Guidance Of Unmanned Aerial Vehicles Research Articles

Air–ground cooperative multi-target searching under an unknown urban environment

The collaboration of heterogeneous multiple robots has been shown to be capable of significantly enhancing the system redundancy, autonomy, robustness, and so on. However, realizing the collaboration with specific tasks in practice often requires the development of sophisticated mechanisms which are envisioned to exploit distinct benefits of the heterogeneous platforms. Thus, we propose a novel air–ground cooperative framework in this paper for the task of multi-target searching under an unknown urban environment. In particular, a group of unmanned aerial vehicles (UAVs) is employed to operate above the urban area to provide surveillance from a global perspective. Under the guidance of UAVs, multiple teams of unmanned ground vehicles (UGVs) are deployed to conduct the target searching missions. The UAVs’ and UGVs’ searching strategies are devised correspondingly leveraging on their own advantageous features. Finally, an ingenious integration of UAVs’ and UGVs’ searching operations is established by a notion of the upper confidence bound. Simulation results are provided to demonstrate the effectiveness of our approach.

Adaptive Collision Avoidance for Multiple UAVs in Urban Environments

The increasing number of unmanned aerial vehicles (UAVs) in low-altitude airspace is seriously threatening the safety of the urban environment. This paper proposes an adaptive collision avoidance method for multiple UAVs (mUAVs), aiming to provide a safe guidance for UAVs at risk of collision. The proposed method is formulated as a two−layer resolution framework with the considerations of speed adjustment and rerouting strategies. The first layer is established as a deep reinforcement learning (DRL) model with a continuous state space and action space that adaptively selects the most suitable resolution strategy for UAV pairs. The second layer is developed as a collaborative mUAV collision avoidance model, which combines a three-dimensional conflict detection and conflict resolution pool to perform resolution. To train the DRL model, in this paper, a deep deterministic policy gradient (DDPG) algorithm is introduced and improved upon. The results demonstrate that the average time required to calculate a strategy is 0.096 s, the success rate reaches 95.03%, and the extra flight distance is 26.8 m, which meets the real-time requirements and provides a reliable reference for human intervention. The proposed method can adapt to various scenarios, e.g., different numbers and positions of UAVs, with interference from random factors. The improved DDPG algorithm can also significantly improve convergence speed and save training time.

Adaptive Dynamic Occupancy Guidance for Air Combat of UAV

The occupancy guidance of Unmanned Aerial Vehicle (UAV) is one of the inevitable stages in the future air combat. Considering both constraints of missile launch condition and UAV flight performance, this paper established a multi-objective optimization model of occupancy guidance for UAV autonomous pursuit of enemy aircraft. The distance, angle and speed of occupancy guidance state are used as optimization variables to construct the advantage evaluation function. Based on the traditional heuristic algorithm, the Gradient Descent-Truncated Symbiotic Organizations Search (GDT-SOS) is designed to achieve rapid convergence of fitness values. By both numerical simulation and field test, the experiments demonstrated that UAV can autonomously achieve continuous, rapid and advantageous occupancy guidance under different dynamic initial conditions, which verifies that GDT-SOS has more effective than the comparison algorithm.

UAV4PE: An Open-Source Framework to Plan UAV Autonomous Missions for Planetary Exploration

Autonomous Unmanned Aerial Vehicles (UAV) for planetary exploration missions require increased onboard mission-planning and decision-making capabilities to access full operational potential in remote environments (e.g., Antarctica, Mars or Titan). However, the uncertainty introduced by the environment and the limitation of available sensors has presented challenges for planning such missions. Partially Observable Markov Decision Processes (POMDPs) are commonly used to enable decision-making and mission-planning processes that account for environmental, perceptional (extrinsic) and actuation (intrinsics) uncertainty. Here, we propose the UAV4PE framework, a testing framework for autonomous UAV missions using POMDP formulations. This framework integrates modular components for simulation, emulation, UAV guidance, navigation and mission planning. State-of-the-art tools such as python, C++, ROS, PX4 and JuliaPOMDP are employed by the framework, and we used python data-science libraries for the analysis of the experimental results. The source code and the experiment data are included in the UAV4PE framework. The POMDP formulation proposed here was able to plan and command a UAV-based planetary exploration mission in simulation, emulation and real-world experiments. The experiments evaluated key indicators such as the mission success rate, the surface area explored and the number of commands (actions) executed. We also discuss future work aimed at improving the UAV4PE framework, and the autonomous UAV mission planning formulation for planetary exploration.

Separate Guidance and Control Design for Autonomous Path-following

This paper addresses the separate guidance and control design for an unmanned aerial vehicle (UAV) to follow the desired path with arbitrary curvature. The proposed design is appealing since it neither requires the information about the path curvature to be implemented nor necessitates tracking a moving virtual target. Such a feature makes our design simple and elegant. We first show that it is sufficient to nullify the UAV's perpendicular distance with respect to the path in order for it to track and follow the desired path and that there is no restriction on the UAV's initial configuration. This makes the proposed design globally valid. Next, we augment the UAV's guidance system with an autopilot obtained by approximating its higher-order dynamics as a second-order transfer function. We demonstrate the efficacy of the proposed design through simulations for various scenarios.

Emergency Planning for UAV-Controlled Crowd Evacuations

One of the most challenging problems during disasters involving crowds is pedestrian evacuation. It is very important in such situations to improve survival rates by getting all or most of the people out in the shortest possible time. Technological intervention through augmenting the evacuation process using an unmanned aerial vehicle (UAV) exhibits great potential in improving survival rates, but the exploration of this method is still in its infancy. Therefore, this study explores the potential of utilizing UAVs for crowd management during emergency evacuations. We conducted a rigorous study, using a simulation model featuring four UAVs and differing numbers of pedestrians, with use of two UAV guidance approaches: partial guidance and full guidance. The experimental results suggest that exploiting UAVs in crowd evacuation and following the partial guidance approach can lead to a more efficient evacuation process.

THE MODELS OF AUTO TARGETING STRIKE UAV AT THE MOVING TARGET

A special kind of use of UAVs as transport is their use to deliver the destruction means to the target on the battlefield, i.e. as strike unmanned aerial vehicles (UAVs), capable to inflict missile and bomb strikes on the enemy.
 Of particular interest are such drones, which do not require constant aiming at the target by the flight operator (ground pilot) with the help of a video transmitting system, but are targeted independently. Such UAVs have onboard a target sighting sensor (coordinator), with which the onboard control system determines the angular position of the target relative to the UAV. It is necessary to determine the most optimal targeting algorithms to be used in onboard control systems.
 UAVs due to the successful combination of cheap goods, technological availability and availability of other technical indicators (load capacity, efficiency and duration of flight, maneuverability) can be used as vehicles for cargo delivery. For example, UAVs are used to deliver ammunition and medicine to victims.
 The mechanism of the primary notification to the on-board control system of the UAV of the coordinates of the target to be hit, ie the task of "primary targeting", remains not completely solved by the technical and organizational task when using strike UAVs. Targeting methods using the on-board video broadcast system and direct commands of the ground pilot require, firstly, the installation of appropriate payload on board with an understanding of the loss of this equipment in combat use of UAVs, and secondly, effective avoidance of radio suppression by the enemy.
 The article presents the results of modeling with the help of Simulink models of the processes of guidance of a strike UAV with targeting on a moving aim by two methods: direct guidance and dynamic bias. The trajectories of UAV flights during diving to the target were obtained. The changes in the main working parameters during the guidance process were obtained as well. The conclusion about the advantage of the method of dynamic bias is made.

Drone Deep Reinforcement Learning: A Review

Unmanned Aerial Vehicles (UAVs) are increasingly being used in many challenging and diversified applications. These applications belong to the civilian and the military fields. To name a few; infrastructure inspection, traffic patrolling, remote sensing, mapping, surveillance, rescuing humans and animals, environment monitoring, and Intelligence, Surveillance, Target Acquisition, and Reconnaissance (ISTAR) operations. However, the use of UAVs in these applications needs a substantial level of autonomy. In other words, UAVs should have the ability to accomplish planned missions in unexpected situations without requiring human intervention. To ensure this level of autonomy, many artificial intelligence algorithms were designed. These algorithms targeted the guidance, navigation, and control (GNC) of UAVs. In this paper, we described the state of the art of one subset of these algorithms: the deep reinforcement learning (DRL) techniques. We made a detailed description of them, and we deduced the current limitations in this area. We noted that most of these DRL methods were designed to ensure stable and smooth UAV navigation by training computer-simulated environments. We realized that further research efforts are needed to address the challenges that restrain their deployment in real-life scenarios.

Impact of Submesoscale Eddies on the Transport of Suspended Matter in the Coastal Zone of Crimea Based on Drone, Satellite, and In Situ Measurement Data

High-resolution measurements by unmanned aerial vehicles (UAVs), satellite and in-situ data are used to study the optical structure of coastal submesoscale eddies and their influence on the transport of total suspended matter (TSM). It is shown that submesoscale cyclones cause intense cross-shelf TSM fluxes and its subsequent accumulation in their cores, similar to what is observed in large mesoscale anticyclones. UAV measurements make it possible to observe intensive dynamic processes on the periphery of eddies, seen as a periodic structure with scales of 20–100 m; the spiral structure of TSM transport; and convergence of TSM in the cores of cyclones. Using UAV guidance, detailed measurements of the distribution of hydrological and hydro-optical characteristics of one of these eddies, with a diameter of about 2 km, were obtained during a scientific cruise in October 2019. Warm brackish water with a high TSM content and the descent of all isosurfaces were observed in the core of this cyclone, located near the topographic slope. We suggest that the probable reason for the convergence in such submesoscale cyclones is the interaction of radial currents with the topographic slope, which causes downwelling in the coastal part of eddies.

UAV Path Planning Based on Multi-Layer Reinforcement Learning Technique

Unmanned aerial vehicles (UAVs) have been widely used in many applications due to its small size, swift mobility and low cost. Therefore, the study of guidance, navigation and control (GNC) system of UAV has becoming a popular research direction. Path planning plays an important role in the GNC system. In this paper, a multi-layer path planning algorithm based on reinforcement learning (RL) technique is proposed. Compared to the classic Q-learning, the proposed multi-layer algorithm has a distinct advantage that it collects both global and local information which greatly improves overall performance. The proposed RL algorithm has two layers, the higher layer deals with the local information, which could be considered as a short-term strategy. The lower layer deals with the global information, which could be considered as a long-term strategy. Both the higher layer and lower layer are working in harmony to plan a collision-free path. B-spline curve approach is applied for on-line path smoothing. Simulation results in different scenarios prove the effectiveness of multi-layer Q-learning algorithm.

Integrity Analysis for GPS-Based Navigation of UAVs in Urban Environment

The increasing use of Unmanned Aerial Vehicles (UAVs) in safety-critical missions in both civilian and military areas demands accurate and reliable navigation, where one of the key sources of navigation information is presented by Global Navigation Satellite Systems (GNSS). In challenging conditions, for example, in urban areas, the accuracy of GNSS-based navigation may degrade significantly due to user-satellite geometry and obscuration issues without being noticed by the user. Therefore, considering the essentially dynamic rate of change in this type of environment, integrity monitoring is of critical importance for understanding the level of trust we have in positioning and timing data. In this paper, the dilution of precision (DOP) coefficients under nominal and challenging conditions were investigated for the purpose of integrity monitoring in urban environments. By analyzing positioning information in a simulated urban environment using a software-based GNSS receiver, the integrity monitoring approach based on joint consideration of GNSS observables and environmental parameters has been proposed. It was shown that DOP coefficients, when considered together with a number of visible satellites and cut-off elevations specific to the urban environment carry valuable integrity information that is difficult to get using existing integrity monitoring approaches. This has allowed generating indirect integrity measures based on cut-off elevation and satellite visibility that can be used for UAV path planning and guidance in urban environments.

Analytical Solution for Nonlinear Three-Dimensional Guidance With Impact Angle and Field-of-View Constraints

An analytical 3-D guidance law with impact angle and field-of-view (FOV) constraints considering nonlinear coupled dynamics is proposed. As a stepping stone, the guidance model is transformed to a set of nonlinear differential equations in terms of the relative range variable. To meet the desired impact angles in the pitch and yaw planes, two cubic polynomials including eight coefficients are developed with respect to the relative range for creating reference line-of-sight (LOS) profiles. The unknown coefficients are explicitly solved by initial and terminal conditions on the LOS angles and LOS rates in the mutually orthogonal planes. Then, the analytical 3-D impact angle guidance (IAG) law is derived via formulating the second-order LOS dynamics of the transformed model. Moreover, the relation between the seeker's look angle and the reference LOS profiles is developed, such that the achievable impact angle set can be obtained to handle the FOV limit. Numerical simulations with comparison study and a realistic model are conducted to verify effectiveness and robustness of the guidance law. Its feasibility is additionally validated by applying it to the guidance of unmanned aerial vehicles landing on surface moving carriers.

A Semi-Physical Platform for Guidance and Formations of Fixed-Wing Unmanned Aerial Vehicles.

Unmanned Aerial Vehicles (UAVs) have multi-domain applications, fixed-wing UAVs being a widely used class. Despite the ongoing research on the topics of guidance and formation control of fixed-wing UAVs, little progress is known on implementation of semi-physical validation platforms (software-in-the-loop or hardware-in-the-loop) for such complex autonomous systems. A semi-physical simulation platform should capture not only the physical aspects of UAV dynamics, but also the cybernetics aspects such as the autopilot and the communication layers connecting the different components. Such a cyber-physical integration would allow validation of guidance and formation control algorithms in the presence of uncertainties, unmodelled dynamics, low-level control loops, communication protocols and unreliable communication: These aspects are often neglected in the design of guidance and formation control laws for fixed-wing UAVs. This paper describes the development of a semi-physical platform for multi-fixed wing UAVs where all the aforementioned points are carefully integrated. The environment adopts Raspberry Pi’s programmed in C++, which can be interfaced to standard autopilots (PX4) as a companion computer. Simulations are done in a distributed setting with a server program designed for the purpose of routing data between nodes, handling the user inputs and configurations of the UAVs. Gazebo-ROS is used as a 3D visualization tool.

The problem of reliable design of vector-field path following in the presence of uncertain course dynamics

Abstract Reliable guidance of fixed-wing Unmanned Aerial Vehicles (UAVs) is challenging, as their high maneuverability exposes them to several dynamical changes and parametric uncertainties. Reliability of state-of-the-art guidance methods is often at stake, as these methods heavily rely on precise UAV course dynamics, assumed in a decoupled first-order form with known time constant. To improve reliability of guidance for fixed-wing UAVs, this work proposes a novel vector field law that can handle uncertain course time constant and state-dependent uncertainty in the course dynamics arising from coupling. Stability is studied in the Lyapunov framework, while reliability of the proposed method is tested on a software-in-the loop UAV simulator. The simulations show that, in the presence of such uncertainty, the proposed method outperforms the standard vector field approaches.

Sliding Mode Guidance and Control for UAV Carrier Landing

A guidance and control system is proposed for automatic carrier landing of an unmanned aerial vehicle (UAV). To this end, the position and attitude of the carrier is predicted by considering the sea state so that the UAV guidance loop can be compensated during the final approach phase. The landing kinematics and reference trajectory are determined using the predicted touchdown position. Nonlinear guidance and control laws are designed based on a sliding-mode control scheme to make the UAV robustly follow the reference trajectory.

A UAV Guidance System Using Crop Row Detection and Line Follower Algorithms

In precision agriculture, activities such as selective spraying of agrochemicals are essential to maintaining high productivity and quality of agricultural products. The use of unmanned aerial vehicles (UAVs) to perform this activity reduces soil compaction, compared to the use of heavy machinery, and helps to reduce the waste of these artificial substances through a punctual and self-regulating application. This work proposes an entire guiding system for use on UAVs (hardware and software) based on image processing techniques. The software part consists of two algorithms. The first algorithm is the Crop Row Detection which is responsible for the correct identification of the crop rows. The second algorithm is the Line Filter that is responsible for generating the driving parameters sent to the flight controller. In the field experiments performed on the proposed hardware, the algorithm achieved a detection rate of 100% of the crop rows for images with resolutions above 320 × 240. The system performance was measured in laboratory experiments and reached 31.22 FPS for images with small resolution, 320 × 240, and 1.63 FPS for the highest resolution, 1920 × 1080. The main contribution of this work is the design and development of an entire embedded guidance system composed of a hardware and software architectures. Other contributions are: the proposed filter for the image pretreatment; the filter to remove the false positive lines; and the algorithm for generating the guiding parameters based on detected crop rows.

Auto-landing guidance for unmanned aerial vehicle with engine flame-out

Landing is the most dangerous phase of the entire flight phases. If the total loss of thrust occurs during flight, a powered aircraft converts to a glider, which can use kinetic and potential energy only. For this reason, a proper scheme is needed for safe landing in cases of the total loss of thrust. This paper presents three-dimensional unpowered auto-landing guidance based on trajectory generation, expanding the concept of the energy-to-range ratio. We develop the terminal velocity estimation method for a horizontal plane applied to three-dimensional space; this method is based on the previously suggested terminal velocity estimation method for a vertical plane. Then, we show trajectory generation for landing guidance combining vertical with horizontal waypoints. The proposed auto-landing guidance with trajectory generation is evaluated by numerical simulation.

Fixed-wing UAV guidance law for ground target over-flight tracking

This paper studies the problem of tracking a ground target for a fixed-wing unmanned aerial vehicle (UAV) based on the proposed guidance law. The algorithm ensures that a UAV continuously overflies the target whether it is fixed or moving. The requirements of the UAV flight constraints such as bounded airspeed and acceleration are considered. A Lyapunov function is constructed to prove the stability of the proposed guidance law, and parameter design criteria have been developed. Considering the fixed and moving ground targets, numerical simulations are performed to verify the feasibility and benefits of the proposed guidance algorithm.

Autonomous Landing of a UAV on a Moving Platform Using Model Predictive Control

Developing methods for autonomous landing of an unmanned aerial vehicle (UAV) on a mobile platform has been an active area of research over the past decade, as it offers an attractive solution for cases where rapid deployment and recovery of a fleet of UAVs, continuous flight tasks, extended operational ranges, and mobile recharging stations are desired. In this work, we present a new autonomous landing method that can be implemented on micro UAVs that require high-bandwidth feedback control loops for safe landing under various uncertainties and wind disturbances. We present our system architecture, including dynamic modeling of the UAV with a gimbaled camera, implementation of a Kalman filter for optimal localization of the mobile platform, and development of model predictive control (MPC), for guidance of UAVs. We demonstrate autonomous landing with an error of less than 37 cm from the center of a mobile platform traveling at a speed of up to 12 m/s under the condition of noisy measurements and wind disturbances.

Nash Bargaining Solution based rendezvous guidance of unmanned aerial vehicles

This paper addresses a finite-time rendezvous problem for a group of unmanned aerial vehicles (UAVs), in the absence of a leader or a reference trajectory. When the UAVs do not cooperate, they are assumed to use Nash equilibrium strategies (NES). However, when the UAVs can communicate among themselves, they can implement cooperative game theoretic strategies for mutual benefit. In a convex linear quadratic differential game (LQDG), a Pareto-optimal solution (POS) is obtained when the UAVs jointly minimize a team cost functional, which is constructed through a convex combination of individual cost functionals. This paper proposes an algorithm to determine the convex combination of weights corresponding to the Pareto-optimal Nash Bargaining Solution (NBS), which offers each UAV a lower cost than that incurred from the NES. Conditions on the cost functions that make the proposed algorithm converge to the NBS are presented. A UAV, programmed to choose its strategies at a given time based upon cost-to-go estimates for the rest of the game duration, may switch to NES finding it to be more beneficial than continuing with a cooperative strategy it previously agreed upon with the other UAVs. For such scenarios, a renegotiation method, that makes use of the proposed algorithm to obtain the NBS corresponding to the state of the game at an intermediate time, is proposed. This renegotiation method helps to establish cooperation between UAVs and prevents non-cooperative behaviour. In this context, the conditions of time consistency of a cooperative solution have been derived in connection to LQDG. The efficacy of the guidance law derived from the proposed algorithm is illustrated through simulations.