Swarm Systems Research Articles

Adversarial imitation learning with deep attention network for swarm systems

AbstractSwarm systems consist of a large number of interacting individuals, which exhibit complex behavior despite having simple interaction rules. However, crafting individual motion policies that can manifest desired collective behaviors poses a significant challenge due to the intricate relationship between individual policies and swarm dynamics. This paper addresses this issue by proposing an imitation learning method, which derives individual policies from collective behavior data. The approach leverages an adversarial imitation learning framework, with a deep attention network serving as the individual policy network. Our method successfully imitates three distinct collective behaviors. Utilizing the ease of analysis provided by the deep attention network, we have verified that the individual policies underlying a certain collective behavior are not unique. Additionally, we have analyzed the different individual policies discovered. Lastly, we validate the applicability of the proposed method in designing policies for swarm robots through practical implementation on swarm robots.

Emergent Dynamic Formation through Optical Interactions in a Robot Swarm

Self‐organized formation is a key direction in swarm robotics. It is still challenging to design local interactions toward desired global formations and even more challenging for dynamic formations in a physical robot swarm system. Herein, a self‐organized method for emergent dynamic circling formation in a robot swarm through optical interactions is proposed. First, this method is quantitatively modeled based on the geometrical relations among robots. This model is further adjusted according to the characteristics of the robot swarm system. To demonstrate the effectiveness of this model, the effects of three key parameters of this model are tested on the size and disorder level of the emergent dynamic circling formation. The experimental results are consistent with the model predictions. Overall, a robot swarm system, in the physical environment, is quantitatively controlled to emerge a dynamic circling formation in this article. This work advances the swarm robotics for quantitatively designing local interactions among robots to reliably emerge dynamic global patterns.

Research on Target Allocation for Hard-Kill Swarm Anti-Unmanned Aerial Vehicle Swarm Systems

In response to the saturated attacks by low, slow, and small UAV swarms, there is currently a lack of effective countermeasures. Counter-UAV swarm technology is an important issue that urgently requires breakthroughs. This paper conducts research on a mid–short-range hard-kill counter-swarm scenario where fewer swarms confront multiple swarms and stronger swarms confront weaker swarms. The requirement is for counter-swarm UAVs to quickly penetrate the swarm at mid–short range and collide with as many incoming UAVs as possible to destroy them. To address the sparse solution space problem, an improved genetic algorithm that integrates multiple strategies is adopted to calculate the spatial density distribution of the incoming swarm. A baseline is identified through gradient descent that maximizes the density integral in a straight-line direction. Based on this baseline, the solution space for single strikes on the swarm is filtered. During the solution process, an elite strategy is introduced to prevent the overall degradation of the population performance. Additionally, the feasibility of the flight trajectory needs to be assessed. A piecewise cubic spline interpolation method is used to optimize the flight trajectory, minimizing the maximum curvature. Ultimately, multiple counter-swarm UAV targets within the swarm and their corresponding trajectories are obtained.

AI-based autonomous UAV swarm system for weed detection and treatment: Enhancing organic orange orchard efficiency with agriculture 5.0

Weeds significantly threaten agricultural productivity by competing with crops for nutrients, particularly in organic farming, where chemical herbicides are prohibited. On Spain’s Mediterranean coast, organic citrus farms face increasing challenges from invasive species like Araujia sericifera and Cortaderia selloana, which further complicate cover crop management. This study introduces a swarm system of unmanned aerial vehicles (UAVs) equipped with neural networks based on YOLOv10 for the detection and geo-location of these invasive weeds. The system achieves F1-scores of 0.78 for Araujia sericifera and 0.80 for Cortaderia selloana. Using GPS and RTK, the UAVs generate KML files to guide diffuser drones for precise, localized treatments with organic products. By automating the detection, treatment, and elimination of invasive species, the system enhances both productivity and sustainability in organic farming. Additionally, the proposed solution addresses the high labor costs associated with manual weeding by reducing the need for human intervention. A comprehensive economic analysis indicates potential savings ranging from 1810 to 2650 € per hectare, depending on farm size. This innovative approach not only improves weed control efficiency but also promotes environmental sustainability, offering a scalable solution for organic and conventional agriculture alike.

The influence of neighbor selection on self-organized UAV swarm based on finite perception vision.

Recently, vision-based unmanned aerial vehicle (UAV) swarming has emerged as a promising alternative that can overcome the adaptability and scalability limitations of distributed and communication-based UAV swarm systems. While most vision-based control algorithms are predicated on the detection of neighboring objects, they often overlook key perceptual factors such as visual occlusion and the impact of visual sensor limitations on swarm performance. To address the interaction problem of neighbor selection at the core of self-organizing UAV swarm control, a perceptually realistic finite perception visual (FPV) neighbor selection model is proposed, which is based on the lateral visual characteristics of birds, incorporates adjustable lateral visual field widths and orientations, and is able to ignore occluded agents. Based on the FPV model, a neighbor selection method based on the Acute Angle Test (AAT) is proposed,
which overcomes the limitation that the traditional neighbor selection mechanism can only interact with the nearest neighboring agents. A large number of Monte Carlo simulation comparison experiments show that the proposed FPV+AAT neighborselection mechanism can reduce the redundant communication burden between large scale self-organized UAV swarms, and outperforms the traditional neighbor selection method in terms of order, safety, union, connectivity, and noise resistance.

Enhancing UAV Swarm Tactics with Edge AI: Adaptive Decision Making in Changing Environments

This paper presents a drone system that uses an improved network topology and MultiAgent Reinforcement Learning (MARL) to enhance mission performance in Unmanned Aerial Vehicle (UAV) swarms across various scenarios. We propose a UAV swarm system that allows drones to efficiently perform tasks with limited information sharing and optimal action selection through our Efficient Self UAV Swarm Network (ESUSN) and reinforcement learning (RL). The system reduces communication delay by 53% and energy consumption by 63% compared with traditional MESH networks with five drones and achieves a 64% shorter delay and 78% lower energy consumption with ten drones. Compared with nonreinforcement learning-based systems, mission performance and collision prevention improved significantly, with the proposed system achieving zero collisions in scenarios involving up to ten drones. These results demonstrate that training drone swarms through MARL and optimized information sharing significantly increases mission efficiency and reliability, allowing for the simultaneous operation of multiple drones.

Combat Effectiveness Evaluation of UAV Intelligent Swarm Based on AHP-Fuzzy Comprehensive Evaluation

In order to evaluate the combat effectiveness of UAV intelligent swarm, this paper puts forward the establishment of UAV intelligent swarm combat effectiveness evaluation index system based on availability, credibility and combat ability, analyzes the factors affecting the combat effectiveness of UAV intelligent swarm, establishes the effectiveness evaluation model, determines the weight of each factor by AHP, and evaluates the effectiveness index by fuzzy evaluation matrix. Using this method to evaluate the combat effectiveness of UAV intelligent swarm, the evaluation result is objective and reasonable, which provides theoretical reference for the optimal design and use of UAV intelligent swarm system.

Revolutionizing swarm dynamics: the role of receptive fields in enhancing convergence and stability

The classic Vicsek model, while influential in understanding swarm behavior, has limitations in achieving motion consensus and convergence speed, especially under varying conditions of density and noise. This study aims to introduce a novel receptive field mechanism to the Vicsek model to enhance its performance in terms of motion consensus and convergence speed within swarms. The modified model divides a particle’s surrounding area into excitation and inhibition zones based on distinct functions. This structural modification is designed to enrich evolutionary behavior and improve consensus convergence capabilities. Experimental outcomes indicate that the proposed model achieves faster convergence rates towards motion consensus under various density and noise conditions compared to traditional models. Specifically, while classic Vicsek models fail to converge to an overall polarization state under high noise levels and exhibit quasi-periodic oscillations, the enhanced model demonstrates stable convergence without oscillatory behavior across both low- and high-noise environments. The findings highlight the superior evolutionary consistency characteristics of the improved model, offering new theoretical and practical insights into the stability and controllability of swarms. This advancement presents significant implications for the development of more robust swarm systems.

Data-Driven Fault Detection and Diagnosis for UAV Swarm Systems

With the widely application and promotion of Unmanned aerial vehicle (UAV) technology, UAV swarms are widely used in military and civilian fields by taking advantage of group collaboration. It has huge economic and national defense value. The safety and reliability requirements of the UAV swarm system are extremely strict, and real-time fault detection and diagnosis is one of its important supporting technologies. In this paper, a fault diagnosis method based on statistical model and improved broad learning system (BLS) is proposed. The behavior characteristics of the UAV swarm system under normal and different failure modes are characterized by multivariate data statistical analysis, and the improved BLS model is adopted to achieve accurate and fast fault diagnosis. And a high-fidelity simulation verification platform is developed to verify the rationality and effectiveness of the proposed method.

Multiscale Spatial–Temporal Bayesian Graph Conv-Transformer-Based Distributed Fault Diagnosis for UAVs Swarm System

Multiscale Spatial–Temporal Bayesian Graph Conv-Transformer-Based Distributed Fault Diagnosis for UAVs Swarm System

Enhancing Inter-AUV Perception: Adaptive 6-DOF Pose Estimation with Synthetic Images for AUV Swarm Sensing

The capabilities of AUV mutual perception and localization are crucial for the development of AUV swarm systems. We propose the AUV6D model, a synthetic image-based approach to enhance inter-AUV perception through 6D pose estimation. Due to the challenge of acquiring accurate 6D pose data, a dataset of simulated underwater images with precise pose labels was generated using Unity3D. Mask-CycleGAN technology was introduced to transform these simulated images into realistic synthetic images, addressing the scarcity of available underwater data. Furthermore, the Color Intermediate Domain Mapping strategy is proposed to ensure alignment across different image styles at pixel and feature levels, enhancing the adaptability of the pose estimation model. Additionally, the Salient Keypoint Vector Voting Mechanism was developed to improve the accuracy and robustness of underwater pose estimation, enabling precise localization even in the presence of occlusions. The experimental results demonstrated that our AUV6D model achieved millimeter-level localization precision and pose estimation errors within five degrees, showing exceptional performance in complex underwater environments. Navigation experiments with two AUVs further verified the model’s reliability for mutual 6D pose estimation. This research provides substantial technical support for more complex and precise collaborative operations for AUV swarms in the future.

Theory and experiment on distributed output formation tracking of unmanned aerial and ground vehicle swarm systems over jointly connected digraphs

This paper studies the distributed output formation tracking control problem of the unmanned aerial vehicle (UAV) and unmanned ground vehicle (UGV) swarm systems, where the UAV swarm cooperatively tracks the output trajectory of the UGV in a formation under directed jointly connected communication networks. A three-layer formation tracking protocol is proposed. Firstly, a novel distributed observer using neighbor interactions is designed for each UAV to estimate the states of the UGV with parameterized inputs over periodic jointly connected digraphs. Next, based on the observation result and output regulation theory, a virtual reference system that tracks the trajectory of the UGV in the desired formation is constructed to generate reference states for each UAV. Then based on the differential flatness of UAVs, a geometric controller is utilized for UAVs to track the reference states and form the formation. An algorithm to determine the gain matrices of the protocol is also presented while the convergence of the system is analyzed. Finally, an experiment platform with three quadrotor UAVs and one UGV is built. The effectiveness of the proposed protocol is validated both by the simulation and experiment.

Localized Control of the Swarming of Kinesin-Driven Microtubules Using Light.

The swarming of self-propelled cytoskeletal filaments has emerged as a new aspect in the field of molecular machines or robotics, as it has overcome one of the major challenges of controlling the mutual interaction of a large number of individuals at a time. Recently, we reported on the photoregulated swarming of kinesin-driven cytoskeletal microtubule filaments in which visible (VIS) and ultraviolet (UV) light triggered the association and dissociation of the swarm, respectively. However, systematic control of this potential system has yet to be achieved to optimize swarming for further applications in molecular machines or robotics. Here, we demonstrate the precise and localized control of a biomolecular motor-based swarm system by varying different parameters related to photoirradiation. We control the reversibility of the swarming by changing the wavelength or intensity of light and the number of azobenzenes in DNA. In addition, we regulate the swarming in local regions by introducing different-sized or shaped patterns in the UV light system. Such a detailed study of the precise control of swarming would provide new perspectives for developing a molecular swarm system for further applications in engineering systems.

HGEE: Learning for Trajectory Prediction with Heterogeneous Graph Interaction and External Embedding of Unmanned Swarm Systems in Adversarial Environment

Trajectory prediction of unmanned swarm systems, serving as the foundation for behavioral and intentional cognition, has attracted extensive attention and made considerable progress in adversarial research. The influence of heterogeneous interaction relationships and external factors is crucial for trajectory prediction. Consequently, this paper proposes the Heterogeneous Graph with External Embedding (HGEE) network. We model the latent variables as multi-layer heterogeneous graphs based on prior knowledge of different interaction relationships and propose a method for calculating edge embeddings for heterogeneous graphs. Furthermore, we introduce a method that combines external environmental feature with historical observational trajectory data as the input for the decoder, enabling the model to learn the impacts of obstacles, targets, and desired formations on trajectories. We demonstrate that our approach surpasses state-of-the-art models in interaction inference and trajectory prediction through experiments on our proposed formation datasets based on consensus theory, across five evaluation metrics.

Integral Event-triggered Time-varying Formation Tracking for UAV Swarm Systems With Switching Directed Topologies

Integral Event-triggered Time-varying Formation Tracking for UAV Swarm Systems With Switching Directed Topologies

Dynamic event-triggered time-varying formation control for heterogeneous unmanned swarm systems with scaling attacks

This paper addresses the problem of time-varying formation control for heterogeneous unmanned swarm systems (USSs) consisting of multiple unmanned aerial vehicles (UASs) and unmanned surface vehicles (USVs) based on a directional switching topology. Firstly, a novel USSs model is established, involving the effect of multiple network attacks, network bandwidth constraints and input saturation. In particular, a scaling attack model is proposed to account for denial-of-service (DoS) and deception attacks. Secondly, a distributed controller based on a dynamic event triggering mechanism is designed to reduce the communication load by adjusting the communication frequency between individuals, and the Zeno phenomenon is effectively avoided. Finally, the Lyapunov function is used to obtain sufficient conditions for system stability and bounds on the frequency and duration of network attacks. The experiments and comparison results verify the proposed strategy's effectiveness and superiority.

Optimal saturated information load analysis for enhancing robustness in unmanned swarms system

Saturated information load is defined as the information received by a unmanned aerial vehicle (UAV) node in a swarm network reaches the overload limit of its processing capability. When a UAV swarm performs a mission in an uncertain and adversarial complex environment, overloading of UAVs will lead to information diversion, which may cause other UAVs to experience overloading and diversion as well, affecting the transmission efficiency and robustness of the entire swarm network, which in turn affects the information sensing ability, execution ability, and coordination ability of the swarm in performing the mission. Therefore, this paper proposes a saturated information load-based UAV swarm network topology modelling method, which sets the saturated information load of the nodes in the network model in order to reasonably allocate network resources and optimise the network topology. In addition, through robustness experiments of complex networks and comparative analysis of different saturated information loads and three typical modelling methods, the saturated information load-based network structure modelling method has outstanding advantages and performance in terms of network connectivity, network communication efficiency, and destruction resistance.

Disturbance rejection based consensus controller for swarm system

This article introduces a robust controller designed to facilitate consensus among leader-follower agents in a swarm system. The method employs algebraic graph theory to establish a consensus framework and incorporates a sliding surface-based uncertainty and disturbance estimator (SUDE) to mitigate the effects of nonlinearities, external disturbances, and parametric uncertainty inherent in swarm systems. The stability of the SUDE controller is examined using Lyapunov theory. Importantly, this approach operates without prior knowledge of disturbance bounds, relying instead on the dynamics of disturbance error. Simulation studies demonstrate the effectiveness of the proposed control methodology, showing that the SUDE controller expedites consensus achievement while ensuring smooth and non-chattering control behavior. The proposed SUDE controller demonstrates a notable improvement in achieving position and velocity consensus in both the X and Y directions, with a 75% and 66% faster convergence rate, respectively, when compared to the conventional sliding mode controller (SMC). Additionally, the performance of the proposed controller is assessed under various external disturbances and parametric uncertainty conditions. Comparative analyses using root mean square error (RMSE) and integral absolute time error (IATE), both based on consensus error, confirm the effectiveness of the proposed SUDE consensus controller.

Incremental swarm coordination control with self-triggered-organized topology and predictive-based control method

This work focuses on coordinating a swarm of multiple mobile robots with incremental joining units using an incremental-self-organized formation (ISOF) scheme. The ISOF scheme enables the assimilation of newly joined robots and autonomously establishes a feasible communication topology between the incremental and existing robots. To avoid redundant generation, the new communication topology is expanded and updated based on the existing topology while preserving the original structure. To realize the real-time update of control model and topology for generalized swarm system using the ISOF scheme, an incremental-self-triggered updating mechanism (ISTUM) is designed, and specific triggering conditions are defined. With the updated communication topology and control model, the incremental swarm system, subject to predefined constraints, can be controlled to achieve the desired formation objective using a model-predictive-control-based (MPC-based) protocol. The proposed control scheme aims to minimize the need for manual intervention in swarm reconstruction when incremental behaviour occurs, thereby enhancing the flexibility of swarm coordination. Simulation results demonstrate the effectiveness of the developed control scheme.

Bionic Swarm Control Based on Second-Order Communication Topology.

In this article, we propose bionic swarm control based on second-order communication topology (SOCT) inspired by the migration of birds, which solves the difficulty in constructing communication topologies and high-computational complexity in controlling large-scale swarm systems. To realize bionic swarm control, there are three problems supposed to be solved. First, the adjacency matrix and the Laplacian matrix in traditional methods cannot be applied to SOCT directly, which should be redesigned. Second, sub-swarm systems formed based on 2-order communication topology (2-OCT) and independently distributed with each other also need to be put forward to reduce computational complexity. At last, the followers in 1-order communication topology (1-OCT) are set as the leaders of sub-swarm systems in 2-OCT. As a result, coupling in large-scale swarm systems would be reduced. The bionic swarm controller is designed through the backstepping method. In this case, the stability of bionic swarm controller is proven by the designed Lyapunov function. The simulations show the efficiency of the designed bionic swarm controller. And the tracking-containment control based on SOCT with 42 swarm members is realized.