Safe and reliable control method for full-state constrained rigid manipulator based on deterministic learning

Abstract

The invention discloses a full-state constraint rigid mechanical arm safe and reliable control method based on defining learning. The method includes the following steps that a kinetic model and an expected period track of a full-state constraint rigid mechanical arm are established; a state converter is established; a non-affine model is configured into an affine model; the track error of the angle position of the converted mechanical arm is defined; a compensation track error signal is designed; a neural network controller based on defining learning is designed; and a static neural network controller based on experiential knowledge is established. By means of the method, it can be ensured that the track error is finally converged into a small neighborhood of zero, the mechanical arm can also be limited to run in a given safe work zone, a defining learning theory is adopted, uncertain closed-loop dynamic learning of the rigid mechanical arm is achieved, the converged neural network weight is stored, redundancy training is avoided due to the stored experience knowledge, the responding speed of a system is increased, and the tracking performance on the transient state process of the rigid mechanical arm is improved.

Description

technical field [0001] The invention relates to the field of full-state limited control of rigid manipulators, in particular to a safe and reliable control method for full-state limited rigid manipulators based on deterministic learning. Background technique [0002] At present, the robotic arm is the most widely used automatic mechanical device in the field of robotics. In addition to being mainly used in industrial manufacturing, its application devices can be found in commercial agriculture, medical rescue, entertainment services, military security, and even space exploration. . With the increase of application scenarios and the increase of task complexity, people have higher and higher requirements for the work efficiency and work quality of the robotic arm. At present, there are relatively few studies on the safety control of robotic arms. When the robotic arm interacts with people, its high rigidity is likely to cause harm to surrounding personnel. Therefore, it is n...

AI Technical Summary

Problems solved by technology

On this basis, the present invention adopts the command filter and successfully solves the loop construction problem of the full-state limited rigid manipulator controller that may be generated by the traditional push-back design through the method of designing the compensation signal

Images