Autonomous Vehicle Self-Position Estimation via Dynamic Range Adjustment
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional autonomous vehicles expand the range of estimable self-position regardless of error size, leading to decreased reliability in self-position estimation.
Innovation Solution
An autonomous vehicle equipped with a first sensor for environmental information and a control unit that includes a first estimation unit using a probabilistic method and a second estimation unit using a matching method, where the control unit adjusts the second estimate range based on the first estimate value to improve self-position estimation accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Area of stationary object
If the autonomous vehicle expands the range of estimable self-position when environmental information is insufficient, then the coverage of position estimation is improved, but the reliability of self-position estimation decreases
Solution Approach 1:
The patent applies dynamics by making the second estimate range adjustable rather than fixed. The control unit dynamically changes the second estimate range based on the first estimate value from the probabilistic method. When the first estimate value indicates high confidence, the second estimate range is reduced; when confidence is low, the range is expanded. This dynamic adjustment resolves the contradiction by adapting the estimation range to current conditions, maintaining reliability while preserving coverage when needed.
Solution Approach 2:
The patent changes the parameter of the estimate range based on the first estimate value. By using the output from the probabilistic method (first estimation unit) as a basis for adjusting the matching method's search range (second estimation unit), the system transforms a static parameter into a variable one that responds to environmental conditions and information quality, thereby resolving the reliability-coverage tradeoff.
2Device complexity
If the autonomous vehicle uses a fixed second estimate range for matching method calculation, then the computational complexity is reduced, but the accuracy of self-position estimation decreases
Solution Approach 1:
The system dynamically adjusts the second estimate range based on the first estimate value, creating an adaptive computational approach. Rather than using a fixed range that must be conservative to ensure accuracy, the system can expand or contract the search range based on current confidence levels, optimizing the balance between computational load and estimation accuracy in real-time.
Solution Approach 2:
The probabilistic method (first estimation unit) performs preliminary estimation before the matching method is applied. This preliminary action provides information about the likely position and confidence level, which is then used to pre-adjust the second estimate range for the matching method. This preliminary estimation step enables the system to focus computational resources more efficiently, reducing the overall complexity while maintaining or improving accuracy.
Data Source
AI summary
An autonomous vehicle includes: a first sensor which obtains environmental information on surrounding environment of the autonomous vehicle; and a control unit which controls a drive unit based on a self position. The control unit includes: a first estimation unit which calculates a first estimate value indicating an estimated self position, by estimation using a probabilistic method based on the environmental information; and a second estimation unit which calculates a second estimate value indicating an estimated self position, by estimation using a matching method based on the environmental information, and the control unit changes, according to the first estimate value, a second estimate range for the calculating of the second estimate value by the second estimation unit, and controls the drive unit using the second estimate value as the self position.