In modern warfare and civilian applications, unmanned aerial vehicles (UAVs) have become increasingly pivotal due to their versatility and cost-effectiveness. My research focuses on enhancing the operational capabilities of UAVs through the development of a ground-based control system that enables coordinated drone formation flight. This system integrates hardware and software components to monitor and control multiple drones simultaneously, ensuring precise formation keeping and robust performance. The core innovation lies in the design of a formation keeping controller using linear control theory, which has been validated through simulations. This article elaborates on the system architecture, control strategies, and implementation details, emphasizing the importance of ground control in managing complex drone formation operations.
The proliferation of UAV technology has led to a growing demand for multi-drone协同 operations, where drone formation flight plays a critical role in tasks such as surveillance, reconnaissance, and environmental monitoring. A ground control system serves as the nerve center for these operations, providing real-time data processing, communication, and command execution. My work addresses the challenge of designing a scalable system that can handle multiple drones while maintaining formation integrity. By leveraging ground-based测控, I aim to achieve high precision in drone formation control, reducing reliance on onboard autonomy and enabling human oversight. This approach not only enhances reliability but also facilitates adaptive responses to dynamic environments.
The ground control system is structured around five key modules: central processing, human-machine interaction, wireless communication, command acquisition, and a command center. The central processing module consists of four computers, each dedicated to a specific platform—multi-drone status monitoring, digital mapping, trajectory tracking, and mission control. These computers handle data exchange, processing, and visualization, ensuring comprehensive situational awareness for operators. The human-machine interaction module features interfaces built using MFC and third-party controls, with a digital map implemented by invoking Google Earth’s COM API via VC++6.0 for enhanced geospatial context. This setup allows operators to monitor drone formation parameters, such as position, speed, and altitude, in real-time.
Command acquisition is facilitated through input devices like joysticks and keyboards, which capture flight and payload control instructions. The wireless communication module employs data radios to establish links between the ground station and drones, utilizing a time-division multiple access (TDMA) scheme for efficient multi-point connectivity. Specifically, a fixed-slot TDMA protocol is used, where each drone in the drone formation is assigned a predetermined time slot for communication, minimizing interference and ensuring synchronized data transmission. The command center acts as a higher-level指挥 hub, receiving位置 data from the ground station and issuing task directives. This architecture supports scalable drone formation management, enabling control over multiple UAVs from a single ground station.
| Module | Components | Function |
|---|---|---|
| Central Processing | 4 computers with displays | Data processing and interface display |
| Command Acquisition | Instruction cards, keyboards, joysticks | Capture operator inputs for drone control |
| Wireless Communication | Data radios (ground and airborne terminals) | Establish reliable links for drone formation communication |
Software development centers on modular design to ensure flexibility and robustness. The human-machine interface includes multiple monitoring platforms that display critical parameters for drone formation flight, such as relative distances and velocities. Serial communication is implemented using timers to manage data exchange between the ground station and drones, while Socket-based network applications handle communication with the command center. The TDMA-based wireless networking module coordinates access among drones, with a synchronization process where the ground station broadcasts clock signals, and drones respond in assigned slots. This ensures orderly communication within the drone formation, reducing latency and packet loss. The integration of these software elements allows for seamless control of drone formation dynamics, supporting both autonomous and manual intervention modes.
To achieve effective drone formation flight, I developed a kinematic model for relative motion between drones, focusing on a leader-follower configuration. This model simplifies the control problem by considering two drones in a planar scenario, with the follower’s velocity aligned along its body-fixed x-axis. The relative motion equations describe the dynamics between the leader and follower in a drone formation, accounting for position, velocity, and heading differences. These equations form the basis for designing a formation keeping controller that maintains desired spacing and orientation within the drone formation.
The relative motion is expressed in terms of the follower’s reference frame, where the leader’s velocity and heading are projected onto this frame. Let \( V_W \) and \( V_L \) denote the velocities of the follower and leader, respectively, and \( \psi_W \) and \( \psi_L \) their heading angles. The relative distance between them is \( R_{WL} \), with the heading error defined as \( \psi_E = \psi_L – \psi_W \). The components of relative distance along the follower’s x and y axes are \( x \) and \( y \). The kinematic equations for horizontal motion are:
$$ \dot{x} = V_L \cos \psi_E – V_W + \dot{\psi}_w y $$
$$ \dot{y} = V_L \sin \psi_E – \dot{\psi}_w x $$
For vertical motion, the height difference \( z \) between drones in the drone formation is governed by:
$$ \dot{z} = \dot{h}_L – \dot{h}_W $$
These equations capture the essential interactions in a drone formation, enabling the design of control laws for formation keeping. By linearizing these dynamics around equilibrium points, I applied linear control theory to develop a proportional-integral (PI) controller that ensures stable drone formation flight. The controller operates in three channels: longitudinal, lateral, and vertical, each addressing specific error signals to adjust the follower’s speed, heading, and altitude.
The formation keeping controller employs a two-level strategy: upper-level control for overall drone formation navigation (e.g., collective height and speed) and lower-level control for individual drone adjustments (e.g., relative spacing). This hierarchical approach simplifies the complexity of managing multiple drones in a formation. For the longitudinal channel, the error signal \( \Delta X \) combines spacing and velocity deviations:
$$ \Delta X = k_x (x – x_c) + k_V (V_L – V_W) = k_x \Delta x + k_V \Delta V $$
where \( x_c \) is the desired longitudinal spacing, \( \Delta x \) and \( \Delta V \) are errors, and \( k_x \) and \( k_V \) are gain coefficients. The PI control law outputs a speed command \( V_{Wc}(t) \) for the follower:
$$ V_{Wc}(t) = K_{xp} \Delta X + K_{xi} \int_0^t \Delta X \, dt $$
Similarly, for the lateral channel, the error \( \Delta Y \) includes spacing and heading errors:
$$ \Delta Y = k_y (y_c – y) + k_{\psi} (\psi_L – \psi_W) = k_y \Delta y + k_{\psi} \Delta \psi $$
with the heading command \( \psi_{Wc}(t) \) given by:
$$ \psi_{Wc}(t) = K_{yp} \Delta Y + K_{yi} \int_0^t \Delta Y \, dt $$
For the vertical channel, the error \( \Delta Z \) is based on height difference:
$$ \Delta Z = k_z (z_c – z) = k_z \Delta z $$
and the height command \( h_{Wc}(t) \) is:
$$ h_{Wc}(t) = K_{zp} \Delta Z + K_{zi} \int_0^t \Delta Z \, dt $$
These control laws are implemented in the ground station software, which processes real-time data from the drone formation to compute corrective actions. The gains are tuned to ensure rapid response and minimal steady-state error, as summarized in the table below.
| Parameter | Value | Description |
|---|---|---|
| \( k_x \) | -5 | Longitudinal spacing gain |
| \( k_V \) | 8 | Velocity error gain |
| \( K_{xp} \) | 25.1 | Proportional gain for longitudinal control |
| \( K_{xi} \) | 10.3 | Integral gain for longitudinal control |
| \( k_y \) | -2.1 | Lateral spacing gain |
| \( k_{\psi} \) | 10 | Heading error gain |
| \( K_{yp} \) | 1.5 | Proportional gain for lateral control |
| \( K_{yi} \) | 0.01 | Integral gain for lateral control |
| \( k_z \) | 10 | Vertical spacing gain |
| \( K_{zp} \) | 1 | Proportional gain for vertical control |
| \( K_{zi} \) | 0.1 | Integral gain for vertical control |
Simulation studies were conducted to validate the controller’s performance in maintaining drone formation integrity. The scenario involved two drones: a leader and a follower, with initial conditions set to a height of 3000 m, leader speed of 60 m/s, follower speed of 50 m/s, and desired lateral and longitudinal spacings of 500 m. The leader’s maneuvers, such as changes in speed, heading, or altitude, acted as disturbances, while the controller adjusted the follower’s states to maintain formation. Results demonstrated that the PI controller effectively stabilized the drone formation, with zero steady-state error and fast convergence. The speed response showed the follower matching the leader’s velocity within seconds, while the heading and height responses indicated precise tracking of desired trajectories. These outcomes confirm the feasibility of using linear control for drone formation keeping, especially when integrated with ground-based测控 systems.
The “human-in-the-loop” control aspect complements the autonomous formation keeper by allowing operator intervention when anomalies occur. In practice, the ground station displays real-time drone formation data, such as relative distances and flight paths. If deviations exceed thresholds—for instance, if spacing errors arise due to environmental factors—operators can use input devices to issue corrective commands. This hybrid approach enhances robustness, as human judgment can address unforeseen scenarios that automated controllers might not handle optimally. The ground station software processes these manual inputs alongside controller outputs, ensuring seamless integration for reliable drone formation management. This combination of autonomy and human oversight is crucial for complex missions where drone formation flight must adapt to dynamic conditions.
Further analysis of the drone formation control system involves considering scalability to larger formations. The TDMA communication protocol supports multiple drones by allocating time slots, but as the number of drones increases, latency may become a concern. To address this, dynamic slot allocation could be explored, though the fixed scheme sufficed for the tested scenarios. Additionally, the kinematic model assumes planar motion; extending it to three dimensions would involve more complex equations but could improve accuracy for vertical maneuvers in drone formation flight. The control gains listed in the table were optimized through iterative simulation, but adaptive tuning methods might enhance performance under varying operational conditions. Future work could integrate machine learning techniques to predict and compensate for disturbances, further solidifying the drone formation control system.
In conclusion, the ground-based control system for drone formation flight represents a significant advancement in multi-UAV operations. By designing a robust hardware and software architecture, I enabled simultaneous monitoring and control of multiple drones, with a formation keeping controller that ensures stable and precise coordination. The PI control laws, validated through simulations, demonstrate effective error reduction in speed, heading, and altitude channels. The inclusion of “human-in-the-loop” capabilities adds a layer of resilience, making the system adaptable to real-world challenges. This research underscores the potential of ground测控 in enhancing drone formation capabilities, paving the way for more sophisticated applications in military and civilian domains. As drone technology evolves, such integrated systems will be vital for achieving reliable and efficient drone formation flight in complex environments.