Autonomous Driving Path Determination Using Adaptive Likelihood Field
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional path determining methods for autonomous driving vehicles face a trade-off between optimization and computational efficiency, generating a large number of candidate paths which increases computation and reduces efficiency, and fail to effectively reduce the number of paths to select the optimal one.
Innovation Solution
A path determining apparatus and method that periodically generates autonomous driving paths, detects an adaptive likelihood field (ALF) based on road and obstacle information, applies weights to each path to determine the final path, using a path generator, ALF detector, weight setter, and path determiner to prioritize paths with low collision probability.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If a large number of candidate paths are generated to find a stable and optimal path, then path optimization is improved, but the amount of computation increases and computational efficiency is lowered
Solution Approach 1:
The patent extracts and removes paths with low probabilities from the candidate paths early in the process. By calculating probabilities based on road shape information and obstacle information, the system eliminates unlikely paths before they consume computational resources in subsequent optimization steps, thereby maintaining path optimization quality while reducing overall computation
Solution Approach 2:
The patent performs preliminary probability calculation and path filtering before the main path optimization process. By pre-assessing each candidate path's likelihood based on road and obstacle data, the system prepares a reduced set of high-probability paths for subsequent optimization, avoiding the need to process all candidate paths through computationally intensive algorithms
2Productivity
If the number of candidate paths is reduced to improve computational efficiency, then processing speed is improved, but the ability to select the optimal path is compromised
Solution Approach 1:
The patent implements a feedback mechanism where path probabilities are calculated based on road shape information and obstacle information, and this probability feedback is used to guide subsequent path generation and optimization. The system continuously refines path probabilities as new information becomes available, ensuring that the reduced set of candidate paths still contains the optimal path
Solution Approach 2:
The patent changes the parameter of path probability by incorporating multiple factors including road shape matching degree, obstacle distance, and obstacle type. By adjusting and refining these probability parameters dynamically, the system ensures that even with fewer candidate paths, the optimal path is identified through more accurate probability assessment
Data Source
AI summary
A path determining apparatus for an autonomous driving vehicle includes a path generator periodically generating a plurality of autonomous driving paths, an adaptive likelihood field (ALF) detector detecting an ALF on a road based on road information and obstacle information, a weight setter applying the ALF detected by the ALF detector to each of the autonomous driving paths generated by the path generator to set a weight with respect to each of the autonomous driving paths, and a path determiner determining a final autonomous driving path based on the weight set with respect to each of the autonomous driving paths by the weight setter.