Autonomous Driving Test Scenario Prioritization via Weighted Risk Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The existing methods for testing autonomous driving systems face significant pressure and inefficiency due to the need to test tens of thousands of scenarios without distinguishing their varying degrees of importance, leading to prolonged testing times and reduced efficiency.
Innovation Solution
A method and apparatus that analyze scenario description information to determine scenario risk, probability, and complexity, calculating a scenario weight to prioritize testing scenarios, thereby optimizing test periods and evaluating driving scores through weighted averaging.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If all testing scenarios are tested without distinction, then testing coverage is comprehensive, but testing time and resource consumption increase significantly
Solution Approach 1:
The patent segments the testing scenarios into different priority levels based on risk, probability, and complexity. By dividing the comprehensive test set into high-priority, medium-priority, and low-priority groups, the system can focus testing resources on critical scenarios while reducing overall testing time without sacrificing essential coverage.
Solution Approach 2:
The patent introduces a scenario weight parameter that combines risk level, probability level, and complexity level to quantify the importance of each testing scenario. This parameter transformation enables the system to prioritize scenarios objectively, allowing comprehensive coverage to be achieved with reduced time by focusing on high-weight scenarios.
2Reliability
If all testing scenarios are tested without distinction, then no critical scenarios are missed, but testing efficiency decreases
Solution Approach 1:
The patent transforms multiple qualitative attributes (risk, probability, complexity) into a quantitative scenario weight parameter. This enables efficient prioritization and selection of testing scenarios, significantly improving testing efficiency while maintaining reliable coverage of critical scenarios through the weighted scoring system.
Solution Approach 2:
The patent applies different testing intensities and resource allocations to different scenarios based on their calculated weights. High-weight scenarios receive focused attention and more rigorous testing, while low-weight scenarios receive minimal testing, optimizing the overall testing efficiency while ensuring critical areas are thoroughly covered.
3Productivity
If testing focuses only on high-priority scenarios, then testing efficiency improves, but comprehensive coverage may be compromised
Solution Approach 1:
The patent uses the scenario weight parameter (derived from risk, probability, and complexity) to determine an optimal testing mix. By setting appropriate weight thresholds and testing depth for different priority levels, the system achieves both high efficiency and comprehensive coverage, ensuring that even low-priority scenarios receive baseline testing while critical scenarios receive intensive testing.
Data Source
AI summary
A method, an electronic device and a computer-readable storage medium for testing an autonomous driving system which relate to the technical field of autonomous driving are proposed. An embodiment for testing the autonomous driving system includes: obtaining scenario description information of a testing scenario; analyzing the scenario description information, and determining a scenario risk, a scenario probability and a scenario complexity corresponding to the testing scenario; obtaining a scenario weight of the testing scenario according to the scenario risk, scenario probability and scenario complexity; determining a test period corresponding to the scenario weight, where the test period is used for the autonomous driving system being tested in the testing scenario. The technical solution may reduce the testing pressure of the autonomous driving system and improve the testing efficiency of the autonomous driving system.