Pith. sign in

REVIEW 3 major objections 5 minor 48 references

Stream Function-Based Navigation for Complex Quadcopter Obstacle Avoidance

T0 review · 3 major / 5 minor · reviewed 2026-08-06 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read A quadcopter can avoid complex, fast-moving obstacles in real time by treating them as rigid surfaces in an artificial fluid flow and steering along the resulting streamlines.

desk verdict A genuinely integrative reactive-avoidance framework with consistent math and honest experiments, but the close-proximity safety claim outruns the constant-acceleration model and the slack-relaxed HOCBF. read the letter →

arxiv 2507.06787 v2 pith:K6FFIMBP submitted 2025-07-09 cs.RO cs.SYeess.SY

Sean Smith , Emmanuel Witrant , Ya-Jun Pan This is my paper · ORCID
classification cs.ROcs.SYeess.SY
keywords quadcopternavigationvortexpanelmethodstreamfunctionmodelpredictivecontrolhigher-orderbarrieradaptiveKalmanfilterobstacleavoidanceminimumboundingellipse
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

This paper claims that a LiDAR-equipped quadcopter can navigate collision-free through unknown, partially observed environments by representing each detected obstacle as a two-dimensional rigid surface in an inviscid, incompressible flow, and then steering along streamlines of that flow. The vortex panel method supplies the velocity field, and a higher-order control barrier function baked into a model predictive controller takes over at close range to handle obstacles that accelerate sharply. An adaptive Kalman filter estimates the obstacle's position, velocity, acceleration, and elliptical shape, and propagates that estimate through the prediction horizon. If the claim holds, drones could avoid arbitrarily shaped, moving obstacles without a global map, using only onboard sensing and real-time computation.

What carries the argument

The load-bearing object is the stream function $\psi$ of a two-dimensional inviscid, incompressible flow, whose level sets are collision-free streamlines. Obstacle surfaces are discretized into panels carrying unknown vortex strengths $\gamma_j$; the boundary condition that each surface is a streamline, closed by a Kutta condition that fixes the surface stream function $\psi_s$, gives a linear system whose solution defines the velocity field at any point. The second mechanism is the HOCBF constraint $-\frac{\Delta p^T}{\|\Delta p\|}\Delta a \le \Upsilon$, with $b(\eta)=\|\Delta p\|-r$, which is imposed inside a receding-horizon MPC so that the control input keeps the relative position outside the inflated safety radius $r_{si}=r_s+\Lambda_0\sigma_p$ over the whole horizon. The third is the AKF with adaptive forgetting factor, which turns noisy LiDAR scans and minimum-bounding-ellipse fits into predicted obstacle states $x_O=[p_O,v_O,a_O]$ used by the MPC.

What would settle it

Run the complex-dynamic or 3D dynamic Gazebo scenarios with an obstacle that executes a sharp lateral step or an abrupt turn whose acceleration is not constant; a collision-free rate below the reported 90%, or a violation of the barrier function $b(\eta)\ge 0$ before any collision, would show that the constant-acceleration assumption and the fixed inflation factor do not carry the safety claim.

Watch

Extended reading notes

Core claim

The central claim is that the VPM-MPC-HOCBF-AKF pipeline is the first application of the panel method to drones that works with real-time onboard sensing in a moving environment, and that this pipeline lets a quadcopter avoid rapidly accelerating obstacles in close proximity. In the reported evaluation, the full system achieves a 100% collision-free rate over ten runs in static and simple dynamic simulation cases, 90% in complex dynamic and 3D dynamic simulation cases, and 100% in static hardware experiments; the ablation-style baselines that drop the HOCBF or the adaptive filter collide far more often. The mechanism is that the VPM generates a smooth reference velocity field that is free of local-minima traps, while the MPC recedes over a five-second horizon and enforces the higher-order barrier constraint $b(\eta)=\|\Delta p\|-r\ge 0$, so the acceleration command, not just the position or velocity, reacts to the predicted relative motion.

Load-bearing premise

The whole safety argument depends on the assumption that an obstacle moving on a curved or sideways path can be treated as moving with constant acceleration over the next five seconds, with a fixed safety-inflation factor covering any error.

Editorial extensions

If this is right

  • A drone using this pipeline can handle concave and convex obstacles without a global map, and the VPM velocity field avoids the local-minima entrapment that sinks the artificial potential field baseline.
  • The MPC-HOCBF layer extends the 2D panel method to 3D by making vertical acceleration adjustments, so 3D avoidance is obtained without the computational cost of full 3D panel methods.
  • The system maintains collision-free operation against obstacles reaching roughly 7-8 m/s with 13-20 m/s^2 acceleration in simulation, at a 90% success rate over ten runs.
  • Replacing the HOCBF with a plain Euclidean-norm MPC constraint drops the success rate to 40% in the complex dynamic case and 30% in the 3D dynamic case, showing that the higher-order constraint and the AKF prediction carry the fast-obstacle scenarios.
  • Static hardware flights on a COEX Clover with 360-degree LiDAR confirm the approach works outside simulation, with the VPM computed offboard because the onboard Raspberry Pi lacks the compute.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • An inference the authors leave implicit is that the fixed inflation factor $\Lambda_0=2$ is the weakest tuning knob: it grows the safety radius but cannot bend the HOCBF constraint in the direction of a lateral model error, so adaptive or distributionally robust uncertainty scaling is the natural next step.
  • A second inference is that the reported 90% success rate in the hardest dynamic cases makes a hardware trial with accelerating obstacles the decisive test the paper does not run; the static hardware results alone do not yet confirm the dynamic claim on real hardware.
  • A third inference is that offloading the panel-method computation to an FPGA or GPU, which the authors suggest in the discussion, would let the VPM update fast enough to run onboard and would remove the ground-station dependency seen in the experiments.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

3 major / 5 minor

Summary. The manuscript proposes a layered stream-function navigation architecture for a quadcopter. A vortex panel method (VPM) builds a 2D potential-flow velocity field from LiDAR-detected obstacle contours, with a Kutta-condition extension (VPM-A) and bounds on the surface stream function that aim to guarantee convergence to the sink (VPM-B). Minimum bounding ellipses (MBEs) parameterize detected obstacles, an adaptive Kalman filter (AKF) estimates obstacle position, velocity, and acceleration, and a model predictive controller with higher-order control barrier function (HOCBF) constraints tracks the VPM reference while enforcing a distance-based safety condition. The full VPM-MPC-HOCBF-AKF system is evaluated in PX4/Gazebo simulation against APF, MPC, and MPC-AKF baselines and in static indoor hardware tests with a COEX Clover quadcopter. The paper claims the first real-time onboard panel-method navigation for drones and reports collision-free rates of 100% for simple dynamic simulation and 90% for complex dynamic and 3D dynamic simulation, with 100% for static hardware tests.

Significance. If the central claims hold, the work is significant because it integrates potential-flow path generation, receding-horizon control, and online obstacle-state estimation under partial sensing into one framework. The algebraic developments that are present, in particular the influence-coefficient system (8)-(10), the VPM-B stream-function bounds (13), and the HOCBF derivative sequence (27)-(30), are internally consistent, and the comparison against external baselines together with the static hardware validation is a genuine strength. The contribution is weakened, however, by the gap between the close-proximity safety claim and the uncertainty model: the constant-acceleration obstacle propagation (35) and the scalar radius inflation (47) do not address directional acceleration-estimation error, and the slack variable in (34) means the HOCBF forward-invariance theorem is not actually enforced. The 90% success rates over only ten runs, with no failure analysis, are consistent with a tuned safety margin rather than with a robustness certificate.

major comments (3)
  1. [V-D, Eq. (47), and VI-C] The constant-acceleration propagation in (35) combined with the isotropic safety-radius inflation r_si = r_s + Lambda0*sigma_p in (47) does not bound the dominant model-mismatch term. The HOCBF constraint (30) depends on the obstacle acceleration estimate only through the directional projection (Delta p^T/||Delta p||) Delta a; if the true acceleration differs from the AKF estimate by delta a, the constraint error is (Delta p^T/||Delta p||) delta a, which is unaffected by increasing r_si. The 3D dynamic tests in Section VI-C use obstacle accelerations of 13.6-19.8 m/s^2 along torus and Lissajous trajectories over a 5 s horizon, so the constant-acceleration assumption is violated by the test design itself. The 90% success rate in Table III, with no analysis of the failing runs, is consistent with a tuned margin rather than a safety guarantee. To support the central close-proximity claim, report the failure cases, quantify the worst-case constraint violation, and either incorporate a directional uncertainty set into (30) or provide a robustified version of the HOCBF constraint.
  2. [IV-E, Eq. (34)] The slack variable sigma(s) in (34) relaxes the HOCBF inequality, so the conditions of Theorem 1 in [11] are not satisfied and forward invariance of C1 n C2 is not guaranteed for the closed-loop MPC. The paper itself shows the consequences in Figs. 8 and 10, where Gamma(eta) and b(eta) are only mostly preserved, and Section VI-C reports intervals with b <= r. The language in Sections I and IV-C that the framework ensures avoidance or makes the system safe is therefore stronger than the mathematics supports. Please either soften the guarantee language to an empirical claim or replace the slack with a robust or terminal constraint that restores the invariance argument.
  3. [V-B, Eq. (37)] The AKF transition matrix in (37) is typeset in a way that does not allow the reader to verify the model. The state vector x = [p_i_O, v_i_O, a_i_O, Sigma]^T in R^9 and the measurement z = [p_i_O, Sigma]^T in R^5 are consistent if p_i_O, v_i_O, and a_i_O are two-dimensional, but the displayed A(k) matrix mixes incomplete rows with block entries such as O6x3 and an ellipsis, and it cannot be checked against the linear trajectory model (35). Since the AKF output enters the HOCBF constraint (30), please provide the exact 9x9 transition matrix and the corresponding 5x9 measurement matrix in a clean, unambiguous form.
minor comments (5)
  1. [Table III] The VPM*-MPC-HOCBF-AKF entry in the Static column reads [0*,100]%; the asterisk and the two-valued bracket are not explained in the text or the table caption, and the corresponding discussion in Section VI-A2 does not resolve the notation.
  2. [VI-C, text near Fig. 10] The sentence that begins 'From Fig. 9(b), Sphere O1 approaches from the left' appears to refer to Fig. 10, not Fig. 9, since Fig. 9 shows trajectories and Fig. 10 shows the b(eta) and Gamma(eta) profiles.
  3. [VI-B, trajectory name] The trajectory name is misspelled as 'Lemniscate of Bernouli' in Section VI-B; it should be 'Bernoulli'.
  4. [Figs. 8 and 10] The captions and text refer to a function Gamma(eta), but the formulation in (27) defines both Gamma_1 and Gamma_2; please specify which function is plotted.
  5. [VII-C] The claim that the proposed system runs in real time should be reconciled with the statement in Section VII-C that VPM computations on the Raspberry Pi 4B are performed offboard on a ground station computer rather than on the drone; the 'onboard sensing' claim in contribution 2 refers to sensing, but the offboard computation is a relevant qualification for the real-time claim.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the VPM, MPC-HOCBF, and AKF derivations are externally sourced or self-contained, and the safety outcomes are measured rather than recovered from fitted parameters.

full rationale

The paper's derivation chain is not circular. The VPM part is built from standard potential-flow equations (1)-(10) and external references: the Kutta-condition closure is attributed to [12] and the VPM-B convergence argument to [13], both by non-overlapping authors. The MPC-HOCBF part imports the HOCBF framework from [11] (Definition 2, Theorem 1) and derives the collision-avoidance inequality (30) directly from the relative-degree-two translational dynamics (20)-(21); no safety set, barrier function, or control law is defined in terms of the later simulation or experimental outcomes. The AKF is a standard Kalman filter with adaptive noise covariance updates taken from [47]. Table III reports measured collision-free frequencies over ten runs, not quantities recovered from the tuned parameters; the safety-radius inflation in Eq. (47) is a design input rather than a relabeled output. The only author-overlapping citation is [48], used in the hardware discussion as a pointer to the authors' prior work on robust position control; it is not load-bearing for the central avoidance claim and does not exclude alternative approaches. The use of a slack variable in constraint (34) and the 'mostly preserved' forward-invariance profiles in Figs. 8 and 10 are honest empirical limitations, not instances of circularity: the paper explicitly reports 90% rather than 100% success in the hardest dynamic cases. Therefore no enumerated circularity pattern is exhibited.

Assumptions & free parameters 7 free parameters · 6 assumptions · 1 invented entities

The central claim rests on premises the paper does not fully pay for. The flow-model premise turns real 3D aerodynamics and sensing into a 2D inviscid flow, the standard cost of stream-function methods. The convergence premise (inequality (11)) comes from Fahimi's textbook [13], not a proof in this paper, and is invoked for a multi-obstacle, partially observed, quasi-steady setting beyond the cited one. The safety premise is the largest: the HOCBF theorem [11] guarantees forward invariance only without slack, while the MPC uses a slack variable to preserve feasibility, so the guarantee is 'mostly preserved' rather than assured. The obstacle-motion premise is a constant-acceleration model whose mismatch is absorbed by a manually scaled covariance inflation. Seven hand-tuned parameters (xi, beta1, beta2, Lambda0, mu, kappa, l_kutta) plus three AKF constants (alpha_min, alpha_max, rho) shape the reported success rates. No new physical entities are postulated; the virtual surface of Algorithm 1 is an algorithmic construct evaluated only through the flows it produces.

free parameters (7)
  • xi (stream function control parameter) = ±0.3, ±0.4, ±0.5; sign chosen by obstacle side
    User-defined parameter in Eq (14); the sign and magnitude are selected per scenario to route the drone left or right around an obstacle (Section III-C2), and the hardware success rate varies from 100% (xi=±0.3) to 80% (xi=±0.5).
  • beta1, beta2 (HOCBF gains) = 4,4 (2D sim); 1,2 (3D sim and hardware)
    Tuned manually per Section IV-D; they determine when the safety constraint activates, and the paper states that optimal selection is out of scope.
  • Lambda0 (uncertainty scaling factor) = 2
    Scales the Mahalanobis uncertainty term in the safety radius (47); justification is only that it 'can be set based on Gaussian distribution uncertainty' (Section V-D).
  • mu, kappa, l_kutta (VPM-A surface transformation parameters) = mu=0.3, kappa=0°/10°, l_kutta=0.8 m (sim) / 0.15 m (hw)
    Algorithm 1 parameters that place the virtual surface offset and the Kutta point; chosen per platform with no sensitivity analysis.
  • alpha_min, alpha_max, rho (AKF adaptive forgetting factor) = 0.7, 1, 1.5 (sim); 0.7, 1, 1.2 (hw)
    Eq (43) tuning constants that control how quickly the adaptive covariance responds to MBE shape fluctuations.
  • Sink/source strengths and uniform flow magnitude (|zeta_g|, zeta_w, Q_infinity) = not stated numerically
    The sink strength |zeta_g| sets the circulation budget in inequality (11) and the convergence bounds in (13), while Q_infinity sets flow speed; none of these values are reported, leaving the main convergence parameter unspecified.
  • Base safety margin rs and body radius ra = not stated numerically
    Fig. 3 defines r = ra + rs and Eq (26) uses it as the HOCBF safety radius; the absolute values are never given, though the Table II closest approaches (around 0.77 m against obstacles of 1.5 m radius) suggest the scalar radius is small relative to the obstacle size.
assumptions (6)
  • domain assumption Global convergence guarantee of VPM-B via inequality (11), |zeta_g| > zeta_s > -|zeta_g|
    Invoked in Section III-C2 with the global-minimum reasoning from textbook [13] and the saddle-point property cited from [25]; not proven here for multiple obstacles, partial 2D LiDAR views, and the vortical, non-harmonic flow produced by vortex panels.
  • domain assumption Obstacle surfaces behave as 2D rigid surfaces in an inviscid, incompressible flow, so drone trajectories are streamlines of a potential flow
    Section III-A/B sets up the navigation-as-flow model; real 3D aerodynamics, wake effects, and sensor noise are only handled through simulator noise and controller robustness rather than through the model itself.
  • domain assumption Quadcopter translational dynamics are well approximated by the double integrator p_dot = v, v_dot = u
    Section IV-A/C replace the 6-DOF dynamics (19) with (20)-(21); the HOCBF forward-invariance theorem applies to this surrogate, while the hardware wall-effect disturbances noted in VII-A1 show the real vehicle deviates from it.
  • domain assumption The constant-acceleration obstacle model (35), open-loop propagated in (45), adequately predicts obstacle motion over the horizon, with covariance inflation (47) covering the model error
    Sections V-A and V-D: the HOCBF constraint (30) is only correct if the predicted obstacle state is near the true state; Lemniscate, torus, and Lissajous trajectories are not constant-acceleration, so safety depends on the ad hoc Lambda0*sigma_p inflation.
  • ad hoc to paper The Kutta condition, applied at the trailing edge of a detected surface, acts as a safety mechanism guiding the drone around the obstacle
    Section III-C1 re-purposes an aerodynamic smoothness condition as a collision-avoidance heuristic; its failure modes are visible in the VPM-A dynamic results, which achieve only 70% (Case A) and 30% (Case B) collision-free rates (Table III).
  • ad hoc to paper A scalar safety distance r = ra + rs adequately describes a safety set for obstacles modeled as MBE ellipses
    Figure 3 and Eq (26) use a circular safety set while the MBE state (ra, rb, theta) enters only the AKF; the mapping from the oriented ellipse to the scalar radius r is never specified, so the shape-aware parameterization does not directly shape the safety constraint.
invented entities (1)
  • Virtual rigid surface Z_tilde_k^T with offset and Kutta point Z_kutta
    purpose: A shifted replica of the detected obstacle surface used to make the flow, and therefore the drone's path, divert before reaching the physical obstacle
    Algorithmic construct from Algorithm 1; its safety benefit is assessed through trajectories generated by the same flow it defines, so the evidence is internal to the framework and no external falsifiable handle is offered.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Stream Function-Based Navigation for Complex Quadcopter Obstacle Avoidance." pith.science (2026). https://pith.science/paper/K6FFIMBP

@misc{pith2026250706787,
  author       = {Pith},
  title        = {Pith review of: Stream Function-Based Navigation for Complex Quadcopter Obstacle Avoidance},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/K6FFIMBP}},
  note         = {Machine review of arXiv:2507.06787}
}
read the original abstract

This article presents a novel stream function-based navigational control system for obstacle avoidance, where obstacles are represented as two-dimensional (2D) rigid surfaces in inviscid, incompressible flows. The approach leverages the vortex panel method (VPM) and incorporates safety margins to control the stream function and flow properties around virtual surfaces, enabling navigation in complex, partially observed environments using real-time sensing. To address the limitations of the VPM in managing relative distance and avoiding rapidly accelerating obstacles at close proximity, the system integrates a model predictive controller (MPC) based on higher-order control barrier functions (HOCBF). This integration incorporates VPM trajectory generation, state estimation, and constraint handling into a receding-horizon optimization problem. The 2D rigid surfaces are enclosed using minimum bounding ellipses (MBEs), while an adaptive Kalman filter (AKF) captures and predicts obstacle dynamics, propagating these estimates into the MPC-HOCBF for rapid avoidance maneuvers. Evaluation is conducted using a PX4-powered Clover drone Gazebo simulator and real-time experiments involving a COEX Clover quadcopter equipped with a 360 degree LiDAR sensor.

Figures

Figures reproduced from arXiv: 2507.06787 by the authors.

Figure 1
Figure 1. Structural overview of the vortex panel method (VPM) in a 2D [PITH_FULL_IMAGE:figures/full_fig_p004_1.png] view at source ↗
Figure 2
Figure 2. Control architecture for the proposed VPM-MPC-HOCBF-AKF [PITH_FULL_IMAGE:figures/full_fig_p005_2.png] view at source ↗
Figure 3
Figure 3. Geometric representation of a detected obstacle. The radial distance [PITH_FULL_IMAGE:figures/full_fig_p007_3.png] view at source ↗
Figures from the paper (8 more)
Figure 4
Figure 4. Figure 4: Clover Gazebo simulation showing global trajectory profiles generated [PITH_FULL_IMAGE:figures/full_fig_p009_4.png]
Figure 6
Figure 6. Figure 6: Clover Gazebo simulation showing global trajectory profiles generated [PITH_FULL_IMAGE:figures/full_fig_p010_6.png]
Figure 7
Figure 7. Figure 7: Clover Gazebo simulation illustrating global trajectory profiles and [PITH_FULL_IMAGE:figures/full_fig_p011_7.png]
Figure 9
Figure 9. Figure 9: Clover Gazebo simulation showing global 3D trajectory profiles and [PITH_FULL_IMAGE:figures/full_fig_p012_9.png]
Figure 10
Figure 10. Figure 10: Clover Gazebo simulation showing evolution profiles of functions [PITH_FULL_IMAGE:figures/full_fig_p012_10.png]
Figure 11
Figure 11. Figure 11: COEX Clover 4.2 platform: 1. Raspberry Pi 4 Model B, 2. Laser [PITH_FULL_IMAGE:figures/full_fig_p013_11.png]
Figure 13
Figure 13. Figure 13: (a) Hardware experiment demonstrating the global trajectory and [PITH_FULL_IMAGE:figures/full_fig_p013_13.png]
Figure 14
Figure 14. Figure 14: (a) Hardware experiment demonstrating the global 3D trajectory [PITH_FULL_IMAGE:figures/full_fig_p014_14.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

48 extracted references · 45 canonical work pages

  1. [11]

    High-order control barrier functions,

    W. Xiao and C. Belta, “High-order control barrier functions,” IEEE Transactions on Automatic Control, vol. 67, no. 7, pp. 3655–3662, 2022

    work page 2022

  2. [1]

    A guidance and maneu- vering control system design with anti-collision using stream functions with vortex flows for autonomous marine vessels,

    H. Zhou, Z. Ren, M. Marley, and R. Skjetne, “A guidance and maneu- vering control system design with anti-collision using stream functions with vortex flows for autonomous marine vessels,” IEEE Transactions on Control Systems Technology , vol. 30, no. 6, pp. 2630–2645, 2022

    work page 2022

  3. [2]

    Autonomous robot path plan- ning based on swarm intelligence and stream functions,

    C. Hu, X. Wu, Q. Liang, and Y . Wang, “Autonomous robot path plan- ning based on swarm intelligence and stream functions,” in Evolvable Systems: From Biology to Hardware. Berlin, Germany: Springer, 2007, pp. 277–284

    work page 2007

  4. [3]

    Harmonic potential field path planning for high speed vehicles,

    R. Daily and D. M. Bevly, “Harmonic potential field path planning for high speed vehicles,” in American Control Conference . IEEE, 2008, pp. 4609–4614

    work page 2008

  5. [4]

    Using stream functions for complex behavior and path generation,

    J. Sullivan, S. Waydo, and M. Campbell, “Using stream functions for complex behavior and path generation,” in AIAA Guidance, Navigation, and Control Conference and Exhibit , 2003, p. 5800

    work page 2003

  6. [5]

    Implementa- tion of quadrotor path planning using fluid flow equations,

    A.-R. Merheb, H. Noura, M. A. Mannah, and A. Haddad, “Implementa- tion of quadrotor path planning using fluid flow equations,” in IEEE 3rd International Multidisciplinary Conference on Engineering Technology (IMCET), 2021, pp. 90–95

    work page 2021

  7. [6]

    Implementation studies of robot swarm navigation using potential functions and panel methods,

    A.-R. Merheb, V . Gazi, and N. Sezer-Uzol, “Implementation studies of robot swarm navigation using potential functions and panel methods,” IEEE/ASME Transactions on Mechatronics , vol. 21, no. 5, pp. 2556– 2567, 2016

    work page 2016

  8. [7]

    Implementation of robot formation control and navigation using real-time panel method,

    A.-R. Merheb, Y . Atas, V . Gazi, and N. Sezer-Uzol, “Implementation of robot formation control and navigation using real-time panel method,” in IEEE/RSJ International Conference on Intelligent Robots and Systems , 2010, pp. 5001–5006

    work page 2010

Show all 48 references
  1. [8]

    Mobile robot motion framework based on enhanced robust panel method,

    J. Velagi ´c, L. Vukovi ´c, and B. Ibrahimovi ´c, “Mobile robot motion framework based on enhanced robust panel method,” International Journal of Control, Automation and Systems , vol. 18, no. 5, pp. 1264– 1276, 2020

    2020

  2. [9]

    Panel method based guidance for fixed wing micro aerial vehicles,

    Z. Bilgin, M. Bronz, and I. Yavrucuk, “Panel method based guidance for fixed wing micro aerial vehicles,” in International Micro Air Vehicle Conference, 2022

    2022

  3. [10]

    Two-dimensional sails in inviscid flow,

    P. Jackson, “Two-dimensional sails in inviscid flow,” Journal of ship research, vol. 28, no. 1, pp. 11–17, 1984

    1984

  4. [12]

    A potential flow design method for multicomponent airfoil sections,

    J. L. Kennedy and D. Marsden, “A potential flow design method for multicomponent airfoil sections,” Journal of Aircraft, vol. 15, no. 1, pp. 47–52, 1978

    1978

  5. [13]

    Fahimi, Autonomous Robots

    F. Fahimi, Autonomous Robots. Modeling, Path Planning, and Control . Springer, 2009

    2009

  6. [14]

    Real-time obstacle avoidance for manipulators and mobile robots,

    O. Khatib, “Real-time obstacle avoidance for manipulators and mobile robots,” The international journal of robotics research, vol. 5, no. 1, pp. 90–98, 1986. 15

    1986

  7. [15]

    Collaborative target tracking for swarming mavs using potential fields and panel methods,

    O. Uzol, I. Yavrucuk, and N. Sezer Uzol, “Collaborative target tracking for swarming mavs using potential fields and panel methods,” in AIAA Guidance, Navigation and Control Conference and Exhibit , Honolulu, Hawaii, USA, 2008

    2008

  8. [16]

    Urban air mobility guidance with panel method: Experimental evaluation under wind disturbances,

    Z. Bilgin, I. Yavrucuk, and M. Bronz, “Urban air mobility guidance with panel method: Experimental evaluation under wind disturbances,” Journal of Guidance, Control, and Dynamics , vol. 47, no. 6, pp. 1080– 1096, 2024

    2024

  9. [17]

    Modified robust panel method for mobile robot path planning in partially unknown static and dynamic environments,

    B. Ibrahimovi ´c and J. Velagi ´c, “Modified robust panel method for mobile robot path planning in partially unknown static and dynamic environments,” in 3rd IEEE Conference on Control and Fault-Tolerant Systems (SysTol), 2016, pp. 51–58

    2016

  10. [18]

    A 3-d potential panel method for robot motion planning,

    Y . Zhang and K. P. Valavanis, “A 3-d potential panel method for robot motion planning,” Robotica, vol. 15, pp. 421 – 434, 1997

    1997

  11. [19]

    Dynamic control barrier function-based model predictive control to safety-critical obstacle-avoidance of mobile robot,

    Z. Jian, Z. Yan, X. Lei, Z. Lu, B. Lan, X. Wang, and B. Liang, “Dynamic control barrier function-based model predictive control to safety-critical obstacle-avoidance of mobile robot,” in IEEE International Conference on Robotics and Automation (ICRA) , 2023, pp. 3679–3685

    2023

  12. [20]

    Safe control for navigation in cluttered space using multiple lyapunov-based control barrier functions,

    I. Jang and H. J. Kim, “Safe control for navigation in cluttered space using multiple lyapunov-based control barrier functions,” IEEE Robotics and Automation Letters , 2024

    2024

  13. [21]

    Safe control with learned certificates: A survey of neural lyapunov, barrier, and contraction methods for robotics and control,

    C. Dawson, S. Gao, and C. Fan, “Safe control with learned certificates: A survey of neural lyapunov, barrier, and contraction methods for robotics and control,” IEEE Transactions on Robotics , vol. 39, no. 3, pp. 1749– 1767, 2023

    2023

  14. [22]

    Obstacle avoidance strategy using onboard stereo vision on a flapping wing mav,

    S. Tijmons, G. De Croon, B. Remes, C. De Wagter, and M. Mulder, “Obstacle avoidance strategy using onboard stereo vision on a flapping wing mav,” IEEE Transactions on Robotics, vol. 33, no. 4, pp. 858–874, 2017

    2017

  15. [23]

    Robust collision avoidance for mobile robots in the presence of moving obstacles,

    J. A. Ricardo and D. A. Santos, “Robust collision avoidance for mobile robots in the presence of moving obstacles,” IEEE Control Systems Letters, vol. 7, pp. 1584–1589, 2023

    2023

  16. [24]

    Fluid motion planner for nonholo- nomic 3-d mobile robots with kinematic constraints,

    D. Lau, J. Eden, and D. Oetomo, “Fluid motion planner for nonholo- nomic 3-d mobile robots with kinematic constraints,” IEEE Transactions on Robotics, vol. 31, no. 6, pp. 1537–1547, 2015

    2015

  17. [25]

    Real-time obstacle avoidance using harmonic potential functions,

    J.-O. Kim and P. Khosla, “Real-time obstacle avoidance using harmonic potential functions,” 1992

    1992

  18. [26]

    Sensor-based 2-d potential panel method for robot motion planning,

    Y . Zhang and K. P. Valavanis, “Sensor-based 2-d potential panel method for robot motion planning,” Robotica, vol. 14, no. 1, p. 81–89, 1996

    1996

  19. [27]

    Collision avoidance of arbitrarily shaped deforming objects using collision cones,

    V . Sunkara, A. Chakravarthy, and D. Ghose, “Collision avoidance of arbitrarily shaped deforming objects using collision cones,” IEEE Robotics and Automation Letters , vol. 4, no. 2, pp. 2156–2163, 2019

    2019

  20. [28]

    Real-time dynamic obstacle avoidance for robot manipulators based on cascaded nonlinear mpc with artificial potential field,

    T. Zhu, J. Mao, L. Han, C. Zhang, and J. Yang, “Real-time dynamic obstacle avoidance for robot manipulators based on cascaded nonlinear mpc with artificial potential field,” IEEE Transactions on Industrial Electronics, pp. 1–11, 2023

    2023

  21. [29]

    Nonlinear mpc for collision avoidance and control of uavs with dynamic obstacles,

    B. Lindqvist, S. S. Mansouri, A.-a. Agha-mohammadi, and G. Niko- lakopoulos, “Nonlinear mpc for collision avoidance and control of uavs with dynamic obstacles,” IEEE Robotics and Automation Letters , vol. 5, no. 4, pp. 6001–6008, 2020

    2020

  22. [30]

    Model predictive control for aerial col- lision avoidance in dynamic environments,

    M. Castillo-Lopez, S. A. Sajadi-Alamdari, J. L. Sanchez-Lopez, M. A. Olivares-Mendez, and H. V oos, “Model predictive control for aerial col- lision avoidance in dynamic environments,” in26th IEEE Mediterranean Conference on Control and Automation (MED) , 2018, pp. 1–6

    2018

  23. [31]

    Nonlinear model predictive control-based guidance algorithm for quadrotor trajectory tracking with obstacle avoidance,

    C. Zhao, D. Wang, J. Hu, and Q. Pan, “Nonlinear model predictive control-based guidance algorithm for quadrotor trajectory tracking with obstacle avoidance,”Journal of Systems Science and Complexity, vol. 34, no. 4, pp. 1379–1400, 2021

    2021

  24. [32]

    Multitrajectory model pre- dictive control for safe uav navigation in an unknown environment,

    D. Saccani, L. Cecchin, and L. Fagiano, “Multitrajectory model pre- dictive control for safe uav navigation in an unknown environment,” IEEE Transactions on Control Systems Technology , vol. 31, no. 5, pp. 1982–1997, 2023

    1982

  25. [33]

    A perception-aware nmpc for vision-based target tracking and collision avoidance with a multi-rotor uav,

    A. Dmytruk, G. Silano, D. Bicego, D. B. Licea, and M. Saska, “A perception-aware nmpc for vision-based target tracking and collision avoidance with a multi-rotor uav,” in IEEE International Conference on Unmanned Aircraft Systems (ICUAS) , 2022, pp. 1668–1673

    2022

  26. [34]

    Dual model predictive control of multiple quadrotors with formation maintenance and collision avoidance,

    S. Wang, Y . Wang, Z. Miao, X. Wang, and W. He, “Dual model predictive control of multiple quadrotors with formation maintenance and collision avoidance,” IEEE Transactions on Industrial Electronics , vol. 71, no. 12, pp. 16 037–16 046, 2024

    2024

  27. [35]

    Decentralized nonlinear model predictive control for safe collision avoidance in quadrotor teams with limited detection range,

    M. Goarin, G. Li, A. Saviolo, and G. Loianno, “Decentralized nonlinear model predictive control for safe collision avoidance in quadrotor teams with limited detection range,” arXiv preprint arXiv:2409.17379 , 2024

    2024 arXiv

  28. [36]

    Robust model predictive control with control barrier functions for autonomous surface vessels,

    W. Wang, W. Xiao, A. Gonzalez-Garcia, J. Swevers, C. Ratti, and D. Rus, “Robust model predictive control with control barrier functions for autonomous surface vessels,” in IEEE International Conference on Robotics and Automation (ICRA) , 2024, pp. 6089–6095

    2024

  29. [37]

    Robust vision-based obstacle avoidance for micro aerial vehicles in dynamic environments,

    J. Lin, H. Zhu, and J. Alonso-Mora, “Robust vision-based obstacle avoidance for micro aerial vehicles in dynamic environments,” in IEEE International Conference on Robotics and Automation (ICRA), 2020, pp. 2682–2688

    2020

  30. [38]

    Model predictive contouring control for collision avoidance in unstructured dynamic environments,

    B. Brito, B. Floor, L. Ferranti, and J. Alonso-Mora, “Model predictive contouring control for collision avoidance in unstructured dynamic environments,” IEEE Robotics and Automation Letters , vol. 4, no. 4, pp. 4459–4466, 2019

    2019

  31. [39]

    Dynamic obstacle avoidance for uavs using mpc and gp-based motion forecast,

    E. Olcay, H. Meeß, and G. Elger, “Dynamic obstacle avoidance for uavs using mpc and gp-based motion forecast,” in IEEE European Control Conference (ECC), 2024, pp. 1024–1031

    2024

  32. [40]

    Flexible active safety motion control for robotic obstacle avoidance: A cbf-guided mpc approach,

    J. Liu, J. Yang, J. Mao, T. Zhu, Q. Xie, Y . Li, X. Wang, and S. Li, “Flexible active safety motion control for robotic obstacle avoidance: A cbf-guided mpc approach,” IEEE Robotics and Automation Letters , 2025

    2025

  33. [41]

    High-order control barrier function based robust collision avoidance formation tracking of constrained multi-agent sys- tems,

    D. Liu and J. Fu, “High-order control barrier function based robust collision avoidance formation tracking of constrained multi-agent sys- tems,” in International Conference on Neural Information Processing . Springer, 2023, pp. 253–264

    2023

  34. [42]

    Control barrier function-based collision avoidance guidance strategy for multi-fixed-wing uav pursuit-evasion environment,

    X. Lv, C. Peng, and J. Ma, “Control barrier function-based collision avoidance guidance strategy for multi-fixed-wing uav pursuit-evasion environment,” Drones, vol. 8, no. 8, p. 415, 2024

    2024

  35. [43]

    Information fusion estimation-based path following control of quadrotor uavs subjected to gaussian random disturbance,

    Q. Xu, Z. Wang, and Z. Zhen, “Information fusion estimation-based path following control of quadrotor uavs subjected to gaussian random disturbance,” ISA transactions, vol. 99, pp. 84–94, 2020

    2020

  36. [44]

    Barriernet: Differentiable control barrier functions for learning of safe robot control,

    W. Xiao, T.-H. Wang, R. Hasani, M. Chahine, A. Amini, X. Li, and D. Rus, “Barriernet: Differentiable control barrier functions for learning of safe robot control,” IEEE Transactions on Robotics , 2023

    2023

  37. [45]

    Acado toolkit—an open-source framework for automatic control and dynamic optimization,

    B. Houska, H. J. Ferreau, and M. Diehl, “Acado toolkit—an open-source framework for automatic control and dynamic optimization,” Optimal Control Applications and Methods , vol. 32, no. 3, pp. 298–312, 2011

    2011

  38. [46]

    Smallest enclosing disks (balls and ellipsoids),

    E. Welzl, “Smallest enclosing disks (balls and ellipsoids),” in New Results and New Trends in Computer Science: Graz, Austria, June 20– 21, 1991 Proceedings . Springer, 2005, pp. 359–370

    1991

  39. [47]

    Adaptive adjustment of noise covariance in kalman filter for dynamic state estimation,

    S. Akhlaghi, N. Zhou, and Z. Huang, “Adaptive adjustment of noise covariance in kalman filter for dynamic state estimation,” in 2017 IEEE Power & Energy Society General Meeting , 2017, pp. 1–5

    2017

  40. [48]

    Adaptive observer-based super-twisting slid- ing mode control for low altitude quadcopter grasping,

    S. Smith and Y .-J. Pan, “Adaptive observer-based super-twisting slid- ing mode control for low altitude quadcopter grasping,” IEEE/ASME Transactions on Mechatronics, vol. 30, no. 1, pp. 587–598, 2025

    2025

Pith tools

Reviewed August 6, 2026 · model on record in the stance chip above.