Autonomous Vehicle Trajectory Planning Using Pre-stored Coefficients
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing autonomous vehicle navigation systems fail to consider terrain conditions, lateral shift, and total time in determining optimal trajectories, leading to inefficient navigation and energy usage, particularly when dealing with obstacles or changing lanes.
Innovation Solution
A method and system that detect predefined conditions, such as obstacles, by analyzing environment data from sensors, determining minimum lateral shift, and using pre-stored co-efficient value sets to calculate linear and angular velocities, generating trajectories based on position coordinates and orientation, and weighting them by cumulative orientation variation and travel time to determine an optimal navigation path.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If existing trajectory planning techniques are used that depend on trajectories of other moving objects, then the system can generate trajectories based on road cross sections and other objects, but the system fails to consider terrain conditions, lateral shift, and total time which are crucial for navigation
Solution Approach 1:
The patent segments the trajectory planning process into multiple independent components: detecting predefined conditions (obstacles, lane changes), determining minimum lateral shift, calculating co-efficient values, generating multiple trajectory options, and evaluating each trajectory based on cumulative orientation variation and time. This segmentation allows each component to be optimized independently while ensuring comprehensive consideration of all navigation factors including terrain conditions, lateral shift, and total time.
Solution Approach 2:
The patent changes the evaluation parameters from generic cost functions to specific performance metrics including cumulative orientation variation, total time, and energy consumption. By varying these parameters and computing weightage for each trajectory option, the system adapts to different navigation scenarios and vehicle performance characteristics, thereby improving both adaptability and reliability.
2Adaptability or versatility
If the system generates multiple trajectories with different co-efficient value sets, then the navigation can be optimized for specific vehicle performance, but the computational complexity and time required to determine the optimal trajectory increases
Solution Approach 1:
The patent applies preliminary action by pre-calculating and storing co-efficient value sets that correspond to different minimum lateral shift values. When a predefined condition is detected, the system retrieves relevant pre-stored co-efficient values instead of calculating them in real-time. This significantly reduces computational complexity while maintaining the ability to generate vehicle-specific optimized trajectories.
Solution Approach 2:
The patent replaces complex real-time mechanical computation with a lookup-based system using pre-stored co-efficient values. Instead of performing heavy mathematical calculations during navigation, the system substitutes this with efficient data retrieval and comparison operations, thereby reducing computational burden while preserving trajectory optimization capabilities.
3Use of energy by moving object
If the system minimizes lateral shift for obstacle avoidance and lane changes, then energy usage is reduced, but the ability to handle various scenarios such as overtaking and changing lanes is limited
Solution Approach 1:
The patent applies dynamics by making the lateral shift value adaptive rather than fixed. The system determines a minimum lateral shift based on the specific predefined condition detected (obstacle, lane change, overtaking scenario). For each scenario, the system calculates an appropriate lateral shift that balances energy efficiency with maneuverability requirements. This dynamic adjustment allows the vehicle to minimize energy consumption during routine operations while maintaining full capability for complex maneuvers when needed.
Data Source
AI summary
Disclosed subject matter relates to a field of vehicle navigation system that performs a method for determining an optimal trajectory for navigation of an autonomous vehicle. A trajectory determining system associated with the autonomous vehicle may detect an occurrence of a predefined condition related to diversion based on environment data. Further, minimum lateral shift required from a predefined distance for handling the detected predefined condition is determined and co-efficient value sets corresponding to the minimum lateral shift are determined based on pre-stored values that are generated based on a trial run. For each co-efficient value set, velocity and position data is determined at a plurality of time instances based on which one or more trajectories corresponding to each co-efficient value set are generated. Finally, an optimal trajectory is determined among the generated trajectories cumulative variation in orientation and total time required for navigating along the corresponding trajectory.