Loading...
Searching...
No Matches
Class to implement a full order inverse laplacian problem. More...
#include <inverseLaplacianProblemTotalHeatMeasure_CG.H>
Public Member Functions | |
| inverseLaplacianProblemTotalHeatMeasure_CG () | |
| inverseLaplacianProblemTotalHeatMeasure_CG (int argc, char *argv[]) | |
| Construct with argc and argv. | |
| virtual | ~inverseLaplacianProblemTotalHeatMeasure_CG () |
| int | conjugateGradient () |
| Conjugate gradient method. | |
| void | computeGradJ () |
| Computes the gradient of cost function J and its L2 norm. | |
| void | computeSearchStep () |
| Compute the search step beta. | |
| int | conjugateGradientConvergenceCheck () |
| Convergence cher for the conjugate gradient method. | |
 Public Member Functions inherited from inverseLaplacianProblem_CG | |
| inverseLaplacianProblem_CG () | |
| Null constructor. | |
| inverseLaplacianProblem_CG (int argc, char *argv[]) | |
| Construct with argc and argv. Reads the thermocouples dictionary. | |
| virtual | ~inverseLaplacianProblem_CG () |
| Destructor. | |
| void | set_valueFraction () |
| Set valueFraction list values for Robin condition. | |
| void | assignAdjointBC () |
| Set BC of the adjoint problem. | |
| volScalarField | assignAdjointBCandSource () |
| Assign the BC for the adjoint problem and returs the source field. | |
| void | assignSensitivityBC () |
| Set BC and IF of the sensitivity problem. | |
| void | solveAdjoint () |
| Solve adjoint problem. | |
| void | solveSensitivity () |
| Solve sensibility problem. | |
| void | solve (const char *problemID) |
| Solve for direct and sensitivity. | |
| void | defineThermocouplesPlane () |
| Identifies the plane defined by the thermocouples. | |
| void | sensibilitySolAtThermocouplesLocations () |
| Fill the Foam::vector containing the values of the sensitivity solution at the thermocouples locations. | |
| int | conjugateGradient () |
| Conjugate gradient method. | |
| void | computeGradJ () |
| Computes the gradient of cost function J and its L2 norm. | |
| void | searchDirection () |
| Computes the search direction P. | |
| void | computeSearchStep () |
| Compute the search step beta. | |
| void | updateHeatFlux () |
| Updates the heat flux in the conjugate gradient iterations. | |
| int | conjugateGradientConvergenceCheck () |
| Convergence cher for the conjugate gradient method. | |
| int | isInPlane (double cx, double cy, double cz, Foam::vector thermocoupleCellDim) |
| Checks if a cell crosses the interpolation plane. | |
| void | writeFields (label folderNumber, const char *folder) |
| Writes fields to file. | |
| void | thermocouplesInterpolation () |
| Interpolates the thermocouples measurements in the plane defined in readThermocouples() using radial basis functions. | |
| void | thermocouplesInterpolation (DenseMatrix &RBFweights, DenseMatrix &RBFbasis) |
| Interpolates the thermocouples measurements in the plane defined in readThermocouples() using Radial Basis Functions (RBF) Assumes thermocouples are all at the same y coordinate. | |
| void | differenceBetweenDirectAndMeasure () |
| Computes the difference between direct problem solution and measure. | |
| void | restart (word fieldName="all") |
| Restart fields. | |
| Public Member Functions inherited from inverseLaplacianProblem | |
| inverseLaplacianProblem () | |
| Null constructor. | |
| inverseLaplacianProblem (int argc, char *argv[]) | |
| Construct with argc and argv. Reads the thermocouples dictionary. | |
| virtual | ~inverseLaplacianProblem () |
| Destructor. | |
| void | set_g () |
| Set the right g size and fills it with zeros. | |
| volScalarField | list2Field (List< scalar > list, scalar innerField=0.0) |
| Create a field with the hotSide boundary heat flux at the hotSide bounday cells for visualization. | |
| void | set_valueFraction () |
| Set valueFraction list values for Robin condition. | |
| virtual void | solveTrue () |
| Solve the direct problem with the true heat flux as boundary condition and fills the measured temeprature vector. | |
| virtual void | assignDirectBC () |
| Set boundary condition of the direct problem. | |
| void | solveDirect () |
| Solve direct problem. | |
| void | solve (const char *problem) |
| Solve Laplacian problem without source term. | |
| virtual void | readThermocouples () |
| Identifies in the mesh the cells corresponding to the termocouples locations. | |
| Eigen::VectorXd | fieldValueAtThermocouples (volScalarField &field) |
| Interpolates the field value at the thermocouples points. | |
| void | differenceBetweenDirectAndMeasure () |
| Computes the difference between direct problem solution and measures Saves the difference vector in Tdiff. | |
| Foam::vector | cellDim (const faceList &ff, const pointField &pp, const cell &cc, labelList pLabels, pointField pLocal) |
| Compute maximum cell dimension in x, y and z. | |
| void | restart () |
| Restart fields. | |
| Public Member Functions inherited from laplacianProblem | |
| laplacianProblem () | |
| laplacianProblem (int argc, char *argv[]) | |
| Construct with argc and argv. | |
| ~laplacianProblem () | |
| void | truthSolve (List< scalar > mu_now, word folder="./ITHACAoutput/Offline/") |
| Perform a truthsolve. | |
| void | project (label Nmodes) |
| Perform a projection onto the POD modes. | |
| Public Member Functions inherited from reductionProblem | |
| reductionProblem () | |
| Construct Null. | |
| ~reductionProblem () | |
| void | setParameters () |
| Set Parameters Problems. | |
| void | genRandPar () |
| Generate Random Numbers. | |
| void | genRandPar (label tsize) |
| Generate Random Numbers given the dimension of the training set. | |
| void | genEquiPar () |
| Generate Equidistributed Numbers. | |
| void | truthSolve () |
| Perform a TruthSolve. | |
| void | assignBC (volVectorField &s, label BC_ind, Vector< double > &value) |
| Assign Boundary Condition to a volVectorField. | |
| void | assignBC (volScalarField &s, label BC_ind, double &value) |
| Assign Boundary Condition to a volScalarField. | |
| void | reconstructFromMatrix (PtrList< volVectorField > &rec_field2, PtrList< volVectorField > &modes, label Nmodes, Eigen::MatrixXd coeff_matrix) |
| Exact reconstruction using a certain number of modes for vector list of fields and the projection coefficients (volVectorField) | |
| void | reconstructFromMatrix (PtrList< volScalarField > &rec_field2, PtrList< volScalarField > &modes, label Nmodes, Eigen::MatrixXd coeff_matrix) |
| Exact reconstruction using a certain number of modes for vector list of fields and the projection coefficients (volScalarField) | |
| template<typename T , typename G > | |
| void | assignIF (T &s, G &value) |
| Assign internal field condition. | |
| template<typename T > | |
| void | computeLift (T &Lfield, T &liftfield, T &omfield) |
| Homogenize the snapshot matrix, it works with PtrList of volVectorField and volScalarField. | |
| template<typename T > | |
| void | computeLiftT (T &Lfield, T &liftfield, T &omfield) |
| Virtual function to compute the lifting function. | |
| void | liftSolve () |
| Virtual function to compute the lifting function for scalar field. | |
| void | liftSolveT () |
| void | project () |
| General projection operation. | |
| void | writeMu (List< scalar > mu_now) |
| Write out a list of scalar corresponding to the parameters used in the truthSolve. | |
| std::vector< SPLINTER::RBFSpline > | getCoeffManifoldRBF (PtrList< volVectorField > snapshots, PtrList< volVectorField > &modes, word rbfBasis="GAUSSIAN") |
| Constructs the parameters-coefficients manifold for vector fields, based on RBF-spline model. | |
| std::vector< SPLINTER::RBFSpline > | getCoeffManifoldRBF (PtrList< volScalarField > snapshots, PtrList< volScalarField > &modes, word rbfBasis="GAUSSIAN") |
| Constructs the parameters-coefficients manifold for scalar fields, based on RBF-spline model. | |
| std::vector< SPLINTER::BSpline > | getCoeffManifoldSPL (PtrList< volVectorField > snapshots, PtrList< volVectorField > &modes, label splDeg=3) |
| Constructs the parameters-coefficients manifold for vector fields, based on the B-spline model. | |
| std::vector< SPLINTER::BSpline > | getCoeffManifoldSPL (PtrList< volScalarField > snapshots, PtrList< volScalarField > &modes, label splDeg=3) |
| Constructs the parameters-coefficients manifold for scalar fields, based on the B-spline model. | |
Public Attributes | |
| double | gIntegral |
| double | gIntegralWeight = 0 |
| double | gIntegral_meas = 0 |
| Public Attributes inherited from inverseLaplacianProblem_CG | |
| autoPtr< volScalarField > | _lambda |
| Adjoint field. | |
| autoPtr< volScalarField > | _deltaT |
| Sensibility temperature field. | |
| List< scalar > | valueFractionAdj |
| Value fraction for the adjoint Robin BC. | |
| PtrList< volScalarField > | Tfield |
| List of temperature solutions. | |
| PtrList< volScalarField > | lambdaField |
| List of adjoint solutions. | |
| PtrList< volScalarField > | deltaTfield |
| List of sensitivity solutions. | |
| bool | interpolation |
| Flag for interpolation of the temperature measurements. | |
| bool | interpolationPlaneDefined |
| Flag for definition of interpolation plane. | |
| int | cgIter |
| Conjugate Gradient (CG) interations counter. | |
| int | cgIterMax |
| Maximum CG iterations. | |
| bool | saveSolInLists = 0 |
| Flag to save solutions in lists. | |
| Eigen::MatrixXd | Jlist |
| Vector to store the const function J. | |
| double | gradJ_L2norm |
| L2 norm of the gradient of J. | |
| double | Jtol |
| Absolute stopping criterion for the CG. | |
| double | JtolRel |
| Relative stopping criterion for the CG. | |
| double | gamma |
| Conjugate coefficient. | |
| double | gamma_den |
| Denoinator of the conjugate coefficient. | |
| double | beta |
| CG search step size. | |
| List< scalar > | P |
| Search direction. | |
| List< scalar > | gradJ |
| Gradient of the cost function. | |
| Eigen::VectorXd | Tsens |
| Vector of solutions of the sensitivity problem at the thermocouples points. | |
| List< Eigen::MatrixXd > | ArbT |
| Temperature reduced matrix. | |
| List< Eigen::MatrixXd > | ArbLambda |
| Adjoint reduced matrix. | |
| List< Eigen::MatrixXd > | ArbDeltaT |
| Sensitivity reduced matrix. | |
| thermocouplesPlane | interpolationPlane |
| Interpolation plane. | |
| Eigen::VectorXd | cellsInPlane |
| IDs of the cells in the interpolation plane. | |
| Public Attributes inherited from inverseLaplacianProblem | |
| ITHACAparameters * | para |
| autoPtr< volScalarField > | _T |
| Temperature field. | |
| autoPtr< fvMesh > | _mesh |
| Mesh. | |
| autoPtr< simpleControl > | _simple |
| simpleControl | |
| autoPtr< fv::options > | _fvOptions |
| fvOptions | |
| autoPtr< Time > | _runTime |
| Time. | |
| dimensionedScalar | DT |
| Dummy thermal conductivity with unitary value. | |
| double | k |
| Thermal diffusivity [W/(m K)]. | |
| double | H |
| Heat transfer coefficient [W/(m2 K)]. | |
| bool | thermocouplesRead = 0 |
| Flag to know if thermocouples file was read. | |
| int | thermocouplesNum |
| Number of thermocouples. | |
| double | J |
| Cost funtion [K^2]. | |
| double | L2norm |
| double | LinfNorm |
| scalar | homogeneousBC = 0.0 |
| Homogenerous BC. | |
| List< scalar > | homogeneousBCcoldSide |
| List of zeros of the size of coldSide patch. | |
| List< scalar > | Tf |
| Temperature at coldSide [K]. | |
| List< scalar > | refGrad |
| Reference gradient for the Robin BC. | |
| List< scalar > | valueFraction |
| Value fraction for the Robin BC. | |
| label | hotSide_ind |
| Index of the hotSide patch. | |
| label | coldSide_ind |
| Index of the coldSide patch. | |
| List< scalar > | g |
| Heat flux at hotSide [W/m2]. | |
| List< List< scalar > > | gList |
| List of boundary heat fluxes. | |
| List< scalar > | gTrue |
| True heat flux at hotSide [w/m2]. | |
| List< scalar > | faceCellArea |
| List of patch faces areas [m2]. | |
| List< vector > | thermocouplesPos |
| List containing the positions of the thermocouples. | |
| List< int > | thermocouplesCellID |
| List of cells indices containing a thermocouple. | |
| List< int > | thermocouplesCellProc |
| List of incedes of the processors contining each thermocouple. | |
| Eigen::VectorXd | Tmeas |
| Vector of measured temperatures at the thermocouples locations [K]. | |
| Eigen::VectorXd | Tdirect |
| Vector of computed temperatures at the thermocouples locations [K]. | |
| Eigen::VectorXd | Tdiff |
| Difference between computed and measured temperatures at the thermocouples. | |
| label | nProcs |
| Number of processors. | |
| Public Attributes inherited from laplacianProblem | |
| PtrList< volScalarField > | Tfield |
| List of snapshots for the solution. | |
| PtrList< volScalarField > | Tonline |
| List of snapshots for the solution. | |
| volScalarModes | Tmodes |
| List of POD modes. | |
| PtrList< fvScalarMatrix > | operator_list |
| List of operators. | |
| List< scalar > | theta |
| Theta (coefficients of the affine expansion) | |
| PtrList< volScalarField > | nu_list |
| Nu (diffusivity) | |
| label | NTmodes |
| Number of modes reduced problem. | |
| List< Eigen::MatrixXd > | A_matrices |
| A matrices. | |
| Eigen::MatrixXd | source |
| Source vector. | |
| autoPtr< volScalarField > | _T |
| Temperature field. | |
| autoPtr< volScalarField > | _S |
| Source Term. | |
| autoPtr< volScalarField > | _nu |
| Diffusivity. | |
| autoPtr< fvMesh > | _mesh |
| Mesh. | |
| autoPtr< Time > | _runTime |
| Time. | |
| Public Attributes inherited from reductionProblem | |
| label | Pnumber |
| Number of parameters. | |
| label | Tnumber |
| Dimension of the training set (used only when gerating parameters without input) | |
| Eigen::MatrixXd | mu |
| Row matrix of parameters. | |
| Eigen::MatrixXd | mu_range |
| Range of the parameter spaces. | |
| Eigen::MatrixXd | mu_samples |
| Matrix of parameters to be used for PODI, where each row corresponds to a sample point. In this matrix the time dimension is regarded as a parameter for unsteady problems. | |
| double | mu_cur |
| Current value of the parameter. | |
| bool | podex |
| Boolean variable, it is 1 if the POD has already been computed, else 0. | |
| bool | offline |
| Boolean variable, it is 1 if the Offline phase has already been computed, else 0. | |
| IOdictionary * | ITHACAdict |
| dictionary to store input output infos | |
| autoPtr< argList > | _args |
| argList | |
| ITHACAparallel * | paral |
| parallel handling | |
| label | folderN = 1 |
| Counter to save intermediate steps in the correct folder, for unsteady and some stationary cases. | |
| label | counter = 1 |
| Counter used for the output of the full order solutions. | |
| Eigen::MatrixXi | inletIndex |
| Matrix that contains informations about the inlet boundaries. | |
| Eigen::MatrixXi | inletPatch |
| Matrix that contains informations about the inlet boundaries without specifing the direction Rows = Number of parametrized boundary conditions Cols = 1 Example: example.inletIndex.resize(2, 1); example.inletIndex(0, 0) = 0; example.inletIndex(1, 0) = 1; Means that there are two parametrized boundary conditions of which the first row is of patch 0 and the second row of patch 1. | |
| Eigen::MatrixXi | inletIndexT |
Class to implement a full order inverse laplacian problem.
Definition at line 44 of file inverseLaplacianProblemTotalHeatMeasure_CG.H.
| inverseLaplacianProblemTotalHeatMeasure_CG::inverseLaplacianProblemTotalHeatMeasure_CG | ( | ) |
Definition at line 40 of file inverseLaplacianProblemTotalHeatMeasure_CG.C.
| inverseLaplacianProblemTotalHeatMeasure_CG::inverseLaplacianProblemTotalHeatMeasure_CG | ( | int | argc, |
| char * | argv[] ) |
Construct with argc and argv.
Definition at line 42 of file inverseLaplacianProblemTotalHeatMeasure_CG.C.
|
inlinevirtual |
Definition at line 53 of file inverseLaplacianProblemTotalHeatMeasure_CG.H.
| void inverseLaplacianProblemTotalHeatMeasure_CG::computeGradJ | ( | ) |
Computes the gradient of cost function J and its L2 norm.
Definition at line 130 of file inverseLaplacianProblemTotalHeatMeasure_CG.C.
| void inverseLaplacianProblemTotalHeatMeasure_CG::computeSearchStep | ( | ) |
Compute the search step beta.
Definition at line 146 of file inverseLaplacianProblemTotalHeatMeasure_CG.C.
| int inverseLaplacianProblemTotalHeatMeasure_CG::conjugateGradient | ( | ) |
Conjugate gradient method.
| [out] | 1 | if converged within cgIterMax iterations, 0 if not |
Definition at line 51 of file inverseLaplacianProblemTotalHeatMeasure_CG.C.
| int inverseLaplacianProblemTotalHeatMeasure_CG::conjugateGradientConvergenceCheck | ( | ) |
Convergence cher for the conjugate gradient method.
| [out] | 1 | if convergence criteria met, 0 if not |
Definition at line 157 of file inverseLaplacianProblemTotalHeatMeasure_CG.C.
| double inverseLaplacianProblemTotalHeatMeasure_CG::gIntegral |
Definition at line 55 of file inverseLaplacianProblemTotalHeatMeasure_CG.H.
| double inverseLaplacianProblemTotalHeatMeasure_CG::gIntegral_meas = 0 |
Definition at line 57 of file inverseLaplacianProblemTotalHeatMeasure_CG.H.
| double inverseLaplacianProblemTotalHeatMeasure_CG::gIntegralWeight = 0 |
Definition at line 56 of file inverseLaplacianProblemTotalHeatMeasure_CG.H.
1.11.0