Sergey Lyashevskiy

Nonlinear identification of aircraft

In flight dynamics and control, a growing interest in identification is induced by the fact that identification allows one to find the dynamical models for analysis and optimization. Identification plays a key role in nonlinear analysis and control as well as in diagnostics and design of self-repairing reconfigurable algorithms for advanced vehicles. This paper treats the parameter identification issues for nonlinear multivariable aircraft models and reports results in the convergence analysis of the proposed identification algorithm. To handle the nonlinear identification, the model-based framework and Lyapunovs concept are applied. The second method establishes and unifies analytical results in stability. Using Lyapunovs concept, the conditions which ensure the convergence of the parameters are given. The explored identification algorithm has significant advantages for analysis, control and diagnostics of nonlinear continuous-time systems. The unknown parameters of nonlinear aircraft models can be obtained using the developed innovative procedure in the straightforward way. The proposed technique is computationally efficient and the identification can be accomplished on-line in real-time. To support the and to illustrate the practical the methodology, longitudinal derivatives for a high performance aircraft are found employing the available flight data. The modeling results are compared to the experimental data.

Nonlinear error mapping technology in system identification

In system engineering, a growing interest in identification of nonlinear systems is induced by the fact that identification allows one to find the models for consecutive analysis and control. This paper studies the parameter identification features for nonlinear multivariable systems and reports results in the convergence analysis of the proposed identification algorithm. To handle the nonlinear identification, an innovative nonlinear mapping concept is offered, and Lyapunovs theory plays a key role in analysis of parameter convergence. The explored method significantly departures from linear error mapping philosophy, and the offered framework ensures parameter convergence and enhances the sets of initial parameter values where identification can be performed.

Design of robust flight control systems for advanced technology vehicles

Innovative approaches for analysis, design and modeling of advanced flight vehicles are demanded in response to requirements toward substantial performance improvements. Although control theory is thoroughly developed, the calculus of variations and Pontryagins principle are not well suited because these methods exhibit rather poor robustness and are hampered by the limited time available for computations. This paper explores the robust control technology for uncertain nonlinear systems and presents innovative design procedures to enhance the robustness for open-loop unstable systems with constraints. Utilizing the Hamilton-Jacobi theory, the developed procedures allow one to find a straightforward avenue to design the robust controllers. A new control methodology is less conservative than the known results and allows us to solve the nonlinear control problem for uncertain systems. The explored time-optimal and antiwindup techniques are studied for a highly-manoeuvrable missile. The longitudinal and rolling flight control, as well as the guidance problem, are treated and solved. To evaluate the robustness, analytical and simulation results are performed.

On the Application of the Lyapunov and Harmonic Balance Methods for Identification of Nonlinear Systems

This paper addresses the problem of identification of nonlinear dynamical systems. A main difficulty in identification of parameters are that in case of nonlinear proce sses the dynamics is dependent on the initial conditions, inputs and disturbances. Furthermore, the nonlinear systems may exhibit stable and unstable periodic modes. In this paper we pose the J .yapllnov theory and the harmonic balance method towards the identification issues. New straightforward algorith ms arc developed for identification of parameters based 011 the second method and the harmonic linearization concept. The identitication techniques are designed to be computationally e!ficient and reliable. Analytical, experimental and numerical results are presented to illustrate the appli cability and efficiency of the exploited identification procedures. Specifically, the smooth pursuit eye movement system and servnmechanismwi th permanent-m agnet DC motors are identifie d.

DSP-based fuzzy logic control of a positioning system-experimental results

This paper presents some experimental results of a fuzzy logic control system for a high-performance positioning system with real-time digital implementation using digital signal processors (DSP). The PI-like fuzzy logic controller is developed using the fuzzy logic system with fuzzifier and defuzzifier. Seven fuzzy sets with normalized triangular membership functions are used for both input and output variables of the controller. The proposed control system is implemented using a DSP-based real time control testbed. Experimental results for two-dimensional contour following control are presented. The effects of sampling frequency, speed and payload variations are discussed. The results demonstrate the effectiveness of the proposed control system, especially at high speed with payload variations.

Control of discrete-time systems with uncertain parameters

One of the fundamental problems in system design is synthesis of control structures that stabilize uncertain systems and guarantee robust performance. This paper presents a general framework for treating a robust control problem for uncertain discrete-time systems. The reason and importance of designing the digital algorithms is microprocessors realizations of control structures. This paper derives robust strategies via the Lyapunov theory and dynamic programming, and new results are presented. Conditions for the existence of robust controllers and Lyapunov functions are formulated, and design procedures are developed. To illustrate the practical use of the explored design tool, an example of synthesis and analysis of an electric drive with a single-phase induction motor is reported.

Identification of the Smooth Pursuit Eye Movement System

Abstract The purposes of this paper are to estimate the parameters of the human pursuit eye movement system and to describe the identification scheme used. The eye movement system is sufficiently nonlinear that we must necessarily employ a special identification approach which takes account of nonlinearities. The system’s coefficients are estimated using the model reference concept and the gradient method. The identified process parameters are calculated as a solution of nonlinear differential equations.

Feedback Control of Robotic Manipulators with Induction Motors

Abstract Recently, there has been considerable interest in the development of innovative formulations of the robot dynamics and synthesis of the advanced control structures for manipulators with AC motors. In this paper, these issues are treated utilizing advantages of the uncertain modeling setting and robust design framework. The equations of motion for robots with induction motors are derived and the application of a robust control technique is reported. The detailed analysis of the proposed theoretical results is established. The results show that the representation of robotic manipulators as uncertain systems and synthesis of robust algorithms are feasible, straightforward and highly efficient. The controller is implemented on a planar manipulator with two revolute joints actuated by AC motors. Experimental results illustrate the efficiency of the proposed methodology and show the increasing capabilities of the robust control structures for robots with induction motors.

Nonlinear Control for Dynamical Systems with Constraints Based on the Bellman-Lyapunov Method

Abstract We treat the optimization issues for continuous-time dynamical systems whose state variables and control signals are bounded. The objective of this paper is to design of a nonlinear full state feedback control algorithm that satisfies the constraints and minimizes a nonlinear performance integral. The fundamental idea involves using a new class of the costs. The nonquadratic Lyapunov function is obtained by solving the Bellman-Lyapunov equation. The results are illustrated with an example.

Designing of a class of nonlinear robotic manipulators: contributions to control and stability

This paper covers several important topics in control of robotic systems and presents analysis of stability via the Lyapunov concept. Analysis and control issues are considered using the nonlinear model. The rigid-link manipulator with the proportional-integral-differential (PID) controller is asymptotically stable and tracks the desired trajectories. The relationship between control issues and Lyapunovs stability analysis for robots is treated. The essence of the second method is contained in the demonstration of asymptotic stability for nonlinear robots with PID algorithms. Analytical, numerical and experimental studies are performed to design and implement the PID controller as well as to analyze, monitor and demonstrate the dynamic performance of a two-link planar robot. Experimental results illustrate that the procedure is feasible and can be used in systems design.