pymor.algorithms.riccati
¶
Module Contents¶
- pymor.algorithms.riccati.solve_pos_ricc_dense(A, E, B, C, R=None, S=None, trans=False, options=None, default_solver_backend=_DEFAULT_RICC_DENSE_SOLVER_BACKEND)[source]¶
Compute the solution of a Riccati equation.
Returns the solution \(X\) of a (generalized) continuous-time algebraic Riccati equation:
if trans is
False
\[A X E^T + E X A^T + (E X C^T + S^T) R^{-1} (C X E^T + S) + B B^T = 0.\]if trans is
True
\[A^T X E + E^T X A + (E^T X B + S) R^{-1} (B^T X E + S^T) + C^T C = 0.\]
If E is None, it is taken to be identity, and similarly for R. If S is None, it is taken to be zero.
We assume:
A, E, B, C, R, S are real
NumPy arrays
,E is nonsingular,
(E, A, B, C) is stabilizable and detectable,
R is symmetric positive definite, and
\(B B^T - S^T R^{-1} S\) (\(C^T C - S R^{-1} S^T\)) is positive semi-definite if trans is
False
(True
).
If the solver is not specified using the options argument, a solver backend is chosen based on availability in the following order:
slycot
(seepymor.bindings.slycot.solve_ricc_dense
)scipy
(seepymor.bindings.scipy.solve_ricc_dense
)
- Parameters:
A – The matrix A as a 2D
NumPy array
.E – The matrix E as a 2D
NumPy array
orNone
.B – The matrix B as a 2D
NumPy array
.C – The matrix C as a 2D
NumPy array
.R – The matrix B as a 2D
NumPy array
orNone
.S – The matrix S as a 2D
NumPy array
orNone
.trans – Whether the first matrix in the Riccati equation is transposed.
options –
The solver options to use. See:
default_solver_backend – Default solver backend to use (slycot, scipy).
- Returns:
X – Riccati equation solution as a
NumPy array
.
- pymor.algorithms.riccati.solve_pos_ricc_lrcf(A, E, B, C, R=None, S=None, trans=False, options=None, default_dense_solver_backend=_DEFAULT_RICC_LRCF_DENSE_SOLVER_BACKEND)[source]¶
Compute an approximate low-rank solution of a positive Riccati equation.
Returns a low-rank Cholesky factor \(Z\) such that \(Z Z^T\) approximates the solution \(X\) of a (generalized) positive continuous-time algebraic Riccati equation:
if trans is
False
\[A X E^T + E X A^T + (E X C^T + S^T) R^{-1} (C X E^T + S) + B B^T = 0.\]if trans is
True
\[A^T X E + E^T X A + (E^T X B + S) R^{-1} (B^T X E + S^T) + C^T C = 0.\]
If E is None, it is taken to be identity, and similarly for R. If S is None, it is taken to be zero.
If the solver is not specified using the options argument, a solver backend is chosen based on availability in the following order:
slycot
(seepymor.bindings.slycot.solve_pos_ricc_lrcf
),scipy
(seepymor.bindings.scipy.solve_pos_ricc_lrcf
).
Currently, only dense solvers are provided.
- Parameters:
A – The non-parametric
Operator
A.E – The non-parametric
Operator
E orNone
.B – The operator B as a
VectorArray
fromA.source
.C – The operator C as a
VectorArray
fromA.source
.R – The matrix R as a 2D
NumPy array
orNone
.S – The operator S as a
VectorArray
fromA.source
orNone
.trans – Whether the first
Operator
in the positive Riccati equation is transposed.options –
The solver options to use. See:
default_dense_solver_backend – Default dense solver backend to use (slycot, scipy).
- Returns:
Z – Low-rank Cholesky factor of the positive Riccati equation solution,
VectorArray
fromA.source
.
- pymor.algorithms.riccati.solve_ricc_dense(A, E, B, C, R=None, S=None, trans=False, options=None, default_solver_backend=_DEFAULT_RICC_DENSE_SOLVER_BACKEND)[source]¶
Compute the solution of a Riccati equation.
Returns the solution \(X\) of a (generalized) continuous-time algebraic Riccati equation:
if trans is
False
\[A X E^T + E X A^T - (E X C^T + S^T) R^{-1} (C X E^T + S) + B B^T = 0.\]if trans is
True
\[A^T X E + E^T X A - (E^T X B + S) R^{-1} (B^T X E + S^T) + C^T C = 0.\]
If E is None, it is taken to be identity, and similarly for R. If S is None, it is taken to be zero.
We assume:
A, E, B, C, R, S are real
NumPy arrays
,E is nonsingular,
(E, A, B, C) is stabilizable and detectable,
R is symmetric positive definite, and
\(B B^T - S^T R^{-1} S\) (\(C^T C - S R^{-1} S^T\)) is positive semi-definite if trans is
False
(True
).
If the solver is not specified using the options argument, a solver backend is chosen based on availability in the following order:
slycot
(seepymor.bindings.slycot.solve_ricc_dense
)scipy
(seepymor.bindings.scipy.solve_ricc_dense
)
- Parameters:
A – The matrix A as a 2D
NumPy array
.E – The matrix E as a 2D
NumPy array
orNone
.B – The matrix B as a 2D
NumPy array
.C – The matrix C as a 2D
NumPy array
.R – The matrix B as a 2D
NumPy array
orNone
.S – The matrix S as a 2D
NumPy array
orNone
.trans – Whether the first matrix in the Riccati equation is transposed.
options –
The solver options to use. See:
default_solver_backend – Default solver backend to use (slycot, scipy).
- Returns:
X – Riccati equation solution as a
NumPy array
.
- pymor.algorithms.riccati.solve_ricc_lrcf(A, E, B, C, R=None, S=None, trans=False, options=None, default_sparse_solver_backend=_DEFAULT_RICC_LRCF_SPARSE_SOLVER_BACKEND, default_dense_solver_backend=_DEFAULT_RICC_LRCF_DENSE_SOLVER_BACKEND)[source]¶
Compute an approximate low-rank solution of a Riccati equation.
Returns a low-rank Cholesky factor \(Z\) such that \(Z Z^T\) approximates the solution \(X\) of a (generalized) continuous-time algebraic Riccati equation:
if trans is
False
\[A X E^T + E X A^T - (E X C^T + S^T) R^{-1} (C X E^T + S) + B B^T = 0.\]if trans is
True
\[A^T X E + E^T X A - (E^T X B + S) R^{-1} (B^T X E + S^T) + C^T C = 0.\]
If E is None, it is taken to be identity, and similarly for R. If S is None, it is taken to be zero.
We assume:
A and E are real
Operators
,B, C and S are real
VectorArrays
fromA.source
,R is a real
NumPy array
,E is nonsingular,
(E, A, B, C) is stabilizable and detectable,
R is symmetric positive definite, and
\(B B^T - S^T R^{-1} S\) (\(C^T C - S R^{-1} S^T\)) is positive semi-definite if trans is
False
(True
).
For large-scale problems, we additionally assume that
len(B)
andlen(C)
are small.If the solver is not specified using the options argument, a solver backend is chosen based on availability in the following order:
for sparse problems (minimum size specified by
mat_eqn_sparse_min_size
)lrradi
(seepymor.algorithms.lrradi.solve_ricc_lrcf
),
for dense problems (smaller than
mat_eqn_sparse_min_size
)slycot
(seepymor.bindings.slycot.solve_ricc_lrcf
),scipy
(seepymor.bindings.scipy.solve_ricc_lrcf
).
- Parameters:
A – The non-parametric
Operator
A.E – The non-parametric
Operator
E orNone
.B – The operator B as a
VectorArray
fromA.source
.C – The operator C as a
VectorArray
fromA.source
.R – The matrix R as a 2D
NumPy array
orNone
.S – The operator S as a
VectorArray
fromA.source
orNone
.trans – Whether the first
Operator
in the Riccati equation is transposed.options –
The solver options to use. See:
default_sparse_solver_backend – Default sparse solver backend to use (lrradi).
default_dense_solver_backend – Default dense solver backend to use (slycot, scipy).
- Returns:
Z – Low-rank Cholesky factor of the Riccati equation solution,
VectorArray
fromA.source
.