Faulty Agents Research Articles

Simplicial models for the epistemic logic of faulty agents

In recent years, several authors have been investigating simplicial models, a model of epistemic logic based on higher dimensional structures called simplicial complexes. In the original formulation of Goubault et al. (Inf Comput 278:104597, 2021. https://doi.org/10.1016/j.ic.2020.104597), simplicial models are always assumed to be pure, meaning that all worlds have the same dimension. This is equivalent to the standard S5n\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$\\mathbf {S5_n}$$\\end{document} semantics of epistemic logic, based on Kripke models. By removing the assumption that models must be pure, we can go beyond the usual Kripke semantics and study epistemic logics where the number of agents participating in a world can vary. This approach has been developed in a number of papers (van Ditmarsch, WoLLIC 2021, pp 31–46, 2021. https://doi.org/10.1007/978-3-030-88853-4_3; Goubault et al. STACS 2022, pp 33:1–33:20, 2022. https://doi.org/10.4230/LIPIcs.STACS.2022.33; Goubault et al. LICS, pp 1–13, 2023. https://doi.org/10.1109/LICS56636.2023.10175737), with applications in fault-tolerant distributed computing where processes may crash during the execution of a system. A difficulty that arises is that subtle design choices in the definition of impure simplicial models can result in different axioms of the resulting logic. In this paper, we classify those design choices systematically, and axiomatize the corresponding logics. We illustrate them via distributed computing examples of synchronous systems where processes may crash.

Bounding the communication complexity of fault-tolerant common coin tossing

Protocols for tossing a common coin play a key role in the vast majority of implementations of consensus. Even though the common coins in the literature are usually fair (they have an equal chance of landing heads or tails), we focus on the problem of implementing a biased common coin such that the probability of landing heads is \(p \in [0,1]\) . Even though biased common coins can be implemented using fair common coins, we show that this can require significant inter-party communication. In fact, we show that there is no bound on the number of messages needed to generate a common coin of bias p in a way that tolerates even one malicious agent, even if we restrict p to an arbitrary infinite subset of \([0,1]\) (e.g., \(\lbrace 1/2^n : n \in \lbrace 0,1,2,\ldots \rbrace \rbrace\) ) and assume that the system is synchronous. By way of contrast, if we do not require the protocol to tolerate a faulty agent, we can do this. Thus, the cause of the message complexity is the requirement of fault tolerance.

Adaptive distributed fault estimation observer design for nonlinear time-delay multi-agent systems with directed graphs

This paper focuses the development of an adaptive observer-based distributed fault estimation observer (DFEO) for multi-agent nonlinear time-delay systems under a directed communication topology. The process involves constructing a fault estimation observer for each agent based on their relative output estimation errors. An adaptive DFEO design is proposed for multi-agent systems with a directed communication topology. Furthermore, investigating the effect of faulty agents on the criteria of fault detection on healthy agents. The proposed design is represented with Linear Matrix Inequalities (LMIs) to establish a systematic design procedure. The study also includes simulation results to demonstrate the effectiveness of the proposed design techniques.

ASQ-IT: Interactive explanations for reinforcement-learning agents

As reinforcement learning methods increasingly amass accomplishments, the need for comprehending their solutions becomes more crucial. Most explainable reinforcement learning (XRL) methods generate a static explanation depicting their developers' intuition of what should be explained and how. In contrast, literature from the social sciences proposes that meaningful explanations are structured as a dialog between the explainer and the explainee, suggesting a more active role for the user and her communication with the agent. In this paper, we present ASQ-IT – an interactive explanation system that presents video clips of the agent acting in its environment based on queries given by the user that describe temporal properties of behaviors of interest. Our approach is based on formal methods: queries in ASQ-IT's user interface map to a fragment of Linear Temporal Logic over finite traces (LTLf), which we developed, and our algorithm for query processing is based on automata theory. User studies show that end-users can understand and formulate queries in ASQ-IT and that using ASQ-IT assists users in identifying faulty agent behaviors.

Event-Triggered Interval Observer Fault Detection and Isolation for Multiagent Systems.

This article investigates an event-triggered interval observer (ETIO) fault detection and isolation method for multiagent systems. First, an event-triggered mechanism is developed to reduce unnecessary communication transmission. Then, a distributed ETIO is designed by combining an interval observer and the proposed event-triggered mechanism. Furthermore, for achieving the desired tradeoff between the robustness to disturbances and the sensitivity to faults, the ETIO is formulated as a multiobjective optimization with l1 / H∞ performance. Second, a bank of ETIOs are interpreted to isolate the faulty agent on a local agent using only the output information from itself and its neighbors. Comparison result with the existing method is given to highlight the superiority of our methodology. Finally, the multiunmanned aerial vehicles system is utilized as the case research, and specific simulation results are presented.

Diagnosis of intermittent faults in Multi-Agent Systems: An SFL approach

Multi-Agent Systems (MAS) can be found in a wide variety of applications, including industrial systems, transportation, software systems and more. In such systems, agents may experience faults that affect the performance of the whole system. However, faulty agents might not consistently experience their fault, but rather in certain conditions. For example, a robot with a faulty rotating mechanism will appear healthy if it is tasked to only move in a straight line. Those faults are called Intermittent Faults. Such faults may cause the entire system to fail, but not always. Previous work proposed diagnosis algorithms for MAS, assuming faulty agents persistently behave abnormally. To the best of our knowledge, intermittent faults in MAS have not been concretely explored. In this paper we formally present a novel problem called Diagnosis of Intermittent Faults in Multi-Agent Systems (DIFMAS): a group of agents are observed across multiple runs. In each run, the success/failure of the agents and the system is observed, aiming to explain all the failed runs by diagnosing which agent(s) are faulty. The contributions of this paper are: (1) formalizing DIFMAS as a Model-Based Diagnosis problem, (2) solving it by presenting a Spectrum-Based Fault Localization (SFL) based method, called Multi-Run SFL-based Diagnosis Algorithm (MRSD). Experiments demonstrate that MRSD's outperforms competing SFL-based algorithms. Moreover, the algorithm's performance increases if planned interactions are considered.

Privacy-preserving Resilient Consensus for Multi-agent Systems in a General Topology Structure

Recent advances of consensus control have made it significant in multi-agent systems such as in distributed machine learning, distributed multi-vehicle cooperative systems. However, during its application it is crucial to achieve resilience and privacy; specifically, when there are adversary/faulty nodes in a general topology structure, normal agents can also reach consensus while keeping their actual states unobserved. In this article, we modify the state-of-the-art Q-consensus algorithm by introducing predefined noise or well-designed cryptography to guarantee the privacy of each agent state. In the former case, we add specified noise on agent state before it is transmitted to the neighbors and then gradually decrease the value of noise so the exact agent state cannot be evaluated. In the latter one, the Paillier cryptosystem is applied for reconstructing reward function in two consecutive interactions between each pair of neighboring agents. Therefore, multi-agent privacy-preserving resilient consensus (MAPPRC) can be achieved in a general topology structure. Moreover, in the modified version, we reconstruct reward function and credibility function so both convergence rate and stability of the system are improved. The simulation results indicate the algorithms’ tolerance for constant and/or persistent faulty agents as well as their protection of privacy. Compared with the previous studies that consider both resilience and privacy-preserving requirements, the proposed algorithms in this article greatly relax the topological conditions. At the end of the article, to verify the effectiveness of the proposed algorithms, we conduct two sets of experiments, i.e., a smart-car hardware platform consisting of four vehicles and a distributed machine learning platform containing 10 workers and a server.

An anti-consensus strategy based on continuous perturbation updates in opposite directions

In modern society, multi-agent consensus is applied in many applications such as distributed machine learning, wireless sensor networks and so on. However, some agents might behave abnormally subject to external attack or internal faults, and thus fault-tolerant consensus problem is studied recently, among which Q-consensus is one of the state-of-the-art and effective methods to identify all the faulty agents and achieve consensus for normal agents in general networks. To fight against Q-consensus algorithm, this paper proposes a novel strategy, called split attack, which is simple but capable of breaking consensus convergence. By aggregating all the states of neighboring nodes with an extra perturbation, the normal nodes are split into sub-groups and converge to two separate values, so that consensus is broken. Two scenarios, including the introduction of additional faulty nodes and compromise of the original nodes, are considered. More specifically, in the former case, two additional faulty nodes are adopted, each of which is responsible to mislead parts of the normal nodes. While in the latter one, two original normal nodes are compromised to mislead the whole system. Moreover, the compromised nodes selection is fundamentally a classification problem, and thus optimized through CNN. Finally, the numerical simulations are provided to verify the proposed schemes and indicate that the proposed method outperforms other attack methods.

Rational Uniform Consensus with General Omission Failures.

Generally, system failures, such as crash failures, Byzantine failures, and so on, are considered as common reasons for the inconsistencies of distributed consensus and have been extensively studied. In fact, strategic manipulations by rational agents are not ignored for reaching consensus in a distributed system. In this paper, we extend the game-theoretic analysis of consensus and design an algorithm of rational uniform consensus with general omission failures under the assumption that processes are controlled by rational agents and prefer consensus. Different from crashing one, agent with omission failures may crash or omit to send or receive messages when it should, which leads to difficulty of detecting faulty agents. By combining the possible failures of agents at the both ends of a link, we convert omission failure model into link state model to make faulty detection possible. Through analyzing message passing mechanism in the distributed system with n agents, among which t agents may commit omission failures, we provide the upper bound on message passing time for reaching consensus on a state among nonfaulty agents and message chain mechanism for validating messages. Then, we prove that our rational uniform consensus is a Nash equilibrium when n > 2t + 1, and failure patterns and initial preferences are blind (an assumption of randomness). Thus, agents have no motivation to deviate the consensus, which could provide interpretable stability for the algorithm in multiagent systems such as distributed energy systems. Our research strengthens the reliability of consensus with omission failures from the perspective of game theory.

Distributed fault detection for uncertain Lipschitz nonlinear multi‐agent systems in finite frequency domain

AbstractThis article focuses on distributed fault detection for uncertain Lipschitz nonlinear multi‐agent systems (MASs). To handle parameter uncertainties, a finite‐frequency unknown input observer (UIO) is designed to formulate the reference residual system and based on it, a robust UIO corresponding to the fault detection system is designed using an model matching approach. In the UIO constructed for one agent, the unknown input composed of faults in the other agents and disturbances in the whole system is partitioned to decouple part of them, which avoids the constraint of the rank condition of UIO and allows for fault detection for more than one faulty agent. The generated residual is sensitive to the corresponding fault while robust against the nondecoupled unknown input and parameter uncertainties, of which the design conditions are transformed into a set of linear matrix inequalities (LMIs). A numerical simulation is conducted to illustrate the effectiveness of the proposed method.

Discriminator Based Resilient Multi-agent Deep Deterministic Policy Gradient Under Uncertain Faulty Agents

Abstract In recent years, Multi-agent reinforcement learning (MARL) is widely applied in various of fields, to achieve a global goal in a centralized or distributed manner. However, during its application it is crucial to be fault-tolerance as some agents behave abnormal. In this paper, we propose a Resilient Multi-gent Deep Deterministic Policy Gradient (RMADDPG) algorithm to achieve a cooperative task in the presence of faulty agents via centralized training decentralized execution. At training stage, each normal agent observes and records information only from other normal ones, without access to the faulty ones. Meanwhile, a discriminator is generated based on the well-trained actor network to identify each faulty agent via supervised learning. Followed by executing stage, each normal agent selects its action based on local observation according to its actor network and its discriminator, so as to achieve certain system goal. Specifically, RMADDPG offers a scheme to train agents for improved resilience against arbitrary number of faulty agents. Finally, a cooperative navigation experiment is provided to validate the effectiveness of the proposed algorithm.

Integrated active fault‐tolerant formation control for networked multi‐agent systems

This paper focuses on the design and simulation of an integrated reconfigurable control strategy for networked multi-agent systems. The control task is to remain the desired time-varying formation in the presence of actuator faults, and to well address the pairing between fault estimation and reconfigurable formation control. A decentralized fault estimation method using observer based estimator is presented to estimate both additive and multiplicative fault values. The reconfigurable control is constructed by adaptive technique using a distributed strategy. Only the information from neighboring agents as well as the fault values are required. The adaptive gains and fault estimates are designed by analyzing the Lyapunov stability of the overall system. The simulation results validate that the proposed integrated fault-tolerant formation control scheme can recover the original desired formation if there are faulty agents in system.

Performance Analysis of Fault Tolerance Algorithm for Pattern Formation of Swarm Agents

Swarm robotics is a growing field of swarm intelligence inspired by natural swarms for providing a solution for the collaboration among swarm agents to achieve a common goal. One of the challenges in designing a swarm robotic system with desired collective behaviour is to understand the effect of individual behaviour on group performance. This paper considers a system composed of autonomous swarm agents, each one executing its algorithm. We present a generic fault-tolerant pattern formation algorithm, in which different agents from different positions gather and then form a desired geometric pattern specifically a uniform circle and maintain it in the presence of faulty agents. We present an analysis based on simulations and experimental results of the proposed algorithm on swarm forming the pattern, in which failures are deliberately induced to check their fault tolerance capability and pattern formation capability. Our findings show that the formed geometric pattern can be restored and maintained depending on the different parameters in the presence of faulty agents.

Privacy-aware Distributed Diagnosis of Multi-Agent Plans

In many multi-agent systems (MAS), agents are assigned to perform tasks. These tasks require designing plans for the agents, which are called multi-agent plan (MAP). Such systems are prone to failures due to variety of reasons. When a failure occurs, there is a need to diagnose it and identify the faulty agents. This problem has been traditionally addressed in a centralized manner. However, in some systems, agents might not be able to share plans due to privacy or single point of failure reasons. To address this challenge, we propose Distributed Diagnosis of Multi-Agent Plans (DDMAP) algorithms, which identify faulty agents of a MAS in a distributed manner without sharing plans. The contributions of this paper are: (1) formalizing DDMAP as a model-based diagnosis problem, and (2) presenting synchronous, asynchronous and semi-asynchronous distributed algorithms to diagnose the faulty agents. Experiments show that the semi-asynchronous algorithm performs better in terms of run-time while the performance in terms of communication overhead is comparable.

Reinforcement Learning Based Multi-Agent Resilient Control: From Deep Neural Networks to an Adaptive Law

Recent advances in Multi-agent Reinforcement Learning (MARL) have made it possible to implement various tasks in cooperative as well as competitive scenarios through trial and error, and deep neural networks. These successes motivate us to bring the mechanism of MARL into the Multi-agent Resilient Consensus (MARC) problem that studies the consensus problem in a network of agents with faulty ones. Relying on the natural characteristics of the system goal, the key component in MARL, reward function, can thus be directly constructed via the relative distance among agents. Firstly, we apply Deep Deterministic Policy Gradient (DDPG) on each single agent to train and learn adjacent weights of neighboring agents in a distributed manner, that we call Distributed-DDPG (D-DDPG), so as to minimize the weights from suspicious agents and eliminate the corresponding influences. Secondly, to get rid of neural networks and their time-consuming training process, a Q-learning based algorithm, called Q-consensus, is further presented by building a proper reward function and a credibility function for each pair of neighboring agents so that the adjacent weights can update in an adaptive way. The experimental results indicate that both algorithms perform well with appearance of constant and/or random faulty agents, yet the Q-consensus algorithm outperforms the faulty ones running D-DDPG. Compared to the traditional resilient consensus strategies, e.g., Weighted-Mean-Subsequence-Reduced (W-MSR) or trustworthiness analysis, the proposed Q-consensus algorithm has greatly relaxed the topology requirements, as well as reduced the storage and computation loads. Finally, a smart-car hardware platform consisting of six vehicles is used to verify the effectiveness of the Q-consensus algorithm by achieving resilient velocity synchronization.

Observer-based distributed fault detection and isolation for second-order multi-agent systems using relative information

This paper considers fault detection and isolation problem for second-order leader-follower multi-agent systems in a distributed setting. With the existing tracking protocol, an observer is constructed at each agent based on the concept of unknown input observer using relative information between the agent and its neighbors. Fault in position, velocity, both position and velocity dynamics are considered separately. Based on the residual signal generated from the designed observer, two algorithms are proposed so that the faulty agent in network can be detected and isolated simply. Effectiveness of the presented strategy is validated by some application examples.

Observer-Based Distributed Fault Detection for Heterogeneous Multi-Agent Systems

This paper solves the distributed fault detection (FD) problem for heterogeneous multi-agent systems (MAS). For a heterogeneous MAS, we adopt a distributed control law to realise cooperative output regulation (COR) when no fault occurs in the MAS, and propose a state-feedback-based FD scheme, where the adopted distributed control law and proposed FD scheme all utilise state information. Furthermore, we consider the condition that state information is unmeasurable, the output-feedback-based distributed FD scheme is proposed, and the adopted distributed control law also utilises measurement output. Finally, two numerical examples are utilised to verify that the proposed distributed FD schemes could locate and remove the faulty agent in time.

Control for Leader–Follower Consensus of Multi-agent Systems with Actuator Faults Using Decentralized Robust Fault-tolerant Control

The leader–follower consensus tracking is one of the important problems in multi-agent systems (MASs) which can be influenced by actuator faults. The main issue of the paper is fault occurrence, and the challenge is to design an appropriate decentralized fault-tolerant control (FTC) based on fault estimations to minimize the effects of actuator fault and maintain the tracking problem. In this paper, the Luenberger detection filter is adopted for the fault detection process and fault estimations are obtained by recursive least-square (RLS) parameter estimation method. The procedure utilized for FTC is fault hiding technique so that using virtual actuator methodology, the effects of faults are partially improved and a robust controller is combined with a virtual actuator to reduce the tracking error. The main contribution of this paper is designing a robust FTC which not only can compensate for the performance degradation, but also can make the system robust against fault estimation error uncertainties. It is also proved that the tracking error of faulty agents converges to zero asymptotically and the consensus tracking is preserved by using the proposed method. The effectiveness of the proposed method is verified through simulations.

Partial-Nodes-Based Distributed Fault Detection and Isolation for Second-Order Multiagent Systems With Exogenous Disturbances.

In this article, a robust distributed fault detection and isolation (FDI) scheme for second-order multiagent systems (MASs) with exogenous disturbances based on partial agents' absolute state will be proposed. We design a robust control law with disturbance rejection term for each agent, then, an augmented MAS closed-loop system without disturbance term is formed. We construct a bank of observers for FDI in some selected agents for the augmented MAS closed-loop system based on the idea of an unknown input observer (UIO), where the observers for FDI of the entire MAS are updated by the selected agents' absolute states and relative states between themselves and their neighbors, which avoids utilizing all agents' absolute states. We run observers for FDI in the selected agents to generate residuals to detect the faulty agent between themselves and their neighbors. Meanwhile, the existence proof of observers for FDI is also presented. It is worth noting that the remaining healthy agents will reach consensus again under the original robust control law if the communication topology is still connected. Finally, the effectiveness of the proposed scheme is illustrated by a multivehicle example.

Active fault‐tolerant consensus control of Lipschitz nonlinear multiagent systems

SummaryThis article investigates the active fault‐tolerant consensus problem for Lipschitz nonlinear multiagent systems under detailed balanced directed graph and actuator faults. First, a fault detection filter for each agent is designed, and all agents can be divided into two categories: healthy agents and possibly faulty agents. Second, fully distributed adaptive fault‐tolerant consensus protocols for healthy and possibly faulty agents are proposed to achieve state consensus. Third, based on the fault detection method and fault‐tolerant consensus protocols, active fault‐tolerant consensus algorithms are given. Simulation examples are presented to verify the effectiveness of the proposed active fault‐tolerant algorithms.