Pith. sign in

REVIEW 4 major objections 5 minor 46 references

Polar Coordinate-based Differential Evolution for Moving Target Search Using Vision Sensor on Unmanned Aerial Vehicles

T0 review · 4 major / 5 minor · reviewed 2026-08-01 · deepseek-v4-flash

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

Pith's one-line read The paper proposes encoding UAV search paths in polar coordinates — step length and yaw angle — in differential evolution, and claims this raises cumulative detection probability, cuts execution time, and keeps paths flyable.

desk verdict The polar-coordinate DE encoding is a useful, practical trick, but the paper's headline claim about detection probability is undercut by an open-loop belief that is never conditioned on no-detection observations. read the letter →

arxiv 2607.17771 v1 pith:6UABQNAB submitted 2026-07-20 cs.RO

Thu Hang Khuat , Duy-Nam Bui , Thuy Ngan Duong , Manh Duong Phung This is my paper
classification cs.RO
keywords UAVtargetsearchpolarcoordinatesdifferentialevolutionBayesianmovingNVESDdetectionmodelpathplanningcumulativeprobability
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

The paper tries to establish that encoding a UAV's search path as a sequence of polar coordinates — a step length and a yaw angle per segment — rather than as Cartesian waypoint coordinates lets a differential evolution (DE) planner maximize the cumulative probability of detecting a moving target more effectively and more quickly than standard metaheuristics. The authors derive a Bayesian fitness function J that combines a Gaussian belief map of the target's location, a deterministic Markov transition model for its motion, the geometry of the camera footprint, and the NVESD model of human target detection from imagery. The proposed polar DE (PDE) applies UAV kinematic limits as simple bounds on the step length and yaw angle, narrowing the search to flyable paths and avoiding repeated trigonometric conversions during fitness evaluation. Across six simulated scenarios, PDE reports the highest detection probability in five and the shortest execution time in five compared with standard DE, genetic algorithm, particle swarm optimization, grey wolf optimization, and gaining-sharing knowledge; a real quadcopter flight confirms the planned path can be followed. If this holds, the practical consequence is that a modest change in search representation can improve both hit probability and runtime for vision-based UAV search and rescue.

What carries the argument

The central mechanism is the polar-coordinate path encoding and its direct mapping to UAV kinematics. A path is stored as step lengths bounded by speed limits and yaw angles bounded by turn limits, then converted to Cartesian via the recurrence above. This has two effects: kinematic and physical constraints are enforced by clamping scalars rather than by solving constraint equations, and the camera footprint calculation receives the yaw angle directly, eliminating atan2 and square-root conversions needed when starting from raw x,y coordinates. The objective being maximized is the cumulative first-detection probability J, built from the Markov predict-update Bayesian recursion and the NVESD d

What would settle it

Run repeated real or simulated searches where the target's initial position is offset from the belief-map center by, say, one sigma, and where its heading changes at random intervals; compare realized detection frequency with the J value the planner predicts. If realized detection is near zero while predicted J is high, the central claim collapses. A cheaper check is to benchmark PDE against a polar-encoded version of any other metaheuristic; if the speed advantage disappears, the benefit comes from the encoding alone, not from the algorithm.

Watch

Extended reading notes

Core claim

The central claim is that the polar-coordinate representation of the flight path is the key improvement: each candidate path is encoded as a sequence of step lengths and yaw angles, which maps to Cartesian waypoints through a recurrence using cosine and sine of the yaw angle. The mutation and crossover of differential evolution then operate directly on these feasible control parameters, and UAV speed and yaw limits become variable bounds instead of post-hoc constraint checks. The same fitness function J, defined as the sum over time of the probability of first detecting the target (computed by Bayesian prediction-update with the NVESD detection likelihood), is used. The paper claims PDE outp

Load-bearing premise

Everything rests on the premise that the target moves as a deterministic Markov process with a known transition and that the Gaussian belief map with constant sigma_k captures all residual uncertainty — if the target does not start where the belief says or does not follow the assumed heading, the maximized fitness J is not the true detection probability, and the paper concedes there is no guarantee the UAV will find the target.

Editorial extensions

If this is right

  • Any Cartesian-space evolutionary search for UAV paths can be re-expressed in polar coordinates; the comparison shows standard DE in polar form substantially outperforms Cartesian DE, so the representation itself is a transferable improvement.
  • Because speed and yaw limits are encoded as simple bounds, the optimizer's output is physically feasible without a separate path-smoothing or feasibility step, as the real flight path overlap demonstrates.
  • The derived fitness J provides a concrete, comparable objective for moving-target search that couples sensor geometry, target dynamics, and human observer performance; future planners can optimize this same objective with different solvers.
  • Detection probability degrades as map size grows (from about 0.53 to 0.10 across the six scenarios), indicating that a single UAV's detection ceiling is limited; the paper's proposed direction of multi-UAV coordination is a direct corollary of these results.

Reading between the lines

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

  • Untested extension: the reported speedup likely transfers to other metaheuristics. If the real cost is repeated atan2 and sqrt in Cartesian representation, a polar-encoded PSO or GWO should show similar runtime gains, a benchmark the paper does not run.
  • The detection-probability advantage may be sensitive to target model mismatch. Since the field experiment succeeded only because the target started at the last known location and moved along the predicted heading, a natural test is to run PDE with perturbed initial positions or turning targets and compare realized detection frequency against the model's predicted J.
  • The constant sigma_k assumption in the belief update is a deliberate simplification; adapting sigma_k over time would extend the model to stochastic motion and might change the relative ranking of planners, which is a testable variant of the algorithm.
  • A quick falsifying experiment: use a polar-encoded version of the same DE with a different initial population or a simpler search pattern (e.g., lawnmower) and compare J; this would separate the contribution of polar encoding from the contribution of the DE evolutionary search itself.
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

4 major / 5 minor

Summary. This paper proposes a path-planning method, polar coordinate-based differential evolution (PDE), for a UAV searching for a moving target. The target is modeled as a deterministic Markov process with an initial Gaussian belief (Eq. 1); the camera footprint is derived geometrically; the detection likelihood is obtained from the NVESD model (Eq. 7). A recursive Bayesian prediction-update formulation is written down (Eqs. 12–15), and the cumulative detection probability J (Eq. 21) is used as the fitness function. The paper then encodes candidate paths in polar coordinates (step length and yaw angle), applies DE mutation/crossover/selection directly to this representation, and reports simulation comparisons with standard DE, GA, PSO, GWO, and GSK on six scenarios, plus one real-world experiment with a 3DR Solo. The abstract claims that PDE outperforms state-of-the-art algorithms in detection probability and execution time while remaining practical, and the source code is publicly released.

Significance. If the claims were fully supported, the contribution would be a practically useful encoding for evolutionary search planning: polar coordinates make UAV kinematic constraints simple to enforce and reduce the trigonometric overhead of evaluating footprints. The real experiment and the public code are genuine assets, and the paper is clearly written. However, the headline detection-probability claim rests on an internally inconsistent treatment of the Bayesian belief update: the fitness J is computed from an open-loop prior, not from the posterior conditioned on the candidate path's own no-detection events. In addition, the 'state-of-the-art' comparison omits the search-specific algorithms the paper itself cites (motion-encoded PSO, ACO, Bayesian optimization, greedy), and the execution-time advantage is not supported by statistical tests. With a corrected belief update and a more relevant baseline comparison, the central algorithmic idea could still be a valid contribution.

major comments (4)
  1. [§2.3–2.4, Eq. (21), Algorithm 1] The derivation defines the Bayesian update in Eq. (15), but the implemented fitness J never applies this update for the no-detection observations that the candidate path itself would produce. In Eqs. (17)–(21), r_k is evaluated using p(ξ^t_k|ζ_{1:k−1}) from the prediction step, yet Algorithm 1 contains no operation that conditions the belief on the visited footprints and the corresponding 'no detection' events. The belief used in the fitness evaluation remains the constant-variance Gaussian from Eq. (1), propagated open-loop. Consequently J is not the detection probability of the planned path under the paper's own target and sensor model; it is a surrogate that is generally optimistic, especially for paths that revisit previously searched areas. Table 4 therefore does not establish the abstract's claim about detection probability. The authors should either implement the correct Bayesian
  2. [§4.3, Tables 4–5] The comparison is against generic metaheuristics (DE, GA, PSO, GWO, GSK), not against the search-specific planners that the paper itself identifies as the state of the art in Table 1: motion-encoded PSO [12], ACO [13], Bayesian optimization [18], and greedy search [16]. Since the claimed contribution is to moving-target search rather than to generic DE, Table 4 at most shows that PDE beats these generic baselines on the (open-loop) surrogate J. A comparison with at least one dedicated search planner is necessary to support the 'state-of-the-art' wording.
  3. [§4.3, Table 5] The paper states that PDE is fastest in 5/6 scenarios, but the reported times do not clearly support that statement. For example, in Scenario 1 PDE is 246±16 s vs GA 249±33 s; in Scenario 3 PDE is 516±8 s vs GA 505±46 s; in Scenario 5 PDE is 907±85 s vs GA 956±57 s — differences are within or comparable to the reported standard deviations. No statistical test (e.g., paired t-test or Wilcoxon) is given. The execution-time advantage should be either supported with an error-bar analysis or restated as 'comparable to' the other Cartesian-coordinate algorithms.
  4. [§4.1, parameters] The scaling factor F is sampled from [0.1, 0.4], which the authors acknowledge is smaller than the standard DE range, and no sensitivity study is reported for this choice, nor for p_Cr and N_p. Since the PDE's advantage over standard DE could depend on this specific hand-tuning, the robustness of the conclusions to the parameter setting should be demonstrated, or at least a small sensitivity analysis should be added.
minor comments (5)
  1. [§3.2, Eq. (30)] The mapping in Eq. (30) is denoted f: U → ξ, but f was already used for the fitness function in Eqs. (21) and (24). This double use of f is confusing; use a different symbol for the coordinate transformation.
  2. [§4.3, Table 4] The table entries have inconsistent formatting, e.g., '0.0867±003' in Scenario 6 and '0.3215±0.011' in Scenario 4; standard errors or standard deviations should be labeled explicitly and formatted consistently.
  3. [§4.3, text after Table 5] There is a duplicated sentence fragment: 'while the heading angle is used directly in the footprint evaluation without requiring additional coordinate conversion. while the heading angle is used directly...' The sentence should be rewritten.
  4. [§4.4.2] The sentence 'Note that the detection event happened in this experiment is due to the alignment of our setup with the assumptions used in theory... hence there is no guarantee that the UAV will find the target' is an honest limitation, but it also underscores that the field test validates path feasibility rather than detection performance. This caveat should be moved to the experimental-design section.
  5. [§2.1, Eq. (1)] The belief is described as a Gaussian with constant σ_k, but the target is also modeled as following a deterministic known transition (Eqs. 13–14). It would help to state explicitly whether σ_k is meant only to represent initial localization uncertainty rather than motion uncertainty, since under a deterministic transition a Gaussian prior remains Gaussian with the same covariance.

Circularity Check

0 steps flagged · score 2.0 of 10

No significant circularity: PDE is an optimizer comparison on a common model-derived fitness J, with minor non-load-bearing self-citations.

full rationale

The claimed derivation chain is self-contained: J in Eq. 21 is assembled from the Gaussian belief model (Eq. 1), the deterministic Markov transition (Eqs. 12-14), and the NVESD no-detection likelihood (Eqs. 7-11), without fitting any parameter of J to the reported results. All compared algorithms optimize the same J, so reporting which metaheuristic attains higher J is a standard benchmark comparison, not a circular prediction. The self-citations, mainly [12] for a prior PSO formulation and [3,5] for UAV systems, are contextual and are not the sole support for any load-bearing step; the Bayesian and NVESD machinery is also anchored in external work [9,10,26,27,31-33]. No uniqueness theorem or ansatz is imported from the authors' own prior work to force the PDE representation; the polar-coordinate encoding (Eqs. 26-31) is a new algorithmic contribution. The acknowledged limitation in Section 4.4.2 ('there is no guarantee that the UAV will find the target') concerns real-world model mismatch, not circularity. A separate correctness risk exists because Algorithm 1 does not visibly condition the Gaussian belief on no-detection events when evaluating r_k in Eq. 17, so Table 4's J may not equal the Bayesian posterior detection probability; however, this is an internal modeling/implementation issue rather than a circular reduction of the result to its inputs.

Assumptions & free parameters 4 free parameters · 5 assumptions · 0 invented entities

The paper introduces no new physical entities. Its free parameters are DE hyperparameters and the Gaussian belief spread; the NVESD constants and camera geometry are inherited from prior literature. The most load-bearing assumptions are the deterministic target dynamics and the constant-sigma Gaussian belief, which together make the model tractable but also limit real-world validity.

free parameters (4)
  • DE scaling factor F = sampled from [0.1, 0.4]
    Section 4.1: chosen by the authors 'to provide better exploitation' without sensitivity analysis; affects convergence and final fitness.
  • Crossover probability p_Cr = 0.9
    Section 4.1: fixed algorithm hyperparameter; not varied in the study.
  • Population size N_p = 500
    Section 4.1: fixed for PDE and all compared algorithms; affects runtime and solution quality.
  • Max iterations = 100
    Section 4.1: fixed termination condition; not justified by convergence analysis.
assumptions (5)
  • domain assumption Target motion is a deterministic Markov process with known transition (Eqs. 13-14)
    Section 2.1. If the target does not move according to the known kinematic equation, the planned path may not track the true target; the paper acknowledges this in Section 4.4.2.
  • domain assumption The belief distribution remains Gaussian with constant sigma_k (Eq. 1, Steady State framework)
    Section 2.1. Needed to keep the belief map tractable; the update in Eq. 15 with a non-Gaussian no-detection likelihood would generally make the posterior non-Gaussian unless approximate or discretized.
  • domain assumption NVESD model (Eqs. 6-10) with empirical constants 1.51, 0.24, and V50 predicts real detection probabilities
    Section 2.2. The entire fitness objective J is only as good as this external model's fidelity for a small consumer drone camera.
  • domain assumption UAV flies at constant altitude and can follow the planned polar-encoded path (Eqs. 26-30)
    Section 3.2. Kinematic constraints are encoded only as bounds on rho and psi; the actual flight controller may smooth or deviate from the path, as seen in the field test.
  • standard math Cumulative detection probability is computed by multiplying per-step no-detection probabilities (Eqs. 18-21)
    Standard probability chain; implicitly assumes conditional independence of detection events given the target state.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Polar Coordinate-based Differential Evolution for Moving Target Search Using Vision Sensor on Unmanned Aerial Vehicles." pith.science (2026). https://pith.science/paper/6UABQNAB

@misc{pith2026260717771,
  author       = {Pith},
  title        = {Pith review of: Polar Coordinate-based Differential Evolution for Moving Target Search Using Vision Sensor on Unmanned Aerial Vehicles},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/6UABQNAB}},
  note         = {Machine review of arXiv:2607.17771}
}
read the original abstract

In search and rescue operations, there is a period known as the "golden time" during which the probability of finding the target alive is highest. The objective of this work is to propose a new search algorithm for unmanned aerial vehicles (UAVs) with a focus on improving the detection probability and execution time. We approach this problem by first modeling target dynamics as a Markov process and the detection likelihood as a function of image quality and the observer's vision. We then employ Bayesian theory to derive a fitness function representing the probability distribution of the target's location over the search area. Finally, we introduce a new algorithm named polar coordinate-based differential evolution (PDE) to generate a UAV search path that maximizes this fitness function. The PDE algorithm utilizes polar coordinates to incorporate kinematic constraints and maneuver properties of the UAV, allowing for better exploration of the solution space. A series of simulations and comparative analyses have been conducted to evaluate the performance of the proposed algorithm. Experiments involving a real UAV have also been conducted. Results demonstrate that the PDE algorithm outperforms state-of-the-art algorithms in terms of detection probability and execution time across diverse search scenarios while remaining practical for real-world applications. The source code of the algorithm is available at https://github.com/thuhangkhuat/PDE_target_search.

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

46 extracted references

  1. [12]

    Motion-encoded par- ticle swarm optimization for moving target search using UAVs,

    M. D. Phung and Q. P. Ha, “Motion-encoded par- ticle swarm optimization for moving target search using UAVs,”Applied Soft Computing, vol. 97, p. 106705, 2020

    2020

  2. [13]

    Ant colony optimiza- tion for multi-UAV minimum time search in uncer- tain domains,

    S. Perez-Carabaza, E. Besada-Portas, J. A. Lopez- Orozco, and J. M. de la Cruz, “Ant colony optimiza- tion for multi-UAV minimum time search in uncer- tain domains,”Applied Soft Computing, vol. 62, pp. 789–806, 2018

    2018

  3. [18]

    A bayesian approach for con- strained multi-agent minimum time search in uncer- tain dynamic domains,

    P. Lanillos, J. Ya˜ nez Zuluaga, J. J. Ruz, and E. Besada-Portas, “A bayesian approach for con- strained multi-agent minimum time search in uncer- tain dynamic domains,” inProceedings of the 15th Annual Conference on Genetic and Evolutionary Computation, GECCO ’13, (New York, NY, USA), p. 391–398, Association for Computing Machinery, 2013

    2013

  4. [16]

    Planning a multi-sensors search for a moving tar- get considering traveling costs,

    F. Delavernhe, P. Jaillet, A. Rossi, and M. Sevaux, “Planning a multi-sensors search for a moving tar- get considering traveling costs,”European Journal of Operational Research, vol. 292, no. 2, pp. 469– 482, 2021

    2021

  5. [1]

    A novel UAV path planning algorithm to search for floating objects on the ocean surface based on object’s trajectory pre- diction by regression,

    M. Boulares and A. Barnawi, “A novel UAV path planning algorithm to search for floating objects on the ocean surface based on object’s trajectory pre- diction by regression,”Robotics and Autonomous Systems, vol. 135, p. 103673, 2021

    2021

  6. [2]

    Lee, G.-H

    J.-H. Lee, G.-H. Gwon, I.-H. Kim, and H.-J. Jung, “A motion deblurring network for enhancing uav 13 3DR Solo Drone Mission Planner Hero 4 Camera Remote Controller Streaming Monitor (a) The 3DR Solo Drone and its flight controller (b) The initial belief map Fig. 6: Experimental setup (a) The planned search path in the belief map Waypoints UAV trajectory St...

    2023

  7. [3]

    System architecture for real-time sur- face inspection using multiple uavs,

    V. T. Hoang, M. D. Phung, T. H. Dinh, and Q. P. Ha, “System architecture for real-time sur- face inspection using multiple uavs,”IEEE Systems Journal, vol. 14, no. 2, pp. 2925–2936, 2020

    2020

  8. [4]

    Range, endurance, and optimal speed estimates for multi- copters,

    L. Bauersfeld and D. Scaramuzza, “Range, endurance, and optimal speed estimates for multi- copters,”IEEE Robotics and Automation Letters, vol. 7, no. 2, pp. 2953–2960, 2022

    2022

Show all 46 references
  1. [5]

    Enhanced discrete particle swarm opti- mization path planning for uav vision-based surface inspection,

    M. D. Phung, C. H. Quach, T. H. Dinh, and Q. Ha, “Enhanced discrete particle swarm opti- mization path planning for uav vision-based surface inspection,”Automation in Construction, vol. 81, pp. 25–33, 2017

    2017

  2. [6]

    Wind-induced motion blur and 3d digital twinning: Challenges in uav visual data collection on construction sites,

    M. Kamari and J. Kim, “Wind-induced motion blur and 3d digital twinning: Challenges in uav visual data collection on construction sites,” inComput- ing in Civil Engineering 2025: Computational and Intelligent Technologies, pp. 858–866

    2025

  3. [7]

    A multi-sensor approach for precision uav landing on dynamic platforms under challenging conditions,

    C. I¸ silak and H. Oktal, “A multi-sensor approach for precision uav landing on dynamic platforms under challenging conditions,”The Aeronautical Journal, vol. 130, no. 1344, pp. 526–552, 2026

    2026

  4. [8]

    Search- trajectory optimization: Part i, formulation and theory,

    J. Foraker, J. O. Royset, and I. Kaminer, “Search- trajectory optimization: Part i, formulation and theory,”Journal of Optimization Theory and Appli- cations, vol. 169, pp. 530–549, June 2015

    2015

  5. [9]

    Optimal search for a lost target in a 14 Fig. 8: The target within the camera’s field of view bayesian world,

    F. Bourgault, T. Furukawa, and H. F. Durrant- Whyte, “Optimal search for a lost target in a 14 Fig. 8: The target within the camera’s field of view bayesian world,” inSpringer Tracts in Advanced Robotics, pp. 209–222, Springer Berlin Heidelberg, 2006

    2006

  6. [10]

    Recursive bayesian search-and- tracking using coordinated UAVs for lost targets,

    T. Furukawa, F. Bourgault, B. Lavis, and H. Durrant-Whyte, “Recursive bayesian search-and- tracking using coordinated UAVs for lost targets,” inProceedings 2006 IEEE International Conference on Robotics and Automation, 2006. ICRA 2006., pp. 2521–2526, 2006

    2006

  7. [11]

    Aerial vehicle search-path optimiza- tion: A novel method for emergency operations,

    M. Raap, S. Meyer-Nieberg, S. Pickl, and M. Zsifkovits, “Aerial vehicle search-path optimiza- tion: A novel method for emergency operations,” Journal of Optimization Theory and Applications, vol. 172, pp. 965–983, Sept. 2016

    2016

  8. [14]

    Online path planning for AUV ren- dezvous in dynamic cluttered undersea environment using evolutionary algorithms,

    S. MahmoudZadeh, A. Yazdani, K. Sammut, and D. Powers, “Online path planning for AUV ren- dezvous in dynamic cluttered undersea environment using evolutionary algorithms,”Applied Soft Com- puting, vol. 70, pp. 929–945, Sept. 2018

    2018

  9. [15]

    The complexity of decentralized control of markov decision processes,

    D. S. Bernstein, R. Givan, N. Immerman, and S. Zilberstein, “The complexity of decentralized control of markov decision processes,”Mathematics of Operations Research, vol. 27, pp. 819–840, Nov. 2002

    2002

  10. [17]

    Multi-UAV target search using explicit decentralized gradient-based negotiation,

    S. K. Gan and S. Sukkarieh, “Multi-UAV target search using explicit decentralized gradient-based negotiation,” in2011 IEEE International Confer- ence on Robotics and Automation, pp. 751–756, 2011

    2011

  11. [19]

    Theory of optimal search (lawrence d. stone),

    B. O. Koopman, “Theory of optimal search (lawrence d. stone),”SIAM Review, vol. 19, no. 2, pp. 361–364, 1977

    1977

  12. [20]

    Kennedy, R

    J. Kennedy, R. C. Eberhart, and Y. Shi,Swarm Intelligence. Elsevier, 2001

    2001

  13. [21]

    Differential evolution: A recent review based on state-of-the-art works,

    M. F. Ahmad, N. A. M. Isa, W. H. Lim, and K. M. Ang, “Differential evolution: A recent review based on state-of-the-art works,”Alexandria Engineering Journal, vol. 61, pp. 3831–3872, May 2022

    2022

  14. [22]

    An improved differential evolution algo- rithm and its application in optimization problem,

    W. Deng, S. Shang, X. Cai, H. Zhao, Y. Song, and J. Xu, “An improved differential evolution algo- rithm and its application in optimization problem,” Soft Computing, vol. 25, pp. 5277–5298, Jan. 2021. 15

    2021

  15. [23]

    Price, R

    K. Price, R. M. Storn, and J. A. Lampinen,Dif- ferential evolution: a practical approach to global optimization. Springer Science & Business Media, 2006

    2006

  16. [24]

    Bernstain-search dif- ferential evolution algorithm for numerical function optimization,

    P. Civicioglu and E. Besdok, “Bernstain-search dif- ferential evolution algorithm for numerical function optimization,”Expert Systems with Applications, vol. 138, p. 112831, Dec. 2019

    2019

  17. [25]

    Particle swarm optimization or dif- ferential evolution—a comparison,

    A. P. Piotrowski, J. J. Napiorkowski, and A. E. Piotrowska, “Particle swarm optimization or dif- ferential evolution—a comparison,”Engineering Applications of Artificial Intelligence, vol. 121, p. 106008, May 2023

    2023

  18. [26]

    Probabilistic robotics,

    S. Thrun, “Probabilistic robotics,”Communications of the ACM, vol. 45, no. 3, pp. 52–57, 2002

    2002

  19. [27]

    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,” in2017 IEEE power & energy society general meeting, pp. 1–5, IEEE, 2017

    2017

  20. [28]

    Optimal sur- vivor search for a target with conditionally deter- ministic motion under reward criterion,

    K. Iida, R. Hohzaki, and K. Inada, “Optimal sur- vivor search for a target with conditionally deter- ministic motion under reward criterion,”Journal of the Operations Research Society of Japan, vol. 41, no. 2, pp. 246–260, 1998

    1998

  21. [29]

    Uav tra- jectory optimization for minimum time search with communication constraints and collision avoid- ance,

    S. P´ erez-Carabaza, J. Scherer, B. Rinner, J. A. L´ opez-Orozco, and E. Besada-Portas, “Uav tra- jectory optimization for minimum time search with communication constraints and collision avoid- ance,”Engineering Applications of Artificial Intel- ligence, vol. 85, pp. 357–371, 2019

    2019

  22. [30]

    Aerial vehicle search-path optimiza- tion: A novel method for emergency operations,

    M. Raap, S. Meyer-Nieberg, S. Pickl, and M. Zsifkovits, “Aerial vehicle search-path optimiza- tion: A novel method for emergency operations,” Journal of optimization theory and applications, vol. 172, no. 3, pp. 965–983, 2017

    2017

  23. [31]

    Target detection cycle criteria when using the targeting task performance metric,

    J. G. Hixson, E. L. Jacobs, and R. H. Vollmer- hausen, “Target detection cycle criteria when using the targeting task performance metric,” inSPIE Proceedings(R. G. Driggers and D. A. Huckridge, eds.), vol. 5612, pp. 275–283, SPIE, Dec. 2004

    2004

  24. [32]

    New metric for predicting target acquisition performance,

    R. H. Vollmerhausen, E. Jacobs, and R. G. Drig- gers, “New metric for predicting target acquisition performance,”Optical Engineering, vol. 43, no. 11, pp. 2806–2818, 2004

    2004

  25. [33]

    Model- ing the target acquisition performance of active imaging systems,

    R. L. Espinola, E. L. Jacobs, C. E. Halford, R. Vollmerhausen, and D. H. Tofsted, “Model- ing the target acquisition performance of active imaging systems,”Optics express, vol. 15, no. 7, pp. 3816–3832, 2007

    2007

  26. [34]

    An exhaustive review of the metaheuristic algorithms for search and optimization: taxonomy, applications, and open challenges,

    K. Rajwar, K. Deep, and S. Das, “An exhaustive review of the metaheuristic algorithms for search and optimization: taxonomy, applications, and open challenges,”Artificial Intelligence Review, vol. 56, p. 13187–13257, Apr. 2023

    2023

  27. [35]

    Chaotic local search-based differen- tial evolution algorithms for optimization,

    S. Gao, Y. Yu, Y. Wang, J. Wang, J. Cheng, and M. Zhou, “Chaotic local search-based differen- tial evolution algorithms for optimization,”IEEE Transactions on Systems, Man, and Cybernetics: Systems, vol. 51, no. 6, pp. 3954–3967, 2021

    2021

  28. [36]

    Differential evo- lution: A survey of the state-of-the-art,

    S. Das and P. N. Suganthan, “Differential evo- lution: A survey of the state-of-the-art,”IEEE Transactions on Evolutionary Computation, vol. 15, pp. 4–31, Feb. 2011

    2011

  29. [37]

    Optimal trajectory planning for cinematography with multiple unmanned aerial vehicles,

    A. Alc´ antara, J. Capit´ an, R. Cunha, and A. Ollero, “Optimal trajectory planning for cinematography with multiple unmanned aerial vehicles,”Robotics and Autonomous Systems, vol. 140, p. 103778, June 2021

    2021

  30. [38]

    Differential evolution – a simple and efficient heuristic for global optimiza- tion over continuous spaces,

    R. Storn and K. Price, “Differential evolution – a simple and efficient heuristic for global optimiza- tion over continuous spaces,”Journal of Global Optimization, vol. 11, no. 4, p. 341–359, 1997

    1997

  31. [39]

    Differential evolution: A review of more than two decades of research,

    Bilal, M. Pant, H. Zaheer, L. Garcia-Hernandez, and A. Abraham, “Differential evolution: A review of more than two decades of research,”Engineer- ing Applications of Artificial Intelligence, vol. 90, p. 103479, Apr. 2020

    2020

  32. [40]

    Genetic algo- rithms: concepts and applications [in engineering design],

    K. Man, K. Tang, and S. Kwong, “Genetic algo- rithms: concepts and applications [in engineering design],”IEEE Transactions on Industrial Electron- ics, vol. 43, no. 5, pp. 519–534, 1996

    1996

  33. [41]

    A review on genetic algorithm: past, present, and future,

    S. Katoch, S. S. Chauhan, and V. Kumar, “A review on genetic algorithm: past, present, and future,”Multimedia Tools and Applications, vol. 80, p. 8091–8126, Oct. 2020

    2020

  34. [42]

    Particle swarm optimization,

    J. Kennedy and R. Eberhart, “Particle swarm optimization,” inProceedings of ICNN’95 - Inter- national Conference on Neural Networks, vol. 4, pp. 1942–1948 vol.4, 1995

    1942

  35. [43]

    Particle swarm optimization algorithm and its applications: A systematic review,

    A. G. Gad, “Particle swarm optimization algorithm and its applications: A systematic review,”Archives of Computational Methods in Engineering, vol. 29, p. 2531–2561, Apr. 2022

    2022

  36. [44]

    Grey wolf optimizer,

    S. Mirjalili, S. M. Mirjalili, and A. Lewis, “Grey wolf optimizer,”Advances in Engineering Software, vol. 69, pp. 46–61, Mar. 2014

    2014

  37. [45]

    An improved grey wolf optimizer for solving engineering problems,

    M. H. Nadimi-Shahraki, S. Taghian, and S. Mir- jalili, “An improved grey wolf optimizer for solving engineering problems,”Expert Systems with Appli- cations, vol. 166, p. 113917, Mar. 2021

    2021

  38. [46]

    Gaining-sharing knowledge based algo- rithm for solving optimization problems: a novel nature-inspired algorithm,

    A. W. Mohamed, A. A. Hadi, and A. K. Mohamed, “Gaining-sharing knowledge based algo- rithm for solving optimization problems: a novel nature-inspired algorithm,”International Journal of Machine Learning and Cybernetics, vol. 11, pp. 1501–1529, Dec. 2019. 16

    2019

Pith tools

Reviewed August 1, 2026 · model on record in the stance chip above.