning model. The complex iterative numerical computations are replaced by a trained neural network that processes surface images and directly outputs contact distribution estimates, dramatically improving productivity while maintaining manufacturing precision.
3Productivity
If a deep learning model based on CNN is used to estimate contact spot distribution, then productivity is improved, but device complexity increases
Solution Approach 1:
The deep learning model is designed with multi-functionality to handle various contact analysis tasks. A single trained model can estimate contact spot distribution, contact pressure distribution, and analyze different surface types, reducing the need for multiple specialized tools and simplifying the overall system despite the complex internal structure.
Solution Approach 2:
The patent introduces a surface image as an intermediary between the physical contact surface and the digital analysis result. The CNN model processes these image representations to estimate contact distribution, acting as an intelligent mediator that translates visual surface characteristics into quantitative contact metrics without requiring direct complex mechanical modeling.
Data Source
AI summary
An apparatus for estimating contact distribution to quickly estimate contact spot distribution from a contact surface image using a deep learning model based on a convolution neural network (CNN) and a method thereof are disclosed. A method for estimating contact distribution to estimate contact spot distribution between a first contact spot and a second contact spot includes inputting a contact surface image of at least one of the first contact surface and the second contact surface to a deep learning model based on a CNN and estimating contact spot distribution between the first contact surface and the second contact surface from an output of the deep learning model.