Jongho Shin

Robust control of ionic polymer?metal composites

Ionic polymer?metal composites (IPMCs) have been considered for various applications due to their light weight, large bending, and low actuation voltage requirements. However, their response can be slow and vary widely, depending on various factors such as fabrication processes, water content, and contact conditions with the electrodes. In order to utilize their capability in various high-performance microelectromechanical systems, controllers need to address this uncertainty and non-repeatability while improving the response speed. In this work, we identified an empirical model for the dynamic relationship between the applied voltage and the IPMC beam deflection, which includes the uncertainties and variations of the response. Then, four types of controller were designed, and their performances were compared: a proportional?integral?derivative (PID) controller with optimized gains using a co-evolutionary algorithm, and three types of robust controller based on , with loop shaping, and ?-synthesis, respectively. Our results show that the robust control techniques can significantly improve the IPMC performance against non-repeatability or parametric uncertainties, in terms of the faster response and lower overshoot than the PID control, using lower actuation voltage.

Autonomous Flight of the Rotorcraft-Based UAV Using RISE Feedback and NN Feedforward Terms

A position tracking control system is developed for a rotorcraft-based unmanned aerial vehicle (RUAV) using robust integral of the signum of the error (RISE) feedback and neural network (NN) feedforward terms. While the typical NN-based adaptive controller guarantees uniformly ultimately bounded stability, the proposed NN-based adaptive control system guarantees semi-global asymptotic tracking of the RUAV using the RISE feedback control. The developed control system consists of an inner-loop and outer-loop. The inner-loop control system determines the attitude of the RUAV based on an adaptive NN-based linear dynamic model inversion (LDI) method with the RISE feedback. The outer-loop control system generates the attitude reference corresponding to the given position, velocity, and heading references, and controls the altitude of the RUAV by the LDI method with the RISE feedback. The linear model for the LDI is obtained by a linearization of the nonlinear RUAV dynamics during hover flight. Asymptotic tracking of the attitude and altitude states is proven by a Lyapunov-based stability analysis, and a numerical simulation is performed on the nonlinear RUAV model to validate the effectiveness of the controller.

Letters: Model predictive flight control using adaptive support vector regression

This paper explores an application of support vector regression (SVR) to model predictive control (MPC). SVR is employed to identify a dynamic system from input-output data, and the identified model is used for predicting the future states in the MPC framework. In order to deal with constant and dynamic uncertainties, an online adaptation algorithm is designed using the gradient descent (GD) method and the adjusted SVR model is fed to the MPC optimizer. In addition, the convergence property of the adaptation rule and the condition for the convergence of the MPC optimization are discussed using discrete-time Lyapunov stability analysis. Finally, the proposed approach is applied to identification and flight control of an unmanned aerial vehicle (UAV) lateral dynamics.

Adaptive support vector regression for UAV flight control

This paper explores an application of support vector regression for adaptive control of an unmanned aerial vehicle (UAV). Unlike neural networks, support vector regression (SVR) generates global solutions, because SVR basically solves quadratic programming (QP) problems. With this advantage, the input-output feedback-linearized inverse dynamic model and the compensation term for the inversion error are identified off-line, which we call I-SVR (inversion SVR) and C-SVR (compensation SVR), respectively. In order to compensate for the inversion error and the unexpected uncertainty, an online adaptation algorithm for the C-SVR is proposed. Then, the stability of the overall error dynamics is analyzed by the uniformly ultimately bounded property in the nonlinear system theory. In order to validate the effectiveness of the proposed adaptive controller, numerical simulations are performed on the UAV model.

Asymptotic attitude tracking of the rotorcraft-based UAV via RISE feedback and NN feedforward

This paper presents an asymptotic attitude tracking controller for a rotorcraft-based unmanned aerial vehicle (RUAV) using the robust integral of the signum of the error(RISE) feedback and neural network (NN) feedforward terms. Usually, the typical NN-based attitude controller guarantees the uniformly ultimately bounded stability. In this study, semi-global asymptotic tracking of the RUAV is guaranteed by the RISE feedback term and NN feedforward term adapted by the projection method. The controller is basically designed by the linear dynamic model inversion method whose model is obtained by the linearization of the nonlinear RUAV model at the hover flight. Then, the uncertainty generated in the linearization is removed by the RISE feedback and NN feedforward terms. The asymptotic tracking of the attitude states is proven with the Lyapunov stability analysis, and a numerical simulation using the nonlinear RUAV model is performed to validate the effectiveness of the proposed controller.

Formation flight control under communication failure

This paper presents the management of UAV formation flight with respect to varying levels of communication among UAVs. Inter-vehicle communication in the formation is a critical issue because each UAV needs the information on other vehicles for formation. However, since communication is not perfect in reality, the formation performance under communication failure has to be analyzed. This study uses position data measured by sensors for overcoming communication failures in the standard leader-follower structure formulated in the nonlinear model predictive control (NMPC) framework. The perceived and obtained position data of each UAVs through GPS or sensor are noisy. These are estimated by extended Kalman filter. The numerical simulation results support the feasibility of the proposed formation flight method.

Communication in distributed model predictive collision avoidance

This paper presents a model predictive approach for collision avoidance of car-like robots. An optimal problem is formulated in terms of cost minimization under constraints. Information on each robot can be incorporated online in the nonlinear model predictive framework and kinematic constraints are treated by Karush-Kuhn-Tucker(KKT) condition. For distributed collision avoidance of multiple robots with two levels of a communication network, performances are compared. In comparison with different types of communication, how much information the robots share can cause difference in the performance. More successful collision avoidance was possible when the robots share enough amount of information.

Adaptive feedback linearization for an uncertain nonlinear system using support vector regression

This paper explores an adaptive feedback linearization for an uncertain nonlinear system using support vector regression (SVR). SVR, which assures global solution by quadratic programming (QP) problem, is used to learn the nominal dynamics of the input-output feedback-linearized system. Then, an adaptation algorithm of the offline-trained SVR is proposed for eliminating the offline-training error and uncertainties in the control process. In addition, the derivation of the adaptive rule considers the controller singularity problem by utilizing the affine property of the nonlinear system and the concept of the virtual control. Uniformly ultimately bound property of the overall system is analyzed by the Lyapunov stability theory. Simulations using a longitudinal dynamics of the F-16 model validate the performance of the proposed approach.

Adaptive Path-Following Control for an Unmanned Surface Vessel Using an Identified Dynamic Model

This paper proposes a path-following control method for an unmanned surface vessel (USV) based on an identified dynamic model. To handle the USV dynamic model effectively, a three degree-of-freedom model is employed instead of a full nonlinear dynamic model and linearized at specific equilibrium condition. The linearized model is identified with real data from several experiments by utilizing a particle swarm optimization method. Then, based on the identified model, an adaptive control algorithm is proposed to follow several waypoints and velocity command. The proposed control method utilizes virtual control input, dynamic surface control method, and adaptive terms to handle matched and unmatched uncertainties simultaneously. The overall closed-loop stability is analyzed by introducing deadzone errors composed of tracking error and saturation function. Finally, some experiment with a remodeled commercial fishing boat are conducted and analyzed to validate the performance of the proposed methods.

Adaptive inverse control using support vector regression

This paper explores application of support vector regression to adaptive inverse control problems. Support vector regression (SVR) has been proven to generate global solutions contrary to neural networks, because SVR basically solves quadratic programming (QP) problems. With this advantage, a plant model is identified and its inverse model is learned. In addition, adaptive algorithms for compensating the errors between the actual model and identified model are proposed and their convergence property is discussed. Finally, numerical simulation is performed for the validation of the proposed approach using the longitudinal dynamics of unmanned aerial vehicle(UAV).