Robotic belt grinding is commonly utilized for profile recovery of laser remanufactured blades by removing unevenly distributed allowances. Despite various trajectory planning methods that incorporate both blade geometry and physical factors are proposed successively, however, they mainly focus on global trajectory optimization and lack effective strategies for mitigating local overgrinding. To overcome this challenge, in this paper a local trajectory replanning method is presented to grind the remanufactured blade by employing a dual-node force-velocity synergistic (DN-FVS) strategy, which fully controls the high-risk grinding points. Inspired by the model predictive control strategy, a predictive risk quantification model is at first constructed by integrating the effects of curvature, allowance, and normal force distribution, aiming to identify high-risk points and enable feedforward optimization during the trajectory planning stage. Subsequently, both the normal force and velocity at local trajectory points are dynamically optimized using a dual-node interpolation approach combined with force-velocity synergistic strategy, which can effectively mitigate the local overgrinding risks. Experimental results demonstrate the effectiveness of the proposed method compared to state-of-the-art ones, with the average profile accuracy of blade edges enhanced by 17.5% to 0.0396 mm. Additionally, the ground blade surface exhibits superior roughness and smoothness, meeting the repair process requirements.
Xing et al. (Wed,) studied this question.