Autonomous Driving Trajectory Control With Priority-Based Feature Switching
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing ADAS systems require high computing power and complex state machines for processing environment information to generate multiple trajectory lines, leading to low scalability and inefficient feature switching.
Innovation Solution
An intelligent driving method that determines a control point based on surrounding environment and vehicle status information, using preset priority information to generate a control trajectory without pre-processing multiple trajectory lines, reducing computing requirements and enabling smooth feature transitions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If multiple trajectory lines are generated by processing environment information, then different driving features can be supported, but computing power requirements increase
Solution Approach 1:
The patent uses a single control trajectory to serve multiple driving features (ACC, LKA, TJA, ALC, SLC, CAA, LCC, OBF) rather than generating separate trajectory lines for each feature. This universal approach reduces computing power requirements while maintaining support for diverse driving functions.
Solution Approach 2:
The patent merges the generation of multiple feature-specific trajectories into a single control trajectory generation process. By combining environmental information processing and control trajectory generation into one unified workflow, the system eliminates redundant computations and reduces overall computing power demands.
2Adaptability or versatility
If a state machine is used to switch between different features, then feature switching can be implemented, but system complexity increases
Solution Approach 1:
The patent extracts the feature switching logic from a complex state machine architecture and replaces it with a simplified priority-based selection mechanism. This extraction removes the need for complex state transitions while preserving the ability to switch between different driving features dynamically.
Solution Approach 2:
The patent changes the control parameter from complex state machine transitions to simple priority level adjustments. By using priority information associated with different objects in the environment, the system can dynamically switch features through parameter changes rather than complex state transitions, reducing overall system complexity.
Data Source
Figure 1~2
Figure 3
Figure 4(a)~4(f)
AI summary
An intelligent driving method and apparatus, a storage medium, and a computer program are disclosed, which may be used for advanced assisted driving and autonomous driving. The method includes: obtaining surrounding environment information of an ego vehicle at a current moment and status information of the ego vehicle at the current moment; determining a control point at the current moment based on the environment information and the status information; and generating a control trajectory at the current moment based on the control point, a control trajectory at a historical moment, and the status information, where the control trajectory represents a trajectory used to guide traveling of the ego vehicle. In this way, impact of the environment information on features is concentrated on the control point, and the control trajectory at a current moment is generated by using the control point, to meet control requirements in different scenarios. In addition, the environment information does not need to be processed in advance to generate a plurality of trajectory lines, thereby greatly reducing a requirement on a computing power. Switching between different features is implemented based on priorities of a plurality of information sources and corresponding weight information, thereby ensuring smooth transition between the plurality of features or scenarios, and improving performance of an advanced driver assistant system.