Self-position estimation using behavior change filtering
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The accuracy of self-position estimation in moving bodies, such as vehicles, is compromised due to changes in orientation caused by behavioral changes, leading to errors in relative position detection by sensors mounted on these bodies.
Innovation Solution
A self-position estimation method and device that utilizes a combination of sensors like imaging devices, distance measurement devices, gyro sensors, and acceleration sensors to detect and correct for behavioral changes, selecting target position data with minimal error and collating it with map information for accurate position estimation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Difficulty of detecting and measuring
If sensors are mounted in the moving body to detect relative position, then position detection capability is improved, but measurement precision deteriorates due to orientation changes from behavior changes
Solution Approach 1:
The patent introduces behavior change data as an intermediary element that mediates between the sensor detection and position estimation. The behavior change data (obtained from gyro sensors or acceleration sensors) serves as a mediator to correct the relative position data, allowing the system to maintain both sensor mounting for detection capability while compensating for orientation-induced errors to preserve measurement precision
Solution Approach 2:
The patent implements a feedback mechanism where behavior change data is continuously obtained and used to correct relative position data. The correction amount calculated from behavior change data feeds back to adjust the relative position estimation, creating a closed-loop system that maintains accuracy despite orientation changes during movement
2Productivity
If all target position data is used for position estimation, then productivity is improved, but reliability deteriorates due to inclusion of data with large errors
Solution Approach 1:
The patent applies local quality by differentiating the quality of different target position data based on behavior change magnitude. Instead of treating all data uniformly, the system selectively applies correction amounts based on local behavior change characteristics, giving higher weight to data obtained during periods of minimal behavior change and lower weight to data obtained during significant behavior changes
Solution Approach 2:
The patent changes the parameter of data selection by introducing behavior change magnitude as a filtering criterion. The system dynamically adjusts which target position data is used for estimation based on behavior change parameters, excluding or de-emphasizing data obtained during periods when behavior change exceeded predetermined thresholds, thereby improving reliability without sacrificing overall productivity
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
This approach effectively suppresses the reduction in accuracy of position estimation by excluding target position data with significant behavioral change errors, improving the precision of self-position estimation in moving bodies.
Implementation Method 1
a gyro sensor 14 that generates gyro information indicating a yaw rate, a displacement amount in a pitch direction, and a displacement amount in a roll direction
Implementation Method 2
an acceleration sensor 15 that generates acceleration information indicating a lateral G and an acceleration/deceleration in a vehicle front-rear direction
Implementation Method 3
a laser range sensor mounted in the robot
Data Source
Figure 1
Figure 2
Figure 3
AI summary
A self-position estimation method including: detecting a relative position of a target existing around a moving body (1) relative to the moving body (S1); estimating a movement amount of the moving body (S2); correcting the relative position on a basis of the movement amount of the moving body and accumulating the corrected relative position as target position data (S4) ; detecting a behavior change amount of the moving body (S3); selecting, from among the accumulated target position data, the target position data of the relative position detected during a period in which the behavior change amount is less than a threshold value (S5) ; and collating the selected target position data with map information indicating a position on a map of the target to estimate a present position of the moving body (S6).