Identification of Dynamic Systems

Abstract

PrefaceFor many problems of the design, implementation, and operation of automatic control systems, relatively precise mathematical models for the static and dynamic behavior of processes are required. This holds also generally in the areas of natural sciences, especially physics, chemistry, and biology, and also in the areas of medical engineering and economics. The basic static and dynamic behavior can be obtained by theoretical or physical modeling, if the underlying physical laws (first principles) are known in analytical form. If, however, these laws are not known or are only partially known, or if significant parameters are not known precisely enough, one has to perform an experimental modeling, which is called process or system identification. Then, measured signals are used and process or system models are determined within selected classes of mathematical models.The scientific field of system identification was systematically developed since about 1960 especially in the areas of control and communication engineering. It is based on the methods of system theory, signal theory, control theory, and statistical estimation theory and was influenced by modern measurement techniques, digital computations and the need for precise signal processing, control, and automation functions. The development of identification methods can be followed in wide spread articles and books. However, a significant influence had the IFAC-symposia on system identification, which were since 1967 organized every three years around the world, in 2009 a 15 th time in Saint-Malo. The book is intended to give an introduction to system identification in an easy to understand, transparent, and coherent way. Of special interest is an applicationoriented approach, which helps the user to solve experimental modeling problems. It is based on earlier books in German, published in 1971German, published in , 1974German, published in , 1991German, published in and 1992, and on courses taught over many years. It includes own research results within the last 30 years and publications of many other research groups.The book is divided into eight parts. After an introductory chapter and a chapter on basic mathematical models of linear dynamic systems and stochastic signals, part I treats identification methods with non-parametric models and continuous time signals. The classical methods of determining frequency responses with non-periodic VI Preface and periodic test signals serve to understand some basics of identification and lay ground for other identifications methods.Part II is devoted to the determination of impulse responses with auto-and crosscorrelation functions, both in continuous and discrete time. These correlation methods can also be seen as basic identification methods for measurements with stochastic disturbances. They will later appear as elements of other estimation methods and allow directly the design of binary test signals.The identification of parametric models in discrete time like difference equations in Part III is based m...

System Properties and Control of Turbocharged Diesel Engines with High-and Low-Pressure EGR

“…The HP-EGR-valve is excited with an Amplitude modulated Pseudo Random Binary Signal (APRBS) at two different positions of s lp−egr/eth . The APRBS signal is calibrated with a cycle time of 1 s, a signal length of 511 states and uniformly distributed amplitudes between 0 and 1, see [16]. The simulation results reflect p 2i as well asṁ air in static and dynamic characteristics.…”

Development of a mathematical apparatus for determining operator images of the desired quantized transition functions of finite duration

“…In this case, one can perform quantization, i. e. separation of the desired transition function on a set of Heaviside functions [17] with a rising time shift.…”

“…The equation (17) can be written in a generalized form to establish a mathematical interpretation of the transition function of finite duration: c is one level of the transition function of finite duration (Fig. 7).…”

Design of Active Vibration Absorbers Using On-Line Estimation of Parameters and Signals

“…Note that, an immediate consequence is that, the output y = z 1 has relative degree 4 with respect to u and relative degree 2 with respect to f and, therefore, the so-called disturbance decoupling problem of the perturbation f (t) from the output y = z 1 , using state feedback, is not solvable (Isidori, 1995).…”