Research On Surface Tracking Method Of Industrial Robot Based On Model Predictive Control

As machining accuracy requirements increase,industrial robots performing contact work tasks need to respond quickly and correctly to complex working environments in order to achieve good results,which usually involves the stable control of contact forces.For example,in the grinding process,the surface quality can be easily affected by the unstable contact force in the normal direction.However,in the process of force tracking control,it is difficult to achieve force tracking control of complex surfaces by dynamically compensating for the unknown environment based on past contact force changes,and the speed of control is influenced by the selection of parameters in the force control method.In order to solve the problem that normal control speed is easy to overshoot and normal contact force is not stable in uncertain environment,the surface tracking method of industrial robot based on predictive model control is put forward in this article.Taking advantage of Staubli TX90 industrial robot,this method is verified experimentally.The results show that the method can effectively reduce the occurrence of normal contact force mutation,and the force variance decreases by 88.50%.The smoothing method of attitude rotation Angle can effectively reduce the fluctuation range of normal contact force,the force variance is reduced by 35.28%,and the overall force control effect is good.The following is the main results content of this paper:(1)A method to realize constant force tracking of robot surface is presented.The method includes two off-line trajectory planning methods,the trajectory of the end tool can be calculated according to the geometric information of the workpiece model,and the Cartesian velocity of the end tool can be calculated by combining the current pose of the robot.Meanwhile,the update the formula of the normal damping coefficient and the contact model are established,and the normal damping coefficient can be adjusted online according to the attitude change of the end tool,and the normal contact force between the workpiece and the end tool can be enabled indirect measurement.Moreover,this method can realize the conversion from force to velocity and from velocity to position.By using this method,the position type industrial robot can obtain the required joint angles and realize the constant force tracking of the free-form surface.(2)A speed correction algorithm based on MPC is proposed.According to the feedback of real time force signal,and based on the established contact model and the updating formula of normal damping coefficient,the algorithm can be optimized by rolling to realize the online correction of normal control velocity,and then realize the real time correction of cartesian velocity of the terminal tool,eventually get the right speed for force control to avoid overshoot.The test results prove that the test results are better with this algorithm compared to without it,and the algorithm is effective in improving the normal control velocity overshoot on the continuous path for both constant and changing attitude of the end tool,and reduce the force fluctuation range.(3)An attitude smoothing strategy is proposed.Based on the off-line trajectory planning and velocity correction algorithm,this strategy corrects the attitude rotation Angle,avoids the abrupt change of the end tool attitude,reduces the abrupt change of normal contact force,and further improves the stability of force control.The experimental results show that this strategy can effectively improve the normal contact force and make the mean strength closer to the expected value.