Map-Based Vehicle Localization Without Encoder Feedback
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing autonomous vehicles face challenges in localization without relying on encoder outputs, particularly when high-speed movement or wheel slippage occurs, leading to unreliable position estimation.
Innovation Solution
A vehicle system that uses an external sensor to scan the environment, storing environmental maps and employing a localization device with a processor to match scan data against the maps, predicting initial location and attitude based on historical data for efficient matching, thereby eliminating the need for encoder outputs.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If encoder outputs are used for localization, then position estimation can be obtained, but the system becomes unreliable during high-speed movement or wheel slippage
Solution Approach 1:
The patent extracts the localization function from dependence on encoder outputs by using only external sensor scan data and environmental maps. The localization device matches scan data against the environmental map to estimate vehicle location and attitude without requiring encoder information, thereby eliminating the reliability issues associated with encoder-based methods during wheel slippage or high-speed movement.
Solution Approach 2:
The patent replaces the mechanical measurement system (encoders attached to motors or driving wheels) with an optical/sensing-based system (external sensors scanning the environment). This substitution allows localization to be performed through environmental feature matching rather than mechanical rotation measurement, making the system reliable during conditions that cause mechanical measurement failures.
2Reliability
If encoder outputs are used for localization, then position estimation is available, but the device complexity increases
Solution Approach 1:
The patent removes the encoder components from the localization system, keeping only the external sensors that are already present for environmental scanning. By extracting the localization function to rely solely on scan data and environmental map matching, the system eliminates the need for additional encoder hardware and associated complexity.
3Reliability
If traditional matching without prediction is used, then localization can be performed, but the matching time increases
Solution Approach 1:
The patent performs preliminary action by determining predicted values of current location and attitude based on historical estimated locations and attitudes before performing the matching operation. This prediction step provides an optimized starting point for the matching algorithm, reducing the search space and computational iterations required to achieve accurate localization, thereby decreasing matching time while maintaining or improving accuracy.
Solution Approach 2:
The patent implements feedback by using historical estimated locations and attitudes to generate predicted values that inform the current matching process. This feedback mechanism allows the system to leverage past localization results to accelerate current localization, creating a progressively more efficient system that maintains high accuracy while reducing computational time.
Data Source
AI summary
A device includes an external sensor to scan an environment so as to periodically output scan data, a storage to store an environmental map, and a location estimation device to match the sensor data against the environmental map read from the storage so as to estimate a location and an attitude of the vehicle. The location estimation device determines predicted values of a current location and a current estimation of the vehicle in accordance with a history of estimated locations and estimated attitudes of the vehicle, and performs the matching by using the predicted values.