In this paper, the parameter optimization of the hybrid-tandem gas metal arc welding (GMAW) process was studied. The hybrid-tandem GMAW process uses an additional filler-wire with opposite polarity in contrast to the conventional tandem process. In this process, more process parameters and the relationship between the parameters causing strong nonlinearity should be considered. The analysis of variance-based Gaussian process regression (ANOVA-GPR) method was implemented to construct surrogate modeling, which can express nonlinearity including uncertainty of weld quality. Major parameters among several process parameters in this welding process can be extracted by use of this novel method. The weld quality used as a cost function in the optimization of process parameters is defined by characteristics related to penetration and bead shape, and the sequential quadratic programming (SQP) method was used to determine the optimal welding condition. This approach enabled sound weld quality at a high travel speed of 1.9 m/min, which is difficult to achieve in the hybrid-tandem GMAW process.

A monitoring technique for detecting changes in the root gap of butt joints during the flux-cored arc welding (FCAW) was proposed. FCAW experiments were conducted for both increasing and decreasing root gap conditions, and current and voltage were measured during the root-pass welding. The measured time series signals were used as input data for training Random Convolution Kernel Transform (ROCKET) algorithm, which consists of a feature extractor with multiple random kernels, and a linear classifier. A univariate model using current and voltage, respectively, and a multivariate model using both were compared, and the multivariate model showed the highest classification accuracy of 96.2%. Moreover, the classification errors were investigated by correlating the geometry of the root bead with the measured signals.

This paper proposes a robust inverse-optimal control for flexible joint robot manipulator in order to cope with parameter uncertainties. In the first step, a virtual control input was designed as PED form with 2 gain tuning conditions. In the second and third steps, the other virtual input and actual control torque were made using the saturation-type nonlinear control based on the Lyapunov's second method. The controller does not need information of acceleration and jerk of motor side as well as link side. In addition, the one is able to apply to actual robot controller by reducing calculation time caused by nonlinear matrix inequality (NLMI) solving. The robustness was shown in the simulation result through 2-DOF robot manipulator which has flexible joint.