Hyo-Sang Shin

Rendezvous and Standoff Target Tracking Guidance Using Differential Geometry

This paper proposes UAV rendezvous and standoff tracking guidance laws against a moving target using differential geometry. Searching and subsequent tracking of moving ground based target is one of the primary capabilities of cooperative UAVs. In performing such missions, UAVs are to approach a target and keep a certain distance, known as a standoff distance. This allows target tracking without being noticed and acquisition of accurate target information. In this study, standoff target tracking is proposed using the solution of differential geometry between the UAV and the target. The proposed algorithm brings several advantages along with its inherent simplicity: rigorous stability, explicit use of a target velocity, and tuning parameter reduction. The feasibility and performance of the proposed approach is not only mathematically analysed, but also verified through realistic scenarios.

A novel actuation concept for a multi rotor UAV

This paper proposes a novel strategy to improve the performance and fault tolerance of multi-rotor vehicles. The proposed strategy uses dual axis tilting propellers and thus enables three different actuation mechanisms, namely, gyroscopic torques, thrust vectoring and differential thrusting. Unlike the conventional quadrotor, the proposed strategy offers a wider range of control torques by combining the three actuation mechanisms. Conventional quadrotors cannot be reconfigured if one of rotors fails. However, the proposed strategy is still able to reconfigure the vehicle with complete failure of one rotor and a pair of adverse motors. In order to prove this concept, a dual axis tilting UAV is first designed and prototyped. Next, a mathematical representation of the prototyped vehicle is modelled and verified using experiments. Then, a control system is developed based on a PD controller and pseudoinverse control allocator and validated through tests on a rig and flight tests. The tests show that the vehicle is faster than a conventional counterpart and that it can resist the failure of two rotors. Finally, this paper suggests how to lead further substantial improvements in performance.

Nonlinear Model Predictive Control for Multiple UAVs Formation Using Passive Sensing

In this paper, nonlinear model predictive control (NMPC) is addressed to develop formation guidance for multiple unmanned aerial vehicles. An NMPC algorithm predicts the behavior of a system over a receding time horizon, and the NMPC generates the optimal control commands for the horizon. The first input command is, then, applied to the system and this procedure repeats at each time step. The input constraint and state constraint for formation flight and inter-collision avoidance are considered in the proposed NMPC framework. The performance of NMPC for formation guidance critically degrades when there exists a communication failure. In order to address this problem, the modified optimal guidance law using only lineof-sight, relative distance, and own motion information is presented. If this information can be measured or estimated, the proposed formation guidance is sustainable with the communication failure. The performance of this approach is validated by numerical simulations.

Energy Optimal Waypoint Guidance Synthesis for Antiship Missiles

Planar waypoint guidance synthesis methods for antiship missiles (ASMs) using optimal guidance laws are proposed. The energy optimal trajectory optimization problem with waypoint constraints is converted to an unconstrained optimization problem of finding the optimal boundary conditions at waypoints for the guidance laws. An optimal guidance law (OGL) for a 1st-order lag ASM with terminal constraints on the impact angle and lateral acceleration is newly proposed for this purpose. The proposed method produces the energy optimal trajectory with high numerical efficiency. If the ASM is approximated by a lag-free system, optimal boundary conditions become waypoint passing angles which can be simply determined from a set of linear algebraic equations. Since there are no time-consuming numerical optimizations in this approach, the energy optimal trajectory passing through all the waypoints can be generated in real time.

Coordinated standoff tracking of groups of moving targets using multiple UAVs

This paper presents a coordinated standoff tracking methodology of groups of moving targets using multiple UAVs. The vector field guidance approach is first applied to track a group of targets for a single UAV by defining a variable standoff orbit to be followed, which can keep all targets within the field-of view of the UAV. A new feedforward term is included in the guidance command considering variable standoff distance, and the convergence of the vector field to the standoff orbit is analysed and enhanced by adjusting radial velocity using two active measures associated with vector field generation. Moreover, for multiple group tracking by multiple UAVs, a two-phase approach is proposed as a suboptimal solution for an NP-hard problem, consisting of target clustering/assignment and cooperative standoff group tracking with online local replanning. Lastly, localisation sensitivity to the group of targets is investigated for different angular separations between UAVs and sensing configurations. Numerical simulations are performed using randomly moving ground vehicles with four UAVs to verify the feasibility and benefit of the proposed approach.

Coordinated standoff tracking of moving target groups using multiple UAVs

This paper presents a methodology for coordinated standoff tracking of moving target groups using multiple unmanned aerial vehicles (UAVs). The vector field guidance approach for a single UAV is first applied to track a group of targets by defining a variable standoff orbit to be followed, which can keep all targets within the field-of-view of the UAV. A new feedforward term is included in the guidance command considering variable standoff distance, and the convergence of the vector field to the standoff orbit is analyzed and enhanced by adjusting radial velocity using two active measures associated with vector field generation. Moreover, for multiple group tracking by multiple UAVs, a two-phase approach is proposed as a suboptimal solution for a Non-deterministic Polynomial-time hard (NP-hard) problem, consisting of target clustering/assignment and cooperative standoff group tracking with online local replanning. Lastly, localization sensitivity to the group of targets is investigated for different angular separations between UAVs and sensing configurations. Numerical simulations are performed using randomly moving ground vehicles with multiple UAVs to verify the feasibility and benefit of the proposed approach.

UAV Obstacle Avoidance using Differential Geometry Concepts

Abstract This paper investigates the application of differential geometry to UAV collision avoidance based on differential geometry. The Differential Geometry approach has been useful in determining a guidance algorithm to produce a constant curvature evasion manoeuvre to avoid collision with an Aircraft flying in a straight line. Simulation results presented in the paper demonstrated the effectiveness of the proposed UAV collision avoidance algorithms.

Behaviour recognition of ground vehicle using airborne monitoring of unmanned aerial vehicles

This paper proposes a behaviour recognition methodology for ground vehicles moving within road traffic using unmanned aerial vehicles in order to identify suspicious or abnormal behaviour. With the target information acquired by unmanned aerial vehicles and estimated by filtering techniques, ground vehicle behaviour is first classified into representative driving modes, and then a string pattern matching theory is applied to detect suspicious behaviours in the driving mode history. Furthermore, a fuzzy decision-making process is developed to systematically exploit all available information obtained from a complex environment and confirm the characteristic of behaviour, while considering spatiotemporal environment factors as well as several aspects of behaviours. To verify the feasibility and benefits of the proposed approach, numerical simulations on moving ground vehicles are performed using realistic car trajectory data from an off-the-shelf traffic simulation software.

UAV Conflict Detection and Resolution for Static and Dynamic Obstacles

Conflict Detection and Resolution (CD&R) for Unmanned Aerial Vehicles (UAVs) has become important consideration with remarkable development of sensor technology and autonomous control system. Since CD&R deals with safety of UAVs and aircraft, it has to be validated by rigorous analytical verification. This paper investigates the application of differential geometry to UAV CD&R algorithm for non-cooperating intruders such as static and dynamic obstacles. Two-dimensional geometry is used to detect a single conflict and multiple conflicts, and find the condition for resolution guidance. The existence and stability of the conflict resolution, which controls either heading angle only or heading angle and ground speed simultaneously, are then analysed. Whilst there exist region which the UAV can not resolve due to its physical and operational limitations as every CD&R algorithms, the resolution guidance can find the region and prove local stability using Lyapunov theory. The analysis and resolution guidance design do not rely on local linearisation and can be shown that the derived conflict resolution is equivalent concepts to Proportional Navigation Guidance (PNG) and Traffic alert and Collision Avoidance Systems (TCAS). The paper finishes with numerical simulation in two dimensions, illustrating the performance and properties of the proposed CD&R algorithm.

Communication, Ocean, and Meteorological Satellite Orbit Determination Analysis Considering Maneuver Scheme

Orbit determination (OD) analysis for geostationary Communications, Ocean, and Meteorological Satellite (COMS) is presented. Since the orbital longitude of COMS is close to that of satellite tracking site, geometric singularity affects observability. For OD, we fix velocity increment by the wheel off-loading maneuver, and the azimuth angle tracking bias is not estimated because of observability problem. Final epoch of the propagated OD based on different data arc length is used for two-day orbit prediction. The difference between truth and 48-hour predicted orbit that contains the OD error shows 4-18 km Root-Sum-Squares (RSS) in 3-D sense (one sigma) in spite of the singularity problem. Thus an operational Orbit Determination and prediction (ODP) system for COMS fulfills the requirement for 18 km RSS (one sigma) predicted positioning knowledge.