Method and system for controlling robot on basis of trajectory optimization based on reward or cost optimization

WO2025187883A8PCT designated stage Publication Date: 2025-10-02NAVER CORP
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
PCT/KR2024/013765
Authority / Receiving Office
WO · WO
Patent Type
Applications
Current Assignee / Owner
Priority Date
2024-03-04
Filing Date
2024-09-11
Publication Date
2025-10-02

AI Technical Summary

Technical Problem

Existing robot control methods struggle to optimize trajectories efficiently while adhering to hardware constraints and safety requirements, leading to suboptimal performance and potential safety hazards.

Method used

A method and system that utilize trajectory optimization by determining a cost function and updating actions using gradient descent or ascent algorithms, with a Straight Through Estimator (STE)-based approach for non-differentiable sections, to minimize or maximize cost/reward under constraints such as hardware limitations and safety considerations.

Benefits of technology

Enables accurate and efficient robot control that optimizes movement and service provision by minimizing costs and maximizing rewards while satisfying essential constraints, ensuring safe and precise operation.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure KR2024013765_02102025_PF_FP_ABST
    Figure KR2024013765_02102025_PF_FP_ABST
Patent Text Reader

Abstract

Provided is a method in which a cost function for determining the cost incurred or reward obtained as a robot performs possible actions in a first state of the robot is determined, the actions are updated to minimize / maximize the cost / reward on the basis of a gradient value calculated for the cost function under a constraint related to control of the robot, and the robot is controlled on the basis of at least one of the updated actions.
Need to check novelty before this filing date? Find Prior Art