Efficient algorithm for matrix spectral factorization
Spectral factorization represents an essential mathematical tool for solving optimal filtering problems or performing synthesis of linear electrical n-ports in the frequency domain among other important areas of applications. Since continuous-time and discrete-time systems have to be distinguished from each other problems are as follows: Find matrices A so that known positive real polynomial and symmetric matrices B are factored according to \(B(p)=A(p)A^ T(-p)\) in the continuous-time case or \(B(z)=A(z)A^ T(\frac{1}{z})\) in the discrete-time case. The spectral factors A exist and are unique up to a constant orthogonal matrix. In the past various approaches were proposed, among them a suitable partition of all roots of det B or iterative solutions which make appeal of matrix Riccati equations, perform triangular factorization of Toeplitz matrices or are based on Newton's method. All those methods suffer from certain deficiencies. The paper suggests a new algorithm for the spectral factorization of polynomial matrices. Though Newton's method is applied again, the algorithm is more efficient. This improvement in performance is obtained by using fully the structure of the equations to be solved in each iteration. The theory of symmetric polynomial equations represents the mathematical basis of the investigations. The resulting algorithm was implemented in PL/I language and tests are performed using 8 bytes floating-point format. A precision of 8 decimal digits was usually met after 6-8 iterations. Further, it turns out, the farther from the stability boundary the roots of det B lie, the better the computations proceed. The algorithm works even with roots just on the boundary: Before approaching the singularity and then collapsing, the result is calculated to several decimal digits. In such cases the rate of convergence is degraded from a quadratic to a geometric one.
- A Normal form for a Matrix under the Unitary Congruence Group
- Algorithms for Triangular Decomposition of Block Hankel and Toeplitz Matrices with Application to Factoring Positive Matrix Polynomials
- Bauer-type factorization of positive matrices and the theory of matrix polynomials orthogonal on the unit circle
- Computer Algorithm for Spectral Factorization of Rational Matrices
- Factorization of the Covariance Generating Function of a Pure Moving Average Process
- scientific article; zbMATH DE number 3887557 (Why is no real title available?)
- scientific article; zbMATH DE number 3837893 (Why is no real title available?)
- scientific article; zbMATH DE number 3930238 (Why is no real title available?)
- scientific article; zbMATH DE number 3673938 (Why is no real title available?)
- scientific article; zbMATH DE number 3521438 (Why is no real title available?)
- scientific article; zbMATH DE number 3550074 (Why is no real title available?)
- scientific article; zbMATH DE number 3054885 (Why is no real title available?)
- The Factorization of Matricial Spectral Densities
- Implementing the Iakoubovski-Merino spectral factorization algorithm using state-space methods
- On a Schur-algorithm based approach to spectral factorization: State- space formulae
- A spectral factorization algorithm for discrete-time descriptor systems via generalized eigenproblems
- Effective fast algorithms for polynomial spectral factorization
- Risk-sensitive filtering, prediction and smoothing for discrete-time singular systems
- Polynomial factorization through Toeplitz matrix computations
- Optimal and self-tuning white noise estimators with applications to deconvolution and filtering problems
- Optimal correction of an indefinite estimated MA spectral density matrix
- Fast separable factorization and applications
- Full rank positive matrix symbols: interpolation and orthogonality
- Discrete \(J\)-spectral factorization of possibly singular polynomial matrices
- Spectral factorization of non-symmetric polynomial matrices
- Factorization of moving-average spectral densities by state-space representations and stacking
- On multivariable matrix spectral factorization method
- A constructive algorithm for 2-D spectral factorization with rational spectral factors
- Filters and matrix factorization
- From full rank subdivision schemes to multichannel wavelets: a constructive approach
- Discrete-time symmetric polynomial equations with complex coefficients.
- A Newton-Raphson method for moving-average spectral factorization using the Euclid algorithm
- scientific article; zbMATH DE number 3882332 (Why is no real title available?)
- scientific article; zbMATH DE number 3864388 (Why is no real title available?)
- scientific article; zbMATH DE number 4215808 (Why is no real title available?)
- Design of reduced-order optimal estimators directly in the frequency domain
- Homotopy solution of polynomial equations inH∞optimal control
- Optimal reduced-order estimators in the frequency domain: the discrete-time case
- scientific article; zbMATH DE number 4022480 (Why is no real title available?)
- LQG regulation with disturbance measurement feedforward
- Spectral and Inner-Outer Factorizations of Rational Matrices
- Newton schemes for polynomial spectral factorization: scalar case
- A polynomial approach to minimax frequency domain optimization of multivariable feedback systems
- Wiener-Hopf and spectral factorization of real polynomials by Newton's method
- An Iterative Method for Matrix Spectral Factorization
- Feedforward control is dual to deconvolution
- scientific article; zbMATH DE number 3521438 (Why is no real title available?)
- scientific article; zbMATH DE number 3550074 (Why is no real title available?)
- Stable spectral factorization with applications to the estimation of time series models
- Generalized predictive control (GPC) with long-range predictive identification (LRPI) for multivariable anaesthesia
- A variational approach to the polynomial optimization of multivariable control with coloured noise at the controller output
- On the block decomposition and spectral factors of \(\lambda\)-matrices
- scientific article; zbMATH DE number 6415466 (Why is no real title available?)
- A New Method of Matrix Spectral Factorization
- Numerical algorithms for polynomial plus/minus factorization
- A parametrisation for dissipative behaviours–the matrix case
- On the generation of all spectral factors (Corresp.)
- Optimal parallel model solutions using the polynomial approach
- A Newton method for canonical Wiener-Hopf and spectral factorization of matrix polynomials
- Discrete J-spectral factorization.
- Discrete-time average positivity and spectral factorization in a behavioral framework.
- Bauer's spectral factorization method for low order multiwavelet filter design
- Matrix spectral factorization for SA4 multiwavelet
- Detecting rigid convexity of bivariate polynomials
- Algorithms for multidimensional spectral factorization and sum of squares
This page was built for publication: Efficient algorithm for matrix spectral factorization
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q1071712)
