Time-domain Sensitivity Analysis Research Articles

Dual Null Detection Points Removal and Time-Domain Sensitivity Analysis of a Self-Injection-Locked Radar for Small-Amplitude Motion Sensing

Compared with a conventional Doppler radar in which target displacement results in phase modulation, target displacement in a self-injection-locked oscillator (SILO)-based radar results in frequency modulation. Since the instantaneous phase is the time integral of frequency, the SILO-based radar is considered to have higher sensitivity than the conventional Doppler radar. However, a direct comparison between the SILO-based radar and the conventional Doppler radar is still needed to quantitatively understand the sensitivity difference. This article provides a rigorous time-domain analysis of the sensitivity of the SILO-based radar with a delay-based frequency demodulator for small-amplitude periodic motions. The analysis follows an approach similar to the time-domain analysis for the conventional Doppler radar. It reveals that two null detection point mechanisms could arise in SILO-based radar systems. The target distance may introduce one null detection point, and the delay added in the frequency demodulator may introduce another. Furthermore, this article also proposes an SILO-based radar with dual null detection points’ removal. Unlike the Doppler radar, where <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$I/Q$ </tex-math></inline-formula> demodulation eliminates the null detection point introduced by the target distance, <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$I/Q$ </tex-math></inline-formula> demodulation in the SILO-based radar only eliminates the null detection point introduced by the delay cable. To overcome the other null detection point mechanism, a phase shifter is introduced after the LNA. Experiments were carried out to validate the proposed time-domain analysis and demonstrate that both the techniques need to be used simultaneously to remove both null detection points in the system. The sensitivity comparison was performed using a conventional Doppler radar and an SILO-based radar built with identical microwave components.

Topology optimization of dispersive plasmonic nanostructures in the time-domain.

Topology optimization techniques have been applied in integrated optics and nanophotonics for the inverse design of devices with shapes that cannot be conceived by human intuition. At optical frequencies, these techniques have only been utilized to optimize nondispersive materials using frequency-domain methods. However, a time-domain formulation is more efficient to optimize materials with dispersion. We introduce such a formulation for the Drude model, which is widely used to simulate the dispersive properties of metals, conductive oxides, and conductive polymers. Our topology optimization algorithm is based on the finite-difference time-domain (FDTD) method, and we introduce a time-domain sensitivity analysis that enables the evaluation of the gradient information by using one additional FDTD simulation. The existence of dielectric and metallic structures in the design space produces plasmonic field enhancement that causes convergence issues. We employ an artificial damping approach during the optimization iterations that, by reducing the plasmonic effects, solves the convergence problem. We present several design examples of 2D and 3D plasmonic nanoantennas with optimized field localization and enhancement in frequency bands of choice. Our method has the potential to speed up the design of wideband optical nanostructures made of dispersive materials for applications in nanoplasmonics, integrated optics, ultrafast photonics, and nonlinear optics.

Nonlinear breathing crack identification from time-domain sensitivity analysis

Breathing cracks are often encountered in engineering structures and detection of such cracks at the earliest stage is important for safety and serviceability of the structures. In this paper, a new breathing crack identification approach based on the time-domain sensitivity analysis is proposed. For the forward problem, the crack is simply treated as a nonlinear oscillator and how to model a structure with breathing cracks is specified. As regards the inverse crack identification, it is formulated as a nonlinear least-squares optimization problem and a new sensitivity-based approach is developed to get the solution. In doing so, the sensitivity analysis of the possibly non-differentiable breathing crack models is necessarily proceeded through a smoothing strategy. Moreover, to enhance the convergence, the trust-region constraint is introduced and the Tikhonov regularization is naturally called to tackle the constraint. Numerical examples are studied to verify the feasibility and efficiency of the proposed approach in breathing crack identification.

Dynamic regulation characteristics of pumped-storage plants with two generating units sharing common conduits and busbar for balancing variable renewable energy

Pumped-storage plants (PSPs) are becoming increasingly important for balancing variable renewable energy (VRE) in power systems. A large portion of PSPs consist of multiple generating units (GUs) with shared water conduits and busbar. In this paper, an integrated transfer function (ITF) model for such PSPs in generation mode under primary frequency control (PFC) is proposed. Based on a real PSP in China, this ITF model is validated with both simulation results using the method of characteristics (MOC) as well as data from on-site measurements. For a series of governor parameter (Kp and Ki) settings under different governor control modes, stability limits expressed by gain and phase margins are evaluated and time domain simulations for balancing wind power variations are carried out to assess the quality of regulation. The results reveal that the dynamic regulation characteristics of the studied PSP system is clearly effected by the governor control mode. Besides, due to its complexity, a general range of the gain and phase margins is not viable for the studied PSP system under different governor control modes. Therefore, a thorough time domain sensitivity analysis using both step and realistic load disturbance as inputs is required for the studied PSPs for balancing VRE.

Time-Domain Sensitivity Analysis of Delayed Partial Element Equivalent Circuits

In this paper, we present the time-domain sensitivity analysis of delayed partial element equivalent circuit models using direct and adjoint techniques. The adjoint sensitivity equations are derived in the form of delayed differential equations. The two methods to compute the time-domain sensitivity are compared with the perturbation approach demonstrating very good accuracy and speedup.

Adjoint-KHN Equivalent Realization of Current Mode Universal Biquad Employing Third Generation Current Conveyor

The classical “Kerwin-Huelsman-Newcomb” (KHN) filter has been a very popular high performance multifunction biquadratic analog filter. Derived from the adjoint signal flow graph of this KHN circuit a new multi-input universal filter has been presented utilising for the first time all third generation current conveyors. This new universal biquadratic filter employs all grounded passive elements, realizes all the standard five filtering functions namely “Low pass”, “High pass”, “Band pass”, “Band reject” and “All pass” and produces all the outputs without requiring any additional active and /or passive component. The workability of the proposed circuit configuration has been verified by PSPICE simulation along with its non-ideal analysis, time domain analysis, sensitivity analysis and Monte-Carlo analysis. The proposed universal filter with explicit current output may be suitably employed in all current-mode application situations of classical KHN filter such as; FM demodulators, phase-locked-loops etc.

Thermal sensitivity analysis of on‐chip interconnects

A new method for time-domain sensitivity analysis of on-chip coupled interconnects with non-uniform substrate temperature profiles is presented. The parametric macromodelling is performed via multivariate rational approximation using magnitude vector fitting algorithms. The waveform relaxation algorithm with transverse partitioning is employed for efficient sensitivity analysis of coupled interconnects. With these techniques, the thermal sensitivity of on-chip global interconnects arising from typical substrate thermal profiles are investigated.

Generalized Time-Domain Adjoint Sensitivity Analysis of Distributed MTL Networks

In this paper, a generalized approach is presented for time-domain adjoint sensitivity analysis of lossy multiconductor transmission lines (MTLs) with respect to electrical and/or physical parameters. While the new approach provides all the advantages of an adjoint sensitivity analysis, it is independent of the specifics of the MTL macromodel used. Several numerical examples are presented to demonstrate the validity and accuracy of the proposed approach.

Sensitivity Analysis of Lossy Multiconductor Transmission Lines Based on the Passive Method of Characteristics Macromodel

This paper describes an efficient approach for time domain sensitivity analysis of lossy high speed interconnects in the presence of nonlinear terminations. The sensitivity information is derived using the recently developed passive method of characteristics. An important feature of the proposed method is that the sensitivities are obtained from the solution of the original network, leading to significant computational advantages. Numerical examples are presented to illustrate the validity of the proposed approach.

Time-Domain Green's Function-Based Sensitivity Analysis of Multiconductor Transmission Lines With Nonlinear Terminations

Time-domain sensitivity analysis of multiconductor transmission lines is derived from the dyadic Green's function of the 1-D wave propagation problem. The rational nature of the Green's function permits the generation of a time-domain macromodel for the computation of transient voltage sensitivities with respect to both electrical and physical parameters, completely avoiding similarity transformation. Through simulations of multiconductor transmission lines with linear and nonlinear terminations, comparisons are made with the perturbative approach and standard frequency-domain techniques for multiconductor transmission lines.

Time‐domain sensitivity analysis of planar structures using first‐order one‐way wave‐equation boundaries

AbstractAn efficient adjoint variable method technique is developed for time‐domain sensitivity analysis of planar structures with transmission‐line modeling complemented by a first‐order one‐way wave‐equation absorbing boundaries. A backward‐running adjoint simulation is derived and solved. The validity of the technique is illustrated through three microstrip circuits. The examples demonstrate the efficiency and accuracy of the technique in comparison with the classical finite‐difference approaches to the estimation of the response sensitivities. Copyright © 2007 John Wiley &amp; Sons, Ltd.

Time domain sensitivity analysis of electromagnetic quantities utilising FEM for the identification of material conductivity distributions

PurposeThis paper aims to present effective methods for computing electromagnetic field sensitivity in the time domain versus conductivity perturbations in finite elements.Design/methodology/approachTwo‐dimensional cases in linear, isotropic media are considered and two effective methods for sensitivity analysis of a magnetic vector potential in the time domain are described.FindingsThe paper finds that the convergence of numerical identification algorithm depends on exact measurement of magnetic flux density. For identification of real cracks the application of data filtering and TSVD regularization of Gauss‐Newton algorithm is necessary.Practical implicationsThe resultant gradient information may be used for solving inverse problems such as the identification of material conductivity distributions.Originality/valueThe algorithms described are based on known methods from established circuit theory – incremental circuit and adjoint circuit, these have been expanded to apply in electromagnetic field theory.

An adjoint-based approach to computing time-domain sensitivity of multiport systems described by reduced-order models

This paper describes a new algorithm for obtaining time-domain sensitivity information of large linear networks via projection-based model-order reduction (MOR) techniques. The proposed algorithm enables adjoint-based time-domain sensitivity analysis techniques to be applied to a large network containing multiple linear multiport subnetworks. In addition, the proposed algorithm can also be used to obtain frequency-domain sensitivity information. The efficiency and accuracy of the proposed work is demonstrated through optimizing the time-domain waveforms of a high-speed interconnect circuit with nonlinear terminations.

Delay-extraction-based sensitivity analysis of multiconductor transmission lines with nonlinear terminations

An efficient approach is presented for time-domain sensitivity analysis of lossy multiconductor transmission lines with respect to electrical and/or physical parameters. The proposed method employs delay extraction prior to approximating the matrix exponential stamp of the line and guarantees macromodel passivity. A delay-extraction-based equivalent circuit for sensitivity analysis is derived in a closed form, leading to significant computational advantages.

Behavioral simulation for analog system design verification

Synthesis of analog circuits is an emergent field, with efforts focused at the cell level. With the growing trend of mixed ASIC designs that contain significant portions of analog sections, compatible design methodologies in the analog domain are necessary to complement those in the digital domain. The synthesis process requires an associated verification process to ensure that the designs meet performance specifications at the onset. In this paper we present a behavioral simulation methodology for analog system design verification and design space exploration. The verification task integrates with analog system-level synthesis for an integrated synthesis-verification process that avoids expensive post synthesis simulation by invoking external simulators. Thus rapid redesign at the architectural level can be undertaken for design parameter variation and during optimization. The verification suite is composed of a repertoire of analysis modes that include time and frequency domain analysis, sensitivity analysis and distortion analysis. Besides verification of design specifications, these analysis modes are also used to generate metrics for comparison of various architectural choices that could realize a given set of specifications. The implementation is in the form of a behavioral simulator, ARCHSIM. >

Time domain sensitivity of high‐speed VLSI interconnects

AbstractAn adjoint method based on the asymptotic wave‐form evaluation (AWE) technique is described for the time domain sensitivity analysis of VLSI interconnect networks that contain linear lumped and distributed elements. the network sensitivity, which is approximated by a set of poles and residues sensitivities, can be evaluated efficiently with respect to all network components and interconnect parameters. Owing to the recursive relationship between the time moments of distributed networks, the conventional adjoint approach could not be applied directly. Instead, an alternative approach is presented which solves this problem by introducting the concept of adjoint moments for networks with distributed elements. Examples and comparisons with the sensitivity determined by perturbation are presented. the proposed technique is useful for the identification of critical circuit elements and the performance optimization of high‐speed VLSI designs.

A switched capacitor circuit analysis program and its applications

A program developed for analyzing arbitraryN-phase switched capacitor circuits is presented. Application of the program is given. The program simulation facilities, such as time-domain analysis, frequence-domain analysis, sensitivity analysis and multiple objective optimization are described. The algorithm used in the program is introduced briefly. Several examples are given to show the application of the program.

Time-domain sensitivity analysis of dynamic sense amplifier of an n-MOS dynamic RAM

The time-domain sensitivity analysis of a dynamic sense amplifier (DSA) is presented, based on an adjoint system solution. An admissible range of parameter variations for correct latching is computed from the sensitivity analysis. It has been shown that the sense amplifier sensitivity to the load capacitance and the gate to source capacitances of the latching transistors is dependent on the slope of the latching signal. It has also been shown that for a perfectly balanced sense amplifier there exists an optimal slope of the latching signal for which the sense amplifier performance is insensitive to the variations in gate to source capacitances.

Dual-mode control of a time-varying boiler model with parameter and state estimation

A time-varying mathematical model of an oil-fired boiler unit, with estimated parameters, is used to represent a poorly defined multivariable physical process subjected to input disturbances and measurement noise.The model is simulated in discrete form and is used to generate on-line control action related to a quadratic performance criterion.A parameter-estimation technique is included in the overall control logic, based on a time-domain sensitivity analysis of the system-transition and -driving matrixes, and is used for continuously updating the model to the process system which is subjected to continuous change and random input disturbances.A form of predictive on—off and bounded linear control is included within the control programme to give suboptimal performance based on minimisation of a two-period and current-performance index, respectively. Relatively close model following and/or improved control can be obtained with inputs based on the model reference system, and some slight improvement in performance is possible using dual-mode control action.