Research On Dynamics Parameter Identification And Collision Detection Of Special Mechanical Arm

The special manipulator is the key mechanism in a special test,which carries the model movement to achieve the test purpose.During the movement of the manipulator,it may collide with the surrounding environment or another manipulator,so it is necessary to carry out effective collision detection for the special manipulator.Among the existing collision detection methods,the dynamic model-based collision detection method is suitable for the manipulator because it does not need any external sensors,but the effect of this method depends on the accuracy of the dynamic model.Therefore,it is necessary to identify the dynamic parameters of the manipulator.The existing identification methods generally take the dynamic parameters and friction parameters as the minimum parameter set for one-time identification,which is not suitable for the special manipulator in this paper.Therefore,this paper designs a method to identify friction parameters and dynamic parameters separately,and the dynamic parameters are identified step by step.The dynamic parameters are obtained through identification experiments.On this basis,the existing collision detection methods are used to realize the collision detection of special manipulator.The main contents of this paper are as follows:Firstly,the structure of the special manipulator is analyzed.Considering that there are two joints in the manipulator,which are equivalent to the rotating joints of the linear variable arc mechanism,the kinematics of the linear variable arc mechanism is analyzed separately.Then,the kinematics model of the whole manipulator is built by D-H method.After that,the dynamic equations of the manipulator are derived by Lagrange method,and the simulation results are verified by Matlab and Adams.Then,a method is designed to identify friction parameters and dynamic parameters separately,and to identify dynamic parameters step by step.For the identification of friction parameters,the relationship between joint friction and joint velocity can be extracted to complete the identification under the specific motion condition.For the identification of dynamic parameters,the established dynamic equations are linearized and the minimum set of dynamic parameters is obtained.Due to the limitation of the motion range and speed of the manipulator,the regression matrix corresponding to the minimum parameter set is ill conditioned.Therefore,the minimum parameter set is divided into two parts for identification.One part of the parameters is identified by single joint motion,and the remaining parameters are identified by designing the optimal excitation trajectory.Through the identification experiment,the dynamic parameters of the manipulator are obtained.Finally,the identified dynamic parameters are applied to the existing collision detection algorithm,and the effectiveness of the algorithm is verified by simulation,and finally the collision detection of the manipulator is realized.