Yong-Kyu Song

Design of an Aerolevitation Electric Vehicle for High-Speed Ground Transportation System

An aerolevitation electric vehicle, acting as a tracked wing-in-ground-effect vehicle, is conceptually designed to match the design requirements. The aerodynamic interaction between the vehicle and its track is investigated using a combination of approaches. A boundary-element method is used to study the effect of steady, nonplanar ground effect on the vehicle. The more complicated flow characteristics are investigated using a Navier-Stokes computation. The data obtained from the numerical simulations are compared with the data measured from wind-tunnel tests. The results computed using the boundary-element method agree with the measured data. The longitudinal and lateral stability derivatives are estimated, and a guidance and control system is designed using intelligent techniques based on the estimated stability derivates.

Autonomous Formation Flight Tests of Multiple UAVs

In this work, autonomous formation flight tests of multiple UAVs are experimentally studied. After a guidance and control system for a UAV is designed and tested, PID formation controller for follower UAV is tested using longitudinal and lateral distance feedback. It is shown that more stable and efficient formation guidance system is obtained by using position and attitude of the leader aircraft, which is exploited to calculate virtual waypoint for follower. In order to improve transient response during turn, part of roll command of the leader is added to the guidance command. Finally, autonomous formation flight test results of 3 UAVs are shown by using the best guidance algorithm suggested.

IMPLEMENTATION OF A NEURAL NETWORK STATE ESTIMATOR-BASED AUTOPILOT FOR SKID-TO-TURN MISSILES

High performance missile autopilots often require full states and assume some dynamic inversion. In relation to these problems, this paper discusses the implementation of a neural network state estimator-based autopilot for Skid-to-Turn missiles. The missile model exhibits all the nonlinearties including the coupling between pitch and yaw channels under the assumption of perfect roll-stabilization. In addition to neural network state estimator, a partial linearizing controller, which is a key element in the dynamic inversion, is also constructed through neural networks. With some dynamic compensators, the autopilot is implemented in a hybrid manner and shows good simulation results.

Design of an Aero-levitation Electric Vehicle for the High-speed Ground Transportation System

An Aero-levitation Electric Vehicle (AEV) as a Tracked Wing-In-Ground-effect vehicle (TWIG) is proposed as a promising form of next generation transport. The development of the High-speed Ground Transportation System (HGTS) requires the careful consideration of the aerodynamic interaction of the AEV with the Guideway. It is investigated through the following multidisciplinary cooperation: aerodynamic analysis using the potential flow theory and the NavierStokes computation, wind-tunnel measurements using the scaled model system, and dynamic simulations for the given flight missions. The recent results from these multidisciplinary studies will be summarized and addressed. * Professor, School of Mechanical Engineering,

Compensation of Magnetometer in the Navigation System for Unmanned Helicopter using an Electric Motor

GNSS and ARS are the most common sensors in low-end UAVs. However, these sensors are vulnerable to built-in errors and cannot measure the body heading independently. The GNSS/INS cannot fully compensate the IMU errors in initial alignment process and rectilinear flights. For an unmanned helicopter, a magnetometer can be more useful than any other sensors to obtain heading information. However, the electric motor which drives small helicopter UAV keeps the magnetometer from reading the pure magnetotelluric vector. This paper shows the effects of electric motor on the magnetometer readings, and presents a method to compensate the effects. The results are verified with flight test data. The simulation and experimental results in this paper proves that aiding GNSS/INS with magnetometer increases observability and improves accuracy.

Multi-sensor Fusion Filter for the Flight Safety System of a Space Launch Vehicle

Threat due to malfunction of space launch vehicles is significant since it is bigger and flights longer range than military missiles or scientific rockets. It is necessary to implement a flight safety system to minimize the possible hazard. Design objective of the tracking filter for the flight safety system is different from conventional tracking filters since estimation reliability is more emphasized than estimation accuracy. In this paper, a fusion tracking filter was implemented for processing multi-sensor data from a space launch vehicle. The filter performance is evaluated by analyzing the error of the estimated position and instantaneous impact point. Also a fault detection algorithm is implemented to guarantee fusion filter`s reliability under any sensor failure and verified to maintain stability successfully.

System Identification of a Small Unmanned Rotorcraft

In this paper, Recursive Least Squares (RLS) and Fourier Transform Regression (FTR) methods for estimating stability and control derivatives of small unmanned helicopter are evaluated together with MMLE technique. Flight data simulated by using a commercial small-scale helicopter model are exploited to estimate the parameters with accuracies for hover and cruise modes. The performances of the system identification methods are also compared by analyzing the responses of the reconstructed systems using estimated derivatives.In this paper, Recursive Least Squares (RLS) and Fourier Transform Regression (FTR) methods for estimating stability and control derivatives of small unmanned helicopter are evaluated together with MMLE technique. Flight data simulated by using a commercial small-scale helicopter model are exploited to estimate the parameters with accuracies for hover and cruise modes. The performances of the system identification methods are also compared by analyzing the responses of the reconstructed systems using estimated derivatives.

System Identification of a Small Unmanned Air Vehicle Using Neural Networks

In this paper system identification of a small UAV via neural networks is tried and the estimated parameters are then compared to those obtained by Fourier Transform Regression and Maximum Likelihood Estimation Techniques. With the estimated parameters a linear system is constructed and simulated to compare to the flight data. The results show that parameter identification using neural networks is comparable to the existing techniques

Flight Control Experiment of High-Speed Aero-Levitation Electric Vehicle Scale-Model in Wind-Tunnel

An experimenal study on flight control of high-speed AEV(Aero-levitation Electric Vehicle) scale model in wind-tunnel is conducted. The AEV is to fly at very low altitude in predesigned track so that it is always under the wing-in-ground effect. The experiment is intended to fly the scale model to follow the predesigned altitude schedule while holding its attitude (pitch, roll, and yaw). Especially, the altitude changes for climb, cruise, and descent with constant pitch angle are most important maneuvers. The experiment shows that the required mission flights can be performed with appropriate sensors, processors, and actuators.

A study on the control surface/actuator fault detection, identification, and accommodation system for aircraft

In this study a control surface/actuator fault detection, identification, and accommodation system for aircraft is designed. This fault tolerant control system tries to return aircraft to its stable trim condition in a short time. The control system is designed using neural networks with Extended Back Propagation Algorithm which shows fast convergence. F-4 aircraft with possible stabilator or aileron failure/stuck is simulated with the proposed scheme.