Despite remarkable achievements in artificial intelligence, the deployability of learning-enabled systems in high-stakes real-world environments still faces persistent challenges. For example, in safety-critical domains like autonomous driving, robotic manipulation, and healthcare, it is crucial not only to achieve high performance but also to comply with given constraints. Furthermore, adaptability becomes paramount in non-stationary domains, where environmental parameters are subject to change. While safety and adaptability are recognized as key qualities for the new generation of AI, current approaches have not demonstrated effective adaptable performance in constrained settings. Hence, this paper breaks new ground by studying the unique challenges of ensuring safety in nonstationary environments by solving constrained problems through the lens of the meta-learning approach (learning to learn). While unconstrained meta-learning already encounters complexities in end to end differentiation of the loss due to the bi-level nature, its constrained counterpart introduces an additional layer of difficulty, since the constraints imposed on task-level updates complicate the differentiation process. To address the issue, we first employ successive convex-constrained policy updates across multiple tasks with differentiable convex programming, which allows meta-learning in constrained scenarios by enabling end-to-end differentiation. This approach empowers the agent to rapidly adapt to new tasks under nonstationarity while ensuring compliance with safety constraints. We also provide a theoretical analysis demonstrating guaranteed monotonic improvement of our approach, justifying our algorithmic designs. Extensive simulations across diverse environments provide empirical validation with significant improvement over established benchmarks.

In a multirobot system, a number of cyber-physical attacks (e.g., communication hijack, observation per-turbations) can challenge the robustness of agents. This robust-ness issue worsens in multiagent reinforcement learning because there exists the non-stationarity of the environment caused by simultaneously learning agents whose changing policies affect the transition and reward functions. In this paper, we propose a minimax MARL approach to infer the worst-case policy update of other agents. As the minimax formulation is computationally intractable to solve, we apply the convex relaxation of neural networks to solve the inner minimization problem. Such convex relaxation enables robustness in interacting with peer agents that may have significantly different behaviors and also achieves a certified bound of the original optimization problem. We eval-uate our approach on multiple mixed cooperative-competitive tasks and show that our method outperforms the previous state of the art approaches on this topic.

The main challenge of multiagent reinforcement learning is the difficulty of learning useful policies in the presence of other simultaneously learning agents whose changing behaviors jointly affect the environment's transition and reward dynamics. An effective approach that has recently emerged for addressing this non-stationarity is for each agent to anticipate the learning of other agents and influence the evolution of future policies towards desirable behavior for its own benefit. Unfortunately, previous approaches for achieving this suffer from myopic evaluation, considering only a finite number of policy updates. As such, these methods can only influence transient future policies rather than achieving the promise of scalable equilibrium selection approaches that influence the behavior at convergence. In this paper, we propose a principled framework for considering the limiting policies of other agents as time approaches infinity. Specifically, we develop a new optimization objective that maximizes each agent's average reward by directly accounting for the impact of its behavior on the limiting set of policies that other agents will converge to. Our paper characterizes desirable solution concepts within this problem setting and provides practical approaches for optimizing over possible outcomes. As a result of our farsighted objective, we demonstrate better long-term performance than state-of-the-art baselines across a suite of diverse multiagent benchmark domains.

A fundamental challenge in multiagent reinforcement learning is to learn beneficial behaviors in a shared environment with other simultaneously learning agents. In particular, each agent perceives the environment as effectively non-stationary due to the changing policies of other agents. Moreover, each agent is itself constantly learning, leading to natural non-stationarity in the distribution of experiences encountered. In this paper, we propose a novel meta-multiagent policy gradient theorem that directly accounts for the non-stationary policy dynamics inherent to multiagent learning settings. This is achieved by modeling our gradient updates to consider both an agent’s own non-stationary policy dynamics and the non-stationary policy dynamics of other agents in the environment. We show that our theoretically grounded approach provides a general solution to the multiagent learning problem, which inherently comprises all key aspects of previous state of the art approaches on this topic. We test our method on a diverse suite of multiagent benchmarks and demonstrate a more efficient ability to adapt to new agents as they learn than baseline methods across the full spectrum of mixed incentive, competitive, and cooperative domains.

This paper investigates reinforcement learning with constraints, which are indispensable in safety-critical environments. To drive the constraint violation to decrease monotonically, we take the constraints as Lyapunov functions and impose new linear constraints on the policy parameters’ updating dynamics. As a result, the original safety set can be forward-invariant. However, because the new guaranteed-feasible constraints are imposed on the updating dynamics instead of the original policy parameters, classic optimization algorithms are no longer applicable. To address this, we propose to learn a generic deep neural network (DNN)-based optimizer to optimize the objective while satisfying the linear constraints. The constraint-satisfaction is achieved via projection onto a polytope formulated by multiple linear inequality constraints, which can be solved analytically with our newly designed metric. To the best of our knowledge, this is the first DNN-based optimizer for constrained optimization with the forward invariance guarantee. We show that our optimizer trains a policy to decrease the constraint violation and maximize the cumulative reward monotonically. Results on numerical constrained optimization and obstacle-avoidance navigation validate the theoretical findings.

Mixed-type decision-making is ubiquitously required in robotic systems and has attracted significant research interests. Examples include, but not limited to, the integrated task and motion planning and optimal control of hybrid systems involving both continuous and discrete dynamic behaviors. For decision-making of robotic systems to improve operational efficiency, safety, and/or mission success rate, they involve both discrete variables representing task allocation or transitions between discrete modes and continuous variables representing trajectories of the planned motion or states governed by differential equations. This paper formulates a class of mixed-type decision-making problems with polynomial objective and constraints as quadratically constrained quadratic programming (QCQP) problems and a nonconvex optimization method based on alternating direction method of multipliers is proposed to solve the QCQP. The proposed optimization method consists of three sequential subproblems, all of which admit closed-form solutions. Moreover, convergence proof of the optimization algorithm is provided. Two representative problems, traveling salesman with obstacle avoidance and rendezvous and docking of a charging station with distinct phase constraints, are described and solved via the proposed method. Numerical simulations as well as experimental verification of both problems are presented and compared with a state-of-art method to validate the effectiveness, efficacy and robustness of the nonconvex optimization method.

The complexity of multiagent reinforcement learning (MARL) in multiagent systems increases exponentially with respect to the agent number. This scalability issue prevents MARL from being applied in large-scale multiagent systems. However, one critical feature in MARL that is often neglected is that the interactions between agents are quite sparse. Without exploiting this sparsity structure, existing works aggregate information from all of the agents and thus have a high sample complexity. To address this issue, we propose an adaptive sparse attention mechanism by generalizing a sparsity-inducing activation function. Then a sparse communication graph in MARL is learned by graph neural networks based on this new attention mechanism. Through this sparsity structure, the agents can communicate in an effective as well as efficient way via only selectively attending to agents that matter the most and thus the scale of the MARL problem is reduced with little optimality compromised. Comparative results show that our algorithm can learn an interpretable sparse structure and outperforms previous works by a significant margin on applications involving a large-scale multiagent system.