Abstract:We propose a learning-based Control Barrier Function (CBF) to reduce conservatism in collision avoidance of car-like robots. Traditional CBFs often use Euclidean distance between robots' centers as safety margin, neglecting headings and simplifying geometries to circles. While this ensures smooth, differentiable safety functions required by CBFs, it can be overly conservative in tight environments. To address this limitation, we design a heading-aware safety margin that accounts for the robots' orientations, enabling a less conservative and more accurate estimation of safe regions. Since the function computing this safety margin is non-differentiable, we approximate it with a neural network to ensure differentiability and facilitate integration with CBFs. We describe how we achieve bounded learning error and incorporate the upper bound into the CBF to provide formal safety guarantees through forward invariance. We show that our CBF is a high-order CBF with relative degree two for a system with two robots whose dynamics are modeled by the nonlinear kinematic bicycle model. Experimental results in overtaking and bypassing scenarios reveal a 33.5 % reduction in conservatism compared to traditional methods, while maintaining safety. Code: <a class="link-external link-https" href="https://github.com/bassamlab/sigmarl" rel="external noopener nofollow">this https URL</a>
What problem does this paper attempt to address?
The problem that this paper attempts to solve is that in vehicle - like robots such as self - driving vehicles, the traditional safety distance calculation method based on control barrier functions (CBF) is too conservative, especially in narrow or dense environments, which limits the efficient navigation ability of robots. Traditional methods usually simplify the geometry of robots to circles and use the Euclidean distance between the centers of robots as the safety margin, ignoring the actual shape and orientation of robots. Although this simplification ensures the smoothness and differentiability of safety analysis, it may lead to overly conservative robot behavior in complex environments and the inability to complete tasks efficiently.
To overcome these limitations, the paper proposes a learning - based control barrier function (CBF) that combines a minimum translation vector (MTV) safety margin considering the robot's orientation. Through this method, the safety area can be estimated more accurately, the conservatism in collision avoidance can be reduced, and the robot can navigate more efficiently without sacrificing safety. Specifically, the main contributions of the paper include:
1. **Proposing a non - differentiable, orientation - considering safety margin**: The MTV - based method takes into account the orientation and geometry of vehicle - like robots and provides a more accurate safety area estimate. And by training a differentiable neural network to approximate this safety margin, its differentiability is ensured so as to be integrated with the CBF framework.
2. **Establishing the theoretical basis of high - order CBF**: It is proved that the proposed learning - based and MTV - based safety margin is a high - order CBF with a relative degree of two, which is suitable for a two - robot system described by a nonlinear dynamic bicycle model.
3. **Verifying the theoretical results through numerical experiments**: Numerical simulations were carried out in overtaking and detouring scenarios, showing that compared with the traditional center - to - center (C2C) safety margin - based method, the conservatism of the new method is reduced by 33.5% while maintaining safety.
Through these contributions, the paper aims to improve the safety and efficiency of vehicle - like robots in dynamic environments.