Target Power Spectral Density Research Articles

Minimax Robust MIMO Radar Waveform Design

We consider the problem of minimax robust waveform design for multiple-input multiple-output (MIMO) radar based on mutual information (MI) and minimum mean-square error (MMSE) estimation for target identification and classification. Recognizing that a single, precise characterization of target power spectral density (PSD) is rare in practice, we assume the PSD lies in an uncertainty class of spectra bounded by known upper and lower bounds, which markedly relaxes the required target a priori knowledge. Based on this band model, we develop minimax robust waveforms for MIMO radar under both MMSE and MI criteria. These robust waveforms bound the worst-case performance at an acceptable limit, and also could insure performance will be sufficiently good for any PSD in the uncertainty class. Our findings also indicate that the MI and MMSE criteria lead to different minimax robust waveform solutions, which is in contrast to the case of the perfectly known target PSD

Non-Gaussian Simulation: Cumulative Distribution Function Map-Based Spectral Correction

Methods for stochastic simulation of sample functions have increasingly addressed the preservation of both spectral and probabilistic contents to offer an accurate description of the dynamic behavior of system input for reliability analysis. This study presents an efficient, flexible and easily applied stochastic non-Gaussian simulation method capable of reliably converging to a target power spectral density function and marginal probability density function, or a close relative thereof. Several existing spectral representation-based non-Gaussian simulation algorithms are first summarized. The new algorithm is then presented and compared with these methods to demonstrate its efficacy. The advantages and limitations of the new method are highlighted and shown to complement those of the existing algorithms.

Simulation of earthquake ground motions by a deterministic approach

A numerical method is presented for the computation of artificial earthquake records consistent with any arbitrarily specified target response spectra or power spectral density requirements. The proposed algorithm does not create new time histories but rather modifies, on the basis of an iterative deterministic approach, existing records to fit specific design requirements. The efficiency of the algorithm and the accuracy of the fitting process are substantially improved on the basis of a predictor–corrector type approach. The associated integrated computer code can, among its many other features, produce artificial earthquake records consistent with a number of compatible target response spectra at various damping levels, perform quadratic baseline correction, and calculate correlation factors for up to three earthquake records.

Soil–structure interaction in fuel handling building

This paper presents an accurate three-dimensional seismic soil–structure interaction analysis for large structures. The method is applied to the fuel building in nuclear power plants. The analysis is performed numerically in the frequency domain and the responses are obtained by inverse Fourier transformation. The size of the structure matrices is reduced by transforming the equation of motion to the modal coordinate system. The soil is simulated as a layered media on top of viscoelastic half space. Soil impedance matrices are calculated from the principles of continuum mechanics and account for soil stiffness and energy dissipation. Effects of embedment on the field equations is incorporated through the scattering matrices or by simply scaling the soil impedance. Finite element methods are used to discretize the concrete foundation for the generation of the soil interaction matrices. Decoupling of the sloshing water in the spent fuel pools and the free-standing spent fuel racks is simulated. The input seismic motions are defined by three artificial time history accelerations. These input motions are generated to match the ground design basis response spectra and the target power spectral density function. The methods described in this paper can handle arbitrary foundation layouts, allows for large structural models, and accurately represents the soil impedance. Time history acceleration responses were subsequently used to generate floor response spectra at applicable damping values.