State Prediction System Trajectory Stability

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Existing methods for predicting the state of objects in robotic environments often result in unstable predictions due to varying conditions, leading to frequent changes in prediction results.

Innovation Solution

A state prediction system comprising a first trajectory generation element, a second trajectory generation element, and a trajectory specification element that evaluates the degree of approximation of candidate trajectories to a reference trajectory, specifying stable candidate trajectories based on specified ranks and periods to enhance prediction stability.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If existing methods are used to predict the state of objects, then prediction can be performed quickly, but the prediction results become unstable and vary with high frequency

Engineering Contradiction:

Improvestability of prediction resultsVSAvoidfrequency of prediction result changes

Core Design Contradiction:

ReliabilityVSLoss of time

Solution Approach 1:

The system performs preliminary evaluation of candidate trajectories by comparing them against a reference trajectory before final selection. This preliminary action filters out unstable predictions early, ensuring only trajectories with sufficient approximation degree proceed to specification, thereby reducing high-frequency variations in prediction results.

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The system implements feedback by continuously evaluating the degree of approximation between candidate trajectories and the reference trajectory. This feedback mechanism allows the system to adjust trajectory selection based on stability criteria, preventing specification of trajectories that would cause high-frequency variations in prediction results.

Inventive Principle:

Principle #23Feedback

2Measurement precision

If existing methods are used to predict the state of objects, then prediction can be performed with simple models, but prediction precision becomes unstable

Engineering Contradiction:

Improveprediction precisionVSAvoidcomplexity of trajectory evaluation system

Core Design Contradiction:

Measurement precisionVSDevice complexity

Solution Approach 1:

The system segments the trajectory evaluation process into distinct components: reference trajectory generation, candidate trajectory generation, degree of approximation evaluation, and specification. This segmentation allows each component to be optimized independently, achieving stable prediction precision without requiring excessive overall system complexity.

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

The system changes parameters by introducing a degree of approximation threshold and specification period duration as control parameters. These parameter changes enable stable prediction precision by filtering out unstable candidate trajectories based on quantifiable criteria, rather than relying on complex model structures.

Inventive Principle:

Principle #35Parameter changes

Data Source

PatentUS10166677B2State prediction system

Publication Date: 2019.01.01 HONDA MOTOR CO LTD

AI summary

To provide a system that can enhance stability of a result of predicting the state of an object. At least one candidate trajectory having a degree of approximation to a reference trajectory generated based on a current state of the object being in a specified rank or higher over a first specification period is specified as “a first candidate trajectory”. At least one candidate trajectory, extending from a last time point of being the first candidate trajectory to before elapse of a second specification period, is specified as “a second candidate trajectory”. Accordingly, it becomes possible to enhance stability of the specification result of the candidate trajectory as a result of predicting the state of the object.