Autonomous Robot Position Recognition Using Voronoi Space Models
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing position recognition methods for autonomous mobile robots using grid maps are inefficient, as they either fail with a small number of estimated positions or require excessive calculation time with a large number of positions, and necessitate unnecessary movement to converge on a current position.
Innovation Solution
The method involves dividing a grid map into multiple spaces using a generalized Voronoi graph, extracting learning data, generating space models, and recognizing the current position based on actual range scan data without continuous movement by using techniques like spectral clustering and ray casting algorithms.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If the number of estimated positions is increased to improve position recognition accuracy, then position recognition reliability is improved, but calculation time increases
Solution Approach 1:
The grid map is divided into multiple spaces based on generalized Voronoi graphs, and each space is processed independently. This segmentation allows the system to reduce the number of estimated positions needed while maintaining recognition accuracy, as each space can be recognized separately rather than requiring comprehensive coverage of the entire map.
Solution Approach 2:
The system pre-generates space models for each divided space before actual position recognition occurs. By preparing these models in advance, the calculation time during actual operation is reduced, as the system only needs to compare sensor data against pre-computed models rather than performing full calculations in real-time.
2Measurement precision
If the autonomous mobile robot continuously moves to converge on current position, then position recognition accuracy is improved, but energy consumption increases and unnecessary movement occurs
Solution Approach 1:
The system performs position recognition while the robot is stationary by comparing sensor data against pre-generated space models. This eliminates the need for continuous movement to converge on position, as the recognition can be performed at any location within a space, thereby reducing energy consumption and avoiding unnecessary movement.
Solution Approach 2:
The patent replaces the mechanical approach of moving the robot to converge on position with a computational approach using space models and sensor data comparison. This substitution allows position recognition to be performed without physical movement, eliminating energy waste while maintaining accuracy.
Data Source
AI summary
In some example embodiments, a position recognition method of an autonomous mobile robot may include: dividing a grid map into a plurality of spaces; extracting, by a processor, learning data of each of the plurality of spaces; generating, by the processor, space models of each of the plurality of spaces using the extracted learning data; and/or recognizing, by the processor, a current position of the autonomous mobile robot based on the extracted learning data, the space models, and actual range scan data input through the autonomous mobile robot.