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

Abstract

Abstract Aerial manipulation provides adaptable solutions for executing tasks in constrained environments, particularly in civil engineering and disaster response. This paper presents a UAV-based aerial manipulation system for the precise handling and transportation of long objects, such as pipes, in uncertain conditions. Compared to single- or dual-arm systems, which are difficult to scale and maintain, the proposed design features a modular two-finger gripper, enhancing scalability and reliability while reducing mechanical complexity. To address challenges such as positional drift, which can compromise mission success, the system employs a SO-BFBEL controller controller to enhance stability and precision and it is compared with DNN-MRFT-based PID and Fuzzy SMC controller. Experimental results demonstrate that the SO-BFBEL controller reduces the position tracking error up to 50% and compensates for wind disturbances and battery discharge fluctuations more effectively than conventional methods. Additionally, the SO-BFBEL controller can help to conserve battery life during manipulation phases which can boost the operational efficiency without incurring any additional costs.

Keywords

• Quadrotor
• Flight control
• Fuzzy neural network
• Two finger gripper
• Wind disturbance
• Brain emotional learning