Research On Robot Path Planning And Force Control Method For Complex Surface Scanning

Gears,blades,crankshafts and complex body structural parts are the key components of high-end equipment.Mechanical performance testing is an important support for manufacturing process optimization and quality control of the above key components.At present,micro magnetic sensors which can detect the mechanical properties of parts have been developed.Because the surface of blade and other parts is often a free-form surface which is difficult to express mathematically,the curvature of different parts is very different,and a small displacement will bring a large attitude change,which is difficult to apply to the automatic scanning process.Based on the background of complex surface scanning,this paper studies from the following aspects.In the aspect of path planning,the scanning path planning method of complex surface is proposed,including the extraction of topological relationship and curvature feature of complex surface,variable step scanning rule of complex surface,and path point search method of complex surface.The results show that the average curvature of the scanning path obtained by this method is smaller than that of the traditional method,and the curvature fluctuation is smoother.Based on the above path planning results,in the joint space of the robot,the cubic B-spline interpolation method is selected to ensure the smooth motion of the robot,and the interpolation results are optimized for the shortest time.In order to improve the computational efficiency,improvements are made on the basis of the classic genetic algorithm,including: the improvement of crossover and mutation strategies,the adaptive changes of crossover rate and mutation rate,and the introduction of group elite retention strategies and simulated annealing Strategy.Experiments show that the improved algorithm converges extremely fast,which greatly improves the efficiency.In the aspect of force control and path tracking,based on the impedance control model,the equivalent stiffness estimation method of environment is added,the parameter adaptation is realized by using fuzzy controller,and the path tracking is realized by using model predictive control.In order to prevent violent collision between robot and complex surface,virtual force and force feed-forward control strategy are introduced to reduce the impact force,so that the contact force can be quickly stabilized.Finally,experiments verify the rationality and versatility of the above-mentioned theories.In the actual scanning process,the robot did not have large oscillations and collisions,the path tracking accuracy was high,the contact force was always kept constant,and the scheduled scanning work could be completed.