Research on posture optimization and accuracy compensation technology in robotic side milling

Abstract Weak stiffness and poor posture accuracy are two key problems that need to be sorted out in robotic side milling. Firstly, an index for evaluating the cutting plane stiffness was proposed based on the closeness of uniformly distributed chords between the evaluated ellipse and the ideal circle. A novel comprehensive posture optimization model was constructed by considering the kinematic characteristics and milling stability. To improve the posture accuracy, a full closed-loop compensation system with laser tracker was built. Combined with the analysis of error similarity and reducer backlash, mapping posture error to joint space was regarded as the problem of selecting joint angles under certain constraints, and it was solved by introducing an improved grey wolf optimization algorithm. Experimental results show that the posture error can be less than 0.05° through two compensations. The stiffness index is effective and the machining error can be reduced by around 15% after posture optimization.