Mohamed H. El-Shafey

Application of a new frequency estimation technique to power systems

A new technique for estimating the frequency contents in a signal is applied to power systems. The technique uses two sets of measured samples: samples from the signal itself and samples from a filtered signal obtained by passing the original signal through a continuous-time filter. The estimates of the frequencies contained in the signal are obtained from the eigenvalues of the generalized eigenproblem of two matrices formed from the two sets of samples. The technique assumes estimating multiple sinusoids which makes it suitable in cases with large harmonic distortion. Results show that by the proper choice of the sampling time, the eigenproblem condition is improved such that accurate estimates become attainable in case of noise. Further improvement in the estimate accuracy is achievable by utilizing correlated samples. The technique is also applied to estimating varying frequency.

A new architecture for accurate dot product of floating point numbers

Many techniques were proposed to improve the accuracy of floating point operations such as addition, multiplication, and dot product. The purpose of such technique is to reduce the effect of rounding error. This paper introduces an efficient hardware implementation for accurate dot product. The proposed implementation was configured as a custom instruction in the ALTERA NiosII soft processor core. The computed result from the proposed method is as accurate as other algorithms and faster than them. Another advantage is that it has a linear time complexity, without any limitation on the vector length.

Accurate floating-point operation using controlled floating-point precision

Rounding and accumulation of errors when using floating point numbers are important factors in computer arithmetic. Many applications suffer from these problems. The underlying machine architecture and representation of floating point numbers play the major role in the level and value of errors in this type of calculations. A quantitative measure of a system error level is the machine epsilon. In the current representation of floating point numbers, the machine epsilon can be as small as 9.63E-35 in the 128 bit version of IEEE standard floating point representation system. In this work a novel solution that guarantees achieving the desired minimum error regardless of the machine architecture is presented. The proposed model can archive a machine epsilon of about 4.94E-324. A new representation model is given and a complete arithmetic system with basic operations is presented. The accuracy of the proposed method is verified by inverting a high order, Hilbert matrix, an ill-conditioned matrix that cannot be solved in the traditional floating point standard. Finally some comparisons are given.

Effect of the Sampling Time on the Accuracy of Estimating the Frequencies of Multiple Sinusoids

The accuracy of the estimates of the frequencies of multiple sinusoids in noise depends on many factors. Among these factors are the signal to noise ratio, the estimation technique, the number of sinusoids and their relative values. An estimation technique was suggested recently that yields the required frequencies as the generalized eigenvalues of two matrices formed from samples of the sinusoids and a filtered signal of these sinusoids. It was found that the sampling time has a profound effect on the accuracy of the resulting estimates. Recent interest in eigenstructure techniques in engineering applications led to increasing interest in the mathematical literature addressing the generalized eigenproblem. Recent studies exposed the relation between the condition of the generalized eigenproblem and the condition of a corresponding Vandermonde matrix. In this paper it is shown that the condition of the Vandermonde matrix can be influenced by the choice of the sampling time, consequently improving the generalized eigenproblem condition and making the frequency estimation technique more immune to noise

ESPRIT condition in signal parameter estimation

ESPRIT is known of high resolution in the general problem of signal parameter estimation. It can be applied to a wide variety of problems including accurate detection and estimation of sinusoids in noise, and estimation of signal direction-of-arrival. ESPRIT comprises the solution of two eigenvalue problems. The first is to obtain the eigen-decomposition of the signal correlation matrix. Based on the rotational invariance property of the eigenvectors of the matrix obtained in the first step, the second eigen-problem is formed from different rows of these eigenvectors. In this paper it is shown that the second eigen-problem is a generalized eigen-problem and the accuracy of ESPRIT estimates depends mainly on the condition of this generalized eigen-problem. It is shown that the problem condition depends on the sampling time of the correlation matrix. Numerical results illustrates the impact of the sampling time on the problem condition and consequently on the estimates accuracy.

Improving model predictive controller by using non-uniform sampled model

The problem respected in this paper is to design a model predictive controller (MPC), for systems with fast and slow dynamics (e.g. active suspension in road vehicles). Fast sampling rates needed by these systems will lead to high levels of control force generated by MPC. A new non-uniform sampled model technique is used to improve of the MPC performance and decrease control levels. A quarter car active suspension system is considered in simulation results to illustrate the effectiveness of the proposed technique compared with uniform sampled model, to enhance the vehicle performance regarding ride comfort and road handling with minimum control effort.

Towards Real-Time Networked Embedded Generalized Predictive Control for Automotive Active Suspension Systems

Abstract This paper emphasizes the implementation challenge of real-time embedded generalized predictive control (GPC) when applied to a class of automotive active suspension systems. A real-time experimental environment is specially designed in order to verify parameters tuning in closed-loop with the virtual process environment. Real-time experimentation carried on this environment shows the necessity of high prediction and control horizons setting to achieve acceptable comfort performance. A control algorithm is decomposed into parallel real-time computations using a set of systolic array based processors. A systemC model is constructed to verify the proposed implementation. Simulation and profiling results show the feasibility and the efficiency of GPC in future hardware implementation.

Application of eigen- decomposition of signals to harmonic noise detection and cancellation, in electric power systems

A new technique for estimating the frequency contents in a signal is applied to a three-phase shunt active power filter (APF) to compensate the harmonics and reactive power introduced by the nonlinear loads. The technique utilizes two sets of measured samples: samples from the signal itself and samples from a filtered signal obtained by passing the original signal through a continuous-time filter. The estimates of the frequencies contained in the signal are obtained from the eigenvalues of the generalized eigenproblem of two matrices formed from the two sets of samples. Upon detecting the harmonic components of the load currents by the proposed technique, an APF injects these harmonic current components with same magnitude but opposite phase into the power system to eliminate these harmonics and keep the utility line current sinusoidal and in phase with the utility voltage. Simulation results with different nonlinear loads are presented to confirm the validity of the proposed technique.

Fuzzy Controller for Mobile Robots using Evolvable Hardware

In this paper, a new method for solving the kinematics of a mobile robot and obstacles avoidance is proposed. The method uses a genetic algorithm (GA) in solving the kinematics. Using this technique, the robot controller can adjust itself autonomously even with a malfunction in the robot. The new method was implemented and tested with different type of robot malfunctions. The results showed that, using the new method, the robot succeeds to perform its task even with a malfunction.