Vehicle Localization Using Static Objects and Curvature Without GPS
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing vehicle navigation systems rely on satellite signals, which can become weak or lost in areas with poor reception, leading to inaccurate localization and impaired autonomous driving capabilities.
Innovation Solution
Utilizing a combination of on-board sensors, such as cameras, radars, and wheel speed sensors, to determine vehicle location based on static objects and lane sensing information, allowing for precise localization even in the absence of satellite navigation signals.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If satellite navigation information is used for vehicle localization, then localization can be achieved under normal conditions, but localization accuracy deteriorates when satellite signal strength is weak or lost
Solution Approach 1:
The patent introduces static objects (trees, buildings, poles) and lane markers as intermediary reference elements. When satellite signals are weak or lost, the system uses sensors to detect these static objects and lane markers, then calculates vehicle position based on distances and angles to these intermediaries, enabling continuous localization without satellite dependency
Solution Approach 2:
The system pre-stores map information including locations of static objects and lane markers before satellite signal loss occurs. This preliminary data preparation allows the vehicle to switch to map-based localization immediately when satellite signals become unavailable, maintaining continuous positioning capability
2Measurement precision
If multiple sensing systems are integrated for localization, then localization precision is improved, but device complexity increases
Solution Approach 1:
The patent merges satellite navigation information, sensor data (cameras, radars, wheel speed sensors), and pre-stored map information into a unified localization system. The system fuses these multiple data sources to calculate vehicle position, achieving centimeter-level precision while managing complexity through integrated processing
Solution Approach 2:
The system dynamically changes processing parameters based on signal availability. When satellite signals are strong, it uses satellite-based localization; when weak or lost, it switches to sensor-based detection of static objects and lane markers, adjusting the localization method according to environmental conditions
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
Enables accurate vehicle localization with centimeter-level precision, maintaining autonomous driving functionality in environments with weak or no satellite signal, reducing computational burden and ensuring safe navigation.
Implementation Method 1
a satellite receiver configured to receive a satellite signal having at least a threshold signal strength, the satellite signal including satellite navigation information
Implementation Method 2
a sensor configured to generate a first sensor return indicating at least a first static object and a second static object
Implementation Method 3
a wheel speed sensor configured to determine a first travel distance of the vehicle since the first location
Implementation Method 4
a lane monitoring system configured to generate lane sensing information, the lane sensing information reflecting a curvature of the vehicle
Data Source
AI summary
An advanced driver assistance system (ADAS) controls a vehicle using two or more of: static objects, a vehicle heading angle, or a vehicle curvature at least when a satellite navigation signal is weak or lost. Distances from the vehicle to the static objects are measured by sensor(s) to be the same in both coordinate systems, and the vehicle location is found as an intersection between circles in the ground coordinate system. The ADAS can use sums of transverse and longitudinal distances, with a vehicle heading angle or a vehicle curvature, to determine the vehicle location with a weak or lost satellite signal, such as using a linear vehicle model or a curvature-based vehicle model.