An Intelligent Hybrid Artificial Neural Network-Based Approach for Control of Aerial Robots

Abstract

In this work, a learning model-free control method is proposed for accurate trajectory tracking and safe landing of unmanned aerial vehicles (UAVs). A realistic scenario is considered where the UAV commutes between stations at high-speeds, experiences a single motor failure while surveying an area, and thus requires to land safely at a designated secure location. The proposed challenge is viewed solely as a control problem. A hybrid control architecture -an artificial neural network (ANN)-assisted proportional-derivative controller -is able to learn the system dynamics online and compensate for the error generated during different phases of the considered scenario: fast and agile flight, motor failure, and safe landing. Firstly, it deals with unmodelled dynamics and operational uncertainties and demonstrates superior performance compared to a conventional proportionalintegral-derivative controller during fast and agile flight. Secondly, it behaves as a fault-tolerant controller for a single motor failure case in a coaxial hexacopter thanks to its proposed sliding mode control theory-based learning architecture. Lastly, it yields reliable performance for a safe landing at a secure location in case of an emergency condition. The tuning of weights is not required as the structure of the ANN controller starts to learn online, each time it is initialised, even when the scenario changes -thus, making it completely model-free. Moreover, the simplicity of the neural network-based controller allows for the implementation on a low-cost low-power onboard computer. Overall, the real-time experiments show that the proposed controller outperforms the conventional controller.

Adaptive fault‐tolerant robust control based on radial basis function neural network for a class of mechanical systems with input constraints

“…Neural networks have successfully solved many practical problems in the fields of pattern recognition, 41 intelligent robots, 42 automatic control, 43 prediction 44 estimation, 45 etc. Neural network control provides new ideas to solve the control problems of complex, nonlinear and uncertain systems.…”

PencilNet: Zero-Shot Sim-to-Real Transfer Learning for Robust Gate Perception in Autonomous Drone Racing

“…As technical difficulties hinder the extension of batteries' capacity, there is a strong interest in increasing the speed of robots to expand both operating range and capabilities of the systems [6]. This is even more crucial for aerial systems, like multicopters [7], that are capable of reaching high speeds in a short time due to their agile nature [8]. Significant signs of progress for faster quadrotor flights have been made through addressing drone racing [9], previously thought of as merely an entertainment application.…”

Adaptive finite-time fault-tolerant control for quadrotor trajectory tracking systems under a novel smooth event-triggered mechanism

“…It is significant for maintaining satisfactory tracking performance when a fault occurs. Researchers have proposed fault-tolerant control techniques involving sliding mode controllers (SMC) [6] [7], extended state observers (ESOs) [8], and neural networks (NNs) [9]- [12]. To handle the uncertain inertia and sudden actuator failures, [13] proposed a novel integral sliding mode fault-tolerant controller (ISMFTC), and the adaptive techniques were used to compensate for unknown disturbances.…”