A foot-type robot control method and device, electronic equipment and storage medium

By employing a closed-loop mechanism that combines generative motion prior models, dynamic models, residual reinforcement learning, and projection constraint modules, the problem of discrepancies between planned and actual actions in existing legged robot control methods is solved, achieving high-precision, stable, and robust motion control.

CN122185241APending Publication Date: 2026-06-12HARBIN INSTITUTE OF TECHNOLOGY (SHENZHEN) (INSTITUTE OF SCIENCE AND TECHNOLOGY INNOVATION HARBIN INSTITUTE OF TECHNOLOGY SHENZHEN)
View PDF 0 Cites 0 Cited by

Patent Information

Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
HARBIN INSTITUTE OF TECHNOLOGY (SHENZHEN) (INSTITUTE OF SCIENCE AND TECHNOLOGY INNOVATION HARBIN INSTITUTE OF TECHNOLOGY SHENZHEN)
Filing Date
2026-05-12
Publication Date
2026-06-12

AI Technical Summary

Technical Problem

Existing control methods for legged robots rely on pre-set dynamic models, which leads to discrepancies between planned and actual actions, affecting control accuracy and making it difficult to meet the requirements for high-precision and robust autonomous motion control in complex environments.

Method used

By acquiring the robot's state vector and morphological parameter encoding, a candidate action sequence is generated using a generative motion prior model. This sequence is then combined with a dynamic model for forward simulation and risk assessment to select the optimal action sequence. Finally, a residual reinforcement learning model is used to refine the sequence, and an executable action sequence is generated through a projection constraint module.

Benefits of technology

It significantly improves the motion control accuracy, stability and robustness of legged robots in complex and unknown environments, reduces the impact of environmental disturbances, and ensures the physical feasibility of motion execution.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN122185241A_ABST
    Figure CN122185241A_ABST
Patent Text Reader

Abstract

The application provides a kind of foot type robot control method, device, electronic equipment and storage medium, it is related to robot control technical field, the foot type robot control method includes: obtaining the state vector of foot type robot, form parameter coding and control condition, input training good generative motion prior model, generate multiple candidate action sequences;Again candidate action sequence is substituted into the dynamics model that is constructed in advance, corresponding prediction state is obtained;Then risk assessment is carried out in combination with candidate action sequence and prediction state, obtains each sequence evaluation value, selects the minimum as optimal action sequence;Then the action residual correction quantity of trained residual reinforcement learning model is output, and the optimal action sequence is corrected to obtain the modified action sequence;Finally, the robot executable action sequence is obtained by the preset projection constraint module projection processing.The application can effectively improve the efficiency of foot type robot control.
Need to check novelty before this filing date? Find Prior Art