Robot Motion Trajectory Planning With Reduced-Dimensional Joint Models

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing robot technologies are limited by simple motion trajectories, restricting their ability to perform complex actions.

Innovation Solution

A method and apparatus for generating motion trajectories that involve constructing a reduced-dimensional joint model of a robot, determining an action sequence based on this model, and optimizing trajectory points using a cost function and constraint conditions related to the action sequence.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If a full-dimensional joint model is used to determine motion trajectory, then the accuracy of trajectory representation is improved, but the computational complexity and difficulty of trajectory determination increases

Engineering Contradiction:
Improvetrajectory representation accuracyVSAvoidtrajectory determination complexity
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The motion trajectory determination process is segmented into two stages: first determining an action sequence based on a reduced-dimensional joint model, then generating the complete trajectory. This segmentation allows the complex full-dimensional problem to be broken down into a simpler reduced-dimensional subproblem followed by trajectory generation, reducing overall computational complexity while maintaining accuracy

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent transitions from working directly in full-dimensional joint space to working in a reduced-dimensional space by eliminating non-motion dimensions. This dimensionality reduction simplifies the trajectory determination problem while the action sequence framework ensures the reduced model still captures essential motion characteristics needed for accurate trajectory generation

Inventive Principle:
Principle #17Another dimension (Dimensionality change)

2Productivity

If a reduced-dimensional joint model is used to determine motion trajectory, then the efficiency of trajectory generation is improved, but the complexity of motion representation decreases

Engineering Contradiction:
Improvetrajectory generation efficiencyVSAvoidmotion representation complexity
Core Design Contradiction:
ProductivityVSDevice complexity

Solution Approach 1:

The motion representation is segmented into action sequences (high-level behavioral descriptors) and detailed trajectories (low-level execution paths). The reduced-dimensional joint model efficiently determines the action sequence, which captures essential motion characteristics without requiring full-dimensional complexity, thus improving efficiency while preserving necessary motion information

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The action sequence is determined in advance as a preliminary step before generating the detailed trajectory. This preliminary determination using the reduced-dimensional model prepares the essential motion framework ahead of time, enabling more efficient trajectory generation while the action sequence itself serves as a compact representation of the motion's essential complexity

Inventive Principle:
Principle #10Preliminary action

Data Source

PatentEP4549106A1Method and apparatus for generating motion trajectory, robot, electronic device and storage medium
Publication Date: 2025.05.07 BEIJING XIAOMI ROBOT TECH CO LTD
  • EP4549106A1 patent drawingFigure 1
  • EP4549106A1 patent drawingFigure 2
  • EP4549106A1 patent drawingFigure 3~4

AI summary

The disclosure relates to a method and an apparatus for generating motion trajectory, a robot, an electronic device and a storage medium. The method includes: determining (101) a first joint model of a robot, where the quantity of dimensions of the first joint model is smaller than the quantity of dimensions of a full-dimensional joint model of the robot; determining (102) an action sequence of a target motion based on the first joint model, where the action sequence includes a plurality of action stages and switching events between the adjacent action stages; and determining (103) a trajectory of the target motion according to the action sequence.