Showing papers in "International Journal of Numerical Modelling-electronic Networks Devices and Fields in 2018"

Attitude determination by combining arrays of MEMS accelerometers, gyros, and magnetometers via quaternion‐based complementary filter

Abstract: In this study, an Attitude and Heading Reference System (AHRS) consisting of 5 modules is designed where each module has a triaxial gyroscope, accelerometer, and magnetometer. First, a method based on the Levenberg-Marquardt algorithm (LMA) is utilized to correct the bias error, scale factor and axes nonorthogonality. Also, the data from the 5 modules of AHRS are ensemble averaged to reduce the adverse effects of high-frequency noises. Then, the obtained trends are used in an orientation estimation algorithm based on a complementary filter algorithm. In this algorithm, the dynamical accelerations are first decreased via a low-pass filter. Afterwards, it is determined using an algorithm whether the system is experiencing magnetic distortion or not. If distortion is verified, magnetometers' data are discarded, as it will introduce noticeable error in estimating heading angle. In this case, the heading angle starts to diverge; however, employing 5 modules in the system decreases divergence rate noticeably, such that after 2 minutes in quasi-static conditions, the pitch, roll, and heading angles' errors decrease, respectively, from 36.195°, 23.201°, and 12.541° when only 1 module is used to 2.511°, 3.972°, and 0.984° when all 5 modules are used. Moreover, in dynamical conditions, these errors decrease from 37.916°, 13.633°, and 13.071° to 6.514°, 5.961°, and 0.284°. Once the distortion is removed from magnetic field, the magnetometers' data are used again to correct the heading error. The obtained results show that the root mean square (RMS) errors of pitch, roll, and heading angles in quasi-static conditions are 0.536°, 0.323°, and 0.601°, whereas in dynamical condition, they are 1.267°, 1.535°, and 0.994°, respectively.

Principal component analysis for efficient characterization of stochastic electromagnetic fields

Abstract: The propagation of stochastic electromagnetic fields can be accurately modeled using the auto- and cross-correlation spectra of the field components. In this work, we introduce the framework of principal component analysis for reducing the computational cost of handling stochastic electromagnetic fields described by correlation matrices. We consider noisy electromagnetic fields originating from stationary random processes with Gaussian probability distribution. The amount of data obtained by 2-dimensional near-field measurements and by determining the auto- and cross-correlation information of electromagnetic interference can become burdensome for further processing even for problems of moderate size. For obtaining the correlation data, 2 measurement probes have to scan a defined grid of measurement points. For each pair of points, the spatial correlations need to be calculated, and hence, the data obtained scales quadratically with the number of sampling points. To reduce the amount of data, we project the given data obtained by measurement or simulation onto directions of maximum variation, the so called principal components, and keep only those principal components which contribute most to the total variance. In this way, the memory demand for storage and further computation of stochastic electromagnetic fields can be reduced significantly.

Electrothermal formulation of the partial element equivalent circuit method

Abstract: This paper presents a combined electrothermal formulation of the partial element equivalent circuit (PEEC) method. It is shown that the distributed thermal problem can be formulated as an equivalent thermal network that shares the same topological and geometrical properties of the PEEC equivalent model. Temperature-dependent thermal heat conduction and convection models are adopted thus leading to a time-varying thermal network, which is studied by using pertinent techniques along with the electrical equivalent circuit. The proposed electrothermal PEEC solver is used to analyze 2 problems of interest and verified through comparison with a commercial software.

Graphics processing unit-based solution of nonlinear Maxwell's equations for inhomogeneous dispersive media

Abstract: Summary A new approach is developed for fast solution of complex dynamic problems in nonlinear optics. The model consists of the nonlinear Maxwell's equations coupled with time-dependent electron density equation. The approach is based on the Finite-Difference Time-Domain and the auxiliary differential equation methods for frequency-dependent Drude media with a time-dependent carrier density, changing due to Kerr, photoionization, avalanche, and recombination effects. The system of nonlinear Maxwell-Ampere equations is solved by an iterative fixed-point procedure. The proposed approach is shown to remain stable even for complex nonlinear media and strong gradient fields. Graphics-processing-units technique is implemented by using an efficient algorithm enabling solution of massively 3-dimensional problems within reasonable computation time.

Impact of source pocket doping on RF and linearity performance of a cylindrical gate tunnel FET

Abstract: The paper presents a cylindrical gate tunnel (CGT) field effect transistors (FETs) with a highly doped pocket layer introduced in the source region. The presence of pocket doped layer in the source provides higher lateral electric field and band-to-band tunneling (BTBT) generation rate in the vicinity of tunneling junction which in turn increases the drain current and transconductance significantly. Also, the linearity and radio frequency (RF) performance of the CGT FET with source pocket doping (CGTS) have been extensively investigated. The different linearity and RF figure of merits such as gmn, VIP2, VIP3, IIP3, ZCP, 1-dB compression point, GBWP, TFP, unity gain cut-off frequency, and maximum oscillation frequency of the present device are extracted and compared with the results of conventional CGT. The results exhibit superior linearity and RF performance along with improved current carrying capability of the proposed device. Thus, the device can be one of the possible contenders to replace bulk MOSFET in high-frequency microwave applications. The accuracy of both the devices is validated by TCAD Sentaurus simulator.

A spiral planar inductor: An experimentally verified physically based model for frequency and time domains

Abstract: The results of a comprehensive investigation of spiral inductors are presented. A physically based model includes the self and mutual inductances, the capacitances between adjacent turns, the substrate capacitance, and the ohmic loss effects proposed. Heterogeneous simulation scheme, including circuit and device models during time and frequency domains, is successfully implemented in VHDL-AMS language and is simulated numerically in Simplorer platform. The model has been validated during frequency domain with measurement data of spirals having different geometries and various sizes. Moreover, the fabricated spiral inductors are evaluated in DC-DC power converter to benchmark the model accuracy. Simulation and experimental results show excellent agreement. Validity domain is discussed.

Advancements in the identification of passive RC networks for compact modeling of thermal effects in electronic devices and systems

Abstract: We deal with the problem of identifying the parameters of an equivalent lumped RC multiport network from time domain tabulated data of a corresponding distributed real eigenvalues problem, as obtained from either measurements or accurate simulation tools. A novel step response matrix identification procedure is proposed, based on the combination of an outer nonlinear least squares iteration for the relocation of time constants, and an inner convex programming cycle for the identification of the corresponding residue matrix terms, able to guarantee a priori the passivity property of the equivalent RC network. The structure of the algorithm and its principal functions connections are shown, and the mathematical features of the proposed formulation of the identification problem are accurately described and commented. Moreover, the possibility of getting direct synthesis of a Foster generalized concretely passive multiport, as a consequence of the optimal identification of a passive real eigenvalues model from data, is discussed. Since the technique is well suited (although not limited) to the reduced analysis of typical electro-thermal problems, a set of significant case studies in this area is considered. In this way the algorithm present implementation is validated, also comparing it to previous well assessed methods, evidencing its large potential and value.

Adaptive controller with RBF neural network for induction motor drive

Abstract: In this paper, the radial basis function neural network-based model reference adaptive speed control for vector controlled induction motor drive system is presented. The speed control of induction motors is challenging because of their complex mathematical model, non-linear structure, and time varying dynamics. The radial basis function neural network is used to compensate the non-linearity which comes from the non-linear state equations of induction motor model. Neural network parameters are online updated via gradient descent algorithm to minimize the error. The drive system has been tested under various operating conditions. This paper demonstrated benefits of the proposed control approach by comparing the algorithm to conventional PI controllers. The results show that the proposed controller ensures good robustness and stable operation of the system under variable speed and variable loads than the PI controller.

Waveport modeling for the DGTD simulation of electromagnetic devices

Abstract: The issue of waveport modeling for the numerical simulation of electromagnetic devices using the discontinuous Galerkin time-domain (DGTD) method is investigated. A new implementation of the waveport boundary condition (WPBC) is presented along with a fast evaluation of the required convolution. The performance of the WPBC for the DGTD simulation is then compared with that of the first-order absorbing boundary condition and the uniaxial perfectly matched layer (UPML). Two numerical examples are presented to demonstrate the waveport modeling for the DGTD simulation of electromagnetic devices based on the WPBC and UPML.

Explicit Time Integration of Transient Eddy Current Problems

Abstract: For time integration of transient eddy current problems commonly implicit time integration methods are used, where in every time step one or several nonlinear systems of equations have to be linearized with the Newton-Raphson method due to ferromagnetic materials involved. In this paper, a generalized Schur-complement is applied to the magnetic vector potential formulation, which converts a differential-algebraic equation system of index 1 into a system of ordinary differential equations (ODE) with reduced stiffness. For the time integration of this ODE system of equations, the explicit Euler method is applied. The Courant-Friedrich-Levy (CFL) stability criterion of explicit time integration methods may result in small time steps. Applying a pseudo-inverse of the discrete curl-curl operator in nonconducting regions of the problem is required in every time step. For the computation of the pseudo-inverse, the preconditioned conjugate gradient (PCG) method is used. The cascaded Subspace Extrapolation method (CSPE) is presented to produce suitable start vectors for these PCG iterations. The resulting scheme is validated using the TEAM 10 benchmark problem.

Passivity verification and enforcement—A review paper

Abstract: Today's advanced high-speed technology requires building behavioral models of systems, from measured/simulated data, with increasingly higher operating frequencies. Different research efforts have come to light in literature for macromodeling, based on such types of data. Ensuring passivity is one of the most fundamental issues affecting system macromodeling, because system-level performance can be unstable if even a single component of the system becomes nonpassive. Thus, from the computer-aided design perspective, generation and provision of macromodels, while preserving passivity, is a significant challenge. In this paper, we carefully review novel techniques of passive macromodeling as well as their passivity verification and enforcement, from the early days to the present. We critically review the prominent methods in literature developed for macromodeling, and for verification and enforcement of passivity, and emphasize on their strengths and shortcomings.

Fast redesign and geometry scaling of multiband antennas using inverse surrogate modeling techniques

Abstract: Redesign of antenna structures for various operating conditions or substrate parameters is a practically important problem in antenna engineering. At the same time, it is challenging because finding an optimum set of geometry parameters for a specified operating/material conditions requires going through an entire design optimization process. The problem is even more challenging for multiband antennas where 2 or more bands need to be independently controlled and the joint influence of some of antenna dimensions on several bands has to be taken into account. In this work, we propose a simple technique for rapid dimension scaling of multiband antennas, which uses inverse surrogate modeling techniques. The inverse surrogate model constructed using a set of reference designs optimized for several values of operating frequencies allows us to predict optimum dimensions of the antenna structure of interest corresponding to a required set of operating frequencies. The model is constructed at the level of coarse-discretization electromagnetic simulations; therefore, it undergoes enhancement so as to be reliably used to redesign the antenna at the level of its high-fidelity electromagnetic model. A correction procedure is further applied to accommodate the initial scaling errors. Our approach is demonstrated using a triple-band uniplanar dipole antenna.

Nodal discontinuous Galerkin method for high-temperature superconductors modeling based on the H-formulation

Abstract: There is growing interest in superconducting machinery design in AC regime such as generators, motors, and magnets. High-temperature superconductors are used as tapes or cables for the machine windings. Therefore, AC losses have to be evaluated efficiently for an optimal design. Moreover, non-linearities, arising from the power law characterizing the electrical behaviour of superconductor, have to be also dealt with. The development of efficient numerical methods is therefore critical to model high-temperature superconductors. Although numerous methods have already been proposed, the finite element method applied on the time-dependent curl-curl–based H-formulation remains the mostly used. It uses edge elements of Nedelec to naturally ensure the continuity of the tangential components of H. To take into account non-linearities from superconductors, a linearisation of the superconductors constitutive power law E=ρ(J)J was implemented. Discontinuous Galerkin finite element method provides an interesting alternative to edge elements for the time-dependent H-formulation. Indeed, the curl-curl operator is written as a div operator and the interior penalty approach defines numerical fluxes at the interfaces. Those fluxes will ensure the continuity of the tangential components of H. The resistivity ρ(J) is explicitly evaluated at the previous iteration with such an approach leading to convergence. In this paper, both numerical approaches implementation will be presented. We will also compare the numerical results from those methods applied to 3D modelling of simple superconducting geometries in AC regime.

Acquisition of PN sequences using multilayer perceptron neural network adaptive processor for multiuser detection in spread-spectrum communication systems

Abstract: A non-coherent serial acquisition scheme for direct sequence spread spectrum communication systems is analyzed and discussed in this paper. The adaptive thresholding based on constant false alarm rate and multilayer perceptron neural network (MLP-NN) techniques are combined to improve the performance of code division multiple access systems. One of the most important problems in code acquisition of pseudo-noise sequences for multiuser detection is the presence of interferences caused by the multiple access technique and multipath replicas. To solve this problem, an MLP-NN is trained and adapted to work as a constant false alarm rate detector using the error back propagation gradient descendent algorithm. It is named MLP-NN adaptive processor. The performance of this proposed algorithm is presented using the serial search acquisition system, which is chosen because of its simple hardware implementation. The performance of the MLP-NN adaptive processor algorithm in homogeneous and non-homogenous environments for additive white Gaussian noise and Rayleigh fading channels is evaluated via computer simulations. The obtained results are compared to other serial acquisition schemes using the cell-averaging adaptive processor, the order statistics adaptive processor, and the automatic censoring adaptive processor algorithms.

Parallel preconditioners and high order elements for microwave imaging

Abstract: This paper combines the use of high order finite element methods with parallel preconditioners of domain decomposition type for solving electromagnetic problems arising from brain microwave imaging. The numerical algorithms involved in such complex imaging systems are computationally expensive since they require solving the direct problem of Maxwell's equations several times. Moreover, wave propagation problems in the high frequency regime are challenging because a sufficiently high number of unknowns is required to accurately represent the solution. In order to use these algorithms in practice for brain stroke diagnosis, running time should be reasonable. The method presented in this paper, coupling high order finite elements and parallel preconditioners, makes it possible to reduce the overall computational cost and simulation time while maintaining accuracy.