AI Multi-Pass Welding Control for Geometric Tolerance Variation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The steel construction industry faces a shortage of skilled welders due to a generational shift in expertise, and existing automation methods struggle to overcome complex geometric tolerances, steel plate deformation, assembly errors, and the interdependence of welding parameters in multi-pass weld steel structures.
Innovation Solution
An intelligent automated multi-pass welding method utilizing artificial intelligence (AI) technology, which includes measuring the welding space with an optical instrument, inputting reference geometry information into AI models to obtain control parameters for a welding device and robotic arm, and iteratively forming and assessing weld passes until completion.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Extent of automation
If manual welding is used, then skilled welders can handle complex geometric tolerances and steel plate deformation, but there is a shortage of skilled welders due to generational shift
Solution Approach 1:
The system employs optical measuring instruments to continuously scan the welding space and obtain real-time geometry information. This measurement data is fed back to the AI model, which dynamically adjusts welding parameters based on detected geometric variations, assembly errors, and steel plate deformation, enabling automated welding to adapt to complex conditions previously requiring skilled welders
Solution Approach 2:
The AI model dynamically optimizes welding parameters (current, voltage, speed, angle) based on real-time geometric measurements and multi-pass welding state. This allows the automated system to adjust parameters in response to geometric tolerances and deformation, achieving adaptability without requiring skilled welders
2Extent of automation
If depth cameras or LiDARs are used to identify weld passes, then automated welding can be achieved, but the massive amount of data generated limits production efficiency
Solution Approach 1:
The system extracts only the essential geometric features and welding-relevant information from the measurement data using the AI model, rather than processing all raw data. This selective extraction reduces computational burden while maintaining accurate weld pass identification and welding parameter optimization
Solution Approach 2:
The system creates simplified digital representations (virtual models) of the welding space and weld passes based on optical measurements. These digital copies enable fast AI-based analysis and decision-making without requiring processing of massive raw measurement datasets, improving production efficiency
3Productivity
If multi-pass welding is performed with AI control, then production efficiency improves, but the interdependence of welding parameters increases system complexity
Solution Approach 1:
The AI model serves as a universal control system that integrates optimization of multiple interdependent welding parameters (current, voltage, speed, angle, positioning) into a single decision-making framework. This multi-functional AI controller manages the complexity of parameter interdependence while improving production efficiency across all welding operations
Data Source
AI summary
An automated multi-pass welding method is provided. An optical measuring instrument measures a welding space to obtain reference geometry information, and then the reference geometry information is inputted into an AI model to obtain control parameters that are used by a welding device and a robotic arm as operation settings to perform welding. The optical measuring instrument measures the welding space after the welding to obtain post-welding geometry information for a computerized control device to generate a classification result. When the computerized control device determines to form a next weld pass based on the classification result, the aforesaid actions are repeated until the computerized control device determines to stop welding.