Robots have gradually been applied to all aspects of our life.How to further improve the quality of robot has become a very interesting research direction.Robot is a complex,multi-input multi-output(MIMO),nonlinear,multi-coupling system.To improve the performance of the robot,advanced control systems are needed.The purpose of this thesis is to establish a complete and practical control system for serial multi-degree-of-freedom robots.We not only apply the latest research results in this field,but also propose an advanced control strategy.In the establishment of multi-degree-of-freedom robot kinematics,the screw method based on the screw theory is used to establish the kinematic model.The method describes the motion of a full rigid body globally,and thus provides a complete geometric description,avoiding the singularity problem caused by the application of the local coordinate system description by the D-H parameter method.At the same time,the inverse kinematics is solved by using the product of exponential(POEs)formula.It is easy to determine the conditions for generating the number and details of the multiple solutions,to clarify its geometric meaning,to avoid the drawbacks of abstracted mathematical symbols,and to simplify the analysis and solution process of the kinematics.A trajectory planning method based on homogeneous coordinate matrix is adopted in robot trajectory planning.The motion of the robot in Cartesian space is decomposed into translational motions and rotational motions,and then the trajectory equations of robot in Cartesian space are obtained by solving the translational motions and the rotational motions,respectively.The method overcomes the shortcomings caused by the singularity of the robot and the Euler angle algorithm in the original trajectory planning method of industrial robots.At the same time,it has the characteristics of intuitive concept,accurate planning path and strong operability.In the aspect of robot dynamics modeling,we make full use of the advantages of the screw method to establish the robot kinematics model and apply the screw theory to the Lagrangian method to obtain robot dynamics.Compared with the traditional Lagrangian dynamic modeling that used D-H parameters,it has the characteristics of simple operation,clear meaning,and intuitiveness.In terms of robot control,we propose a new hybrid control method based on the existing control method – Nonlinear PD Terminal slide model control(NPD-TSMC).We have applied the NPD-TSMC to the established dynamics of robot system to form a multi-degree-of-freedom robot control system,and carried out simulation experiments,with the comparisons of the existing PD-SMC,NPD-SMC and PD-TSMC control methods.The simulation experiments show that the proposed NPD-TSMC control method has better trajectory tracking performance than the existing PD-SMC,NPD-SMC and PD-TSMC control methods.Finally,for the proposed NPD-TSMC control method,we use the established robot control system to carry out related analysis and research about its control parameters on the system effects of performances for different trajectories. |