2. POD for Proper Orthogonal Decomposition¶
2.1. What is it?¶
The Proper Orthogonal Decomposition (POD) is a technique used to decompose a matrix and characterize it by its principal components which are called modes [AnindyaChatterjee2000]. To approximate a function , only a finite sum of terms is required:
The function have an infinite representation. It can be chosen as a Fourier series or Chebyshev polynomials, etc. For a chosen basis of function, a set of unique time-functions arise. In case of the POD, the basis function are orthonormal. Meaning that:
The principle of the POD is to choose such that the approximation of is the best in a least squares sense. These orthonormal functions are called the proper orthogonal modes of the function.
When dealing with CFD simulations, the size of the domain is usually smaller than the number of measures, snapshots, . Hence, from the existing decomposition methods, the Singular Value Decomposition (SVD) is used. It is the snapshots methods [Cordier2006].
The Singular Value Decomposition (SVD) is a factorization operation of a matrix expressed as:
with diagonalizes , diagonalizes and is the singular value matrix which diagonal is composed by the singular values of . Knowing that a singular value is the square root of an eigen value. and are eigen vectors of respectively and which form an orthonormal basis. Thus, the initial matrix can be rewritten:
being the rank of the matrix. If taken , an approximation of the initial matrix can be constructed. This allows to compress the data as only an extract of and need to be stored.