Adaptive Path Following for Heavy-Duty Vehicle Off-Tracking Control
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing path following algorithms for heavy-duty vehicles, such as trucks and semitrailers, fail to effectively adjust the preview distance parameter based on lateral deviation and vehicle speed, leading to inadequate control effort and poor off-tracking performance, especially when close to the target path.
Innovation Solution
The method adjusts the preview distance based on multiple criteria, including lateral deviation and longitudinal velocity, using tuning parameters to maintain consistent control effort and improve path following performance, particularly through vector field guidance methods.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If traditional path following algorithms are used, then implementation is simple, but they do not account for vehicle dynamics and cannot track paths accurately
Solution Approach 1:
The algorithm transitions from static path following to dynamic path following by continuously adapting to changing vehicle states and road conditions. The controller dynamically adjusts path following commands based on real-time vehicle dynamics data, allowing accurate tracking while maintaining computational efficiency through incremental updates rather than complete recalculation.
Solution Approach 2:
The system implements feedback by continuously monitoring vehicle dynamics parameters (lateral acceleration, yaw rate, steering angle) and using this information to adjust path following commands. The feedback loop compares actual vehicle behavior with desired path behavior and corrects deviations, improving accuracy without requiring overly complex algorithms.
2Manufacturing precision
If path following commands are generated without considering vehicle dynamics, then computational load is low, but tracking accuracy deteriorates in dynamic conditions
Solution Approach 1:
The system performs preliminary calculations by pre-computing ideal path following commands based on geometric path parameters. These pre-computed commands serve as a foundation that is then adjusted with minimal additional computations based on real-time vehicle dynamics, reducing the computational energy required for real-time control while maintaining accuracy.
Solution Approach 2:
The algorithm changes parameters incrementally by adjusting path following commands based on variations in vehicle dynamics parameters. Rather than recalculating entire trajectories, the system makes parameter adjustments proportional to changes in vehicle state, reducing computational energy while improving tracking accuracy under dynamic conditions.
3Adaptability or versatility
If fixed path following commands are used, then control is simple, but the vehicle cannot adapt to changing road conditions
Solution Approach 1:
The control system transitions from fixed to dynamic path following by continuously adapting commands based on real-time vehicle dynamics and road conditions. The system maintains simplicity through structured algorithms that process dynamic data in a systematic manner, achieving adaptability without excessive complexity.
Solution Approach 2:
The algorithm achieves multi-functionality by using a unified control framework that handles multiple functions: path following, vehicle dynamics compensation, and road condition adaptation. This universal approach avoids the need for separate specialized controllers for each function, maintaining control system simplicity while enhancing adaptability.
Data Source
Figure 1~2
Figure 3~4
Figure 5~6
AI summary
A method for controlling a heavy-duty vehicle (100) to follow a reference path (P), the method comprising obtaining (S1) the reference path (P) to be followed by the vehicle (100), determining (S2) a goal point (G) along the path (P) to be used as a steering reference from a vehicle location (x) in vicinity of the path (P), where the goal point (G) is distanced along the path (P) by a preview distance (Dp) measured from a reference location (x, Go) associated with the vehicle location (x), where the preview distance (Dp) is determined at least partly based on a lateral deviation (y) of the vehicle location (x) from the reference path (P), such that the preview distance (Dp) increases with an increasing lateral deviation (y) from the reference path (P), and decreases with a decreasing lateral deviation (y), and controlling (S3) the vehicle (100) on the basis of the goal point (G).