Autonomous Vehicle Trajectory Planning With Sensor Fusion and NMPC
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Autonomous all-terrain vehicles face challenges in navigating unknown dynamic environments with heavy or difficult payloads, requiring effective obstacle detection and trajectory planning to avoid collisions while following a goal in complex terrains.
Innovation Solution
A system comprising a first sensor (image capture) and a second sensor (LiDAR) for obstacle detection, with a processor performing matching and modeling existence probabilities, a trajectory planner using non-linear model predictive control (NMPC), and a low-level PID controller to implement vehicle actuator commands for efficient path planning.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If multiple sensor types are used for obstacle detection, then measurement precision is improved, but device complexity increases
Solution Approach 1:
The system segments the detection task by assigning different sensor types (image capture sensor and LiDAR sensor) to detect different aspects of obstacles. The image capture sensor detects visual features while the LiDAR sensor detects range and depth information, dividing the complex detection problem into manageable parts that can be processed independently and then integrated.
Solution Approach 2:
The system merges data from multiple sensor types through sensor fusion. The processor combines the detection results from the image capture sensor and LiDAR sensor to create a comprehensive obstacle model, merging complementary information to achieve higher precision than any single sensor could provide alone.
2Reliability
If sensor fusion and matching algorithms are implemented, then reliability of obstacle tracking is improved, but computational requirements and device complexity increase
Solution Approach 1:
The system implements feedback through the matching algorithm that continuously compares detected obstacles with tracked obstacles. The Hungarian matching algorithm provides feedback by identifying correspondences between sensor detections and existing track hypotheses, allowing the system to refine its tracking estimates and maintain reliable obstacle tracking through iterative correction.
Solution Approach 2:
The system performs preliminary action by pre-defining the matching framework and probability models before actual tracking begins. The Bayesian probability model and Hungarian matching algorithm are established in advance, allowing the processor to efficiently evaluate and update track hypotheses without complex real-time computations during critical tracking operations.
3Adaptability or versatility
If non-linear model predictive control is used for trajectory planning, then navigation performance in dynamic environments is improved, but computational complexity increases
Solution Approach 1:
The system applies dynamics by using non-linear model predictive control that can adapt to changing environmental conditions and vehicle states. The NMPC framework dynamically adjusts the trajectory plan based on current obstacle positions, vehicle velocity, and environmental constraints, allowing the controller to handle the dynamic and uncertain nature of off-road navigation while optimizing navigation performance.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
Enables autonomous vehicles to efficiently navigate and follow targets while mitigating collisions with static and dynamic obstacles in unknown environments, optimizing speed and steering while minimizing control effort.
Implementation Method 1
The first sensor may be an image capture sensor and may detect an obstacle and a goal
Implementation Method 2
The second sensor may be a light detection and ranging (LiDAR) sensor and may detect the obstacle and the goal
Data Source
AI summary
An autonomous vehicle capable of trajectory prediction may include a first sensor, a second sensor, a processor, a trajectory planner, a low-level controller, and vehicle actuators. The first sensor may be of a first sensor type and may detect an obstacle and a goal. The second sensor may be of a second sensor type and may detect the obstacle and the goal. The processor may perform matching on the obstacle detected by the first sensor and the obstacle detected by the second sensor, model an existence probability of the obstacle based on the matching, and track the obstacle based on the existence probability and a constant velocity model. The trajectory planner may generate a trajectory for the autonomous vehicle based on the tracked obstacle, the goal, and a non-linear model predictive control (NMPC). The low-level controller may implement the trajectory for the autonomous vehicle by driving vehicle actuators.