Joint Motion Planning for Autonomous Vehicles
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Autonomous vehicles face challenges in planning safe trajectories while avoiding collisions and considering passenger comfort, especially at high speeds in complex environments, as existing motion planning methods fail to effectively account for the intentions and trajectories of multiple vehicles and road constraints.
Innovation Solution
A method for joint motion planning and trajectory estimation using spline functions to model the intentions and trajectories of vehicles, determining a cost function that incorporates road constraints and passenger comfort, and generating control signals to minimize speed and steering changes, allowing for the simultaneous planning of the host vehicle's and nearby vehicles' trajectories.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If the host vehicle travels at high speed to maintain productivity, then the productivity is improved, but the difficulty of detecting and measuring other vehicles' intentions and trajectories increases
Solution Approach 1:
The system performs preliminary trajectory estimation for multiple potential intentions of other vehicles before the host vehicle reaches the decision point. By pre-computing cost functions for various intention scenarios (lane changes, straight continuation, turns), the system prepares multiple candidate trajectories in advance, allowing for faster response at high speeds without sacrificing estimation accuracy.
Solution Approach 2:
The trajectory estimation system dynamically adapts to changing conditions by continuously updating cost functions based on real-time sensor data and observed vehicle behaviors. The system adjusts the weighting of different intention probabilities and re-evaluates candidate trajectories as the host vehicle moves, enabling accurate trajectory prediction even at high speeds where conditions change rapidly.
2Reliability
If multiple vehicle intentions and trajectories are considered to improve safety, then the reliability is improved, but the device complexity increases
Solution Approach 1:
The motion planning system is segmented into modular components: a cost function generator that creates individual cost functions for each vehicle and intention type, a trajectory evaluator that assesses candidate paths, and a optimization module that selects the best trajectory. This segmentation allows the system to handle multiple vehicles independently through standardized cost function templates, reducing overall complexity while maintaining comprehensive safety analysis.
Solution Approach 2:
The system uses parameterized cost functions that can be adjusted based on vehicle type, intention probability, and environmental context. By changing parameters such as weighting factors for different safety constraints, prediction horizons, and intention probabilities, the system can adapt to different scenarios without requiring fundamentally different algorithms, thus managing complexity while improving reliability.
3Reliability
If passenger comfort is considered in addition to collision avoidance, then the reliability is improved, but the device complexity increases
Solution Approach 1:
The system merges collision avoidance requirements and passenger comfort requirements into a unified cost function framework. Safety constraints (collision avoidance, lane keeping) and comfort constraints (smooth acceleration, minimal steering changes) are combined as weighted terms in the same optimization objective. This merging allows simultaneous consideration of both requirements without requiring separate complex systems, as the unified cost function can be optimized in a single trajectory selection process.
4Manufacturing precision
If spline functions are used to describe trajectory, speed, and acceleration for each possible intention, then the manufacturing precision is improved, but the loss of time increases
Solution Approach 1:
The system computes spline function parameters partially by focusing only on the most likely intentions based on initial probability assessments. Instead of fully computing all possible intention trajectories with high precision, the system computes detailed spline parameters only for top candidate intentions, using coarser approximations for less likely scenarios. This partial computation approach maintains sufficient accuracy for decision-making while significantly reducing computation time.
Data Source
AI summary
Aspects of the disclosure are related to a method, apparatus and system for joint motion planning and trajectory estimation, comprising: determining a cost function to describe system kinematics comprising trajectories, speeds, and accelerations of a host vehicle and of one or more other vehicles for each possible intention of the host vehicle and of the other vehicles, wherein the trajectories are described with spline functions; and determining jointly the trajectories of the host vehicle and of the other vehicles.