Two algorithms for short-term electricity demand forecasting in the regional electricity distribution network are presented. Several approaches - feedforward neural network, adaptive modelling and fuzzy modelling - ar...
详细信息
Two algorithms for short-term electricity demand forecasting in the regional electricity distribution network are presented. Several approaches - feedforward neural network, adaptive modelling and fuzzy modelling - are applied to the forecast. Two different models are designed. A one hour forecasting is based on the General Regression Neural Network (GRNN) model and Principle Component Analysis. The multi-model with adaptive features and fuzzy reasoning is used for a longer-term forecast.
The problems of nonlinear complex system control design are still in the centre of attention and there exists no universal solution by now. The aim of this paper is to propose the control algorithm for nonlinear syste...
详细信息
The problems of nonlinear complex system control design are still in the centre of attention and there exists no universal solution by now. The aim of this paper is to propose the control algorithm for nonlinear systems consisting of several subsystems. The subsystem nonlinearities are assumed to be known and the state variables are accessible only locally. The interactions between subsystems can be unknown, but bounded. The proposed control algorithm for such class of complex systems is based on the feedback linearization technique with automatic adjusting of the subsystem decay rate. The control performances are illustrated by an example of nonlinear system consisting of two interconnected subsystems.
Presents theoretical and experimental investigations into the dynamic modelling and characterisation of a flexible manipulator system. A constrained planar single-link flexible manipulator is considered. A dynamic mod...
详细信息
Presents theoretical and experimental investigations into the dynamic modelling and characterisation of a flexible manipulator system. A constrained planar single-link flexible manipulator is considered. A dynamic model of the system is developed based on finite element methods incorporating structural damping. Performance of the algorithm in describing the dynamic behaviour of the system is assessed in comparison to an experimental test-rig in the time and frequency domains.
This paper deals with the supervisory control of the discrete event dynamic systems (DEDS). It contains descriptions of the different approaches to the supervisory control synthesis, their comparison and analysis. Fur...
详细信息
This paper deals with the supervisory control of the discrete event dynamic systems (DEDS). It contains descriptions of the different approaches to the supervisory control synthesis, their comparison and analysis. Further and implementation of the method based on the place invariants of a Petri net is described. It also outlines possibilities for further enhancements of the considered methods.
In this paper, a normalized stochastic gradient algorithm is proposed for learning in feedforward neural networks. By using a new methodology based on the martingale convergence results, asymptotic properties of the a...
详细信息
ISBN:
(纸本)0780355121
In this paper, a normalized stochastic gradient algorithm is proposed for learning in feedforward neural networks. By using a new methodology based on the martingale convergence results, asymptotic properties of the algorithm are analyzed. It is proved that, in general, the sequence of the algorithm states converges with probability one to the set of zeroes of the gradient of the criterion function locally on the event where it is bounded. Then, these results are applied to learning in multilayer perceptrons.
This is a contribution to PDS 2000 International Conference concerning methods of using Matlab computing environment to solve a wide range of engineering problems and to make courseware for engineering students. The f...
详细信息
This is a contribution to PDS 2000 International Conference concerning methods of using Matlab computing environment to solve a wide range of engineering problems and to make courseware for engineering students. The following subjects of Theory of the Dynamics Systems and of Signal Processing are used as examples in this paper. The purpose of the paper is to make technical students familiar with Matlab language environment, which is becoming more and more commonly used in engineering problem solving.
A practical adaptive control application is reported. In coal-burned thermal plant, the air supply rate is controlled using a simple rion-linear model of carbon monoxide generation. Parameters of the model are on-line...
详细信息
A practical adaptive control application is reported. In coal-burned thermal plant, the air supply rate is controlled using a simple rion-linear model of carbon monoxide generation. Parameters of the model are on-line estimated using the standard least-squares estimation technique. This model is used for the the optimal air/fuel ratio re-calculation, which should ensure that the steady state CO concentration would exceed a given limit with only a small probability (cautious optimization). Both the limit and probability threshold are the two principal input parameters. To avoid covariance wind-up, the regularized forgetting algorithm is used. To model the fact the parameter variations are caused mainly by the fuel quality fluctuations, the forgetting is directional. The output data are provided by CO concentration sensors, which may get saturated or clogged. Mathematical solutions to both problems is presented.
作者：
Branislav HrúzSlovak University of Technology
Faculty of Electrical Engineering and Information Technology Department of Automatic Control Systems llkovièova 3 812 19 Bratislava Republic Phone: +421 7 60291698
The scheduling problem is stated and formally specified in the paper. The system where the problem is to be solved is represented by a timed Petri net pertaining to a new class suitable for it. Then a method based on ...
详细信息
The scheduling problem is stated and formally specified in the paper. The system where the problem is to be solved is represented by a timed Petri net pertaining to a new class suitable for it. Then a method based on the defined Petri net class for the scheduling problem solution is presented. The method is illustrated on an example.
The paper interests in continuous diagnostics of the heart of the old people by the transmission of ECG signal through voice channel of mobile telephone. The ECG signal is measured during active life, it is compressed...
详细信息
The paper interests in continuous diagnostics of the heart of the old people by the transmission of ECG signal through voice channel of mobile telephone. The ECG signal is measured during active life, it is compressed through 24 hours and stored in the memory to be possible a long term ECG diagnostics. The time compression is supported by wavelet analysis in special processor. The implementation of "Wavelet Analyser" uses custom design and FPGA circuitry. The ECG signal is compressed in amplitude, and it is equipped by redundancy of error control code before transmission by the voice channel of mobile telephone line.
作者：
V. VeselýI. SekajDepartment of Automatic Control Systems
Faculty of Electrical Engineering and Information Technology Slovak University of Technology Ilkovičova 3 812 19 Bratislava Slovak Republic Tel : 0421 7 602 91 667
The robust output feedback stabilization is tackled using a genetic algorithm optimisation procedure for uncertain linear parametric systems in the time domain. The necessary and sufficient conditions for static outpu...
详细信息
The robust output feedback stabilization is tackled using a genetic algorithm optimisation procedure for uncertain linear parametric systems in the time domain. The necessary and sufficient conditions for static output feedback stabilizability of linear systems are the basics of the proposed robust controller design procedure. The nature of genetic algorithms predetermines them for optimising also objectives that involve computation of feedback control.
暂无评论