Work Machine Position Fusion for Changing Terrain Landmarks
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The existing scan matching navigation techniques face challenges in accurately estimating the position of a work machine, especially when the detection data of non-contact sensors is not properly matched with map data, particularly in environments with natural structures like banks that change shape due to weather or ground leveling, leading to decreased estimation accuracy.
Innovation Solution
A work machine control system that integrates position data from multiple sensors, including a position sensor, a first non-contact sensor for detecting banks, and a second non-contact sensor for detecting landmarks, to determine an integrated position of the work machine, thereby enhancing estimation accuracy even when detection accuracy deteriorates.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If scan matching navigation using non-contact sensor detection data of banks is employed, then position estimation can be performed without GNSS, but estimation accuracy deteriorates when detection data does not properly match map data due to natural shape changes
Solution Approach 1:
The patent combines multiple position estimation results from different calculation units (first position calculation unit using bank detection data, second position calculation unit using additional detection data, and third position calculation unit using map data) into a single integrated position estimate. This merging approach ensures that even when one data source becomes unreliable due to shape changes, the overall position estimation remains accurate by relying on other consistent data sources.
2Device complexity
If only a single non-contact sensor detecting banks is used, then the system structure remains simple, but position estimation accuracy deteriorates when the bank shape changes due to weather or ground leveling
Solution Approach 1:
The patent segments the position estimation function into multiple independent calculation units, each processing different types of detection data. The first position calculation unit processes bank detection data, while the second and third units process additional detection data and map data respectively. This segmentation allows the system to maintain simplicity in individual components while achieving robustness through their coordinated operation.
Solution Approach 2:
The patent changes the parameters used for position estimation by incorporating multiple types of detection data beyond just bank shape detection. When bank shape changes occur, the system switches to or integrates with alternative parameters from other detection sources, thereby maintaining estimation accuracy despite variations in the primary detection target.
3Measurement precision
If multiple position calculation methods are integrated, then position estimation accuracy is maintained under various conditions, but the control system complexity increases
Solution Approach 1:
The integrated position determination unit continuously monitors the reliability of detection data from different sources and dynamically adjusts the weighting or selection of position calculation methods based on current conditions. When bank shape changes are detected, the system provides feedback to reduce reliance on bank detection data and increase reliance on alternative data sources, thereby maintaining accuracy without requiring manual intervention or complex reconfiguration.
Data Source
AI summary
A work machine control system includes: a position sensor detecting a position of a work machine traveling on a traveling path; first and second non-contact sensors provided in the work machine and detecting a position of an object around the work machine; a map data creation unit creating map data based on detection data of the position sensor and detection data of the first non-contact sensor; a first position calculation unit calculating a first position of the work machine by matching the map data and the detection data of the first non-contact sensor; a second position calculation unit calculating a second position of the work machine based on detection data of the second non-contact sensor; and an integrated position determination unit determining an integrated position of the work machine by integrating the first and second positions.