Ste

ANNALS OF THE UNIVERSITY OF CRAIOVA

Series: AUTOMATION, COMPUTERS, ELECTRONICS and MECHATRONICS
ISSN: 1841-0626

    Year 2004 Volume 1(28) no. 2

  1. STABILITY ANALYSIS OF CELLULAR NEURAL NETWORKS WITH TIME DELAYS AS LARGE SCALE SYSTEMS

    Daniela Danciu

    Abstract: Cellular Neural Networks (CNN) are artificial neural networks displaying multidimensional arrays of cells and local interconnections among the cells. Since in a CNN all the cells are identically, the qualitative behavior of the entire network can be studied via stability results obtained for the interconnected systems. The finite switching speed of amplifiers and communications time between cells of the VLSI technology CNN implementations introduce time lags that may lead to oscillations or to the instability of the network. The aim of the paper is to obtain sufficient conditions for the asymptotic stability of a cellular neural network displaying interaction delays. Within the framework of the qualitative theory of the large-scale composite systems these conditions are based on the properties of the Liapunov functions (functionals).
    Keywords: cellular neural network, interconnected systems, large-scale systems, time delays, asymptotic stability.

  2. SIGNAL DETECTION USING NEURAL NETWORK

    Eugen Iancu

    Abstract: In this paper, the author proposes some practical behaviour in detection of signal process in presence of perturbations. This problem is an inherently complex one, which can affect the helpful information. Because of this reason, in the paper it is considered one of the frequently case: when data transmission uses the pulse code modulation (PCM) in base band of channel. For signal processing it is proposed a neural network, which has the ability to recognize and to reconstruct the data.
    Keywords: Signal processing, data transmission, neural network.

  3. TRUETIME LIBRARY FOR SIMULATING COMPLEX AND DISTRIBUTED REAL-TIME CONTROL SYSTEMS

    Cosmin Ionete

    Abstract: This article is focused in introducing a new MATLAB/Simulink-based tool for simulation of real-time complex and distributed control systems. This introduction is made by examples, to facilitate the understanding of main skills of the TrueTime tool. After a short presentation, we present an example of a more complex case of control for three DC-servos using different priority strategies and we introduce the Real Time Network by the use of a PID control law implementation in a distributed CAN network architecture. We shall emphasize the differences between the classical simulation and real-time simulation. Also is presented the influence o real-time priority strategies and network parameters on the real control system. The flexibility of the TrueTime Simulink Library extension is underlined by using both MATLAB and Simulink implementation of the PID control law.
    Keywords: Real time control systems, Event based simulation, Shared resources, Real time kernel, Feedback scheduling, Networked control systems.

  4. COMPARATIVE STUDY OF OF TIME DELAY APPROXIMATION IN DIGITAL CONTROL

    Constantin Marin

    Abstract: In the general framework of model mismatching problems, this paper analyses the effects of the time delay approximation in digital control systems. Usually in literature it is considered that the time delay equals to an integer number of sampling period that is not always true. The discrete time transfer functions are evaluated pointing out the fractional part of the time delay. Considering the dead beat method for controller synthesis, the closed loop behavior and stability is analyzed with respect to this fractional part of time delay. Through some examples, the sensitivity of discrete time control systems to small variation in time delay is pointed out, showing the correctness of the analytical results.
    Keywords: Time-delay systems, discretization, hybrid systems.

  5. ON ADAPTIVE CONTROL OF NONLINEAR DISTRIBUTED PARAMETER BIOREACTORS VIA ORTHOGONAL COLLOCATION

    Emil Petre, Dan Selisteanu

    Abstract: This paper presents the control problem of a class of propagation bioprocesses that are carried out in fixed bed reactors without dispersion. Since the dynamics of these processes are described by partial differential equations for control of these processes one method consists of approximation of their infinitely order associated models by finite order models. A class of nonlinear adaptive controllers are then designed based on these finite order models that consist of a set of ordinary differential equations obtained here by orthogonal collocation method. Computer simulations conducted in the case of a fixed bed reactor are included to illustrate the performances of the presented adaptive system.
    Keywords: Nonlinear systems, Fixed bed bioreactors, Distributed parameter systems, Orthogonal collocation, Adaptive control.

  6. PERFECT SECRECY OF VERNAMíS CIPHER

    Sendrescu Dorin

    Abstract: Cryptography describes a wide field of secret communications by means and methods used for the protection of the passive interception (that is recording of the sent message) and/or active interception (that is information changing or false messages inserted in the transmission channel, between the legal emitter and receptor). Shannonís theory of communication has been the cornerstone in laying the foundations for the design of secure ciphers. It consists of three important parameters: diffusion, confusion and unconditional security. In this paper these principles are presented.
    Keywords: cryptography, diffusion, confusion, security, encoding/decoding