TL;DR
This paper presents two distinct models for multi-unit heavy vehicles: a complex simulation model capturing detailed dynamics and a simplified analytical model for controller design. These models are crucial for developing automated guidance systems for trucks, addressing challenges like inter-unit coupling and complex vehicle configurations.
What Problem Does This Paper Address?
The increasing interest in Automated Highway Systems (AHS) for commercial heavy vehicles stems from potential benefits in operational cost and safety. However, automating these vehicles is more complex than passenger cars due to their multi-unit nature. Heavy vehicles, especially those with multiple trailers, exhibit intricate dynamics, including phenomena like jack-knifing and trailer oscillations, which are not present in single-unit vehicles. Existing models often struggle to capture these complexities comprehensively or provide the simplified forms needed for controller design. This paper aims to bridge this gap by developing both a detailed simulation model and a practical, analytically derived control model for general multi-unit heavy vehicles.
Methods
The authors propose two main modeling approaches:
- Complex Simulation Model: This model treats the heavy vehicle as a system of interconnected rigid bodies. It incorporates longitudinal and lateral dynamics, considering the environment's interaction. A key innovation is the detailed modeling of how roll and pitch motions of adjacent units become coupled due to hitching mechanisms. This coupling is analyzed from a geometric perspective, and its "backward propagation" through the vehicle train is investigated. The model uses Newtonian mechanics for translational motion and Euler's equations for rotational motion, allowing for modular integration of various suspension and tire models.
- Simple Control Model: This model simplifies the dynamics by focusing only on the planar translational and yaw motions of each unit. The authors present a general method for deriving these models without relying on the complex simulation model. They utilize Lagrangian mechanics for yaw motions to avoid explicit handling of constraint forces, resulting in closed-form analytical equations of motion. This simplified representation is ideal for designing controllers for automated guidance.
Datasets
Not discussed in the paper.
Key Insights and Ideas
- Coupling of Roll and Pitch: A significant insight is the geometric modeling of how roll and pitch motions of connected units become coupled, particularly in fifth-wheel connections. Understanding and modeling this "backward propagation" is crucial for accurate simulation and control.
- Two-Tiered Modeling Approach: The paper's strength lies in offering both a high-fidelity simulation model and a simplified analytical model. This dual approach caters to different needs: detailed analysis and simulation versus practical controller development.
- Generalizability: The models are designed for general multi-unit heavy vehicles, accommodating various hitching mechanisms (fifth-wheel, A-dolly, B-dolly, pintle joint) and configurations, making them broadly applicable.
- Lagrangian Mechanics for Yaw Control: The use of Lagrangian mechanics for the yaw motion in the control model is a clever way to derive analytical equations without explicitly dealing with constraint forces, simplifying controller design.
Key Results
The paper successfully derives two sets of models:
- Complex Simulation Model: This model provides a detailed representation of vehicle dynamics, including the coupled roll and pitch motions, suitable for in-depth simulations of longitudinal and lateral dynamics for AHS applications.
- Simple Control Model: This model yields closed-form analytical equations of motion for translational and yaw dynamics, which are essential for designing controllers for automated guidance. An example of a lateral control model for a tractor-semitrailer is presented, demonstrating the practical application of the derived equations.
Limitations and Future Work
- Linear Tyre Model Assumption: In the example lateral control model, a linear tyre model is used. The authors acknowledge that for more accurate control, non-linear tyre models might be necessary.
- Small Motion Assumptions: For certain aspects, like roll and pitch motions, small motion assumptions are made. The paper hints at the need to investigate the relationship between nominal and actual pitch motions, suggesting that deviations from small angles could be a point for further study.
- Future Work: The paper lays the groundwork for developing automated guidance systems. Future work would involve implementing these models in control algorithms, testing them in simulations and real-world scenarios, and potentially extending them to include more complex dynamics or environmental factors. The modular nature of the simulation model also suggests it can be readily adapted with different suspension and tire models.
Summary updated