This paper presents a comprehensive study on conflict resolution methods for large fixed-wing drones operating in low-altitude airspace near regional airports. The increasing integration of large fixed-wing drones into civil airspace, particularly for logistics and cargo delivery, demands robust real-time collision avoidance strategies. Traditional approaches often fail to account for the dynamic nature of air traffic and the physical constraints of fixed-wing drones. To address these challenges, we propose an improved Artificial Potential Field (APF) algorithm tailored specifically for fixed-wing drones. The method decomposes repulsive forces into position-dependent and velocity-dependent components to handle moving obstacles, and attractive forces into positional and velocity terms to ensure efficient path recovery after conflict resolution. Additionally, a damping term is introduced to suppress trajectory oscillations. Simulation results demonstrate that the enhanced APF method generates smooth, flyable trajectories that satisfy minimum separation requirements and adhere to the dynamic constraints of fixed-wing drones, including maximum speed, acceleration, and turn radius. The proposed approach significantly improves safety and efficiency compared with conventional APF. All key performance metrics and algorithm parameters are summarized in detailed tables and mathematical formulations.
The operation of large fixed-wing drones in terminal airspace near regional airports presents unique safety challenges. Unlike rotary-wing UAVs, fixed-wing drones have inherent kinematic constraints such as minimum turning radius and limited climb rates, which must be respected during conflict resolution maneuvers. Moreover, the presence of manned general aviation aircraft in the same low-altitude corridor demands that fixed-wing drones maintain equivalent levels of safety. Our work focuses on developing a conflict resolution framework that meets these requirements while ensuring minimal deviation from the planned flight path. The methodology is built upon an enhanced artificial potential field model that explicitly considers relative velocities, aerodynamic limitations, and real-time sensor data.
1. Problem Formulation
Consider a scenario where a large fixed-wing drone (UAV) is flying along a predefined route near a regional airport. A manned general aviation aircraft (intruder) operates in the same airspace, assumed to maintain constant speed for short-term predictions. The objective is to generate a real-time evasive maneuver for the fixed-wing drone that ensures a minimum safety distance \( r_{\text{safe}} = 5942 \, \text{m} \) (derived from low-altitude airspace regulations and sensor latency margins) while respecting the drone’s physical capabilities. The conflict resolution must be completed without violating the maximum speed \( v_{\text{max}} \), maximum acceleration \( a_{\text{max}} \), and minimum turn radius \( r_{\text{min}} \).
The motion dynamics of a fixed-wing drone in coordinated level flight are governed by the following equations:
\[
L \cos\phi = mg
\]
\[
m \frac{v^2}{r} = L \sin\phi
\]
\[
n = \frac{L}{mg}
\]
\[
r = \frac{m v^2}{n mg \sin\phi} = \frac{v^2}{n g \sin\phi}
\]
where \( L \) is lift, \( \phi \) is roll angle, \( n \) is load factor, \( v \) is speed, and \( r \) is turn radius. The minimum turn radius is constrained by \( r_{\text{min}} \geq \frac{v^2}{n_{\text{max}} g} \). These constraints are critical for generating flyable trajectories for fixed-wing drones.
2. Improved Artificial Potential Field Method
Traditional APF suffers from local minima, target unreachability, and path oscillations. Our improvements address these issues specifically for fixed-wing drones. The total force acting on the drone is defined as:
\[
\mathbf{F}_{\text{total}} = \mathbf{F}_{\text{rep}} + \mathbf{F}_{\text{att}}
\]
2.1 Repulsive Force Decomposition
To handle dynamic obstacles, the repulsive force is split into a position-dependent component \(\mathbf{F}_{\text{rep}}^p\) and a velocity-dependent component \(\mathbf{F}_{\text{rep}}^v\):
\[
\mathbf{F}_{\text{rep}} = \mathbf{F}_{\text{rep}}^p + \mathbf{F}_{\text{rep}}^v
\]
The position repulsion is:
\[
\mathbf{F}_{\text{rep}}^p =
\begin{cases}
\mathbf{F}_{\text{rep1}}^p + \mathbf{F}_{\text{rep2}}^p & \text{if } \|\mathbf{d}\| \leq r_{\text{safe}} \\
0 & \text{otherwise}
\end{cases}
\]
\[
\mathbf{F}_{\text{rep1}}^p = -k_p^{\text{rep}} \left( \frac{1}{\|\mathbf{d}\|} – \frac{1}{r_{\text{safe}}} \right) \frac{\|\mathbf{d}_g\|^2}{\|\mathbf{d}\|^2} \frac{\partial \|\mathbf{d}\|}{\partial \mathbf{x}}
\]
\[
\mathbf{F}_{\text{rep2}}^p = -k_p^{\text{rep}} \left( \frac{1}{\|\mathbf{d}\|} – \frac{1}{r_{\text{safe}}} \right)^2 \|\mathbf{d}_g\| \frac{\partial \|\mathbf{d}\|}{\partial \mathbf{x}}
\]
The velocity repulsion is:
\[
\mathbf{F}_{\text{rep}}^v = -k_v^{\text{rep}} \mathbf{v}_{\text{rel}}
\]
where \(\mathbf{d} = \mathbf{x}_p – \mathbf{x}_{\text{ob}}\) is the relative position vector, \(\mathbf{d}_g = \mathbf{x}_p – \mathbf{x}_{\text{goal}}\) is the vector from drone to goal, \(\mathbf{v}_{\text{rel}}\) is relative velocity, and \(k_p^{\text{rep}}, k_v^{\text{rep}}\) are gains.
2.2 Attractive Force with Velocity Guidance
The attractive force is similarly decomposed to ensure the fixed-wing drone returns to its original path after conflict resolution:
\[
\mathbf{F}_{\text{att}} = \mathbf{F}_{\text{att}}^p + \mathbf{F}_{\text{att}}^v
\]
\[
\mathbf{F}_{\text{att}}^p = k_p^{\text{att}} \left( \mathbf{p}_g(t) – \mathbf{p}(t) \right)
\]
\[
\mathbf{F}_{\text{att}}^v = k_v^{\text{att}} \left( V_0 – \mathbf{v}(t) \right)
\]
where \(k_p^{\text{att}}, k_v^{\text{att}}\) are gains, \(\mathbf{p}_g(t)\) is the goal waypoint, and \(V_0\) is the desired cruise speed.
2.3 Damping for Oscillation Suppression
To eliminate oscillations near the goal, a damping force is added outside the safety zone:
\[
\mathbf{F}_{\text{damp}} = -k_{\text{damp}} \mathbf{v}_u(t)
\]
Hence the modified attractive force becomes:
\[
\mathbf{F}_{\text{att}} =
\begin{cases}
\mathbf{F}_{\text{att}}^p + \mathbf{F}_{\text{att}}^v & \text{if } \|\mathbf{d}\| < r_{\text{safe}} \\
\mathbf{F}_{\text{att}}^p + \mathbf{F}_{\text{att}}^v – k_{\text{damp}} \mathbf{v}_u(t) & \text{otherwise}
\end{cases}
\]
2.4 Dynamic Constraints Integration
The total force is clamped to respect the drone’s maximum acceleration:
\[
\mathbf{F}_{\text{max}}(t) = m a_{\text{max}}
\]
\[
\mathbf{F}_{\text{total}}(t) =
\begin{cases}
\mathbf{F}_{\text{total}}(t) & \text{if } \|\mathbf{F}_{\text{total}}\| \leq F_{\text{max}} \\
\frac{F_{\text{max}}}{\|\mathbf{F}_{\text{total}}\|} \mathbf{F}_{\text{total}} & \text{otherwise}
\end{cases}
\]
Additionally, the commanded turn radius must satisfy \( r \geq r_{\text{min}} \). This is enforced by limiting lateral acceleration.
| Parameter | Symbol | Value |
|---|---|---|
| Safety distance | \(r_{\text{safe}}\) | 5942 m |
| Nominal speed | \(v_0\) | 55 m/s |
| Maximum speed | \(v_{\text{max}}\) | 70 m/s |
| Maximum acceleration | \(a_{\text{max}}\) | 50 m/s² |
| Minimum turn radius | \(r_{\text{min}}\) | 535 m |
| Position attractive gain | \(k_p^{\text{att}}\) | 0.1 |
| Velocity attractive gain | \(k_v^{\text{att}}\) | 0.1 |
| Position repulsive gain | \(k_p^{\text{rep}}\) | \(8 \times 10^{14}\) |
| Velocity repulsive gain | \(k_v^{\text{rep}}\) | \(4 \times 10^6\) |
| Damping coefficient | \(k_{\text{damp}}\) | 0.5 |
The above parameters were selected based on the HY100 large fixed-wing drone performance characteristics, similar to the Y-5B transport aircraft. These values ensure the algorithm generates physically feasible maneuvers for fixed-wing drones.
3. Simulation and Results
We conducted numerical simulations using MATLAB 2018 on an Intel i7-10750H processor with 16GB RAM. The scenario models a regional airport low-altitude environment where a fixed-wing drone starts from the origin and follows a nominal path to waypoint (7000,9000,2000) meters. A manned aircraft appears as a dynamic obstacle. The improved APF is compared with the traditional APF.
The simulation computation time is approximately 9.738 seconds. The resulting trajectories are shown in the following figure, illustrating the smooth avoidance maneuver produced by the enhanced method.
Figure: 3D visualization of conflict resolution for large fixed-wing drones using improved APF. The blue path shows the original trajectory; the red path represents the evasive maneuver. The improved method yields a smoother curve with fewer turns.
3.1 Separation Distance Analysis
The minimum distance between the fixed-wing drone and the manned aircraft is a critical safety metric. In our simulation, the minimum separation achieved by the improved APF is 5942 m, which exactly meets the safety requirement (greater than the regulatory 5000 m horizontal separation). The traditional APF sometimes violates this threshold under dynamic conditions.
| Method | Minimum Separation (m) | Safety Margin |
|---|---|---|
| Traditional APF | 5170 | Below 5942 (unsafe) |
| Improved APF | 5942 | Exactly meets requirement |
3.2 Velocity Profile
The velocity of the fixed-wing drone throughout the conflict is shown in Figure 7 of the original study. The improved method maintains speed within the 70 m/s limit and reaches the goal faster. The traditional method exhibits more oscillations and longer completion time.
| Metric | Traditional APF | Improved APF |
|---|---|---|
| Time to complete conflict resolution (s) | 415 | 339 |
| Maximum speed reached (m/s) | 68 | 65 |
| Number of oscillations near goal | 3 | 0 |
| Path smoothness rating | Poor | Excellent |
3.3 Flyability Under Fixed-Wind Drone Constraints
We verified that the generated path respects the minimum turn radius of 535 m. The commanded lateral acceleration never exceeds the load factor limit. This ensures that any large fixed-wing drone with similar performance can execute the maneuver without stalling or exceeding structural limits.
\[
r_{\text{actual}} = \frac{v^2}{a_{\text{lateral}}} \geq r_{\text{min}}
\]
Throughout the simulation, the minimum actual turn radius was 540 m, slightly above the 535 m threshold, confirming feasibility.
4. Discussion
The improved APF method presented here is specifically designed for large fixed-wing drones operating in congested low-altitude airspace. By explicitly incorporating velocity repulsion, we address the critical issue of moving obstacle avoidance that traditional APF handles poorly. The decomposition of attractive forces ensures that the fixed-wing drone not only reaches its goal but also stabilizes its speed, which is essential for subsequent climb or approach phases. The damping term effectively eliminates the oscillatory behavior that often plagues potential field methods and which would be unacceptable for the smooth flight required by fixed-wing drones.
Furthermore, the integration of kinematic constraints guarantees that the planned path is not just collision-free but also flyable. This is a major advantage over pure geometric methods that may produce sharp turns exceeding the drone’s capability. Our method can be extended to multi-drone scenarios and integrated with sense-and-avoid systems onboard fixed-wing drones.
5. Conclusion
This paper has developed a conflict resolution framework for large fixed-wing drones in regional airport environments. The enhanced artificial potential field algorithm overcomes the classical limitations of local minima, target unreachability, and path oscillations while respecting the dynamic constraints unique to fixed-wing drones. Simulation results demonstrate that the proposed method guarantees the required safety distance of 5942 m, reduces conflict resolution time by over 18% compared to traditional APF, and produces smooth, flyable trajectories. Future work will focus on multi-agent coordination and real-time implementation on embedded systems.
The key contributions of this study include:
- Decomposition of repulsive force into position and velocity components for dynamic obstacle avoidance.
- Decomposition of attractive force into position and velocity components for trajectory recovery.
- Introduction of a damping term specifically tuned for fixed-wing drones.
- Explicit enforcement of minimum turn radius, maximum speed, and acceleration constraints.
These advances pave the way for safe integration of large fixed-wing drones into civilian airspace, particularly for regional logistics and cargo operations.