Output Feedback Control of a Quadrotor UAV Using Neural Networks

Quadrotor unmanned aerial vehicles (UAVs) have been actively used in various fields. However, only the altitude and the attitude in three degrees of freedom can be independently controlled since quadrotor UAVs are underactuated systems. A quad tilt rotor UAV solves the problem of an underactuated system in a general quadrotor UAV. The quad tilt rotor UAV can control both position and attitude independently by tilting the directions of the propellers. However, the flight control system in a wide range of attitudes has not been discussed yet, for example a UAV can fly and hover with a 90 pitch angle, and it can even flip over when the thrust direction is tilted in a wide enough range. In this paper, we present the attitude transition flight control system for pitch angles ranging from 0 to 90 since flight conditions with a 90 pitch angle significantly differs from that in a conventional quadrotor UAV flight. We construct an adequate control system for a flight with a wide range of attitude conditions.

Lyapunov based trajectory tracking controller for a quadrotor UAV with nonholonomic constraints

Accurate and precise trajectory tracking is crucial for unmanned aerial vehicles (UAVs) to operate in disturbed environments. This paper presents a novel tracking hybrid controller for a quadrotor UAV that combines the robust adaptive neuro-fuzzy inference system (ANFIS) and particle swarm optimization (PSO) algorithm. The ANFIS-PSO controller is implemented to govern the behavior of three degrees of freedom quadrotor UAV. The ANFIS controller allows controlling the movement of UAV to track a given trajectory in a 2D vertical plane. The PSO algorithm provides an automatic adjustment of the ANFIS parameters to reduce tracking error and improve the quality of the controller. The results showed perfect behavior for the control law to control a UAV trajectory tracking task. To show the effectiveness of the intelligent controller, simulation results are given to confirm the advantages of the proposed control method, compared with ANFIS and PID control methods.

Accurate trajectory prediction is essential for ensuring flight safety, autonomous navigation, and efficient control of unmanned aerial vehicles (UAVs). However, existing neural network approaches often focus on a single platform, limiting their generalizability across UAVs with different dynamics. This study evaluates three neural architectures for spatiotemporal trajectory prediction using high-frequency flight data. Two UAV configurations, such as a fixed-wing (Cessna) and a multirotor (Phantom), were simulated in XPlane 11 under identical environmental conditions. Flight data of latitude, longitude, and altitude sampled at 80 Hz were preprocessed using a sliding-window segmentation method to preserve temporal dependencies. Backpropagation Neural Network (BPNN), Recurrent Neural Network (RNN), and Long Short-Term Memory (LSTM) models were trained using standardised preprocessing and tuned hyperparameters. Results show that recurrent models outperform the feedforward BPNN in capturing temporal patterns. For the Cessna dataset, the RNN achieved balanced accuracy with a mean-squared error (MSE) of <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$5.5 \times 10^{-8}$</tex>, while the LSTM obtained the lowest MSE (<tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$2.7 \times 10^{-8}$</tex>). In the more dynamic Phantom dataset, the BPNN remained competitive (MSE <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$\approx 1.8 \times 10^{-7}$</tex>), whereas RNN and LSTM maintained superior altitude prediction. These results highlight the suitability of lightweight recurrent architectures for real-time UAV trajectory prediction.

: Unmanned aerial vehicles (UAVs) are critical components of the future naval forces. UAV control and monitoring with autonomous operation will become an absolute necessity and adaptive cooperation of vehicles is the only practical alternative. The objective of this project is to develop and evaluate new methodologies for cooperative (formation) control of multiple unmanned air vehicles. The goal is to have multiple UAVs working together as a group. Instead of separately assigning distinct tasks to each vehicle, the operator would assign tasks to the UAV group, which then determines the best way to accomplish each task, freeing the operator to maintain surveillance over the entire operation. In this project we investigated Path Tracking and obstacle avoidance of UAVs using fuzzy logic method. Algorithms for close formation control of multi-UAVs are developed and simulated. We also investigated fault-tolerant control of single UAVs by neuro-adaptive method. Detailed description of this method is provided in this document. The project has supported S graduate students with 9 technical papers published.

This paper presents the control of an indoor unmanned aerial vehicle (UAV) using multi-camera visual feedback. For the autonomous flight of the indoor UAV, instead of using onboard sensor information, visual feedback concept is employed by the development of an indoor flight test-bed. The indoor test-bed consists of four major components: the multi-camera system, ground computer, onboard color marker set, and quad-rotor UAV. Since the onboard markers are attached to the pre-defined location, position and attitude of the UAV can be estimated by marker detection algorithm and triangulation method. Additionally, this study introduces a filter algorithm to obtain the full 6-degree of freedom (DOF) pose estimation including velocities and angular rates. The filter algorithm also enhances the performance of the vision system by making up for the weakness of low cost cameras such as poor resolution and large noise. Moreover, for the pose estimation of multiple vehicles, data association algorithm using the geometric relation between cameras is proposed in this paper. The control system is designed based on the classical proportional-integral-derivative (PID) control, which uses the position, velocity and attitude from the vision system and the angular rate from the rate gyro sensor. This paper concludes with both ground and flight test results illustrating the performance and properties of the proposed indoor flight test-bed and the control system using the multi-camera visual feedback.

Hybrid modeling and control issues provide a flexible powerful tool for Unmanned Aerial Vehicles (UAVs) as complex control platforms. This paper aims to briefly review recent results in this area. Firstly, existing works on hybrid modeling and control of a single UAV are reviewed. Then multiple UAVs control in a synchronously cooperative task planning, which gives much more maneuverability, has been surveyed in the hybrid framework. Formation control of UAVs is a typical cooperative strategy with considerable applications in multi-agent aerial robotics area and can leverage limited abilities of single UAVs into complicated tasks of group missions. There are a few works in this field which we briefly reviewed their results and explored the possible future directions of these research activities.

Robust fuzzy tracking control of a quad-rotor unmanned aerial vehicle based on sector linearization and interval matrix approaches

Unmanned vehicles (UAVs) are technological tools whose application is carried out in the civil and military areas for surveillance, recognition and intelligence tasks. Given its versatility and features, in this paper the implementation of an autopilot is determined, after an analysis of commercial models. Said device must comply with technical requirements for the automatic flight control of a medium-sized UAV, considering the operation scenarios in the northern border of Ecuador. In addition, the parameters that intervene in the control loops and the relevant simulations are specified. The results obtained from the experimental flights of the aircraft are reflected in graphs, where the stability of the plant properly calibrated for autonomous flight is verified.

An intelligent PID controller based on RBF neural network was designed for longitudinal attitude control of a small unmanned aerial vehicle (UAV) according to the feature of complexity and nonlinear; RBF neural network is used to real-time update the PID parameters on line by its learning function for achieving the longitudinal channel control of the UAV; Simulation results show that the design of the RBF neural network PID controller has better control performance and it better satisfies the requirements of the longitudinal channel flight control of small UAV.

NFTSM-based Fault Tolerant Control for Quadrotor Unmanned Aerial Vehicle with Finite-Time Convergence

Traditional bulky and complex control devices such as remote control and ground station cannot meet the requirement of fast and flexible control of unmanned aerial vehicles (UAVs) in complex environments. Therefore, a data glove based on multi-sensor fusion is designed in this paper. In order to achieve the goal of gesture control of UAVs, the method can accurately recognize various gestures and convert them into corresponding UAV control commands. First, the wireless data glove fuses flexible fiber optic sensors and inertial sensors to construct a gesture dataset. Then, the trained neural network model is deployed to the STM32 microcontroller-based data glove for real-time gesture recognition, in which the convolutional neural network-Attention mechanism (CNN-Attention) network is used for static gesture recognition, and the convolutional neural network-bidirectional long and short-term memory (CNN-Bi-LSTM) network is used for dynamic gesture recognition. Finally, the gestures are converted into control commands and sent to the vehicle terminal to control the UAV. Through the UAV simulation test on the simulation platform, the average recognition accuracy of 32 static gestures reaches 99.7%, and the average recognition accuracy of 13 dynamic gestures reaches 99.9%, which indicates that the system’s gesture recognition effect is perfect. The task test in the scene constructed in the real environment shows that the UAV can respond to the gestures quickly, and the method proposed in this paper can realize the real-time stable control of the UAV on the terminal side.

Ensuring UAV safety: A deep learning-based fault-tolerant approach with control barrier functions

Since one GRS used for Control and Non-Payload Communication (CNPC) of Unmanned Aerial Vehicle (UAV) cannot support all operational ranges of an UAV, the handover technology of CNPC is essential to operate an UAV in a wide range. However, the CNPC physical layer standard of Radio Technical Commission for Aeronautics (RTCA) does not have a reference signal for measuring the signal strength of a neighboring base station used for handover decision in an existing mobile communication system. Therefore, in this paper, we propose a method to transmit measurement signals for stable handover of CNPC. The proposed method starts transmission of the measurement signal when the UAV enters the handover zone for the preventing waste of power and generates the measurement signal using the same data that is transmitted through the currently connected Ground Radio Station (GRS) for the preventing packet loss during the handover operation.

The control performance of quadrotor unmanned aerial vehicles (UAVs) in complex environments can be affected by external disturbances and other factors. In this paper, an adaptive neural network backstepping controller based on the barrier Lyapunov function (BLF) is designed for a quadrotor UAV with internal uncertainties, input–output constraints and external disturbances. Radial basis function neural networks are used to approximate the uncertainties in the dynamic model of the UAV, while the minimum parameter learning method is combined to accelerate the adjustment speed of neural network weights. A robust term is designed to balance the total system disturbance and improve the anti-interference performance. The BLF is used to handle the output constraint so that the constrained parameters cannot break the predefined constraints. An auxiliary system is introduced to solve input saturation and avoid the dependence of tracking error on the input amplitude in the method of approximating input saturation using the smoothing function. The stability of the control system is demonstrated by the Lyapunov method. The simulation results show that the proposed method has high tracking accuracy compared with the backstepping dynamic surface control method, and the input and output are in the predefined range.