Today most of the parallel robot are controlled in joint space where the inverse kinematics problem has to be solved. Inverse kinematics uses geometrical parameter of the robot like joints location and links lengths.Because of the manufacturing and assembly tolerance, the actual geometrical parameters of a robot deviate from theirnominal values. The geometrical errors would result in the errors of the robot if the nominal geometry were used to estimate the pose of the robot. The kinematic calibration is an effective way to improve the absolute accuracy of robots. In this context, an offline robot calibration method based on inertial measurement unit (IMU) is presented in this work. The robot is a 6- degree of freedom parallel robot driven by six electromechanical linear actuator. The principle of the calibration process will be to get constraints equation at various poses of the end-effector.The constraint equations are calculated from the inverse kinematic problem. Then, the constraint equations are solvedby minimizing the sum of the squares of these equations. The Levenberg-Marquardt algorithm is used for the solution of the nonlinear least-squares problem. The performance of the robot is evaluated via pose accuracy and pose repeatability in according to international standard.
News published in Agência FAPESP Newsletter about the scholarship: