Hitch Angle Estimation Using LIDAR Point Cloud Fusion
Solution Overview
Problem
Current systems for autonomous driving of tractor-trailer combinations face challenges in accurately estimating the hitch angle, especially in scenarios where human input is absent and without data communication to external servers, leading to issues with noise and occlusions, which affect the precision and reliability of motion control.
Innovation Solution
The solution employs a multilayer low-noise estimation method using point clouds from LIDAR or RGB-D cameras, combined with speed and orientation measurements, to determine the hitch angle and its rate of change, utilizing a stationary Kalman filter for fusion and outlier rejection, allowing for robust and accurate hitch angle estimation without mechanical contact or external equipment on the trailer.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If point clouds from LIDAR or RGB-D cameras are used for hitch angle estimation, then measurement precision is improved, but noise and occlusions worsen the reliability of the estimation
Solution Approach 1:
The patent introduces a stationary Kalman filter as an intermediary mechanism to process and fuse hitch angle measurements from multiple sources (LIDAR point clouds, RGB-D cameras, and mechanical sensors). This filter mediates between the noisy sensor inputs and the final reliable estimation, reducing the impact of noise and occlusions while maintaining measurement precision through weighted fusion of multiple measurement sources
Solution Approach 2:
The patent combines multiple measurement sources (point clouds from LIDAR, depth information from RGB-D cameras, and mechanical sensor data) into a unified hitch angle estimation system. By merging these diverse data sources and processing them through a stationary Kalman filter, the system achieves both high measurement precision and improved reliability, as the fusion compensates for the weaknesses of individual sensors
2Reliability
If mechanical contact sensors are used for hitch angle measurement, then reliability is improved, but device complexity and maintenance requirements worsen
Solution Approach 1:
The patent replaces mechanical contact sensors with optical sensing systems (LIDAR and RGB-D cameras) that use light-based measurement instead of mechanical contact. This substitution eliminates the mechanical wear and complexity associated with traditional sensors while maintaining measurement reliability through advanced signal processing and data fusion algorithms that compensate for the non-contact measurement approach
3Measurement precision
If external equipment on the trailer is added for hitch angle estimation, then measurement precision is improved, but ease of operation worsens due to installation and configuration requirements
Solution Approach 1:
The patent implements a self-service approach where the tractor-mounted sensors (LIDAR and RGB-D cameras) automatically track and measure the trailer's position and orientation without requiring any equipment installation on the trailer itself. The system uses the tractor's own sensing capabilities to perform all measurements, eliminating the need for external trailer equipment while maintaining high measurement precision through advanced point cloud processing and geometric calculations
Data Source
Figure 1
Figure 2
Figure 3
AI summary
Disclosed is an arrangement and a method for hitch-angle estimation for a tractor-trailer combination. The arrangement comprising one or more sensors for providing one or more point clouds comprising a representation for the trailer as seen from one or more measurement positions at the tractor, a first measurement for the speed of the tractor, and a second measurement for an orientation of the tractor. The arrangement also comprises a computing unit configured to obtain one or more first values indicating the hitch angle between the tractor and the trailer, determined from the one or more point clouds, obtain a second value indicating the rate of change for the hitch angle, determined from the first and the second measurement, and fuse the one or more first values and the second value to provide an estimate for the hitch angle between the tractor and the trailer.