Towards A General-Purpose Motion Planning for Autonomous Vehicles Using Fluid Dynamics

TL;DR

The paper tackles general-purpose motion planning for autonomous driving by reframing trajectory generation as a fluid-flow problem solved with the lattice Boltzmann method to yield a 3D spatiotemporal vector field. Trajectories are sampled along fluid streamlines while respecting nonlinear, nonholonomic vehicle dynamics, and a cost function balancing safety, comfort, and efficiency selects the best candidate in real time. Compared to model predictive control, the fluid-based approach achieves comparable safety and comfort but superior efficiency, feasibility, and computational cost, without requiring predefined maneuvers. This fluid-dynamics perspective yields a scalable, explainable framework with potential extensions to trajectory prediction and data-driven integration, suitable for diverse driving scenarios and operational design domains.

Abstract

General-purpose motion planners for automated/autonomous vehicles promise to handle the task of motion planning (including tactical decision-making and trajectory generation) for various automated driving functions (ADF) in a diverse range of operational design domains (ODDs). The challenges of designing a general-purpose motion planner arise from several factors: a) A plethora of scenarios with different semantic information in each driving scene should be addressed, b) a strong coupling between long-term decision-making and short-term trajectory generation shall be taken into account, c) the nonholonomic constraints of the vehicle dynamics must be considered, and d) the motion planner must be computationally efficient to run in real-time. The existing methods in the literature are either limited to specific scenarios (logic-based) or are data-driven (learning-based) and therefore lack explainability, which is important for safety-critical automated driving systems (ADS). This paper proposes a novel general-purpose motion planning solution for ADS inspired by the theory of fluid mechanics. A computationally efficient technique, i.e., the lattice Boltzmann method, is then adopted to generate a spatiotemporal vector field, which in accordance with the nonholonomic dynamic model of the Ego vehicle is employed to generate feasible candidate trajectories. The trajectory optimising ride quality, efficiency and safety is finally selected to calculate the imminent control signals, i.e., throttle/brake and steering angle. The performance of the proposed approach is evaluated by simulations in highway driving, on-ramp merging, and intersection crossing scenarios, and it is found to outperform traditional motion planning solutions based on model predictive control (MPC).

Figures (10)

• Figure 1: Classification of motion planning design approaches within the general system diagram of an ADS. The motion planner is fed with the output of the perception module and the predictions of other road users' intentions before producing the signals for the low-level control of the Ego vehicle.
• Figure 2: Spatiotemporal trajectory (brown curves) of the EV moving on a) 1D straight line and b) 2D plane with the vertical axis representing time. The 1D example helps visualise that the speed of the vehicle at each time step is given by the gradient of the spatiotemporal trajectory.
• Figure 3: Example illustration of Cartesian ($X,Y$) and Frenet frame ($s,d$) coordinates in a motorway merging driving scenario. One may notice the variations of the curvature $\kappa$ along the route. The parameters $s_e$ and $d_e$ denote the spatial extent of the generated trajectory in Frenet frame coordinates. The parameter $\theta(s)$ stands for the orientation of the Frenet frame with respect to the global coordinate system $(X,Y)$.
• Figure 4: Kinematic variables and parameters of the dynamic bicycle model used in the equations of motion. The parameter $\delta_f$ is the steering angle of the front wheels, $l_i, i\in\{f,r\}$ is the distance between the centre of gravity and the front $(i=f)$ or the rear wheels $(i=r)$, $C_i$ is the cornering stiffness constant for the wheels, and $\mu_i$ is the wheel friction coefficient. $(X,Y)$ is the global coordinate system and $(x,y)$ refers to the local coordinate system.
• Figure 5: The motion planning system diagram developed in this paper illustrating the relations between its various components. The high-level route planning along with the current and future states of surrounding vehicles (trajectory predictions) are fed into the motion planner by other modules.