Aerial manipulation of long objects using adaptive neuro-fuzzy controller under battery variability

Area of Science:

• Robotics and Automation
• Civil Engineering Applications
• Aerial Systems

Background:

• Aerial manipulation offers versatile solutions for tasks in confined spaces, crucial for civil engineering and disaster response.
• Existing single- or dual-arm systems face scalability and maintenance challenges.
• Precise handling of long objects like pipes in uncertain conditions remains a significant challenge.

Purpose of the Study:

• To present a Unmanned Aerial Vehicle (UAV)-based aerial manipulation system for precise handling and transportation of long objects.
• To introduce a modular two-finger gripper design for enhanced scalability and reliability.
• To evaluate the performance of a SO-BFBEL controller against conventional methods for improved stability and precision.

Main Methods:

• Development of a modular two-finger gripper for UAV-based aerial manipulation.
• Implementation of a SO-BFBEL (Sliding Mode Based Feedback Error Learning) controller for enhanced stability and precision.
• Comparative analysis of the SO-BFBEL controller with DNN-MRFT-based PID and Fuzzy SMC controllers.

Main Results:

• The SO-BFBEL controller reduced position tracking error by up to 50% compared to conventional controllers.
• The SO-BFBEL controller demonstrated superior compensation for wind disturbances and battery discharge fluctuations.
• The SO-BFBEL controller contributed to conserving battery life during manipulation tasks.

Conclusions:

• The proposed UAV-based aerial manipulation system with a modular gripper and SO-BFBEL controller offers a scalable and reliable solution for precise long object handling.
• The SO-BFBEL controller significantly enhances system stability, precision, and robustness against environmental disturbances.
• The system improves operational efficiency and conserves battery life, providing cost-effective advantages for civil engineering and disaster response applications.