Autonomous Flight Research Articles

Filling the Gaps: Using Synthetic Low-Altitude Aerial Images to Increase Operational Design Domain Coverage.

A key necessity for the safe and autonomous flight of Unmanned Aircraft Systems (UAS) is their reliable perception of the environment, for example, to assess the safety of a landing site. For visual perception, Machine Learning (ML) provides state-of-the-art results in terms of performance, but the path to aviation certification has yet to be determined as current regulation and standard documents are not applicable to ML-based components due to their data-defined properties. However, the European Union Aviation Safety Agency (EASA) published the first usable guidance documents that take ML-specific challenges, such as data management and learning assurance, into account. In this paper, an important concept in this context is addressed, namely the Operational Design Domain (ODD) that defines the limitations under which a given ML-based system is designed to operate and function correctly. We investigated whether synthetic data can be used to complement a real-world training dataset which does not cover the whole ODD of an ML-based system component for visual object detection. The use-case in focus is the detection of humans on the ground to assess the safety of landing sites. Synthetic data are generated using the methods proposed in the EASA documents, namely augmentations, stitching and simulation environments. These data are used to augment a real-world dataset to increase ODD coverage during the training of Faster R-CNN object detection models. Our results give insights into the generation techniques and usefulness of synthetic data in the context of increasing ODD coverage. They indicate that the different types of synthetic images vary in their suitability but that augmentations seem to be particularly promising when there is not enough real-world data to cover the whole ODD. By doing so, our results contribute towards the adoption of ML technology in aviation and the reduction of data requirements for ML perception systems.

Unmanned Helicopter Airborne Fourier Transform Infrared Spectrometer Remote Sensing System for Hazardous Vapors Detection

The rapid development of unmanned aerial vehicles (UAVs) provides a new application mode for gas remote sensing. Compared with fixed observation and vehicle-mounted platforms, a Fourier transform infrared spectrometer (FTIR) integrated in the UAV can monitor chemical gases across a large area, can collect data from multiple angles in three-dimensional space, and can operate in contaminated or hazardous environments. The unmanned helicopter has a larger payload and longer endurance than the rotary-wing drone, which relaxes the weight, size and power consumption limitations of the spectrometer. A FTIR remote sensing system integrated in an unmanned helicopter was developed. In order to solve the data acquisition and analysis problem caused by vibration and attitude instability of the unmanned helicopter, a dual-channel parallel oscillating mirror was designed to improve the stability of the interferometer module, and a robust principal component analysis algorithm based on kernel function was used to separate background spectrum and gas features. The flight experiment of sulfur hexafluoride gas detection was carried out. The results show that the system operates stably and can collect and identify the target spectrum in real time under the motion and hovering modes of an unmanned helicopter, which has broad application prospects.

Study on Practical Design of Datalink in Interoperable UAV Systems

Uumanned aerial vehicle(UAV) systems have been used in various fields including industry and military. According to increasing the number of UAVs, the attention on interoperable UAV systems is increasing. In this paper, we propose the practical design of datalink in interoperable UAV systems. For practical design, we firstly review the operational scenarios in the interoperable UAV system. We then propose the system model of the datalink in interoperable UAV system. Consequently, the technical components such as the design of the network, the link management, the support of the multicast transmission, the support for autonomous mission and flight safety, and the datalink security are derived and reviewed for the practical design.

The application of artificial intelligence in aerospace engineering

In recent years, there has been considerable interest in applying Artificial Intelligence (AI) in the field of aerospace engineering. However, the existing literature on this topic is not sufficiently comprehensive. This paper is purposed to solve this problem by providing a thorough analysis and overview of the current state of AI in aerospace engineering. The paper is divided into four sections. Firstly, the use of AI in autonomous navigation and flight control is explored, focusing on advanced algorithms and sensor technologies that enable highly autonomous and efficient aircraft navigation. Secondly, the application of AI in image recognition and computer vision is discussed, highlighting its significance in remote sensing and aerospace component quality inspection. The third section examines the integration of AI in unmanned aerial vehicles (UAV), covering the control system and the utilization of machine learning techniques for improved UAV capabilities. Lastly, the paper explores the impact of AI on data analysis and prediction in the aerospace industry, encompassing weather forecasting, resource allocation, and decision-making processes. Finally, this paper gives a general overview of the nowadays application of AI in aerospace engineering.

Controlling a bio-inspired miniature blimp using a depth sensing neural-network camera

Miniature blimps are lighter-than-air vehicles which have become an increasingly common unmanned aerial system research platform due to their extended endurance and collision tolerant design. The UNSW-C bio-inspired miniature blimp consists of a 0.5 m spherical mylar envelope filled with helium. Four fins placed along the equator provide control over the three translatory axes and yaw rotations. A gondola attached to the bottom of the blimp contains all the electronics and flight controller. Here, we focus on using the UNSW-C blimp as a platform to achieve autonomous flight in GPS-denied environments. The majority of unmanned flying systems rely on GPS or multi-camera motion capture systems for position and orientation estimation. However, such systems are expensive, difficult to set up and not compact enough to be deployed in real environments. Instead, we seek to achieve basic flight autonomy for the blimp using a low-priced and portable solution. We make use of a low-cost embedded neural network stereoscopic camera (OAK-D-PoE) for detecting and positioning the blimp while an onboard inertia measurement unit was used for orientation estimation. Flight tests and analysis of trajectories revealed that 3D position hold as well as basic waypoint navigation could be achieved with variance (<0.1 m). This performance was comparable to that when a conventional multi-camera positioning system (VICON) was used for localizing the blimp. Our results highlight the potentially favorable tradeoffs offered by such low-cost positioning systems in extending the operational domain of unmanned flight systems when direct line of sight is available.

A hospitalization mechanism based immune plasma algorithm for path planning of unmanned aerial vehicles

Unmanned aerial vehicles (UAVs) and their specialized variants known as unmanned combat aerial vehicles (UCAVs) have triggered a profound change in the well-known military concepts and researchers from different disciplines tried to solve challenging problems of the mentioned vehicles. Path planning is one of these challenging problems about the UAV or UCAV systems and should be solved carefully by considering some optimization requirements defined for the enemy threats, fuel or battery usage, kinematic limitations on the turning and climbing angles in order to further improving the task success and safety of autonomous flight. Immune plasma algorithm (IP algorithm or IPA) modeling the details of a medical method gained popularity with the COVID-19 pandemic has been introduced recently and showed promising performance on solving a set of engineering problems. However, IPA requires setting the control parameters appropriately for maintaining a balance between the exploration and exploitation characteristics and does not design the particular treatment and hospitalization procedures by taking into account the implementation simplicity. In this study, IP algorithm was supported with a newly designed and realistic hospitalization mechanism that manages when an infected population member enters and discharges from the hospital. Moreover, the existing treatment schema of the algorithm was changed completely for improving the efficiency of the plasma transfer operations and removing the necessity of IPA specific control parameters and then a novel path planner called hospital IPA (hospIPA) was presented. For investigating the performance of hospIPA on solving path planning problem, a set of detailed experiments was carried out over twenty test cases belonging to both two and three-dimensional battlefield environments. The paths calculated by hospIPA were also compared with the calculated paths of other fourteen meta-heuristic based path planners. Comparative studies proved that the hospitalization mechanism making an exact discrimination between the poor and qualified solutions and modified treatment schema collecting the plasma being transferred by guiding the best solution give a tremendous contribution and allow hospIPA to obtain more safe and robust paths than other meta-heuristics for almost all test cases.

Evaluation of the Success of Simulation of the Unmanned Aerial Vehicle Precision Landing Provided by a Newly Designed System for Precision Landing in a Mountainous Area

Unmanned aerial vehicle technology is the most advanced and helpful in almost every area of interest in human work. These devices become autonomous and can fulfil a variety of tasks, from simple imaging and obtaining data to search and rescue operations. The most challenging environment for search and rescue operations is the mountainous area. This article is devoted to the theoretical description and simulation tests of a prototype method of landing the light and the medium-weight UAVs used as supplementary devices for SAR (search and rescue) and HEMS (helicopter emergency medical service) in hard-to-reach mountainous terrains. The autonomous flight of a UAV in mountainous terrain has many specifics, and it is usually performed according to predetermined map points (pins) uploaded directly into the control software of the UAV. It is necessary to characterise each point flown on the chosen flight route line in advance and therefore to know its exact geographical coordinates (longitude, latitude and height of the point above the terrain), and the control system of UAV must react to the change in the weather and other conditions in real time. Usually, it is difficult to make this forecast with sufficient time in advance, mainly when UAVs are used as supplementary devices for the needs of HEMS or MRS (mountain rescue service). The most challenging phase is the final approach and landing of the UAV, especially if a loss of GNSS (global navigation satellite system) signal occurs, like in the determined area of the Little Cold Valley in the Slovak High Tatras—which is infamous for the widespread loss of GNSS signals or communication/controlling connection between the UAV and the pilot-operator at the operational station. To solve the loss of guidance, a new method for guiding and controlling the UAV in its final approach and landing in a determined area is tested. An alternative landing navigation system for UAVs in a specific mountainous environment—the authors’ designed frequency Doppler landing system (FDLS)—is briefly described but thoroughly tested with the help of artificial intelligence. An estimation of dynamic stability is used based on the time recording of the current position of the UAV, with the help of a frequency-modulated or amplitude-modulated signal based on the author’s prototype of a precision landing system designed for mountainous terrain. This solution could overcome the problems of GNSS signal loss. The presented research primarily evaluates the success of the simulation flights for the supplementary UAV. The success of navigating the UAV to land in the mountainous environment at an exact landing point using the navigation signals from the FDLS was evaluated at more than 95%.

Assessment of the effectiveness of stall delay and tip loss corrections for the simulation of small propeller performance with Virtual Blade Model

The powertrain efficiency of most small unmanned air vehicles (SUAVs) depend on the aerodynamic performance of multiple propellers, which must be modelled before its optimisation. To predict propeller performance while capturing interaction effects not accountable for by momentum theories, the Virtual Blade Model (VBM) can be used at much lower costs than conventional Computational Fluid Dynamics (CFD). However, the accuracy of VBM depends on aerofoil polar data that often misrepresent rotational and three-dimensional effects that influence small propellers. In an attempt to expand the capabilities of VBM, it is a common practice to introduce correction models for aerofoil coefficients in its formulation, but the actual effectiveness of such approach for small propeller analysis remains unassessed. In this paper, VBM in OpenFOAM was extended with stall delay and tip loss models, while aerodynamic databases at low Reynolds numbers were built with XFOIL and extrapolated to the full angle of attack range. The accuracy of three VBM versions was validated against wind tunnel measurements of two off-the-shelf small propellers. For the lower-pitch propeller, thrust and power at low advance ratios and efficiency at high advance ratios were significantly improved by the stall delay model. But for the higher-pitch propeller, thrust and power were overpredicted with extended versions of VBM under most operating conditions, which is attributable to an excessive shift of effective angles of attack to the post-stall region by the stall delay model and to uncertainties in extrapolated polar data. The results suggest that without reliable procedures for obtaining polar data, correction models should be implemented in VBM only for the simulation of low-pitch propellers operating at low advance ratios, given their avoidance of the effective angle of attack range where XFOIL and extrapolation methods are mainly expected to fail in predicting lift behaviour.

Geofencing Motion Planning for Unmanned Aerial Vehicles Using an Anticipatory Range Control Algorithm

This paper presents a range control approach for implementing hard geofencing for unmanned air vehicles (UAVs), and especially remotely piloted versions (RPVs), via a proposed anticipatory range calculator. The approach employs turning circle intersection tests that anticipate the fence perimeter on approach. This ensures the vehicle turns before penetrating the geofence and remains inside the allowable operational airspace by accounting for the vehicles’ turning dynamics. Allowance is made for general geozone shapes and locations, including those located at the problematic poles and meridians where nonlinear angle mapping is dealt with, concave geozones, narrow corners with acute internal angles, and transient turn dynamics. The algorithm is shown to prevent any excursions using a high-fidelity simulation of a small remotely piloted vehicle. The algorithm relies on a single tuning parameter which can be determined from the closed-loop rise time in the aircraft’s roll command tracking.

Learning-Observer-Based Adaptive Tracking Control of Multiagent Systems Using Compensation Mechanism

This paper focuses on the adaptive consensus tracking problem for nonlinear multiagent systems (MASs) with unmeasurable state and input saturation. For unmeasurable state and nonlinearities, a learning-observer inspired by the idea of extended state observer (ESO) and the learning ability of neural network (NN) is established to observe the reconstructed system. In view of input saturation, a nonlinear auxiliary system with the same order is constructed to obtain the compensation signals. Different from the previous auxiliary system, the proposed method simultaneously compensates the change of nonlinearity caused by input saturation. Subsequently, a consensus tracking error is designed to ensure that the follower tracks the neighbor compensated output. Moreover, an adaptive consensus tracking control method based on learning-observer and compensation mechanism is proposed under the backstepping framework. The stability analysis is given by virtue of the Lyapunov theory. Finally, the effectiveness and superiority of the proposed control algorithm are verified through the simulation of unmanned air vehicles (UAVs) with one virtual leader and four followers.

Multi-sensor data fusion for autonomous flight of unmanned aerial vehicles in complex flight environments

The flight environment of unmanned aerial vehicles faces various challenges. To effectively navigate and perform tasks, they need to effectively integrate multiple sensors. This study applies the adaptive weighted average method, combined with data from global positioning system, inertial measurement unit, three-dimensional optical detection and ranging, and uses linear Kalman filtering to smooth the merged velocity data. High-order B-spline curves for route planning and applying flight constraint formulas to better adapt are used to the dynamics of unmanned aerial vehicles. The research results indicated that the improved adaptive weighting algorithm had high comprehensive performance for multi-sensor data fusion, with the highest accuracy, robustness, real-time performance, and consistency of 94.2%, 93.7%, 100%, and 95.6%, respectively. The flight path lengths planned by the A* algorithm and higher-order B-spline curve were 15.7 and 16.3 m, respectively, and the flight time was 8.2 and 7.1 s, respectively. The flight path planned by higher-order B-spline curve was further away from obstacles. The use of adaptive weighted fusion and linear Kalman filtering facilitates the fusion of multi-sensor data, and autonomous flight routes planned using high-order B-spline curves can also meet the needs of unmanned aerial vehicle flight in complex flight environments.

Fluid and aero-acoustic analysis of internal bays: Geometric length-to-depth effects

Weapon bays are essential for any bomber, Next Generation Fighter Aircraft, and unmanned combat air vehicles (UCAVs). As flow approaches the leading edge of such cavity, it separates and forms a shear layer. The structure of bays and residing stores thus become prone to intense acoustic and aerodynamic loading– a major design concern. Cavity flow shows significant dependence on its shape. This makes flow actuation possible by only changing the geometric variable of length-to-depth (L/D) ratio. This study compares two bays with an L/D ratio of 5 (deep) and 10 (shallow) in terms of acoustics and affiliated aerodynamic drag. A transient RANS formulation obtained overall sound pressure levels (OASPL) on bay ceilings. An integral-based method was also used to predict far-field noise propagation due to flow fields of both cavities. It is argued that at Mach 0.6, the character of bays changes remarkably when its L/D is altered. Deeper bays show a smoother pressure profile and significantly less drag while operating in shear mode. Shallow bays, on the other hand, show bluff body type drag with uneven pressure distribution, having the potential of causing downwash on residing stores. At the same time, acoustic loads on the ceiling and acoustic signatures in the far-field are comparable for both cavities.

Computational Validation and Assessment of Large Amplitude Transverse Gust Physics

Rotary-wing vehicles, in particular smaller, lighter unmanned and urban air vehicles in urban and shipboard settings, will operate in low-speed flight conditions that are dominated by strong transient aerodynamics. Understanding the physics of these transient aerodynamics, specifically large amplitude transverse gusts and the resulting vehicle response, is crucial to the successful development and certification of safe air vehicles that operate in these environments. A high-fidelity computational study, including validation with experiments, explores sharp-edged or step transverse gusts where the gust velocity induces nonlinear behavior caused by flow separation. The behavior of the maximum lift and leading edge vortex behavior with the gust ratio is presented. Gust responses are observed to depart from Küssner's theory when the leading edge vortex first forms as a distinct feature and breaks away from the wing, resulting in flow nonlinearities. Traditional linear indicial admittance techniques are shown to no longer be valid to predict gust responses when the gust velocity approaches the vehicle flight speed.

Paper] Deep Learning based Hierarchical Object Detection System Adopting a Depth Correction Scheme for High-Resolution Aerial Images Towards Realization of Autonomous UAV Flight

Paper] Deep Learning based Hierarchical Object Detection System Adopting a Depth Correction Scheme for High-Resolution Aerial Images Towards Realization of Autonomous UAV Flight

Fully Actuated System Approach Based Prescribed-Time Fault-Tolerant Formation Control for Unmanned Helicopters Under Fixed and Switching Topologies

Fully Actuated System Approach Based Prescribed-Time Fault-Tolerant Formation Control for Unmanned Helicopters Under Fixed and Switching Topologies

Design of Unmanned Helicopter Distributed Electric Tail Rotor Controller Based on Improved Active Disturbance Rejection Control

Design of Unmanned Helicopter Distributed Electric Tail Rotor Controller Based on Improved Active Disturbance Rejection Control

Estimation of airship states and model uncertainties using nonlinear estimators

This Airships are lighter than air vehicles and due to their growing number of applications, they are becoming attractive for the research community. Most of the applications require an airship autonomous flight controller which needs an accurate model and state information. Usually, airship states are affected by noise and states information can be lost in the case of sensor's faults, while airship model is affected by model inaccuracies and model uncertainties. This paper presents the application of nonlinear and Bayesian estimators for estimating the states and model uncertainties of neutrally buoyant airship. It is considered that minimum sensor measurements are available, and data is corrupted with process and measurement noise. A novel lumped model uncertainty estimation approach is formulated where airship model is augmented with six extra state variables capturing the model uncertainty of the airship. The designed estimator estimates the airship model uncertainty along with its states. Nonlinear estimators, Extended Kalman Filter and Unscented Kalman Filter are designed for estimating airship attitude, linear velocities, angular velocities and model uncertainties. While Particle filter is designed for the estimation of airship attitude, linear velocities and angular velocities. Simulations have been performed using nonlinear 6-DOF simulation model of experimental airship for assessing the estimator performances. 1−𝜎 uncertainty bound and error analysis have been performed for the validation. A comparative study of the estimator's performances is also carried out.

Coupling an Autonomous UAV With a ML Framework for Sustainable Environmental Monitoring and Remote Sensing

Many countries face problems in monitoring plant problems and monitoring large environments, as until now, there is no accurate means of aerial monitoring through which concerned parties can benefit from watersheds, monitor large agricultural areas, and make wise environmental decisions regarding them. This paper describes a pioneering approach to develop smart agriculture using multimission drones equipped with dual cognitive modules (brains) that are powered by a machine learning (ML) framework. The first brain uses deep reinforcement learning (DRL) principles to enable autonomous flight, allowing drones to navigate complex agricultural terrain with agility and flexibility. The second brain is responsible for precise and crucial agricultural tasks: counting trees, detecting water locations, and observing and analyzing plants using the Faster R‐CNN algorithm. The system is linked with the ground station for control and command, as well as it includes an Internet of Things (IoT) infrastructure equipped with sensors that collect soil parameters, which then get sent via 5G Wi‐Fi. The dual architecture of the drones, combined with the ground‐level IoT system, creates a comprehensive framework that not only enhances agricultural technologies but also aligns with environmental conservation goals, embodying a paradigm shift towards a greener and more sustainable future. Obtained results show reasonable results with an accuracy rate of 98%.

Autonomous Flight Strategy Selection and Interval Maintenance for Aircraft With Unknown Flight Intentions

Autonomous Flight Strategy Selection and Interval Maintenance for Aircraft With Unknown Flight Intentions

Automated Flight Technology for Integral Path Planning and Trajectory Tracking of the UAV

In view of the problem that path planning and trajectory tracking are rarely solved simultaneously in the current research, which hinders their practical implementation, this paper focuses on enhancing the autonomous flight planning capability of unmanned aerial vehicles (UAVs) by investigating integrated path planning and trajectory tracking technologies. The autonomous flight process is divided into two sub-problems: waypoint designing/optimizing and waypoint tracking. Firstly, an improved DB-RRT* algorithm is proposed for waypoint planning to make the algorithm have higher planning efficiency, better optimization results, and overcome the defects of accidental and low reliability of single RRT* planning results. Secondly, the scheme of “offline design + online flight” is adopted to lead the UAV to fly online according to the waypoints’ instructions by using the sliding mode guidance based on angle constraint with finite-time convergence so that it can fly to the destination autonomously. In order to check the performance of the proposed algorithm, a variety of simulations are conducted to verify the feasibility of the proposed algorithm.