Autonomous Fishing Drone
Senior Design Project
Embedded Files
This project involves the development of an autonomous fishing drone designed to detect and capture fish in specific aquatic areas like lakes and ponds. The drone integrates advanced navigation systems, machine learning-based fish detection, and structural optimization for efficient operations. Its key features include real-time fish detection, precise navigation, and payload mechanisms, showcasing innovations in both design and automation.
Methodology
Methodology
Mission Execution: The drone autonomously navigates to preselected fishing spots, detects fish using machine learning models, and captures them via a fishing hook.
Training Model: Fish detection relies on datasets labeled with tools like LabelImg and trained in Google Colab, employing real-time image processing techniques.
Path Planning: Autonomous flight paths are generated using geofencing and adaptive navigation algorithms, supported by obstacle avoidance.
Simulation and Testing: The Gazebo simulator and ANSYS software were used for operational validation and structural analysis.
Key Components
Key Components
Mechanical Design: Lightweight quadcopter frame with optimized rotor arms, landing gear, and payload mechanisms.
Avionics: Equipped with Pixhawk 2.4.8 flight controller, Walksnail Avatar HD video telemetry system, and FrSky radio telemetry modules.
Navigation: High-precision Here3+ GPS system for accurate pathfinding.
Fishing Mechanism: A servo-powered grabber mechanism for deploying and retrieving payloads.
Propulsion: Four BLDC motors (Sunnysky V3 X2212) paired with 60A ESCs.
Power Supply: Lithium-polymer battery system optimized for 10 minutes of flight time.
Key Accomplishments
Key Accomplishments
Successful integration of machine learning for real-time fish detection.
Development of lightweight and robust quadcopter design.
Effective testing and validation of the drone’s fishing capabilities using simulation tools.
Efficient navigation and hotspot detection algorithms for autonomous fishing.
Testing And Simulation
Testing And Simulation
Successful integration of machine learning for real-time fish detection.
Development of lightweight and robust quadcopter design.
Effective testing and validation of the drone’s fishing capabilities using simulation tools.
Efficient navigation and hotspot detection algorithms for autonomous fishing.
Analysis
Analysis
Structural Simulations: Finite element analysis of frame components to evaluate stress, deformation, and bending moments under operational loads.
Aerodynamic Evaluation: Computational fluid dynamics (CFD) analysis confirmed efficient airflow, net thrust generation, and pressure distribution for optimal stability.
Thrust Requirements: Propulsion systems were designed with a 2:1 thrust-to-weight ratio, ensuring robust flight performance.
Outcomes
Outcomes
The project successfully demonstrated autonomous fish detection and retrieval by integrating advanced design, machine learning, and structural optimization. It validated the drone's stability and operational efficiency under real-world conditions, laying a strong foundation for further advancements in sustainable fishing practices and aquatic ecosystem monitoring.
Page updated
Report abuse