Logistic Regression

96.19

0.67

0.80

SVM (RBF)

97.28

0.73

0.85

Logistic Regression

96.77

0.6486

0.7869

0.9991

SVM (RBF)

96.28

0.6154

0.7619

0.9943

0

RF-RFE-no_smote

0.998675

1.000000

0.998565

0.999282

1.000000

1

RF-RFE-with_smote

0.998675

1.000000

0.998565

0.999282

1.000000

2

RF-ANOVA-no_smote

0.997351

1.000000

0.997131

0.998563

0.997786

3

RF-ANOVA-with_smote

0.997351

1.000000

0.997131

0.998563

0.997897

4

LR-RFE-with_smote

0.992053

1.000000

0.991392

0.995677

0.996562

5

SVM-RFE-with_smote

0.985430

1.000000

0.984218

0.992046

0.997650

6

LR-ANOVA-with_smote

0.984106

0.998544

0.984218

0.991329

0.995523

7

SVM-RFE-no_smote

0.977483

0.977528

0.998565

0.987935

0.998862

8

SVM-ANOVA-no_smote

0.972185

0.974719

0.995696

0.985096

0.996413

9

SVM-ANOVA-with_smote

0.972185

0.994152

0.975610

0.984794

0.992530

10

LR-RFE-no_smote

0.965563

0.966620

0.997131

0.981638

0.989586

11

LR-ANOVA-no_smote

0.965563

0.967922

0.995696

0.981612

0.990674

DJI Enterprise

DJI Official Website (enterprise.dji.com/dock ↗)

Automated drone docking with battery swap and autonomous operation

Limited to DJI drones; lacks versatility for other drone models

Mohamed Gadalla, Sayem Zafar

International Journal of Hydrogen Energy (2016)

Proposed a hybrid power system (Fuel Cell + PV + Battery) for small UAVs. Demonstrated endurance improvement from 470 min → 970 min. Validated performance through simulations and bench tests.

No adaptive or AI-based control. No support for dynamic wind disturbances. Lacks nonlinear hybrid control for high-speed manoeuvres.

Chunwu Xiao, Bin Wang, Dan Zhao, Chaohui Wang

Thermal Science and Engineering Progress (2023)

Comprehensive review of Lithium battery technologies for electric & hybrid-electric UAVs. Compared battery chemistries, issues, hybrid power systems, and performance limits. Identified challenges in energy density, power density, and charging time.

Traditional PID suffers under wind, payload changes. No fault-tolerant or self-tuning capability.

Attilio Di Nisio, Giulio Avanzini, Daniel Lotano, Donato Stigliano, Anna Lanzolla

Sensors (MDPI), 2023

Developed and validated a battery discharge model for Li-Po UAV batteries. Achieved <0.7% error in predicting capacity under constant and variable loads. Provided experimental setup using current and voltage sensors.

No nonlinear control methods (SMC, backstepping). Limited robustness under extreme aerodynamic conditions.

“Multirotor Aerial Vehicles: Modeling, Estimation, and Control” – IEEE Robotics & Automation Magazine

Provided detailed modelling and classical control (PID/LQR) for quadrotors. Stability proofs and controller comparisons.

“Design and Control of a Miniature Quadrotor” – IEEE ICRA

Developed dynamic modelling and a PID-based flight control system. Showed stable autonomous hovering.

“Quadrotor Helicopter Flight Dynamics and Control” – AAA Guidance, Navigation, and Control Conference

Identified system parameters and provided closed-loop control validation using LQR and PD controllers.

S. Skogestad (2003) — Simple analytic rules for model reduction & PID tuning.

ResearchGate

Provides simple model-reduction + PID tuning rules giving good robustness/performance tradeoffs.

Still oriented to SISO process plants; limited coverage of MIMO/underactuated systems like quadrotors.

Survey: “PID control of quadrotor UAVs” (2023) — ScienceDirect survey.

ScienceDirect

Surveys PID structures for quadrotors (PID variants, fractional, intelligent hybrids) and practical implementations.