Towards forming accuracy improvement in GMA-based additive manufacturing based on dual-vision sensing and decoupling control

Gas metal arc (GMA)-based additive manufacturing (AM) is a promising technique for fabricating large-scale metal parts owing to its high deposition rate and low cost. Nevertheless, ensuring consistency between the actual and target deposition dimensions remains challenging. This study focuses on improving the deposition dimensional accuracy through collaborative monitoring and automatic control of the deposition width and height in GMA-AM. A dual-vision sensing system was designed and integrated with image processing algorithms to extract the deposition width and height accurately, achieving detection errors for the width and height of less than 2.8% and 3.8%, respectively. The coupling relationship between the deposition current and voltage was analyzed using a relative gain matrix, and a decoupler based on feedforward compensators was developed to decouple their interactions. Following decoupling, the current mainly influenced the deposition width, whereas the voltage mainly affected the deposition height. Two fuzzy controllers were constructed to control the forming accuracy of the deposition layers by adjusting the deposition current and voltage in real time. Finally, thin-walled parts were fabricated using constant parameters and closed-loop control to validate the effectiveness of the decoupling control system. The width and height deviations of the 20 th layer under closed-loop control were less than 0.4 and 0.44 mm, respectively, which were significantly smaller than those of the thin-walled part fabricated using constant parameters.