Vehicle Position Estimation Using LIDAR and Map Landmarks
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional methods for estimating the current position of a vehicle in autonomous driving systems are insufficient in accuracy, as they rely mainly on internal sensors and do not provide a high degree of precision for absolute positioning.
Innovation Solution
An estimation device that acquires map information and uses a combination of LIDAR, gyroscope, and vehicle speed sensors to estimate the vehicle's position by cross-referencing distance and angle data with registered landmarks, employing a Bayesian estimation method and extended Kalman filter for accurate positioning, even in challenging conditions like nighttime or obscured road markings.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional methods using only internal sensors are used for position estimation, then the device complexity is low, but the measurement precision of position is insufficient
Solution Approach 1:
The patent combines multiple information sources including LIDAR distance measurements, gyroscope angular velocity data, vehicle speed sensor information, and map database landmarks into a unified position estimation system. This integration of heterogeneous data sources resolves the contradiction by achieving high measurement precision through data fusion while managing device complexity through coordinated processing of multiple sensors.
Solution Approach 2:
The patent introduces map database landmarks as intermediary reference points that mediate between raw sensor data and absolute position estimation. These landmarks serve as a bridge, allowing the system to convert relative sensor measurements into absolute position information without requiring direct complex processing of all sensor inputs simultaneously.
2Measurement precision
If multiple sensors and map information are integrated for position estimation, then the measurement precision improves, but the computational burden increases
Solution Approach 1:
The patent pre-processes and stores landmark information in a map database before runtime position estimation. By preparing reference data in advance and organizing it for efficient retrieval, the system reduces the computational burden during actual position estimation while maintaining high measurement precision through access to pre-processed spatial information.
Solution Approach 2:
The patent extracts and utilizes only the necessary components from multiple data sources for position estimation. Rather than processing all available sensor data equally, the system selectively extracts relevant distance measurements from LIDAR, angular velocity from gyroscope, and corresponding landmark information from the map database, reducing computational burden while maintaining estimation accuracy.
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
The solution enables precise and accurate estimation of the vehicle's position with a low computational burden, effectively addressing the limitations of conventional methods by integrating external data with internal sensors for enhanced accuracy.
Implementation Method 1
a LIDAR which detects a point group of a surface of an object by performing horizontal scanning with intermittently-emitted laser light and receiving the reflective light (scattering light)
Implementation Method 2
output of internal sensors which detect the state of a vehicle
Data Source
Figure 1~2
Figure 3~4
Figure 5
AI summary
Provided is an estimation device that is capable of estimating a current position with a high degree of accuracy. A driving support system includes: a vehicle mounted device 1 that is mounted on a vehicle and performs a control concerning vehicle driving support; a LIDAR 2; a gyroscope sensor 3; and a vehicle speed sensor 4. The vehicle mounted device 1 acquires, from the LIDAR 2, a measurement value ztk indicating a positional relationship between a reference landmark Lk of an index k and the own vehicle. The vehicle mounted device 1 calculates a post estimated value x^t by correcting a prior estimated value x-t estimated, on the basis of the measurement value ztk and a position vector mk of the reference landmark Lk included in a map DB 10, from a moving speed measured by the vehicle speed sensor 4 and an angular rate measured by the gyroscope sensor 3.