Manipulator motion control method and system based on visual real-time teaching and adaptive dmps

Abstract

The invention relates to a motion control method and system of a mechanical arm based on visual real-time teaching and self-adaptive DMPS. First set the teaching object, then control the teaching object to perform demonstration movement, use Kinect to obtain the depth map and combine the PnP algorithm to perform three-dimensional pose positioning and tracking of the teaching object, and establish a space mapping system to map the teaching object pose to the machine At the end of the arm, the control information of each joint of the manipulator is calculated according to inverse kinematics and sent in real time to indirectly control the movement of the manipulator. Finally, the teaching movement information is recorded online and the adaptive DMPS algorithm is used for local linear optimization and learning. The invention gets rid of the shackles of the hardware structure of the mechanical arm and the dependence on complex sensors in the traditional teaching mode, reduces the hardware cost and difficulty of teaching, and the non-contact feature enhances the safety of the teaching process and has a wide range of advantages. Applicability, the adaptive DMPS method proposed by the present invention makes the whole system have good anti-interference performance.

Description

technical field [0001] The invention belongs to the field of mechanical arm motion planning, and in particular relates to a motion control method and system for online visual teaching and learning of a mechanical arm based on a QR code (a type of two-dimensional barcode). Background technique [0002] Motion planning is mainly aimed at the manipulator module with high-dimensional motion space on the robot. Different from planar path planning, it is mainly divided into two categories: joint space trajectory planning and Cartesian space trajectory planning. The former mainly uses spline interpolation in the traditional The latter mainly uses planning methods such as spatial straight lines or spatial arcs. Due to the multi-degree-of-freedom space characteristics of the manipulator itself, these methods not only make the planning and calculation complex in practical application, but also recalculate each pose of the manipulator during the movement process for each new target, es...

AI Technical Summary

Problems solved by technology

[0004] Aiming at the problems of difficult teaching and poor learning anti-interference in the teaching and learning mode in the field of manipulator motion planning, the present invention proposes a manipulator motion control method and system based on visual real-time teaching and self-adaptive DMPS. The number of sensors and the guaranteed accuracy get rid of the shackles of the hardware structure of the manipulator in the traditional teaching mode and the dependence on complex sensors, reducing the hardware cost and difficulty of teaching, and the non-contact feature enhances the safety of the user while also It has wide applicability, and for problems such as interference in the teaching information, the adaptive DMPS method proposed by the present invention makes the whole system have good anti-interference performance

Images