Skip to content

Chapter 10: Swarm Collaboration and Game-Theoretic Adversarial Interactions

The concept of swarms originated from biological research. French zoologist Grasse, based on observations of termite nest-building behavior, first proposed the concept of stigmergy—a mechanism of intermittent coordination among individuals that enables complex intelligent activities without centralized planning or direct communication. This marked the beginning of the emergence and gradual development of autonomous swarm concepts in human understanding. Extending from biological systems to Multi-Agent Systems (MAS), the concept of swarms has continued to evolve and enrich.

10.1 Background and Theory

Intelligent unmanned swarm systems represent a systemic integration and capability leap of the technologies discussed in previous chapters, signifying the advancement of intelligent unmanned systems to a higher level. Such systems must possess self-organization, self-adaptation, and fault-tolerance capabilities to flexibly execute diverse tasks—including cooperative formation, area coverage, target tracking, and game-theoretic adversarial interactions—in dynamic, even unstructured environments.

10.1.1 Swarm System Concept

The concept of a swarm initially emerged from biological research. Inspired by observations of termite nest-building behavior, the term stigmergy was first introduced, describing an indirect coordination mechanism among individuals that accomplishes complex intelligent tasks without centralized planning or direct communication.

alt text

10.1.2 Development of Swarm System Research

The evolution of swarm system research can be divided into three phases:
- Foundational theory and conceptual exploration phase (1990s to early 2000s),
- Practical application and platform development phase (mid-2000s to 2010s),
- Future-oriented and integrated innovation phase (2010s to present).

10.2 Framework and Interfaces

The RflySim toolchain provides robust support and a reusable paradigm for the development and application of intelligent unmanned swarm systems. It supports the swarm collaborative control architecture and interface usage, enabling a complete engineering workflow—from algorithm development and simulation validation to real-platform deployment.

alt text

10.2.1 Swarm Collaborative Control and Formation

RflySim supports one-click initiation of multi-aircraft swarm simulations, offering development environments for both MATLAB/Simulink and Python. It supports software-in-the-loop (SITL) and hardware-in-the-loop (HITL) simulations for multiple aircraft, as well as distributed swarm simulations across multiple computers within a local area network.

alt text

10.2.2 Intelligent Game-Theoretic and Adversarial Interactions

The toolchain enables intelligent unmanned swarms to flexibly perform diverse tasks—including cooperative formation, area coverage, target tracking, and game-theoretic adversarial interactions—in dynamic and even unstructured environments. Compared to single unmanned platforms, this significantly enhances overall reliability, task adaptability, and execution efficiency.

alt text

10.3 Showcase of Representative Cases

Large-scale Swarm Search and Target Localization Simulation:

Vertical Takeoff and Landing (VTOL) Aircraft Multi-Agent Formation:

10.4 Course-Linked Video Lectures

Public Lecture Replay (Session 8: Swarm Collaboration and Game-Theoretic Adversarial Interactions):

10.5 Chapter Experiment Cases

The relevant verification experiments and guided cases for this chapter are stored in the [Installation Directory]\RflySimAPIs\10.RflySimSwarm folder.

10.5.1 Interface Learning Experiments

Stored in the 10.RflySimSwarm\0.ApiExps folder, covering foundational platform interface tutorials and general introductions to various tools.

Experiment 1: Simulink Acceleration Mode Simulation Experiment for Four UAVs

📝 Experiment Overview:
This experiment uses the Simulink interface to control four UAVs to perform a formation flying simulation drawing a circle. By leveraging Simulink’s acceleration mode, simulation efficiency is enhanced, addressing real-time performance and resource consumption issues associated with complex control algorithms, enabling rapid validation of multi-UAV control algorithms.

Experiment 2: Remote Pixhawk Flight Controller Hardware Reboot via UDP

📝 Experiment Overview:
This experiment demonstrates how to remotely reboot Pixhawk flight controller hardware via UDP broadcast, resolving issues such as failure to take off or abnormal flight behavior caused by disordered flight controller parameters in Hardware-in-the-Loop (HITL) simulations, thereby enabling re-initialization of HITL simulations.

Experiment 3: Multi-UAV Terrain Altitude Acquisition Interface

📝 Experiment Overview:
Using the terrain altitude acquisition interface of the RflySim platform, this experiment loads map data via LoadPngData, retrieves terrain altitude at specified coordinates via getTerrainAltData, and automatically configures initial positions for multiple aircraft. Taking 12 UAVs as an example, it demonstrates how to automatically compute terrain altitude information for initial UAV placement based on the current terrain.

Experiment 4: MATLAB-Based ROS2 Control Experiment

📝 Experiment Overview:
Through this experiment, students will master the integration of ROS2 with MATLAB, understand data transmission principles of the MAVLink protocol within the PX4-ROS2-MATLAB control chain, and learn how to build UAV ROS2 control nodes and flight control logic using Simulink, enabling hardware-in-the-loop simulation validation.

Experiment 5: Batch Launch and Co-Simulation of Multiple Vehicle Types

📝 Experiment Overview:
This experiment uses batch scripts to launch and co-simulate multiple vehicle types—including multirotors, ground vehicles, and fixed-wing aircraft—in parallel—validating the stability of multi-vehicle concurrent operation in the RflySim environment and enabling rapid setup of multi-agent scenarios.

Experiment 6: ROS 2 Multi-UAV Swarm Control

📝 Experiment Overview:
This experiment teaches communication between ROS 2 and the PX4 flight control system, and demonstrates how to achieve cooperative multi-UAV control using Python scripts. Students will master Offboard-mode position control and swarm formation flight techniques.

Experiment 7: MAVLink_Full Swarm Control Experiment

📝 Experiment Overview:
Using the MAVLink communication interface of the RflySim platform, this experiment controls the position, velocity, and heading of four UAVs. It introduces the UDP mode of MAVLink_Full communication and demonstrates coordinated swarm control of UAVs.

Experiment 8: RflySim Swarm RflyUdpFast Interface Experiment

📝 Experiment Overview:
This experiment utilizes the RflySim swarm Simulink-RflyUdpFast interface to control multiple UAVs in Offboard mode. Students will learn the FullData and SimpleData modes of UDP communication and methods for multi-UAV cooperative control.

Experiment 9: Four-UAV Point-Mass Model Swarm Experiment

  • 📦 Version Requirement: Free Edition

    📝 Experiment Overview: Implements quadcopter swarm simulation on the RflySim platform using Python point-mass models. By commanding 4 drones to take off, hover for several seconds, and then descend, this experiment verifies the swarm simulation capability of the simplified flight control model.

Experiment 10: Swarm RflyUdpRaw Interface Experiment

📝 Experiment Overview: Controls 3 drones in Offboard mode by invoking the RflySim swarm Simulink-RflyUdpRaw interface. This experiment teaches IP configuration, UDP port setup, and MAVLink encoding methods for multi-drone swarm control.

Experiment 11: Fixed-Wing Point-Mass Model Swarm Experiment

📝 Experiment Overview: Constructs a fixed-wing point-mass model to achieve precise control over velocity, yaw angle, altitude, and position commands, verifying trajectory tracking performance.

Experiment 12: RflySerialRaw Interface for Single Drone Offboard Control

📝 Experiment Overview: Controls a single drone in Offboard mode via the RflySim swarm Simulink-RflySerialRaw interface. This experiment covers SerialRaw serial communication configuration and methods for issuing Offboard commands to multiple drones.

Experiment 13: RflyUdpMavlink Interface for Multi-Drone Offboard Control

📝 Experiment Overview: Controls multiple drones in Offboard mode by invoking the RflySim swarm Simulink-RflyUdpMavlink interface. This experiment teaches the usage of position, velocity, and acceleration control modes.

Experiment 14: RflySim Motion Capture VRPN Data Reception Interface Experiment

📝 Experiment Overview: Uses the RflySim swarm Simulink-RflyVrpnRecv interface to obtain real-time 6-DOF data (position, velocity, and acceleration) of drones and other objects in a motion-capture environment. Configures motion-capture IP, drone IP, and port settings to transmit motion-capture data to the flight controller for real-flight control.

Experiment 15: High-Maneuver Acceleration Control Experiment

📝 Experiment Overview: Implements drone Offboard control and high-maneuver acceleration control via a Simulink model. This experiment covers the usage of mode-switching modules (None/Offboard/Arm/Takeoff/Flying/Land/Disarm) and validates the acceleration control interface.

10.5.2 Basic Usage Experiments

Stored in the 10.RflySimSwarm\1.BasicExps folder, these experiments provide a complete set of supplementary teaching materials for beginners.

Experiment 1: MATLAB Control of UAV Simulation Experiment

📝 Experiment Overview: Introduces RflySim toolchain Simulink communication interface modules—including FullData, SimpleData, RflyUdpRaw, RflySerialRaw, and RflyUdpMavlink modes—and practices communication control and simulation verification for single- and four-drone systems.

Experiment 2: Swarm Light Show Demonstration

📝 Experiment Overview: Import multi-UAV trajectory data into RflySim 3D via Python scripts for 3D visualization preview, demonstrating lighting transformation effects. This is used to evaluate UAV formation flight performance and optimize flight paths.

Experiment 3: Quadrotor Point-Mass Model Swarm Experiment

📝 Experiment Overview: Implement takeoff and circular formation flight of 8 quadrotors using point-mass models on the RflySim platform. This validates the effectiveness of the hardware-in-the-loop (HIL) simulation method combining high-precision 6DOF models with real PX4 flight controllers.

Experiment 4: RflySim 3D Collision Interface Experiment

📝 Experiment Overview: Demonstrate collision effects for UAVs in the 3D engine via RflySim platform APIs, including single/multi-UAV collision detection, Simulink physics engine response, and P-mode communication optimization. This experiment focuses on learning collision detection and collision avoidance strategy design.

Experiment 5: Swarm Formation Control Simulation with Automatic Collision Avoidance

📝 Experiment Overview: Based on MATLAB/Simulink and the RflyUdpFast transmission module, implement 8-quadrotor figure-8 formation flight control with automatic collision avoidance, supporting 1–10 UAVs in formation.

Experiment 6: RflySim 3D Collision Detection API

📝 Experiment Overview: Demonstrate the usage of RflySim platform collision APIs, implementing UAV collision effects in the 3D engine via raycasting. This validates collision detection and response mechanisms, and evaluates flight safety.

Experiment 7: 8-UAV Point-Mass Model Swarm Simulation

📝 Experiment Overview: Implement takeoff and circular flight missions for 8 quadrotors using point-mass models on the RflySim platform. This validates the effectiveness of the hardware-in-the-loop (HIL) simulation method integrating high-precision 6DOF models (CopterSim) with real flight control systems (PX4).

Experiment 8: Single-UAV Control in RflyUdpFast FullData Mode

📝 Experiment Overview: Use the FullData mode of the RflyUdpFast transmission module to receive UAV state information, and implement local position motion control (circular trajectory) for a single UAV via Simulink modeling. This enables software-in-the-loop (SIL) or hardware-in-the-loop (HIL) simulation experiments.

Experiment 9: 8-UAV Figure-8 Formation Flight Control

📝 Experiment Overview: Implement figure-8 formation flight control for 8 quadrotors using MATLAB/Simulink and the RflyUdpFast module. This validates the effectiveness and stability of multi-UAV formation control algorithms.

Experiment 10: Vision-Based UAV Swarm Following Experiment

📝 Experiment Overview: Identify red circular targets via HSV color space segmentation, and implement visual servoing control for two- or four-UAV formations to follow a moving target using a proportional controller. This experiment focuses on learning and practicing vision-based formation control techniques.

Experiment 11: RflyUdpMavlink Real-Time Simulation

  • 📦 Version Requirement: Free Edition

    📝 Experiment Overview:
    Utilize the RflyUdpMavlink library to implement MAVLink message transmission, reception, and parsing, and control UAVs via Simulink S-Functions for Software-in-the-Loop (SIL) and Hardware-in-the-Loop (HIL) simulations.

Experiment 12: RflyUdpMavlink Real UAV Control

📝 Experiment Overview:
Connect Simulink to real UAVs via the RflyUdpMavlink communication interface to implement safety self-checks and status monitoring. Supports both SIL and HIL modes for algorithm validation and flight controller performance testing.

Experiment 13: Dual-UAV Collision MAVLink API Experiment

📝 Experiment Overview:
Use the Python API to control two aircraft in RflySim 3D and simulate collision events, evaluating collision detection accuracy, response mechanisms, and system performance.

Experiment 14: RflyUdpFast FullData Mode Four-UAV Simulation

📝 Experiment Overview:
Employ the RflyUdpFast transmission module in FullData mode of the RflySim toolchain to build a Simulink model for centralized position control of quadrotor UAVs, validating control algorithm effectiveness and analyzing swarm coordination performance.

Experiment 15: RflySim UDP Mode Dual-UAV Collision Experiment

📝 Experiment Overview:
Simulate the takeoff and collision process of two aircraft via the Python API, evaluating RflySim 3D’s collision engine in terms of detection capability, response mechanisms, and system performance, while learning multi-UAV collision simulation methods under P0–P3 communication modes.

Experiment 16: RflyUdpFast FullData Mode Four-UAV Global Coordinate Control Experiment

📝 Experiment Overview:
Learn to use the RflyUdpFast transmission module (FullData mode) of the RflySim toolchain’s Simulink library to receive state information from four UAVs, construct a local position motion control model, and achieve centralized coordinated control and trajectory tracking for quadrotor UAV swarms.

Experiment 17: RflySim UDP Mode Dual-UAV Collision Simulink Experiment

📝 Experiment Overview:
Use MATLAB/Simulink to control two aircraft in collision simulation experiments via UDP mode, evaluating RflySim 3D’s collision detection, response mechanisms, and system performance under collision engine mode.

Experiment 18: RflyUdpFast SimpleData Mode Single-UAV Circular Trajectory Experiment

📝 Experiment Overview:
Learn to use the SimpleData transmission module of the RflySim toolchain to receive UAV state information, and implement single-UAV circular trajectory control via Simulink modeling. Master parameter configuration of the RflyUdpFast module and SIL/HIL experimental methodologies.

Experiment 19: RflyUdpFast SimpleData Mode Four-UAV Simulation

  • 📦 Version Requirement: Free Edition

    📝 Experiment Overview:
    Learn to use the RflyUdpFast transmission module (SimpleData mode) of the RflySim toolchain to receive state information from quadrotor UAVs. Implement local position motion control for a single UAV via Simulink modeling, master the method of converting control commands into offboard signals and sending them to the flight controller, and conduct software-in-the-loop (SITL) and hardware-in-the-loop (HIL) simulation experiments.

Experiment 20: SimpleData Mode Four-UAV Global Coordinate Control

📝 Experiment Overview:
Achieve centralized global coordinate control of a quadrotor UAV swarm using the SimpleData mode of the RflyUdpFast transmission module. The experiment includes building a Simulink control model, subscribing to localPos position data, generating velocity control commands by comparing with a circular reference trajectory, converting these commands into offboard signals via the SimpleCtrl4D module, and performing SITL or HIL simulations.

Experiment 21: FullDataModel Mode UAV Control Experiment

📝 Experiment Overview:
Conduct SITL/HIL experiments using the RflyUdpFast transmission module. Subscribe to localPos position data from the FullData bus, generate velocity control commands by comparing with a desired circular trajectory, and convert them into offboard-mode signals via the vel_ned_full module to control UAV flight.

Experiment 22: RflySerialRaw FullData Mode Single-UAV Control

📝 Experiment Overview:
Receive UAV state data via the RflyUdpFast transmission module and perform Simulink-based simulation of local-position circular motion control for a single UAV.

Experiment 23: RflyUdpMavlink FullData Mode Single-UAV Communication Experiment

📝 Experiment Overview:
Use the Mavlink_simple mode of the RflyUdpMavlink library to receive and send MAVLink messages, achieving the same functionality as the UDP_simple mode of the RflyUdpFast module for single-UAV control, supporting SITL or HIL simulations.

10.5.3 Advanced Development Experiments

Stored in the 10.RflySimSwarm\2.AdvExps folder, these experiments further familiarize users with certain low-level firmware ecosystem configurations.

Experiment 1: MATLAB Centralized Control of Eight-UAV Simulation

📝 Experiment Overview:
Control eight UAVs to perform circular formation flight using the SimpleData mode of the Simulink communication interface and the RflyUdpFast transmission module, mastering MATLAB/Simulink-based centralized swarm control methods.

Experiment 2: Distributed Local Area Network Communication for 8-UAV Simulation

📝 Experiment Overview:
Learn RflySim platform’s distributed LAN multi-UAV cooperative simulation techniques, master configuration methods for broadcast and point-to-point communication, and achieve circular formation flight for 8 UAVs.

Experiment 3: PX4 and Point-Mass Model Hybrid Swarm Control

  • 📦 Version Requirement: Free Edition

    📝 Experiment Overview:
    This experiment implements heterogeneous swarm formation flight using PX4 SITL and point-mass models on the RflySim platform. It covers MAVLink Offboard control procedures, formation following methods, and NED coordinate system applications, with 1 Leader drone guiding 4 Followers to form a diamond-shaped formation.

Experiment 4: Heterogeneous Swarm Cooperative Control (3 Cars + 3 UAVs)

📝 Experiment Overview:
Through coordinated motion between quadrotor UAVs and ground vehicles, this experiment helps understand control logic and communication mechanisms in multi-agent systems.

Experiment 5: BroadNetSwarm_Mat Distributed Local Area Network Broadcast Communication (8-UAV Simulation)

📝 Experiment Overview:
This experiment uses two computers connected via a local area network for co-simulation, enabling 8 UAVs to fly in a circular formation. It introduces RflySim swarm simulation capabilities and MATLAB control program usage.

Experiment 6: 3v3 UAV Regional Defense Combat System

📝 Experiment Overview:
Through a 3v3 UAV regional defense combat scenario, this experiment teaches rule-based behavior decision-making, swarm control principles, and RflySim simulation operations.

Experiment 7: Centralized Control Simulink Simulation for 8-UAV Swarm

📝 Experiment Overview:
Using the RflyUdpFast transmission module to receive UAV states, this experiment performs centralized position motion control of an 8-UAV swarm via Simulink, verifying control algorithm effectiveness and data transmission performance, achieving circular flight for 8 UAVs.

Experiment 8: Collision-Avoidance Speed Control for UAV Swarm

📝 Experiment Overview:
This experiment implements automatic collision-avoidance speed control for UAV swarm formations using MATLAB/Simulink. Based on the artificial potential field algorithm, it enables various formation transitions and obstacle avoidance, supporting both Software-in-the-Loop (SITL) and Hardware-in-the-Loop (HITL) simulation modes.

Experiment 9: UDP-Based 8-UAV Concentric Circle Trajectory Flight

📝 Experiment Overview:
This experiment achieves automatic takeoff and concentric circle trajectory flight for 8 UAVs via UDP communication, validating RflySim’s swarm control and MAVLink UDP communication capabilities.

Experiment 10: Vehicle-UAV Cooperative Tracking Experiment

📝 Experiment Overview:
Through coordinated motion between quadrotor UAVs and ground vehicles, this experiment deepens understanding of multi-agent system control logic and communication mechanisms, while mastering UDP communication, Offboard mode control, and conditional-triggered state machine techniques.

Experiment 11: UDP-Based Multi-UAV Point-Position Formation Control

📝 Experiment Overview:
This experiment implements multi-UAV point-position formation control using MATLAB/Simulink and UDP communication. It covers hierarchical control architecture, PID position control algorithms, coordinate system transformations, and the design and deployment of formation algorithms.

Experiment 12: Swarm Formation Collision-Avoidance Control Experiment

  • 📦 Version Requirement: Free Edition

    📝 Experiment Overview:
    This experiment implements UAV swarm formation control using MATLAB/Simulink, employing automatic collision-avoidance velocity control (in NED coordinate frame). It demonstrates three formations (triangle, horizontal line, vertical line), formation transitions, and obstacle avoidance capabilities in both software-in-the-loop (SIL) and hardware-in-the-loop (HIL) simulations.

Experiment 13: Distributed Local Area Network 8-UAV Cooperative Simulation

📝 Experiment Overview:
Utilizes the RflySim platform to perform cooperative simulation of 8 UAVs drawing a circular formation across two computers connected via a local area network. Distributed swarm control is implemented using MATLAB/Simulink, overcoming performance limitations of a single computer.

Experiment 14: Vehicle-UAV Cooperative Yaw Control Experiment

📝 Experiment Overview:
Learn cooperative control methods for UAVs and ground vehicles within multi-agent systems, and understand PID-based yaw control mechanisms and their application in target tracking.

Experiment 15: MATLAB Multi-UAV Circular Trajectory Formation Control

📝 Experiment Overview:
Implements circular trajectory formation control for 6 UAVs using MATLAB/Simulink. Covers core processes including initialization configuration, trajectory generation, and mode switching, while learning PID control, UDP communication, and parametric design of circular trajectories.

Experiment 16: UAV Swarm Collision-Avoidance Formation Control Experiment

📝 Experiment Overview:
Implements UAV swarm formation control using MATLAB/Simulink, demonstrating transitions among three formations (triangle, horizontal line, vertical line). Employs the artificial potential field algorithm to achieve automatic obstacle avoidance and formation switching, covering both software-in-the-loop and hardware-in-the-loop simulations.

Experiment 17: Multi-UAV Figure-8 Trajectory Tracking Control

📝 Experiment Overview:
Implements figure-8 trajectory tracking control for single or multiple UAVs using MATLAB/Simulink. Covers Lissajous curve trajectory generation algorithms and multi-UAV formation coordination mechanisms.

Experiment 18: Heterogeneous UAV-USV Cooperative Tracking

📝 Experiment Overview:
Learn cooperative control and communication mechanisms in heterogeneous multi-agent systems, master multi-threaded UDP-based communication and simulation synchronization techniques, and validate real-time UAV tracking of USVs.

Experiment 19: Multi-UAV Master-Slave Formation Control

📝 Experiment Overview:
Implements master-slave formation tracking control for multiple UAVs using MATLAB/Simulink. Covers UDP communication mechanisms, master-slave formation control architecture, and modular design methodologies, including both software-in-the-loop and hardware-in-the-loop simulations.

Experiment 20: USV-UUV Cooperative Towing Control

  • 📦 Version Requirement: Free Edition

    📝 Experiment Overview:
    Learn the cooperative control principle of surface unmanned vessels (USVs) and underwater unmanned vehicles (UUVs) based on tether traction, master the calculation of tether dynamics and the application of PX4 Offboard mode in heterogeneous unmanned systems, and verify the traction control performance through RflySim 3D simulation.

Experiment 21: Multi-UAV Multi-Trajectory Formation Control

📝 Experiment Overview:
Implement multi-UAV formation control and multiple trajectory switching via MATLAB/Simulink, supporting formations such as line, triangle, and square, as well as trajectory patterns including figure-8, square, and spiral, with state management and smooth trajectory transition algorithms included.

Experiment 22: RflySim UDP Virtual Structure Formation Control

📝 Experiment Overview:
Implement multi-UAV formation control based on the virtual structure method via MATLAB/Simulink, learning UDP communication mechanisms, virtual leader-based formation algorithms, and PID position control principles, while mastering the hierarchical control architecture and trajectory planning techniques for quadrotor UAVs.

Experiment 23: Vehicle-UAV Cooperative Tracking Experiment

📝 Experiment Overview:
Through cooperative motion between a quadrotor UAV and an unmanned ground vehicle (UGV), understand heterogeneous communication mechanisms and control logic in multi-agent systems. Two UGVs cruise along a rounded-rectangle path, while two UAVs simultaneously track and follow their respective UGVs in real time, intelligently aligning their headings with the front of the UGVs.

10.5.4 Advanced Development Experiments

Stored in the 10.RflySimSwarm\3.CustExps folder, these experiments are designed for advanced users engaged in custom development.

Experiment 1: Quadrotor Distributed Trajectory Tracking Experiment

📝 Experiment Overview:
A quadrotor UAV formation trajectory tracking experiment based on the leader-follower method: UAV #1 vertically takes off to 10 m, followed by UAVs #2, #3, and #4, which take off sequentially with 3-second intervals and maintain formation while tracking the trajectory.

Experiment 2: Distributed Local Area Network 16-UAV Simulation Experiment

📝 Experiment Overview:
Utilize the cluster simulation capability of the RflySim platform to coordinate 16 UAVs across two PCs connected via a local area network to perform circular flight. This experiment covers multi-UAV cooperative simulation configuration and UDP communication techniques.

Experiment 3: Quadrotor Distributed Trajectory Tracking Experiment

📝 Experiment Overview:
A quadrotor UAV formation trajectory tracking experiment based on the leader-follower method, where four UAVs take off sequentially and track a circular trajectory via Python interface control. The algorithm is migrated to WSL to resolve platform compatibility issues.

Experiment 4: Multi-UAV Point-Mass Model Swarm Experiment

📝 Experiment Overview:
Implement takeoff and circular flight for multiple point-mass-model quadrotor UAVs on the RflySim platform, supporting swarm scales of 30, 100, or 200 UAVs, to validate the feasibility of swarm algorithms in software-in-the-loop (SITL) simulation.

Experiment 5: Swarm Intelligence Tutorial

  • 📦 Version Requirement: Full Version

    📝 Experiment Overview:
    Includes three multi-UAV swarm intelligent control experiments: Ant Algorithm-based Path Planning, Olfati-Saber Swarm Obstacle Avoidance Algorithm, and Deep Reinforcement Learning-based UAV Area Defense. Covers core algorithms such as multi-UAV cooperative path planning, obstacle-avoidance aggregation, and intelligent defense.

Experiment 6: Quadrotor Distributed Trajectory Tracking Experiment

📝 Experiment Overview:
A quadrotor formation trajectory tracking experiment based on the leader-follower method, where multiple UAVs are sequentially launched and maintain formation flight via the Python interface. Focuses on learning distributed formation control algorithms and network communication mechanisms.

Experiment 7: 50-UAV Formation Control Hardware-in-the-Loop Simulation

📝 Experiment Overview:
Hardware-in-the-loop simulation of 50 UAVs divided into three groups for formation control, conducted across multiple computers on a local network. Commands are transmitted between master and slave machines via UDP, validating swarm simulation and formation control algorithms.

Experiment 8: 100-UAV Formation Control Hardware-in-the-Loop Simulation

📝 Experiment Overview:
Hardware-in-the-loop simulation of 100 UAVs divided into three groups for formation control, conducted across 10 computers on a local network. Commands are transmitted between master and slave machines via UDP, enabling learning of distributed swarm simulation and formation control methods.

Experiment 9: Quadrotor Distributed Trajectory Tracking Experiment

📝 Experiment Overview:
A quadrotor formation control experiment based on the leader-follower method, where distributed multi-UAV trajectory tracking is implemented via the Python interface. UAV #1 (leader) and UAV #3 (follower) take off sequentially with a 3-second interval and maintain formation flight.

Experiment 10: Quadrotor Distributed Trajectory Tracking Experiment

📝 Experiment Overview:
A quadrotor formation trajectory tracking experiment based on the leader-follower method, where UAV #1 (leader) guides UAVs #2, #3, and #4 (followers) to take off sequentially and perform route-planned flight via the Python interface.

Experiment 11: PX4 + MPSwarm25 Hybrid Swarm Control

📝 Experiment Overview:
Master flight control methods for hybrid swarm systems combining PX4 SITL and point-mass models. Achieves synchronized flight of five independent formations: 5 PX4 SITL UAVs as Leaders and 20 point-mass model UAVs as Followers.

Experiment 12: Ant Algorithm-Based Multi-UAV Path Planning

📝 Experiment Overview:
Plans feasible and near-optimal paths for multiple UAVs that satisfy obstacle and collision-avoidance constraints by applying and optimizing the Ant Algorithm. Focuses on learning the application of swarm intelligence algorithms in trajectory planning.

Experiment 13: Distributed 8-UAV Broadcast Communication Simulation

📝 Experiment Overview:
Demonstrates a distributed simulation system based on RflySim, where 8 UAVs achieve cooperative flight control via UDP broadcast over a local network. Covers distributed system architecture and Git-based deployment methodologies.

Experiment 14: Distributed Local Network Broadcast Communication 16-UAV Simulation

  • 📦 Version Requirement: Full Version

    📝 Experiment Overview:
    This experiment teaches how to perform joint simulation on two computers connected via a local area network (LAN) on the RflySim platform, enabling 16 drones to fly in a circular formation. It covers configuration and operation of distributed swarm simulations.

Experiment 15: Generating Executable (.exe) Files for Swarm Control Programs

📝 Experiment Overview:
This experiment demonstrates converting a Simulink simulation program for 4 drones into an executable (.exe) file. It covers MATLAB/Simulink code generation and Visual Studio compiler configuration, verifying that the generated .exe correctly executes drone simulation tasks and evaluating its performance.

Experiment 16: DistSim Distributed Swarm Communication Test

📝 Experiment Overview:
This experiment implements distributed communication testing between two computers using the DistSim software, covering LAN configuration, node ID assignment, and methods for sending test commands to all computers or specific ones.

Experiment 17: 30-Drone Point-Mass Model Swarm Experiment

📝 Experiment Overview:
This experiment demonstrates takeoff, hover, and synchronized circular flight of 30 point-mass quadrotor models on the RflySim platform, showcasing single-computer swarm simulation capabilities at the scale of hundreds of drones.

Experiment 18: Hardware-in-the-Loop Formation Control for Multi-Drone Systems

📝 Experiment Overview:
Using RflySim’s swarm simulation functionality, this experiment performs joint simulation across multiple computers on a LAN to control 50–100 drones divided into three groups for formation control, introducing hardware-in-the-loop (HITL) simulation techniques.

Experiment 19: Distributed Simulation 8-Drone UAV Experiment

📝 Experiment Overview:
This experiment demonstrates the architecture of a distributed simulation system on the RflySim platform, covering collaborative execution of 8-drone swarm simulations across multiple computers. It includes Git-based code synchronization, JSON configuration-driven setup, and UDP network communication configuration.

Experiment 20: Distributed Trajectory Tracking for Quadrotor Swarms

📝 Experiment Overview:
This experiment implements distributed control and cooperative trajectory tracking for quadrotor UAV swarms on the RflySim platform. It covers multi-drone coordinated control methods, leader-follower formation control strategies, and UDP-based network communication mechanisms.

Experiment 21: 100-Drone Point-Mass Model Swarm Experiment

📝 Experiment Overview:
This experiment uses Python-based point-mass models on the RflySim platform to simulate a 100-drone swarm, completing takeoff, hover, and synchronized circular flight tasks. It evaluates flight performance and control accuracy when integrating high-fidelity models with real flight controllers.

Experiment 22: Olfati-Saber Drone Swarm Obstacle Avoidance Experiment

  • 📦 Version Requirement: Full Version

    📝 Experiment Overview: Implements obstacle avoidance, collision avoidance, and target-point aggregation for 6 UAVs using the Olfati-Saber algorithm, enabling the swarm to traverse obstacle-rich regions and complete collective flight missions.

Experiment 23: DistSim Distributed 16-UAV Cooperative Flight

📝 Experiment Overview: Demonstrates the establishment of a distributed communication system across two computers to coordinate 16 UAVs in synchronized circular flight. Covers distributed system architecture, UDP communication, multi-UAV cooperative control, and automated deployment.

Experiment 24: Embodied Intelligence Reinforcement Learning for Quadrotor Hover Control

📝 Experiment Overview: Uses the Genesis simulation environment and RflySim software-in-the-loop to achieve acceleration-level hover control of a quadrotor UAV via the PPO reinforcement learning algorithm. Covers sample equivalence and parameter tuning methods for transitioning from high-parallel GPU training to low-parallel CPU training, including domain randomization and noise injection to enhance policy generalization.

Experiment 25: Multi-UAV Area Defense via Reinforcement Learning

📝 Experiment Overview: Trains a UAV defense model using the MADDPG deep reinforcement learning algorithm to enable autonomous interception and neutralization of attacking UAVs, achieving multi-agent cooperative control.

Experiment 26: Point-Mass Model 200-UAV Swarm Simulation

📝 Experiment Overview: Based on the RflySim platform, executes takeoff and circular flight missions for 200 point-mass-model UAVs across two computers, showcasing high-fidelity simulation effects achieved by integrating precise models with real flight control systems.

Experiment 27: 10-UAV Circular Flight

📝 Experiment Overview: Implements circular trajectory control for 10 UAVs in Offboard mode using the RflySim cluster Simulink-RflyUDPFast interface, covering multi-UAV formation control and UDP communication.

Experiment 28: Distributed Swarm Simulation Experiment

📝 Experiment Overview: Demonstrates the working principle of a distributed simulation system based on the RflySim platform, establishing a networked environment with Windows and Linux (NX) nodes, and enabling automated Git-based code deployment and multi-node collaborative simulation control.

Experiment 29: Multi-UAV Area Defense Competitive System

📝 Experiment Overview: Through a 5v5 UAV area defense competition system, learns rule-based behavior decision-making mechanisms, masters RflySim simulation platform operation, understands multi-UAV swarm confrontation system architecture, and achieves coordinated control and state management for multiple UAVs.