Multi-modal perception robot adaptive force control follow-up operation system and control method

CN122353604APending Publication Date: 2026-07-10GUANGZHOU HUARUI ZHIPU INTELLIGENT ROBOT CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
GUANGZHOU HUARUI ZHIPU INTELLIGENT ROBOT CO LTD
Filing Date
2026-05-25
Publication Date
2026-07-10

AI Technical Summary

Technical Problem

Existing industrial robots lack multimodal perception capabilities in dynamic environments, have weak adaptability, and poor motion stability, resulting in incomplete assembly, over- or under-cut grinding, and unstable welding quality. Furthermore, existing force control solutions are slow to respond to high-frequency disturbances and sudden changes in workpiece pose, and have poor environmental adaptability.

Method used

The multimodal perception robot adaptive force control follow-up operation system integrates a distributed real-time acquisition unit, an intelligent decision-making unit, and a robotic arm execution unit. It combines a gravity compensation module, a kinematic modeling module, a safety constraint module, a composite control module, and a motion stabilization control module to achieve active perception and adaptive following of workpiece position changes and surface morphology through hierarchical compensation and reinforcement learning.

Benefits of technology

It achieves high-precision dynamic tracking of contact force, reduces tracking error and oscillation frequency, improves motion stability and operation quality, and significantly enhances operational robustness and process consistency under complex working conditions.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN122353604A_ABST
    Figure CN122353604A_ABST
Patent Text Reader

Abstract

The purpose of this disclosure is to provide a multimodal sensing robot adaptive force-controlled servo operation system and control method, including: a distributed real-time acquisition unit, an intelligent decision-making unit, and a robotic arm execution unit; wherein, the distributed real-time acquisition unit is used to collect the status information of any online unit of the industrial robot and acquire sensing data; the intelligent decision-making unit interacts with the distributed real-time acquisition unit to output control commands based on the sensing data; the robotic arm execution unit interacts with the intelligent decision-making unit to complete adaptive servo operation based on drive commands; this disclosure utilizes distributed joint torque sensing and a gravity compensation strategy based on a feedforward current mapping table, combined with... PID The admittance-parallel composite force control algorithm achieves high-precision dynamic tracking of contact force, controlling the tracking error within ±0.5. N Within this range, control precision has been greatly improved.
Need to check novelty before this filing date? Find Prior Art