Hyperspectral Laser Process Monitoring for Real-Time Error Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current laser working systems face challenges in monitoring and controlling the laser welding process due to the complexity of parameterization, limited spatial and wavelength resolution in in-process monitoring, and the inability to perform closed-loop control effectively, leading to unreliable detection and classification of working errors.
Innovation Solution
The implementation of a hyperspectral sensor, such as a hyperspectral camera, captures continuous hyperspectral images of the workpiece, providing spatially and spectrally resolved data that is processed using a deep neural network for real-time monitoring and control, allowing for precise detection and classification of working errors and process optimization.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If post-process inspection is used for quality assessment and classification, then working errors can be detected, but the cost, integration effort and maintenance effort of the laser working systems increases
Solution Approach 1:
The patent replaces complex mechanical inspection systems with a photonic detection system. Specifically, it uses photonic tensor products of spectral signatures to detect and classify working errors, substituting physical inspection mechanisms with optical field-based detection that naturally encodes spatial and spectral information without requiring complex mechanical parameterization
Solution Approach 2:
The patent transforms the inspection approach by changing from spatial-domain or temporal-domain measurements to spectral-domain measurements. By using spectral signatures and photonic tensor products across multiple wavelengths, the system detects working errors through spectral parameter changes rather than through complex spatial parameterization, simplifying the overall system
2Device complexity
If in-process monitoring is performed without spatial resolution and/or without wavelength resolution, then the monitoring process is simplified, but the detection precision and classification accuracy of working errors is reduced
Solution Approach 1:
The patent adds the spectral dimension to in-process monitoring by measuring radiation across multiple wavelengths simultaneously. This creates a spectral signature for each spatial location, enabling precise detection and classification of working errors without increasing temporal or mechanical complexity. The photonic tensor product operates in this extended spectral dimension to extract diagnostic information
3Manufacturing precision
If up to 300 parameters are set for post-process inspection, then comprehensive quality assessment is achieved, but the parameterization complexity and evaluation difficulty increases
Solution Approach 1:
The patent extracts only the essential diagnostic information from the radiation field by computing photonic tensor products of spectral signatures. Instead of measuring and analyzing 300+ individual parameters, the system extracts a compact spectral signature that contains all necessary information for detecting and classifying working errors, eliminating the need for extensive parameterization
Solution Approach 2:
The photonic tensor product approach serves multiple functions simultaneously: it detects working errors, classifies error types, and provides spatial localization, all through a single unified mathematical operation on spectral signatures. This multi-functional approach replaces the need for multiple separate measurement systems and parameters
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
This approach simplifies the monitoring and closed-loop control of laser working processes, enabling reliable and automated detection of errors in real-time, reducing the complexity of parameterization and enhancing the accuracy of quality assessment and process control.
Implementation Method 1
capturing, in each pixel, a spectrum over a broad wavelength range of radiation emitted or reflected by the workpiece
Data Source
AI summary
A laser working system for performing a working process on a workpiece with a laser beam includes: a laser working head for radiating a laser beam into a working region on the workpiece; and a sensor unit for monitoring the working process, the sensor unit having at least one hyperspectral sensor. The sensor unit is designed to capture a hyperspectral image of a region of the workpiece, the hyperspectral image having N times M pixels. The hyperspectral image has two spatial dimensions x and y and a spectral dimension λ. N indicates the number of pixels in the first spatial dimension x, M indicates the number of pixels in the second spatial dimension y, and L indicates the number of spectral bands in the spectral dimension λ. M, N and L are natural numbers. A method for monitoring a working process is also provided.