2019 11th International Symposium on Advanced Topics in Electrical Engineering (ATEE)
This article studies the efficiency of an Ant Colony Optimization (ACO) algorithm for the solving... more This article studies the efficiency of an Ant Colony Optimization (ACO) algorithm for the solving of electromagnetic optimization problems. A modified version of the ACO for continuous optimization, namely ACOR, is used to solve two benchmark electromagnetic problems referring to the optimization of coils configuration. After choosing the appropriate size of the population of ants, the ACOR algorithm performances are compared with results obtained with other evolutionary computation strategies such as Genetic Algorithm (GA) and Particle Swarm Optimization (PSO). For one problem ACOR outperforms the variants of the GA and PSO used at all statistical criteria checked, whereas for the other problem ACOR is ranked as the second best after the PSO based algorithm tested.
2020 International Symposium on Fundamentals of Electrical Engineering (ISFEE)
The paper proposes a new set of Quantum-behaved Particle Swarm Optimization (QPSO) multi-objectiv... more The paper proposes a new set of Quantum-behaved Particle Swarm Optimization (QPSO) multi-objective algorithms with the final goal to use them for the optimization of a two objective electromagnetic benchmark inspired by a real world application. Starting from some various flavors of single-objective QPSO algorithms, known as: classic, with Gaussian attractor and with random mean, the new multi-objective QPSO implementations integrate principles inspired from notorious algorithms such as NSGA II, OMOPSO and ε-MOEA. The proposed algorithms are tested on benchmark problems proposed by scientific communities working in Evolutionary Computation and Computational Electromagnetics.
A natural coupling of a circuit with an electromagnetic device is possible if special boundary co... more A natural coupling of a circuit with an electromagnetic device is possible if special boundary conditions, called Electric Circuit Element (ECE), are used for the electromagnetic field formulation. This contribution shows how these ECE boundary conditions can be implemented into the 3D-finite element method for solving coupled full-wave electromagnetic (EM) field-circuit problems in the frequency domain. The frequency response allows the extraction of a reduced order model of the analyzed device, accounting for all the EM field effects. The implementation is based on a weak formulation that uses the electric field strength E strictly inside the domain and a scalar potential V defined solely at the boundary. Edge elements for E are used inside the three-dimensional domain and nodal elements for V are used on its two-dimensional boundary. The weak formulation is described and implemented in the free environment Open Numerical Engineering LABoratory (onelab) . The validation is carried...
The saltatory conduction is the way the action potential<br> is transmitted along a myelina... more The saltatory conduction is the way the action potential<br> is transmitted along a myelinated axon. The potential diffuses along<br> the myelinated compartments and it is regenerated in the Ranvier<br> nodes due to the ion channels allowing the flow across the membrane.<br> For an efficient simulation of populations of neurons, it is important<br> to use reduced order models both for myelinated compartments<br> and for Ranvier nodes and to have control over their accuracy and<br> inner parameters. The paper presents a reduced order model of this<br> neural system which allows an efficient simulation method for the<br> saltatory conduction in myelinated axons. This model is obtained<br> by concatenating reduced order linear models of 1D myelinated<br> compartments and nonlinear 0D models of Ranvier nodes. The<br> models for the myelinated compartments are selected from a series of<br> spatially distributed...
A distributed evolutionary strategy is used to solve the optimization problem in a subspace of th... more A distributed evolutionary strategy is used to solve the optimization problem in a subspace of the searching space. At each cost function evalu- ation, a deterministic method is used to find the op- timal solution in the complementary subspace, with an imposed accuracy. The proposed method is applied to solve TEAM Workshop problem 22 with the control of the numerical error. In this case the stochastic method acts in the geometrical parameter subspace and an em- bedded deterministic technique is used in the 2D cur- rent densities subspace.
A method to extract macromodels for RF MEMS switches is proposed. The macromodels include both th... more A method to extract macromodels for RF MEMS switches is proposed. The macromodels include both the coupled structural-electric behavior of the switch as well as its RF behavior. The device with distributed parameters is subject to several analyses from which the parameters of the macromodel are extracted, by model reduction. From the coupled structural-electrostatic analysis the parametric capacitance and the effective stiffness coefficients of the switch are extracted. From the RF characteristics in the up stable state, the transmission line parameters are extracted. Finally, all parameters are combined in a Spice circuit model, which is controlled by the MEMS actuation voltage and is excited with the RF signal. The procedure is applied to a capacitive switch. Relative modeling errors with respect to the non-reduced model, considered as reference, of less than 3 % for the RF characteristics and less than 1 % for the mechanical characteristics are obtained.
There are several ways of coupling field and circuit models but, in our opinion, the most natural... more There are several ways of coupling field and circuit models but, in our opinion, the most natural one uses Electric Circuit Element (ECE) boundary conditions (BC) for the field formulation. In our previous work we have successfully implemented full-wave (FW) frequency-domain E- and H-based formulations with ECE BC in the finite element method (FEM). Our interest was only in high frequency (HF) FW applications. At HF these formulations with an unknown field strictly inside the domain and a scalar potential with support only on the boundary, are stable and accurate. Motivated by the current desire of having all-purpose Maxwell’s solvers at hand, in this contribution we investigate their numerical stability when used at very low frequencies (LF). We show that these FW formulations can be used at LF provided that techniques to ensure the numerical stability are applied. If only global (energy based) stability is sought for, then just using the excitation which is essential for FEM migh...
2021 IEEE 27th International Symposium for Design and Technology in Electronic Packaging (SIITME), 2021
This paper is a commented review of our contributions related to the modelling and simulation of ... more This paper is a commented review of our contributions related to the modelling and simulation of interconnects, presented throughout the years in scientific communities that belong mainly to electrical engineering and mathematics. The emphasis is on important topics that we feel that are not yet much used by the electronics packaging community, such as the electric circuit element boundary conditions in the electromagnetic field formulation as well as complexity reduction techniques throughout the modelling flow.
2017 10th International Symposium on Advanced Topics in Electrical Engineering (ATEE), 2017
This contribution proposes a method to extract parametric reduced models that describe the couple... more This contribution proposes a method to extract parametric reduced models that describe the coupled structural-electric behavior of RF MEMS switches. The equivalent capacitance coefficients and the effective elastic coefficients are extracted from coupled structural-electrostatic analysis. Parametric models are built based on the sensitivities of the extracted equivalent coefficients. The method is validated on two benchmarks: one for which experimental values are available, and the other one from the literature. The results show that, for the tested configurations, even a model of order 1 or 2 can catch accurately enough (e.g. relative error of less than 5 %) the pull-in voltage of the full order model, for a variation of the investigated parameter of less than 20 %. Such a reduced parametric model is useful in early stages of the design.
Scalarization approaches are the simplest methods for solving the multiobjective problems. The id... more Scalarization approaches are the simplest methods for solving the multiobjective problems. The idea of scalarization is based on decomposition of multiobjective problems into single objective sub-problems. Every one of these sub-problems can be solved in a parallel manner since they are independent with each other. Hence, as a scalarization approach, systematically modification on the desirability levels of the objective values of multiobjective problems can be employed for solving these problems. In this study, desirability function-based scalarization approach is converted into parallel algorithm and applied into seven benchmark problems. The performance of parallel algorithm with respect to sequential one is evaluated based on execution time on different graphical processing units and central processing units. The results show that even the accuracy of parallel and sequential codes are same, the execution time of parallel algorithm is up to 24.5-times faster than the sequential a...
This paper investigates two methods of exploiting parallel architectures in order to efficiently ... more This paper investigates two methods of exploiting parallel architectures in order to efficiently extract reduced-order models for passive high-frequency integrated devices. Due to the high frequencies involved, all the effects of electromagnetic fields must be taken into account, but this leads to very large models which have to be reduced. We use an efficient technique, based on Vector Fitting, to extract these models with minimal computational effort, and we propose two methods of mapping the tasks on a parallel architecture.
COMPEL - The international journal for computation and mathematics in electrical and electronic engineering, 2020
Purpose The purpose of this paper is to propose a physics-aware algorithm to obtain radio frequen... more Purpose The purpose of this paper is to propose a physics-aware algorithm to obtain radio frequency (RF)-reduced models of micro-electromechanical systems (MEMS) switches and show how, together with multiphysics macromodels, they can be realized as circuits that include both lumped and distributed parameters. Design/methodology/approach The macromodels are extracted with a robust procedure from the solution of Maxwell’s equations with electromagnetic circuit element (ECE) boundary conditions. The reduced model is extracted from the simulations of three electromagnetic field problems, in full-wave regime, that correspond to three configurations: signal lines alone, switch in the up and down positions. Findings The technique is exemplified for shunt switches, but it can be extended for lateral switches. Moreover, the algorithm is able take frequency dependence into account both for the signal lines and for the switch model. For the later, the order of the model is increased until a sp...
Two parallelization techniques, GPGPU and Pthreads for multiprocessor architectures, are used to ... more Two parallelization techniques, GPGPU and Pthreads for multiprocessor architectures, are used to implement a SPSO algorithm in order to solve electromagnetic optimization problems. Several configurations for the GPGPU implementation are tested and a new full parallel minimum branching implementation is proposed. The best GPGPU approaches are then compared with a Pthreads implementation in terms of speed up and solution quality. To test the efficiency of the parallelization techniques two electromagnetic optimization problems were chosen, namely the TEAM22 benchmark and Loney’s solenoid. In the end the paper provides suggestions regarding what parallelization technique should be used considering the implementation features of the optimization function.
2019 11th International Symposium on Advanced Topics in Electrical Engineering (ATEE)
This article studies the efficiency of an Ant Colony Optimization (ACO) algorithm for the solving... more This article studies the efficiency of an Ant Colony Optimization (ACO) algorithm for the solving of electromagnetic optimization problems. A modified version of the ACO for continuous optimization, namely ACOR, is used to solve two benchmark electromagnetic problems referring to the optimization of coils configuration. After choosing the appropriate size of the population of ants, the ACOR algorithm performances are compared with results obtained with other evolutionary computation strategies such as Genetic Algorithm (GA) and Particle Swarm Optimization (PSO). For one problem ACOR outperforms the variants of the GA and PSO used at all statistical criteria checked, whereas for the other problem ACOR is ranked as the second best after the PSO based algorithm tested.
2020 International Symposium on Fundamentals of Electrical Engineering (ISFEE)
The paper proposes a new set of Quantum-behaved Particle Swarm Optimization (QPSO) multi-objectiv... more The paper proposes a new set of Quantum-behaved Particle Swarm Optimization (QPSO) multi-objective algorithms with the final goal to use them for the optimization of a two objective electromagnetic benchmark inspired by a real world application. Starting from some various flavors of single-objective QPSO algorithms, known as: classic, with Gaussian attractor and with random mean, the new multi-objective QPSO implementations integrate principles inspired from notorious algorithms such as NSGA II, OMOPSO and ε-MOEA. The proposed algorithms are tested on benchmark problems proposed by scientific communities working in Evolutionary Computation and Computational Electromagnetics.
A natural coupling of a circuit with an electromagnetic device is possible if special boundary co... more A natural coupling of a circuit with an electromagnetic device is possible if special boundary conditions, called Electric Circuit Element (ECE), are used for the electromagnetic field formulation. This contribution shows how these ECE boundary conditions can be implemented into the 3D-finite element method for solving coupled full-wave electromagnetic (EM) field-circuit problems in the frequency domain. The frequency response allows the extraction of a reduced order model of the analyzed device, accounting for all the EM field effects. The implementation is based on a weak formulation that uses the electric field strength E strictly inside the domain and a scalar potential V defined solely at the boundary. Edge elements for E are used inside the three-dimensional domain and nodal elements for V are used on its two-dimensional boundary. The weak formulation is described and implemented in the free environment Open Numerical Engineering LABoratory (onelab) . The validation is carried...
The saltatory conduction is the way the action potential<br> is transmitted along a myelina... more The saltatory conduction is the way the action potential<br> is transmitted along a myelinated axon. The potential diffuses along<br> the myelinated compartments and it is regenerated in the Ranvier<br> nodes due to the ion channels allowing the flow across the membrane.<br> For an efficient simulation of populations of neurons, it is important<br> to use reduced order models both for myelinated compartments<br> and for Ranvier nodes and to have control over their accuracy and<br> inner parameters. The paper presents a reduced order model of this<br> neural system which allows an efficient simulation method for the<br> saltatory conduction in myelinated axons. This model is obtained<br> by concatenating reduced order linear models of 1D myelinated<br> compartments and nonlinear 0D models of Ranvier nodes. The<br> models for the myelinated compartments are selected from a series of<br> spatially distributed...
A distributed evolutionary strategy is used to solve the optimization problem in a subspace of th... more A distributed evolutionary strategy is used to solve the optimization problem in a subspace of the searching space. At each cost function evalu- ation, a deterministic method is used to find the op- timal solution in the complementary subspace, with an imposed accuracy. The proposed method is applied to solve TEAM Workshop problem 22 with the control of the numerical error. In this case the stochastic method acts in the geometrical parameter subspace and an em- bedded deterministic technique is used in the 2D cur- rent densities subspace.
A method to extract macromodels for RF MEMS switches is proposed. The macromodels include both th... more A method to extract macromodels for RF MEMS switches is proposed. The macromodels include both the coupled structural-electric behavior of the switch as well as its RF behavior. The device with distributed parameters is subject to several analyses from which the parameters of the macromodel are extracted, by model reduction. From the coupled structural-electrostatic analysis the parametric capacitance and the effective stiffness coefficients of the switch are extracted. From the RF characteristics in the up stable state, the transmission line parameters are extracted. Finally, all parameters are combined in a Spice circuit model, which is controlled by the MEMS actuation voltage and is excited with the RF signal. The procedure is applied to a capacitive switch. Relative modeling errors with respect to the non-reduced model, considered as reference, of less than 3 % for the RF characteristics and less than 1 % for the mechanical characteristics are obtained.
There are several ways of coupling field and circuit models but, in our opinion, the most natural... more There are several ways of coupling field and circuit models but, in our opinion, the most natural one uses Electric Circuit Element (ECE) boundary conditions (BC) for the field formulation. In our previous work we have successfully implemented full-wave (FW) frequency-domain E- and H-based formulations with ECE BC in the finite element method (FEM). Our interest was only in high frequency (HF) FW applications. At HF these formulations with an unknown field strictly inside the domain and a scalar potential with support only on the boundary, are stable and accurate. Motivated by the current desire of having all-purpose Maxwell’s solvers at hand, in this contribution we investigate their numerical stability when used at very low frequencies (LF). We show that these FW formulations can be used at LF provided that techniques to ensure the numerical stability are applied. If only global (energy based) stability is sought for, then just using the excitation which is essential for FEM migh...
2021 IEEE 27th International Symposium for Design and Technology in Electronic Packaging (SIITME), 2021
This paper is a commented review of our contributions related to the modelling and simulation of ... more This paper is a commented review of our contributions related to the modelling and simulation of interconnects, presented throughout the years in scientific communities that belong mainly to electrical engineering and mathematics. The emphasis is on important topics that we feel that are not yet much used by the electronics packaging community, such as the electric circuit element boundary conditions in the electromagnetic field formulation as well as complexity reduction techniques throughout the modelling flow.
2017 10th International Symposium on Advanced Topics in Electrical Engineering (ATEE), 2017
This contribution proposes a method to extract parametric reduced models that describe the couple... more This contribution proposes a method to extract parametric reduced models that describe the coupled structural-electric behavior of RF MEMS switches. The equivalent capacitance coefficients and the effective elastic coefficients are extracted from coupled structural-electrostatic analysis. Parametric models are built based on the sensitivities of the extracted equivalent coefficients. The method is validated on two benchmarks: one for which experimental values are available, and the other one from the literature. The results show that, for the tested configurations, even a model of order 1 or 2 can catch accurately enough (e.g. relative error of less than 5 %) the pull-in voltage of the full order model, for a variation of the investigated parameter of less than 20 %. Such a reduced parametric model is useful in early stages of the design.
Scalarization approaches are the simplest methods for solving the multiobjective problems. The id... more Scalarization approaches are the simplest methods for solving the multiobjective problems. The idea of scalarization is based on decomposition of multiobjective problems into single objective sub-problems. Every one of these sub-problems can be solved in a parallel manner since they are independent with each other. Hence, as a scalarization approach, systematically modification on the desirability levels of the objective values of multiobjective problems can be employed for solving these problems. In this study, desirability function-based scalarization approach is converted into parallel algorithm and applied into seven benchmark problems. The performance of parallel algorithm with respect to sequential one is evaluated based on execution time on different graphical processing units and central processing units. The results show that even the accuracy of parallel and sequential codes are same, the execution time of parallel algorithm is up to 24.5-times faster than the sequential a...
This paper investigates two methods of exploiting parallel architectures in order to efficiently ... more This paper investigates two methods of exploiting parallel architectures in order to efficiently extract reduced-order models for passive high-frequency integrated devices. Due to the high frequencies involved, all the effects of electromagnetic fields must be taken into account, but this leads to very large models which have to be reduced. We use an efficient technique, based on Vector Fitting, to extract these models with minimal computational effort, and we propose two methods of mapping the tasks on a parallel architecture.
COMPEL - The international journal for computation and mathematics in electrical and electronic engineering, 2020
Purpose The purpose of this paper is to propose a physics-aware algorithm to obtain radio frequen... more Purpose The purpose of this paper is to propose a physics-aware algorithm to obtain radio frequency (RF)-reduced models of micro-electromechanical systems (MEMS) switches and show how, together with multiphysics macromodels, they can be realized as circuits that include both lumped and distributed parameters. Design/methodology/approach The macromodels are extracted with a robust procedure from the solution of Maxwell’s equations with electromagnetic circuit element (ECE) boundary conditions. The reduced model is extracted from the simulations of three electromagnetic field problems, in full-wave regime, that correspond to three configurations: signal lines alone, switch in the up and down positions. Findings The technique is exemplified for shunt switches, but it can be extended for lateral switches. Moreover, the algorithm is able take frequency dependence into account both for the signal lines and for the switch model. For the later, the order of the model is increased until a sp...
Two parallelization techniques, GPGPU and Pthreads for multiprocessor architectures, are used to ... more Two parallelization techniques, GPGPU and Pthreads for multiprocessor architectures, are used to implement a SPSO algorithm in order to solve electromagnetic optimization problems. Several configurations for the GPGPU implementation are tested and a new full parallel minimum branching implementation is proposed. The best GPGPU approaches are then compared with a Pthreads implementation in terms of speed up and solution quality. To test the efficiency of the parallelization techniques two electromagnetic optimization problems were chosen, namely the TEAM22 benchmark and Loney’s solenoid. In the end the paper provides suggestions regarding what parallelization technique should be used considering the implementation features of the optimization function.
Uploads
Papers by Gabriela Ciuprina