Multi-Information Detection, Fusion Analysis and Decoupling Control in Pulsed MIG Welding

Literature Overview

Published in 2010 by researchers from Lanzhou University of Technology (Key Laboratory of Nonferrous Metal New Materials and Key Laboratory of Nonferrous Metal Alloys and Processing), this study investigates the comprehensive monitoring and intelligent control of pulsed MIG welding processes. Funded by the National Natural Science Foundation (Grant No. 50675093) and the Gansu Provincial Natural Science Foundation, the research addresses the multi-parameter coupling challenge inherent in pulsed welding operations.

Core Technical Content

Pulsed MIG welding employs periodic variations in current magnitude to achieve controlled droplet transfer with reduced heat input compared to continuous GMAW. The multi-information detection framework encompasses:

The fusion analysis methodology integrates these heterogeneous data streams to construct a comprehensive process state representation. The decoupling control strategy addresses the well-known cross-coupling between arc length, wire feed speed, and travel speed, which in conventional control systems leads to oscillatory behavior and inconsistent weld quality.

Interpretation of Technical Points

The decoupling control represents a significant advancement over single-variable feedback systems. In conventional MIG welding, adjusting the arc voltage (and therefore arc length) inevitably affects the wire feed speed and current, creating a coupled dynamic system that is difficult to control independently. The decoupling approach employs mathematical models of the process to calculate compensating signals that neutralize the cross-influence between control variables.

The multi-information fusion provides redundancy and robustness. When one sensor channel is compromised—for example, by spatter obscuring an optical sensor—the remaining channels can still provide adequate process information. This redundancy is critical in production environments where sensor degradation is inevitable.

Application to Cladding Process Control

In weld overlay applications, particularly for clad plate production per ASTM A263 or EN 10028-7, the control of dilution rate is the primary quality objective. The multi-information detection approach can be adapted to monitor dilution indicators in real-time:

• Arc voltage fluctuations indicate changes in arc length, which affect heat input distribution
• Current waveform analysis reveals the actual heat input to the weld pool
• Optical sensing can estimate pool geometry and, by extension, dilution depth

For electroslag welding (ESW) overlay operations used in thick-section cladding, similar multi-parameter monitoring principles apply, with the slag pool temperature, welding current, and travel speed forming the primary control variables.

Key Reflections and Engineering Practice

The research demonstrates that intelligent process monitoring is not merely an academic exercise but a practical necessity for maintaining quality in demanding applications. In pressure vessel fabrication, where overlay welds must pass rigorous non-destructive testing per JB/T 4730 or ASME V, the margin for process deviation is extremely narrow.

The decoupling control concept can be implemented in industrial welding power sources that support adaptive control modes. Modern welding equipment from major manufacturers already incorporates some form of arc length regulation with multi-parameter feedback, though the full decoupling capability described in this research may require custom controller implementation.

The fusion analysis methodology provides a framework for developing condition-based maintenance systems for welding equipment, where sensor data trends can indicate electrode wear, gas supply degradation, or wire feed mechanism problems before they result in quality failures.