[295] | 1 | /*! |
---|
| 2 | \file |
---|
| 3 | \brief TR 2525 file for testing Toy Problem of mpf for Covariance Estimation |
---|
| 4 | \author Vaclav Smidl. |
---|
| 5 | |
---|
| 6 | ----------------------------------- |
---|
| 7 | BDM++ - C++ library for Bayesian Decision Making under Uncertainty |
---|
| 8 | |
---|
| 9 | Using IT++ for numerical operations |
---|
| 10 | ----------------------------------- |
---|
| 11 | */ |
---|
| 12 | |
---|
| 13 | |
---|
| 14 | |
---|
[384] | 15 | #include <estim/particles.h> |
---|
| 16 | #include <estim/ekf_template.h> |
---|
| 17 | #include <stat/functions.h> |
---|
[295] | 18 | |
---|
| 19 | #include "../pmsm.h" |
---|
| 20 | #include "simulator.h" |
---|
| 21 | #include "../sim_profiles.h" |
---|
[357] | 22 | #include "user_info.h" |
---|
| 23 | #include "stat/loggers.h" |
---|
[295] | 24 | |
---|
| 25 | using namespace bdm; |
---|
| 26 | |
---|
| 27 | int main ( int argc, char* argv[] ) { |
---|
| 28 | const char *fname; |
---|
| 29 | if ( argc>1 ) {fname = argv[1]; } |
---|
| 30 | else { fname = "unitsteps.cfg"; } |
---|
[394] | 31 | UIFile F ( fname ); |
---|
[295] | 32 | |
---|
[357] | 33 | double h = 1e-6; |
---|
[295] | 34 | int Ndat; |
---|
| 35 | int Npart; |
---|
[357] | 36 | F.lookupValue ( "ndat", Ndat ); |
---|
| 37 | F.lookupValue ( "Npart",Npart ); |
---|
[295] | 38 | int Nsimstep = 125; |
---|
| 39 | |
---|
[357] | 40 | // Kalman filter |
---|
[295] | 41 | vec Qdiag; |
---|
[357] | 42 | UI::get( Qdiag, F, "dQ" ); //( "1e-6 1e-6 0.001 0.0001" ); //zdenek: 0.01 0.01 0.0001 0.0001 |
---|
| 43 | |
---|
[295] | 44 | vec Rdiag; |
---|
[357] | 45 | UI::get( Rdiag, F, "dR" );// ( "1e-8 1e-8" ); //var(diff(xth)) = "0.034 0.034" |
---|
[295] | 46 | |
---|
[357] | 47 | // internal model |
---|
| 48 | IMpmsm fxu; |
---|
| 49 | // Rs Ls dt Fmag(Ypm) kp p J Bf(Mz) |
---|
| 50 | fxu.set_parameters ( 0.28, 0.003465, Nsimstep*h, 0.1989, 1.5 ,4.0, 0.04, 0.0 ); |
---|
[295] | 51 | // observation model |
---|
| 52 | OMpmsm hxu; |
---|
| 53 | |
---|
| 54 | vec mu0= "0.0 0.0 0.0 0.0"; |
---|
| 55 | chmat Q ( Qdiag ); |
---|
| 56 | chmat R ( Rdiag ); |
---|
| 57 | EKFCh KFE ; |
---|
| 58 | KFE.set_parameters ( &fxu,&hxu,Q,R ); |
---|
| 59 | KFE.set_est ( mu0, chmat ( zeros ( 4 ) ) ); |
---|
| 60 | KFE.set_rv ( rx ); |
---|
| 61 | |
---|
| 62 | RV rQ ( "{Q }","16" ); |
---|
| 63 | EKFCh_chQ KFEp ; |
---|
| 64 | KFEp.set_parameters ( &fxu,&hxu,Q,R ); |
---|
| 65 | KFEp.set_est ( mu0, chmat ( zeros ( 4 ) ) ); |
---|
| 66 | |
---|
| 67 | rwiWishartCh* evolQw = new rwiWishartCh; |
---|
| 68 | evolQw->set_parameters(4, 0.1, sqrt(Qdiag),0.99); |
---|
| 69 | MPF<EKFCh_chQ> M; |
---|
| 70 | M.set_parameters ( evolQw,evolQw,Npart ); |
---|
| 71 | // initialize |
---|
| 72 | chmat Ch0(diag(Qdiag)); |
---|
| 73 | evolQw->condition ( vec(Ch0._Ch()._data(),16) ); //Zdenek default |
---|
| 74 | M.set_statistics ( evolQw->_e() , &KFEp ); |
---|
| 75 | // |
---|
| 76 | |
---|
| 77 | M.set_rv ( concat ( rQ,rx ) ); |
---|
| 78 | |
---|
[357] | 79 | dirfilelog *L = UI::build<dirfilelog> ( F, "logger" );// ( "exp/mpf_test",100 ); |
---|
[295] | 80 | int l_X = L->add ( rx, "xt" ); |
---|
| 81 | int l_D = L->add ( concat ( ry,ru ), "" ); |
---|
| 82 | int l_Q= L->add ( rQ, "" ); |
---|
| 83 | int l_fullQ= L->add ( rQ, "full" ); |
---|
| 84 | |
---|
| 85 | KFE.set_options ( "logbounds" ); |
---|
[317] | 86 | KFE.log_add ( *L,"KF" ); |
---|
[295] | 87 | M.set_options ( "logbounds" ); |
---|
[317] | 88 | M.log_add ( *L,"M" ); |
---|
[295] | 89 | L->init(); |
---|
| 90 | |
---|
| 91 | // SET SIMULATOR |
---|
| 92 | pmsmsim_set_parameters ( 0.28,0.003465,0.1989,0.0,4,1.5,0.04, 200., 3e-6, h ); |
---|
| 93 | vec dt ( 2 ); |
---|
| 94 | vec ut ( 2 ); |
---|
| 95 | vec xt ( 4 ); |
---|
| 96 | vec xtm=zeros ( 4 ); |
---|
| 97 | double Ww=0.0; |
---|
[357] | 98 | vec vecW; |
---|
| 99 | UI::get( vecW, F ,"profile" ); |
---|
[295] | 100 | |
---|
| 101 | mat tQ=diag(Qdiag); |
---|
| 102 | mat tChQ=chol(tQ); |
---|
| 103 | |
---|
| 104 | for ( int tK=1;tK<Ndat;tK++ ) { |
---|
| 105 | //Number of steps of a simulator for one step of Kalman |
---|
| 106 | for ( int ii=0; ii<Nsimstep;ii++ ) { |
---|
| 107 | //simulator |
---|
| 108 | sim_profile_vec01t ( Ww,vecW ); |
---|
| 109 | pmsmsim_step ( Ww ); |
---|
| 110 | }; |
---|
| 111 | ut ( 0 ) = KalmanObs[4]; |
---|
| 112 | ut ( 1 ) = KalmanObs[5]; |
---|
| 113 | xt = fxu.eval ( xtm,ut ) + tChQ.T() *randn ( 4 ); |
---|
| 114 | dt = hxu.eval ( xt,ut ); |
---|
| 115 | xtm = xt; |
---|
| 116 | |
---|
| 117 | //Variances |
---|
| 118 | /* if ( tK==1000 ) tQ ( 0,0 ) *=10; |
---|
| 119 | if ( tK==2000 ) tQ ( 0,0 ) /=10; |
---|
| 120 | if ( tK==3000 ) tQ( 1,1 ) *=10; |
---|
| 121 | if ( tK==4000 ) tQ( 1,1 ) /=10; |
---|
| 122 | if ( tK==5000 ) tQ( 2,2 ) *=10; |
---|
| 123 | if ( tK==6000 ) tQ( 2,2 ) /=10; |
---|
| 124 | if ( tK==7000 ) tQ( 3,3 ) *=10; |
---|
| 125 | if ( tK==8000 ) tQ( 3,3 ) /=10;*/ |
---|
| 126 | |
---|
| 127 | if (tK>1000) {tQ(0,1)=0.5*sqrt(tQ(0,0)*tQ(1,1));tQ(1,0)=tQ(0,1);} |
---|
| 128 | if (tK>2000) {tQ(0,1)=0; tQ(1,0)=tQ(0,1);} |
---|
| 129 | |
---|
| 130 | if (tK>3000) {tQ(2,3)=-0.5*sqrt(tQ(2,2)*tQ(3,3)); tQ(3,2)=tQ(2,3);} |
---|
| 131 | if (tK>4000) {tQ(2,3)=0; tQ(3,2)=tQ(2,3);} |
---|
| 132 | |
---|
| 133 | if (tK>5000) {tQ(0,2)=0.9*sqrt(tQ(0,0)*tQ(2,2)); tQ(2,0)=tQ(0,2);} |
---|
| 134 | if (tK>6000) {tQ(0,2)=0; tQ(2,0)=tQ(0,2);} |
---|
| 135 | |
---|
| 136 | tChQ=chol(tQ); |
---|
| 137 | |
---|
| 138 | //estimator |
---|
| 139 | KFE.bayes ( concat ( dt,ut ) ); |
---|
| 140 | M.bayes ( concat ( dt,ut ) ); |
---|
| 141 | |
---|
| 142 | L->logit ( l_X,xt ); |
---|
| 143 | L->logit ( l_D,concat ( dt,ut ) ); |
---|
| 144 | mat Q=diag(Qdiag); |
---|
| 145 | L->logit ( l_Q,vec(tQ._data(),16) ); |
---|
| 146 | |
---|
| 147 | mat chQ(4,4); |
---|
| 148 | copy_vector(16,M._e()->mean()._data(),chQ._data()); |
---|
| 149 | mat fQ=chQ.T()*chQ; |
---|
| 150 | L->logit ( l_fullQ,vec(fQ._data(),16) ); |
---|
| 151 | |
---|
[317] | 152 | KFE.logit ( *L ); |
---|
| 153 | M.logit ( *L ); |
---|
[295] | 154 | L->step(); |
---|
| 155 | } |
---|
| 156 | L->finalize(); |
---|
| 157 | //Exit program: |
---|
| 158 | |
---|
| 159 | delete L; |
---|
| 160 | return 0; |
---|
| 161 | } |
---|