• Proceedings of the KIEE Conference
• /
• 1982.07a
• /
• pp.116-117
• /
• 1982

• Transactions on Control, Automation and Systems Engineering
• /
• v.4 no.2
• /
• pp.124-129
• /
• 2002

This paper demonstrates that the largest Lyapunov exponent λ of recurrent neural networks can be controlled efficiently by a stochastic gradient method. An essential core of the proposed method is a novel stochastic approximate formulation of the Lyapunov exponent λ as a function of the network parameters such as connection weights and thresholds of neural activation functions. By a gradient method, a direct calculation to minimize a square error (λ - λ$\^$obj/)$^2$, where λ$\^$obj/ is a desired exponent value, needs gradients collection through time which are given by a recursive calculation from past to present values. The collection is computationally expensive and causes unstable control of the exponent for networks with chaotic dynamics because of chaotic instability. The stochastic formulation derived in this paper gives us an approximation of the gradients collection in a fashion without the recursive calculation. This approximation can realize not only a faster calculation of the gradient, but also stable control for chaotic dynamics. Due to the non-recursive calculation. without respect to the time evolutions, the running times of this approximation grow only about as N$^2$ compared to as N$\^$5/T that is of the direct calculation method. It is also shown by simulation studies that the approximation is a robust formulation for the network size and that proposed method can control the chaos dynamics in recurrent neural networks efficiently.

• Korean Journal of Applied Biomechanics
• /
• v.18 no.1
• /
• pp.21-30
• /
• 2008

The purpose of this study was to investigate the variability to compare local dynamic stability via a linear and nonlinear analysis during walking. Twenty four elderly males, 12 healthy elderly and 12 patients with osteoarthritise walked on a treadmill for 100 consecutive strides. Lyapunov exponent and correlation dimension and coefficient variation were calculated for the kinematic parameters to determine the dynamic stability during walking. The linear measures indicated that the healthy elderly demonstrated significantly higher variability in the ankle joint displacement. The nonlinear analysis revealed that COD for the knee joint angle were higher in patient with osteoarthritise. There were no coincidence in results between linear and nonlinear techniques over two groups. In light of nonlinear analysis, it was concluded that patients with osteoathritise showed higher local instability during walking.

• Journal of the Korean Society for Aeronautical & Space Sciences
• /
• v.31 no.2
• /
• pp.63-71
• /
• 2003

This paper presents a robust slewing control of a flexible space structure based on sliding surface design. A sliding surface is designed for a single-axis rest-to-rest slewing in view of target angle, target angular velocity, and root monent of the flexible appendage. In comparison with the Lypunov control law, both controllers guarantee the stability and command tracking capabilities for nominal system. It is also shown that the designed control law provides further robustness to internal/external uncertainties. Extending the results of a single-axis maneuver, a sliding mode control law was sought for an arbitrary three-axis maneuver. Quaternion was used to determine the attitude of a space structure and sliding surfaces were designed for each axis, thereby a robust control law was derived considering the coupling effects between each rotational axis during the maneuver. Several numerical examples were demonstrated to show the effectiveness of the designed control law.

• Journal of the Korea Academia-Industrial cooperation Society
• /
• v.19 no.7
• /
• pp.16-23
• /
• 2018

Chaos is one of the most significant topics in nonlinear science, and has been intensively studied since the Lorenz system was introduced. One characteristic of a chaotic system is that the signals produced by it do not synchronize with any other system. It therefore seems impossible for two chaotic systems to synchronize with each other, but if the two systems exchange information in just the right way, they can synchronize. This paper addresses the problem of synchronization in a modified Lorenz chaotic system based on active control, sliding mode control, and the Lyapunov stability theory. The considered synchronization scheme consists of identical drive and response generalized systems coupled with linear state error variables. For this, a brief overview of the modified Lorenz chaotic system is given. Then, control rules are derived for chaos synchronization via active control and slide mode control theory, with a strategy for solving the chattering problem. The asymptotic stability of the overall feedback system is established using the Lyapunov stability theory. A set of computer simulation works is presented graphically to confirm the validity of the proposed method.

• Journal of the Institute of Electronics Engineers of Korea SC
• /
• v.46 no.1
• /
• pp.28-33
• /
• 2009

In the research of speech recognition, locating the beginning and end of a speech utterance in a background of noise is of great importance. The conventional methods for speech endpoint detection are based on two simple time-domain measurements-short-time energy, and short-time zero-crossing rate, which couldn't guarantee the precise results if in the low signal-to-noise ratio environments. This paper proposes a novel approach that finds the Lyapunov exponent of time-domain waveform. This proposed method has no use for obtaining the frequency-domain parameters for endpoint detection process, e.g. Mel-Scale Features, which have been introduced in other paper. Accordingly, this algorithm is low complexity and suitable for Digital Isolated Word Recognition System.

• Journal of the Korean Institute of Intelligent Systems
• /
• v.13 no.2
• /
• pp.175-179
• /
• 2003

Convergence condition determines performance of iterative learning control (ILC), for example, convergence speed, remaining error, etc. Hence, the performance can be elevated and a feasible set of learning controllers grows if a less conservative condition is obtained. In the frequency domain, the $H_{\infty}$ norm of the transfer function between consecutive errors has been currently used to test convergence of a learning system. However, even if the convergence condition based on the $H_{\infty}$ norm has a clear property about monotonic convergence, it has a few drawbacks, especially in MIMO plants. In this paper, the relation between the condition and the monotonicity of convergence is clarified and a modified convergence condition is found out using a frequency domain Lyapunov equation, which supersedes the conventional one in the frequency domain.

• Journal of IKEEE
• /
• v.23 no.4
• /
• pp.1265-1272
• /
• 2019

We analyze the robustness of the existing predictor feedback controller for discrete-time linear systems with constant input delays against the structured model uncertainty. By modeling the constant input delay with a first-order PdE (Partial difference Equation), we replace the input delay with the PdE states. By applying a backstepping transformation, we build a target system that enables to construct an explicit Lyapunov function. Constructing the explicit Lyapunov function that covers the entire state variables, we prove the existence of an allowable maximum size of the structured model uncertainty to maintain stability and establish the robustness of the predictor feedback controller. The numerical example demonstrates that the stability of closed-loop system is maintained in the presence of the structured model uncertainty, and verifies the robustness of the predictor feedback controller.

• Transactions of the Korean Society of Mechanical Engineers A
• /
• v.36 no.10
• /
• pp.1147-1154
• /
• 2012

In this study, a method for designing blade pitch and generator torque controllers for a wind turbine generator is presented. This method consists of two steps. First, the Lyapunov stability theory is used to obtain nonlinear control laws that can regulate the rotor speed and the power output at all operating ranges. The blade pitch controller is chosen such that it always decreases a positive definite function that represents the error in rotor speed control. Similarly, the generator torque controller always decreases a positive definite function that reflects the error in power output control. Then, the simulation-based optimization technique is used to tune the design parameters. The controller design procedure and simulation results are presented using the widely adopted two-mass model of the wind turbine.

• Journal of the Korean Society for Aeronautical & Space Sciences
• /
• v.40 no.7
• /
• pp.574-581
• /
• 2012

Because of the exposed blade, the UAV using the rotors entail the risks during operation. While a wrapped duct around the fan blades reduces risks, it is a higher thrust performance than the same power load rotor. In this paper, for applying advantages of a ducted fan, the tri-ducted fan air vehicle configuration is proposed. The vehicle has three ducted fans. Two of them are the same shape and size and the third one is the smaller. It is possible to control a rapid attitude stability using thrust vector control. The equations of motion of the tri-ducted fan were derived. Lyapunov control input was applied to the system and stable inputs were derived. A nonlinear simulation was fulfilled by using parameters of a prototype vehicle. It verified a stable attitude and analyzed results.