In multi-agent navigation, agents need to move towards their goal locations while avoiding collisions with other agents and obstacles, often without communication. Existing methods compute motions that are locally optimal but do not account for the aggregated motions of all agents, producing inefficient global behavior especially when agents move in a crowded space. In this work, we develop a method that allows agents to dynamically adapt their behavior to their local conditions. We formulate the multi-agent navigation problem as an action-selection problem and propose an approach, ALAN, that allows agents to compute time-efficient and collision-free motions. ALAN is highly scalable because each agent makes its own decisions on how to move, using a set of velocities optimized for a variety of navigation tasks. Experimental results show that agents using ALAN, in general, reach their destinations faster than using ORCA, a state-of-the-art collision avoidance framework, and two other navigation models.
more »
« less
Neural Graph Control Barrier Functions Guided Distributed Collision-avoidance Multi-agent Control
We consider the problem of designing distributed collision-avoidance multi-agent control in large-scale environments with potentially moving obstacles, where a large number of agents are required to maintain safety using only local information and reach their goals. This paper addresses the problem of collision avoidance, scalability, and generalizability by introducing graph control barrier functions (GCBFs) for distributed control. The newly introduced GCBF is based on the well-established CBF theory for safety guarantees but utilizes a graph structure for scalable and generalizable decentralized control. We use graph neural networks to learn both neural a GCBF certificate and distributed control. We also extend the framework from handling state-based models to directly taking point clouds from LiDAR for more practical robotics settings. We demonstrated the efficacy of GCBF in a variety of numerical experiments, where the number, density, and traveling distance of agents, as well as the number of unseen and uncontrolled obstacles increase. Empirical results show that GCBF outperforms leading methods such as MAPPO and multi-agent distributed CBF (MDCBF). Trained with only 16 agents, GCBF can achieve up to 3 times improvement of success rate (agents reach goals and never encountered in any collisions) on <500 agents, and still maintain more than 50 success rates for >1000 agents when other methods completely fail.
more »
« less
- Award ID(s):
- 2238030
- PAR ID:
- 10554280
- Publisher / Repository:
- Proceedings of The 7th Conference on Robot Learning
- Date Published:
- Format(s):
- Medium: X
- Sponsoring Org:
- National Science Foundation
More Like this
-
-
null (Ed.)We propose a neural network approach for solving high-dimensional optimal control problems. In particular, we focus on multi-agent control problems with obstacle and collision avoidance. These problems immediately become high-dimensional, even for moderate phase-space dimensions per agent. Our approach fuses the Pontryagin Maximum Principle and Hamilton-Jacobi-Bellman (HJB) approaches and parameterizes the value function with a neural network. Our approach yields controls in a feedback form for quick calculation and robustness to moderate disturbances to the system. We train our model using the objective function and optimality conditions of the control problem. Therefore, our training algorithm neither involves a data generation phase nor solutions from another algorithm. Our model uses empirically effective HJB penalizers for efficient training. By training on a distribution of initial states, we ensure the controls' optimality is achieved on a large portion of the state-space. Our approach is grid-free and scales efficiently to dimensions where grids become impractical or infeasible. We demonstrate our approach's effectiveness on a 150-dimensional multi-agent problem with obstacles.more » « less
-
Safe path planning is critical for bipedal robots to operate in safety-critical environments. Common path planning algorithms, such as RRT or RRT*, typically use geometric or kinematic collision check algorithms to ensure collision-free paths toward the target position. However, such approaches may generate non-smooth paths that do not comply with the dynamics constraints of walking robots. It has been shown that the control barrier function (CBF) can be integrated with RRT/RRT* to synthesize dynamically feasible collision-free paths. Yet, existing work has been limited to simple circular or elliptical shape obstacles due to the challenging nature of constructing appropriate barrier functions to represent irregularly shaped obstacles. In this paper, we present a CBF-based RRT* algorithm for bipedal robots to generate a collision-free path through space with multiple polynomial-shaped obstacles. In particular, we used logistic regression to construct polynomial barrier functions from a grid map of the environment to represent irregularly shaped obstacles. Moreover, we developed a multi-step CBF steering controller to ensure the efficiency of free space exploration. The proposed approach was first validated in simulation for a differential drive model, and then experimentally evaluated with a 3D humanoid robot, Digit, in a lab setting with randomly placed obstacles.more » « less
-
In this paper, we study stable coordination in multi- agent systems with directed interactions, and apply the results for distributed topology control. Our main contribution is to extend the well-known potential-based control framework orig- inally introduced for undirected networks to the case of net- works modeled by a directed graph. Regardless of the particular objective to be achieved, potential-based control for undirected graphs is intrinsically stable. Briefly, this can be explained by the positive semidefiniteness of the graph Laplacian induced by the symmetric nature of the interactions. Unfortunately, this energy finiteness guarantee no longer holds when a multi-agent system lacks symmetry in pairwise interactions. In this context, our contribution is twofold: i) we formalize stable coordination of multi-agent systems on directed graphs, demonstrating the graph structures that induce stability for a broad class of coordination objectives; and ii) we design a topology control mechanism based on a distributed eigenvalue estimation algorithm to enforce Lyapunov energy finiteness over the derived class of stable graphs. Simulation results demonstrate a multi-agent system on a directed graph performing topology control and collision avoidance, corroborating the theoretical findings.more » « less
-
In this paper, a density-driven multi-agent swarm control problem is investigated. Robot swarms can provide a great benefit, especially for applications where a single robot cannot effectively achieve a given task. For large spatial-scale applications such as search and rescue, environmental monitoring, and surveillance, a new multi-agent swarm control strategy is necessary because of physical constraints including a robot number and operation time. This paper provides a novel density-driven swarm control strategy for multi-agent systems based on the Optimal Transport theory, to cover a spacious domain with limited resources. In such a scenario, \textit{efficiency} will likely be a key point in achieving an efficient robot swarm behavior rather than uniform coverage that might be infeasible. With the given reference density, pre-constructed from available information, the proposed swarm control method will drive the multi-agent system such that their time-averaged behavior becomes similar to the reference density. In this way, density-driven swarm control will enable the multiple agents to spend most of their time on high-priority regions that are reflected in the reference density, leading to efficiency. To protect the agents from collisions, the Artificial Potential Field method is employed and combined with the proposed density-driven swarm control scheme. Simulations are conducted to validate density-driven swarm control as well as to test collision avoidance. Also, the swarm performance is analyzed by varying the agent number in the simulation.more » « less
- Free Publicly Accessible Full Text
-
Accepted Manuscript1.0
- Conference Paper:
- The DOI is not currently available.
