Robot Orientation Estimation via Similarity-Based Outlier Filtering
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for determining the position of autonomous robots are prone to inaccuracies due to the use of imprecise expected location data, which can lead to incorrect positioning and increased inaccuracy in navigation, especially when combined with potential errors from sensor measurements and map matching processes.
Innovation Solution
A method that compares expected position data with actual location data using a similarity analysis and similarity matrix to identify and correct deviations, ensuring only consistent and accurate position data are used for further positioning, thereby enhancing the accuracy and reliability of robot navigation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Extent of automation
If expected position data from sensor measurements and map matching is used for robot positioning, then the robot can navigate autonomously, but the positioning accuracy deteriorates due to inherent errors in sensor measurements and map matching processes
Solution Approach 1:
The patent implements a feedback mechanism by continuously comparing expected position data with actual position data obtained from independent sensors (GPS, compass, inertial sensors). The system uses this feedback to detect deviations and correct positioning errors, thereby maintaining autonomous navigation capability while improving positioning accuracy through continuous error correction
Solution Approach 2:
The patent introduces an intermediary correction mechanism that acts as a mediator between expected position data and actual robot position. By using independent sensors as intermediaries to verify and correct position data, the system resolves the contradiction between autonomous navigation requirements and positioning accuracy without directly modifying the core navigation system
2Reliability
If multiple sensor devices and data processing steps are used to improve positioning accuracy, then positioning reliability increases, but the system complexity increases
Solution Approach 1:
The patent applies multi-functionality by using a single processing unit that handles multiple tasks: receiving data from various sensors, performing map matching, calculating expected position, obtaining actual position from independent sensors, comparing data, and generating corrections. This universal approach improves positioning reliability through multiple data sources while avoiding the complexity of separate dedicated systems for each function
Solution Approach 2:
The patent combines multiple sensor inputs (odometry sensors, GPS, compass, inertial sensors) and multiple processing functions (map matching, position calculation, error detection, correction) into an integrated positioning system. By merging these elements into a unified processing framework, the system achieves high positioning reliability through diverse data sources while maintaining manageable system complexity through integrated architecture
Data Source
Figure 1
Figure 2
AI summary
The invention relates to a method for determining the orientation of a robot (10) and comprises at least one method step in which expected orientation data (I) pertaining to an expected orientation of the robot (10) are provided and a method step in which actual orientation data (II) pertaining to an actual orientation of the robot (10) are provided. According to the invention, one method step involves the expected orientation data (I) and the actual orientation data (II) being used to ascertain outlier-free orientation data that are used for determining the orientation of the robot (10). The invention further relates to a robot (10) and to an orientation determination apparatus (26).