Embedded Files
We propose a modified RRT path planning algorithm that utilizes Control Barrier Functions (CBF) to ensure feasible and safe paths (e.g. avoid collisions) in continuous time. The main benefit is that CBF can account for known dynamic obstacles and plan around them.
Illustration of the CBF-RRT process for a unicycle model with 3 states. In the first step, an existing vertex in the tree is randomly selected. Note that the steering angle is free.
In the second step, a random steering direction is randomly sampled. This process allows the node to explore many potential directions to avoid local minima.
In the case where there are no nearby obstacles, the safe steer function will drive the system along the randomly selected angle for a pre-determined time.
In the case where there are (dynamic) obstacles nearby, the safe steer function will avoid the obstacles due to the control barrier function constraint in the control synthesis problem.
Robot motion planning is central to real-world autonomous applications, such as self-driving cars, persistence surveillance, and robotic arm manipulation. One challenge in motion planning is generating control signals for nonlinear systems that result in obstacle free paths through dynamic environments. In this paper, we propose Control Barrier Function guided Rapidly-exploring Random Trees (CBF-RRT), a sampling-based motion planning algorithm for continuous-time nonlinear systems in dynamic environments. The algorithm focuses on two objectives: efficiently generating feasible controls that steer the system toward a goal region, and handling environments with dynamical obstacles in continuous time. We formulate the control synthesis problem as a Quadratic Program (QP) that enforces Control Barrier Function (CBF) constraints to achieve obstacle avoidance. Additionally, CBF-RRT does not require nearest neighbor or explicit collision checks during sampling.
A simulation of the CBF-RRT algorithm for the unicycle model with dynamic obstacles. There are no nearby obstacles at the start so the robot can move directly toward the goal.
The obstacles are still far enough that the robot does not have to adjust its trajectory.
The robot has to adjust its trajectory to avoid collisions with the dynamic obstacles while trying to progress toward the goal.
The robot is able to avoid dynamic obstacles and reach its goal location.
In this work, my contributions were in the problem formulation, control barrier and convergence constraints synthesis, and paper writing/editing.
Page updated
Google Sites
Report abuse