Probabilistic Autonomous Driving Control for Obstacle-Aware Path Planning
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current autonomous vehicle control approaches deal with uncertainties using conservative solutions to minimize traffic issues, which can result in reduced vehicle performance, such as lower speeds and milder steering maneuvers.
Innovation Solution
A probabilistic control strategy that monitors the environment around the autonomous vehicle, calculates the probability of potential events, and adjusts the planned path using a contingency model predictive control (MPC) to avoid obstacles while minimizing performance issues.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If conservative control solutions are used to minimize traffic issues, then safety and obstacle avoidance are improved, but vehicle performance deteriorates due to lower speeds and milder steering maneuvers
Solution Approach 1:
The system dynamically changes control parameters based on real-time probability assessments of potential events. Instead of using fixed conservative parameters, the controller adjusts speed limits, steering aggressiveness, and planning horizons according to the calculated likelihood of obstacles or emergencies, allowing optimal performance when safe while maintaining safety when needed
Solution Approach 2:
The control system transitions from static conservative planning to dynamic adaptive planning. The probabilistic event assessment continuously updates the control strategy based on current environmental conditions, sensor data, and predicted trajectories of other road users, enabling the vehicle to adapt its behavior in real-time rather than following predetermined conservative paths
2Reliability
If conservative control solutions are used to avoid obstacles, then collision risk is reduced, but travel time increases due to slower response and milder maneuvers
Solution Approach 1:
The system performs preliminary probabilistic assessment of potential events and prepares multiple contingency plans in advance. By calculating the probability of various scenarios beforehand and having pre-computed escape trajectories ready, the system can execute rapid responses when events materialize, avoiding the time loss associated with reactive conservative planning
3Productivity
If aggressive reactions to potential events are commanded, then response effectiveness is improved, but vehicle stability and safety deteriorate due to excessive maneuvering
Solution Approach 1:
The control system applies different levels of aggressiveness to different aspects of vehicle control based on the specific event and its probability. Instead of uniformly aggressive or conservative control, the system selectively applies aggressive maneuvers only when and where necessary (e.g., sharp steering for high-probability obstacles, gentle braking for low-probability events), maintaining overall vehicle stability while achieving effective response
Data Source
AI summary
A method for controlling an autonomous vehicle includes receiving road data. The road data includes information about a plurality of potential events along the road ahead of the autonomous vehicle. The method further includes determining, in real time, a probability that the plurality of potential events along the road ahead of the autonomous vehicle will occur while the autonomous vehicle moves along the road and determining, in real time, an adjusted planned path using a probabilistic predictive control that takes into account the probability that the plurality of potential events along the road ahead of the autonomous vehicle will occur. Further, the method includes controlling the autonomous vehicle to cause the autonomous vehicle to autonomously follow the adjusted planned path.