Power Control during Remote Laser Welding Using a Convolutional Neural Network

The increase in complex workpieces with changing geometries demands advanced control algorithms in order to achieve stable welding regimes. Usually, many experiments are required to identify and confirm the correct welding parameters. We present a method for controlling laser power in a remote laser welding system with a convolutional neural network (CNN) via a PID controller, based on optical triangulation feedback. AISI 304 metal sheets with a cumulative thickness of 1.5 mm were used. A total accuracy of 94% was achieved for CNN models on the test datasets. The rise time of the controller to achieve full penetration was less than 1.0 s from the start of welding. The Gradient-weighted Class Activation Mapping (Grad-CAM) method was used to further understand the decision making of the model. It was determined that the CNN focuses mainly on the area of the interaction zone and can act accordingly if this interaction zone changes in size. Based on additional testing, we proposed improvements to increase overall controller performance and response time by implementing a feed-forward approach at the beginning of welding.

Design, Setup, and Evaluation of a Compensation System for the Light Deflection Effect Occurring When Measuring Wrought-Hot Objects Using Optical Triangulation Methods

In this paper, we present a system to compensate for the light deflection effect during the optical geometry measurement of a wrought-hot object. The acquired 3D data can be used to analyze the formed geometry of a component directly after a hot forging process without waiting for the needed cooling time to room temperature. This may be used to parameterize the process and to detect defect components early in the production process, among others. The light deflection as the deviation from the linear path of the light is caused by an inhomogeneous refractive index field surrounding the hot object. We present the design and setup for a nozzle-based forced air flow actuator, which suppresses the light deflection effect. The design process includes a simulation of the developing field, as well as of the interaction of the field with an external forced air flow. The cooling effect of the air flow is evaluated, and conclusions are drawn from the conflicting interests of good measurement conditions against the forced cooling of the hot object. The findings are then implemented in the physical setup of the suppression system. The system is evaluated using a previously established method based on optical triangulation and fringe projection. Other occurring effects and their influence on the evaluation are considered and discussed.