Friday, November 20, 2026
Science
No Result
View All Result
  • Login
  • HOME
  • SCIENCE NEWS
  • CONTACT US
  • HOME
  • SCIENCE NEWS
  • CONTACT US
No Result
View All Result
Scienmag
No Result
View All Result
Home Science News Technology and Engineering

Two-Layer Navigation Framework Coordinates Multi-Drone Swarms in Mountainous Wind Fields

November 20, 2026
in Technology and Engineering
Denise Maddox
By Denise Maddox Scienmag Editorial Profile - Mechanical Engineering
Reading Time: 7 mins read
0
Two-Layer Navigation Framework Coordinates Multi-Drone Swarms in Mountainous Wind Fields

Two-Layer Navigation Framework Coordinates Multi-Drone Swarms in Mountainous Wind Fields

65
SHARES
587
VIEWS
Share on FacebookShare on Twitter
ADVERTISEMENT

Researchers Liping Qiao, Yiqing Huang, He Huang, Zhao Jin, and Tao Gao have engineered a two-layer optimization framework to coordinate multi-drone flight paths across complex three-dimensional environments featuring rugged mountainous terrain and strong aerodynamic disturbances. Reported in the Journal of King Saud University Computer and Information Sciences, the system, designated KDAPSO-PIBR-FF, decouples strategic global path generation from high-speed reactive collision avoidance to overcome computational bottlenecks and deadlocks that compromise uncrewed aerial vehicle swarms. Autonomous drone fleets in shared airspace struggle to balance broad corridor exploration with immediate avoidance of moving obstacles, narrow terrain saddles, and wind turbulence. To resolve this conflict, the framework combines a knowledge-guided dual-population global planner that maps energy-efficient baseline routes with an asymmetric, priority-conditioned local force field that resolves emerging conflicts in milliseconds, ensuring continuous separation among collaborative agents without triggering cascading delays or oscillatory deviations.

Operational demands on uncrewed aerial vehicles have expanded rapidly beyond reconnaissance into civilian domains such as precision agriculture, post-disaster logistics, critical infrastructure inspection, and environmental monitoring. However, coordinating multiple uncrewed platforms within constrained operational theaters introduces severe mathematical and physical hurdles. Conventional deterministic path planners, such as A-star search, Voronoi graph partitioning, and dynamic programming, become computationally prohibitive when scaled to high-dimensional state spaces where multiple agents simultaneously alter spatial positions. As state dimensions multiply, memory overhead escalates and grid representations struggle to capture continuous gradients, forcing planners to trade trajectory optimality for computational tractability. Similarly, classical multi-objective metaheuristics and evolutionary swarm algorithms frequently suffer from premature convergence when navigating tightly bounded airspaces, where diversity loss and archive selection bias trap candidate trajectories in sub-optimal local minima or guide them across dangerous, constraint-violating aerodynamic barriers.

To address the limitations of existing global search paradigms, the authors established the Knowledge-guided Dual-population Adaptive Particle Swarm Optimization algorithm, known as KDAPSO, as the macro-level planning foundation. In conventional particle swarm optimization, particles update velocity vectors relative to personal historical bests and a shared global leader, but treating every individual uniformly produces an unresolved competition between broad exploration and localized exploitation. The KDAPSO algorithm mitigates this tension by structurally dividing the swarm into two complementary subpopulations: an exploration population containing sixty percent of the particles, dedicated to sweeping the three-dimensional workspace, and an exploitation population comprising forty percent, focused on refining the Pareto front. Rather than enforcing a static velocity update equation across all search stages, KDAPSO establishes an adaptive operator pool containing three distinct kinematic behaviors, allowing particles to dynamically adapt their evolutionary trajectories according to empirical search success.

The adaptive operator pool provides three distinct kinematic updates tailored for different search phases: a canonical inertia-weight operator for broad corridor scanning, an algebraic constriction-factor formulation regulated by Clerc parameter conditions for localized exploitation, and an asymmetric Lévy-perturbation operator injecting chaotic displacement vectors to dislodge the swarm from local minima traps. Particles select these operators via a success-driven roulette-wheel mechanism, receiving credit rewards whenever mutations introduce new non-dominated solutions into an external archive or substantially reduce constraint violations. To prevent early-stage search history from permanently biasing operator selection probabilities, KDAPSO implements a modular reset mechanism that clears accumulated credit counters every twenty iterations. Furthermore, the global optimizer solves the structural challenge of spherical coordinate trajectory encoding by introducing a coupled segment triplet update, which processes radial step lengths, elevation angles, and azimuth angles as an indivisible geometric block, preventing heading distortions caused by independent coordinate truncation.

A central innovation within KDAPSO is the integration of a segment-wise violation frequency Knowledge Library that prevents the swarm from selecting global leaders situated along the precarious boundaries of hazardous zones. In complex mountainous environments, non-dominated candidate routes stored in the external Pareto archive frequently skirt the edges of physical terrain boundaries or atmospheric disturbance envelopes, where minor perturbations cause immediate safety infractions. The Knowledge Library tracks cumulative historical constraint violations along each individual path segment using an exponential decay factor and structural indicator functions that isolate geometrically over-extended path components as primary risks. When selecting a global leader from the external archive, KDAPSO modulates traditional crowding-distance metrics by applying an exponential penalty proportional to the mean historical violation frequency mapped to each candidate trajectory. This guidance formulation steers the swarm toward functionally safe flight corridors while maintaining population diversity, ensuring that macro-level reference routes maintain adequate safety margins before execution.

Before deploying KDAPSO into physical trajectory environments, the researchers evaluated its generic numerical search capabilities using twenty-nine benchmark functions from the standardized CEC2017 test suite, excluding the numerically unstable second function. Across thirty independent trials with a problem dimension of thirty and three hundred thousand function evaluations, KDAPSO was compared against nine established and advanced metaheuristics, including classical Particle Swarm Optimization, Genetic Algorithms, Ant Colony Optimization, and modern optimizers such as the Walrus Mutation Algorithm and Beluga Whale Optimization. Statistical evaluation via the Friedman test revealed a statistically significant overall performance difference among the ten algorithms, with KDAPSO achieving the lowest average rank of 2.3103 across the complete benchmark suite. Two-sided Wilcoxon signed-rank tests at a significance level of 0.05 confirmed that KDAPSO achieved statistically significant improvements over classical Particle Swarm Optimization on twenty-seven functions, over Ant Colony Optimization and Honey Badger Optimization on twenty-five functions, and retained balanced, highly competitive performance against contemporary competitors.

While KDAPSO optimizes multi-objective baseline routes that balance path length, aerodynamic exposure surrogates, vertical vibration penalties, and kinematic turning angles, real-world execution requires instantaneous adaptation to unforeseen threats. To provide timely and kinematically feasible corrections, the authors formulated the Priority-aware Iterative Best-Response Force Field strategy, designated PIBR-FF, as the reactive local replanning layer. The operational space is modeled with high geometric fidelity using satellite-derived Digital Elevation Models from the Tibetan Plateau, combined with Approximate Signed Distance Functions that represent localized atmospheric wind zones as oriented three-dimensional ellipsoids. Unlike conventional binary obstacle models that simply classify regions as safe or unsafe, the signed distance formulation computes an analytic penetration depth metric and spatial gradient. This enables uncrewed aircraft to perceive turbulent boundaries continuously and evaluate aerodynamic interference without requiring the prohibitive computational overhead of solving full Navier-Stokes atmospheric fluid dynamic simulations during real-time flight.

Standard artificial potential field methods frequently encounter stagnation points and local minima traps when steering vehicles around large ellipsoidal constraints, especially when outward repulsive vectors directly oppose goal-directed attractive forces. The PIBR-FF framework resolves this vulnerability by decomposing wind-avoidance forces into an analytic normal repulsive component and an orthogonal tangential steering component derived via Gram-Schmidt orthogonalization. When a drone approaches a wind zone boundary where the attractive vector opposes the normal gradient, tangential steering activates to guide the vehicle along the ellipsoidal periphery, inducing a smooth sliding motion rather than abrupt reversals. To eliminate deadlocks at extreme stagnation points where net forces cancel out, PIBR-FF incorporates a recovery mechanism that detects minimal spatial displacement and injects an altitude-biased perpendicular perturbation, lifting the drone over spatial barriers. Following fixed-step force-field integration, discrete waypoint sequences are smoothed using cubic B-spline interpolation, producing continuous, kinematically executable detours while strictly enforcing absolute ceiling and ground-clearance constraints.

To eliminate spatial deadlocks when multiple drones navigate converging launch corridors or narrow ravines, the framework establishes an asymmetric spatiotemporal coordination model organized into three operational priority tiers. Aircraft assigned to the highest priority tier optimize trajectories considering only static terrain, wind zones, and dynamic obstacles, completely disregarding other uncrewed aircraft; once finalized, their deterministic spatiotemporal paths serve as fixed constraints for the remainder of the swarm. Medium-priority aircraft resolve mutual conflicts through an alternating iterative best-response coordination protocol, treating trajectories of high-priority platforms and previous iterations of peer paths as dynamic moving obstacles until time-synchronized separation thresholds are satisfied across flight durations. Lowest-priority platforms execute reactive avoidance, treating all higher-tier trajectories as rigid obstacles and adjusting local velocities and lateral positions to weave through remaining safe corridors. This structured hierarchy prevents the chaotic replanning avalanches that typically paralyze decentralized swarms in congested operational sectors.

The integrated KDAPSO-PIBR-FF framework was evaluated in mission simulations involving four collaborative drones across three distinct environments: a six-hundred-meter synthetic workspace with moving obstacles, a one-kilometer mountainous terrain model reconstructed from Tibetan Plateau satellite elevation data, and an expanded two-kilometer co-directional flight theater with steep elevation gradients. Across thirty independent runs, the integrated framework was compared against four baseline configurations: MPPSO paired with modified dynamic window approaches, IACO combined with adaptive rapidly exploring random trees, and both NMOPSO and NSGA-III paired with PIBR-FF. Quantitative evaluations demonstrated that KDAPSO-PIBR-FF consistently achieved the shortest global path lengths, lowest auxiliary energy-related costs, and lowest overall fitness values. Furthermore, while baseline reactive planners like the modified dynamic window approach required approximately 2.58 to 2.59 seconds per local replanning cycle in mountainous terrain, PIBR-FF executed local trajectory corrections within 1.9 to 2.0 milliseconds in the evaluated MATLAB environment, generating compact detours without triggering premature high-altitude climbs.

Comprehensive ablation studies and cross-pairing analyses verified the distinct, complementary contributions of the architectural components within both planning layers. Disabling the adaptive operator pool caused global optimization to stagnate prematurely, increasing final mission costs by eight to eleven percent, while removing the segment-wise Knowledge Library produced severe search oscillations that escalated final costs by over thirty-five percent in complex mountain scenarios. Similarly, omitting the iterative best-response mechanism within the local layer triggered separation violation rates of thirty to forty-three percent in head-on conflict simulations, whereas the complete PIBR-FF architecture recorded zero separation violations across all evaluated head-on runs. Despite these demonstrated capabilities, the researchers noted key operational constraints, including global computational scaling with larger swarm sizes, static initial priority assignments, and reliance on parameterized aerodynamic ellipsoids rather than empirical turbulence data. Future work will investigate dynamically reconfigurable priority assignments, spatial partitioning approximations for massive swarms, and onboard sensor integration to transition the architecture toward physical flight validation.

Subject of Research: Hierarchical global and local trajectory optimization for cooperative multi-UAV flight in complex three-dimensional environments.

Article Title: KDAPSO-PIBR-FF: a hierarchical global–local optimization framework for cooperative multi-UAV path planning in complex 3D environments

Article References: Qiao, L., Huang, Y., Huang, H., Jin, Z., & Gao, T. (2026). KDAPSO-PIBR-FF: a hierarchical global–local optimization framework for cooperative multi-UAV path planning in complex 3D environments. Journal of King Saud University Computer and Information Sciences, 38(9), Article 1015. https://doi.org/10.1007/s44443-026-01426-7

Image Credits: AI Generated

DOI: 10.1007/s44443-026-01426-7

Keywords: Uncrewed aerial vehicles, Drone swarm coordination, Trajectory optimization, Particle swarm optimization, Path planning, Signed distance function, Digital elevation model, Aerodynamic wind fields, Iterative best response, Artificial potential fields, Multi-objective optimization, B-spline smoothing, Mountainous terrain navigation, Collision avoidance, Autonomous navigation, Dynamic obstacle avoidance, Pareto archive

Cite Scienmag News

Denise Maddox. (November 20, 2026). Two-Layer Navigation Framework Coordinates Multi-Drone Swarms in Mountainous Wind Fields. Scienmag. https://scienmag.com/two-layer-navigation-framework-coordinates-multi-drone-swarms-in-mountainous-wind-fields/

Denise Maddox. "Two-Layer Navigation Framework Coordinates Multi-Drone Swarms in Mountainous Wind Fields." Scienmag, 20 November 2026, https://scienmag.com/two-layer-navigation-framework-coordinates-multi-drone-swarms-in-mountainous-wind-fields/. Accessed 20 November 2026.

Denise Maddox. "Two-Layer Navigation Framework Coordinates Multi-Drone Swarms in Mountainous Wind Fields." Scienmag. November 20, 2026. https://scienmag.com/two-layer-navigation-framework-coordinates-multi-drone-swarms-in-mountainous-wind-fields/

Tags: Aerodynamic wind fieldsartificial potential fieldsautonomous navigationB-spline smoothingcollision avoidancedigital elevation modelDrone swarm coordinationdynamic obstacle avoidanceIterative best responseMountainous terrain navigationmulti-objective optimizationPareto archiveparticle swarm optimizationpath planningSigned distance functiontrajectory optimizationuncrewed aerial vehicles
Share26Tweet16
Previous Post

Dynamic Image Fusion Reconstructs Hydrate Reservoir Core Fracture Networks in Three Dimensions

Next Post

Researchers Evaluate Interpretable Machine Learning for Cardiac Risk in Hip Surgery

Related Posts

Inorganic Contaminants in Recycled Plastics Dictate Zeolite Catalyst Deactivation During Cracking
Chemistry

Inorganic Contaminants in Recycled Plastics Dictate Zeolite Catalyst Deactivation During Cracking

November 20, 2026
Griffon Vulture Algorithm Accurately Extracts Seven Nonlinear Parameters in Fuel Cell Simulations
Technology and Engineering

Griffon Vulture Algorithm Accurately Extracts Seven Nonlinear Parameters in Fuel Cell Simulations

November 20, 2026
Morphology Control and Elemental Doping Synergistically Boost Anatase Titanium Dioxide Photocatalysis
Technology and Engineering

Morphology Control and Elemental Doping Synergistically Boost Anatase Titanium Dioxide Photocatalysis

November 20, 2026
Tailored Polymer Interfaces Enable High Filler Loadings for Carbon Capture
Technology and Engineering

Tailored Polymer Interfaces Enable High Filler Loadings for Carbon Capture

November 20, 2026
Geospatial Screening Identifies Key High Suitability Wind Corridors Across Egypt
Technology and Engineering

Geospatial Screening Identifies Key High Suitability Wind Corridors Across Egypt

November 20, 2026
High-Resolution AO-OCT Detects Hidden Retinal Cell Loss in Stable Glaucoma Patients
Technology and Engineering

High-Resolution AO-OCT Detects Hidden Retinal Cell Loss in Stable Glaucoma Patients

November 20, 2026
Next Post
Researchers Evaluate Interpretable Machine Learning for Cardiac Risk in Hip Surgery

Researchers Evaluate Interpretable Machine Learning for Cardiac Risk in Hip Surgery

  • Mothers who receive childcare support from maternal grandparents show more optimized

    Mothers who receive childcare support from maternal grandparents show more parental warmth, finds NTU Singapore study

    27656 shares
    Share 11059 Tweet 6912
  • University of Seville Breaks 120-Year-Old Mystery, Revises a Key Einstein Concept

    1061 shares
    Share 424 Tweet 265
  • Bee body mass, pathogens and local climate influence heat tolerance

    682 shares
    Share 273 Tweet 171
  • Researchers record first-ever images and data of a shark experiencing a boat strike

    546 shares
    Share 218 Tweet 137
  • Groundbreaking Clinical Trial Reveals Lubiprostone Enhances Kidney Function

    531 shares
    Share 212 Tweet 133
Science

Embark on a thrilling journey of discovery with Scienmag.com—your ultimate source for cutting-edge breakthroughs. Immerse yourself in a world where curiosity knows no limits and tomorrow’s possibilities become today’s reality!

RECENT NEWS

  • Five-Layer Climate Framework Connects Physical Hazards to Real-World Human Impacts
  • Consensus Level Restriction Curbs Simulated 51 Percent Attacks in Ethereum Classic
  • Inorganic Contaminants in Recycled Plastics Dictate Zeolite Catalyst Deactivation During Cracking
  • Self Attention Siamese Neural Networks Boost Accuracy in Case Based Reasoning Retrieval

Categories

  • Agriculture
  • Anthropology
  • Archaeology
  • Athmospheric
  • Biology
  • Biotechnology
  • Blog
  • Bussines
  • Cancer
  • Chemistry
  • Climate
  • Earth Science
  • Editorial Policy
  • Marine
  • Mathematics
  • Medicine
  • Pediatry
  • Policy
  • Psychology & Psychiatry
  • Science Education
  • Science News
  • Social Science
  • Space
  • Technology and Engineering

Subscribe to Blog via Email

Enter your email address to subscribe to this blog and receive notifications of new posts by email.

Join 5,151 other subscribers

© 2026 Scienmag - Science Magazine

Welcome Back!

Login to your account below

Forgotten Password?

Retrieve your password

Please enter your username or email address to reset your password.

Log In
No Result
View All Result
  • HOME
  • SCIENCE NEWS
  • CONTACT US

© 2026 Scienmag - Science Magazine

Discover more from Science

Subscribe now to keep reading and get access to the full archive.

Continue reading