Adaptive Vehicle Control Using Probabilistic Motion Planning
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Autonomous and semi-autonomous vehicles face challenges in motion planning due to the non-convexity of the motion-planning problem, leading to locally optimal solutions that may not be globally optimal, and require efficient path planning that balances safety and comfort while handling complex scenarios.
Innovation Solution
An integrated architecture combining a low-rate, long-term sampling-based motion planning algorithm with a high-rate, highly reactive optimization-based predictive vehicle controller, utilizing probabilistic methods and particle filtering to compute confidence bounds and adapt the tracking cost function, allowing for automatic tuning of the tradeoff between safety and comfort objectives.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If sampling-based motion planning methods are used, then computational speed is improved, but path naturalness deteriorates
Solution Approach 1:
The system dynamically adjusts the balance between sampling-based exploration and optimization-based refinement based on computational resources available and mission requirements. The motion planner transitions from purely sampling-based methods to a hybrid approach that incorporates optimization techniques when resources permit, enabling the path to evolve from computationally efficient but less natural to more natural while still feasible paths.
Solution Approach 2:
The patent introduces an intermediate optimization layer that acts as a mediator between the sampling-based motion planner and the final executed path. This intermediary takes the rapidly generated but potentially unnatural paths from sampling methods and applies optimization techniques to smooth them while maintaining feasibility, thus resolving the contradiction between speed and naturalness.
2Productivity
If deterministic sampling methods are used, then computational efficiency is improved, but path quality deteriorates
Solution Approach 1:
The system merges deterministic sampling methods with probabilistic sampling approaches and optimization techniques. By combining these methods, the system retains the computational efficiency of deterministic sampling while incorporating the exploratory benefits of probabilistic methods and the quality improvement from optimization, thus resolving the contradiction between efficiency and path quality.
Solution Approach 2:
The motion planning system uses a composite approach combining multiple planning methodologies (deterministic sampling, probabilistic sampling, optimization) similar to how composite materials combine different substances to achieve superior properties. Each method contributes its strengths to the final path solution, achieving both efficiency and quality.
3Reliability
If optimization-based methods are used for motion planning, then path quality is improved, but computational load increases
Solution Approach 1:
The motion planning process is segmented into distinct stages: an initial sampling-based phase that generates feasible paths quickly, followed by an optional optimization phase that refines path quality. This segmentation allows the system to apply computational resources selectively, using optimization only when needed rather than always, thus reducing overall computational load while maintaining path quality where necessary.
Solution Approach 2:
The system dynamically changes parameters such as optimization iteration count, sampling density, and refinement depth based on computational resource availability and mission criticality. When computational load is high, parameters are adjusted to reduce optimization intensity; when resources are abundant, parameters are increased to enhance path quality, thus resolving the contradiction between quality and computational load.
Data Source
AI summary
A control system controls a vehicle using a probabilistic motion planner and an adaptive predictive controller. The probabilistic motion planner produces a sequence of parametric probability distributions over a sequence of target states for the vehicle with parameters defining a first and higher order moments. The adaptive predictive controller optimizes a cost function over a prediction horizon to produce a sequence of control commands to one or multiple actuators of the vehicle. The cost function balances a cost of tracking of different state variables in the sequence of the target states defined by the first moments. The balancing is performed by weighting different state variables using one or multiple of the higher order moments of the probability distribution.