Abstract:Control barrier functions (CBF) are widely explored to enforce the safety-critical constraints on nonlinear systems recently. There are many researchers incorporating the control barrier functions into path planning algorithms to find a safe path, but these methods involve huge computational complexity or unidirectional randomness, resulting in arising of run-time. When safety constraints are satisfied, searching efficiency, and searching space are sacrificed. This paper combines the novel motion planning approach using rapid exploring random trees (RRT) algorithm with model predictive control (MPC) to enforce the CBF with dynamically updating constraints to get the safety-critical resolution of trajectory which will enable the robots not to collide with both static and dynamic circle obstacles as well as other moving robots while considering the model uncertainty in process. Besides, this paper first realizes application of CBF-RRT in robot arm model for nonlinear system.
What problem does this paper attempt to address?
The problem that this paper attempts to solve is how to efficiently find a safe path in robot path planning while avoiding collisions with static and dynamic obstacles. Specifically, the paper proposes a method that combines the Rapidly - exploring Random Tree (RRT) algorithm, Model Predictive Control (MPC), and Control Barrier Function (CBF) to achieve safe path planning for multi - robot systems in the presence of linear and nonlinear systems as well as dynamic obstacles. In addition, the paper also explores the possibility of applying this method to robotic arm models, especially for path - planning problems in nonlinear systems.
### Main Contributions
1. **Application in Multi - robot Systems**: For the first time, the paper uses the CBF - RRT method to generate feasible and safe paths for four robots and tracks these paths through MPC while avoiding collisions with other robots and static/dynamic circular obstacles.
2. **Application in Nonlinear Systems**: The CBF - RRT is applied to a nonlinear robotic arm model for the first time to perform object - grasping tasks.
3. **Algorithm Optimization**: By improving the sampling method of the RRT algorithm, the number of vertices and edges is reduced, and the search efficiency is improved.
4. **Balance between Safety and Real - time Performance**: By dynamically updating the constraint conditions, the efficiency of path search is improved while ensuring safety.
### Method Overview
- **CBF - RRT**: Use the Control Barrier Function (CBF) to define the safe set and combine the Rapidly - exploring Random Tree (RRT) algorithm to generate feasible paths.
- **MPC**: Use Model Predictive Control (MPC) to track the reference path generated by CBF - RRT while avoiding collisions.
- **Multi - robot Strategy**: Adopt the "priority robot" strategy, gradually consider the obstacles that each robot needs to avoid, and ensure the safe path planning of the multi - robot system.
### Numerical Experiments
The paper verifies the effectiveness of the proposed method through multiple numerical experiments, including path planning of a single robot in a static - obstacle environment, path planning of four robots in static and dynamic - obstacle environments, and path planning of a two - link robotic arm and the left arm of the Baxter robot.
### Formula Summary
- **System Dynamics**:
\[
\dot{x}=A x + B u
\]
\[
\dot{x}_{\text{obs}} = C u
\]
- **Exponential Control Barrier Function (ECBF)**:
\[
\inf_{u \in U}[\dot{h}(x)+k h(x)] \geq 0 \quad \forall x \in \text{Int}(C)
\]
- **Safe Set**:
\[
C_i=\{x \in \mathbb{R}^2: h_i(x) \geq 0\}
\]
\[
h_i(x)=(x_1 - x_{1,\text{obs},i})^2+(x_2 - x_{2,\text{obs},i})^2 - r_{\text{obs},i}^2
\]
\[
\dot{h}_i(x)=2(x_1 - x_{1,\text{obs},i}) u \cos \theta+2(x_2 - x_{2,\text{obs},i}) u \sin \theta
\]
- **Optimization Problem in CBF - RRT**:
\[
\min_{u \in U}\|u - u_{\text{ref}}\|^2
\]
\[
\text{s.t.} \quad \dot{h}_i(x)+k h_i(x) \geq 0, \quad i = 1, \ldots, N_{\text{obs}}
\]
\[
u \leq u \leq u
\]
Through these methods and experiments, the paper demonstrates the effectiveness and robustness of CBF - RRT - MPC in multi - robot systems and robotic arm path planning.