Robot collision detection method based on second-order generalized momentum observer

Abstract

The invention discloses a robot collision detection method based on a second-order generalized momentum observer. The method comprises the following steps: firstly, constructing a robot kinetic model, and recognizing kinetic parameters; secondly, establishing a generalized momentum observer, determining a gain coefficient, establishing a friction model in a targeted manner, and identifying model parameters; and thirdly, a second-order generalized momentum observer is established through a second-order damping model, a PD regulator is connected in series on the basis of the second-order damping model, and a threshold value is set. According to the robot collision detection method based on the second-order generalized momentum observer, adjustable parameters are increased, system controllability is improved, short transition time is easier to obtain, system collision detection delay is reduced, and the application value is high. The method can be applied to the collision detection technology.

Description

technical field [0001] The invention relates to the technical field of collision detection, in particular to a robot collision detection method based on a second-order generalized momentum observer. Background technique [0002] Robot collaboration is widely used in various production fields. Robots may collide with humans or the environment due to misoperation and machine failure. To ensure the safety of humans and machines, collision detection has become an indispensable functional module for collaborative robots. [0003] Collision detection methods can be mainly divided into external torque monitoring based on dynamic model and external torque monitoring without dynamic model, and external torque monitoring based on dynamic model is the mainstream research direction at this stage. [0004] In the prior art, the energy-based observer was first proposed, which reflects the change of external torque through the change of system energy. However, this method cannot detect e...

AI Technical Summary

Problems solved by technology

However, this method cannot detect external forces at rest, nor can it detect external forces perpendicular to the direction of robot motion, and the monitoring signal directly reflects energy rather than external torque information

Later, a speed-based external torque observer was proposed. This method solved the problem that the specific external force of the energy observer could not be detected. However, due to the inverse operation of the inertia matrix, there is a nonlinear coupling relationship between the monitoring signal and the external torque.

After that, an external torque observer based on generalized momentum was proposed. This method successfully achieved decoupling. The output of the observer directly reflected the external torque changes of each joint. The calculation was simple and no inverse was required, but there was a delay in detection.

Images