Skip to content

Reorder DistributedGenerator variables #378

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
pelesh merged 3 commits into from
Apr 30, 2026
Merged

Reorder DistributedGenerator variables #378

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
3 commits
Select commit Hold shift + click to select a range
6888da9
Re-order DistributedGenerator variables
alexander-novo Apr 22, 2026
3dc67a5
Fix DGTest
alexander-novo Apr 29, 2026
6e597c3
Update Enzyme DG test
alexander-novo Apr 30, 2026
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
228 changes: 120 additions & 108 deletions GridKit/Model/PowerElectronics/DistributedGenerator/DistributedGenerator.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -43,17 +43,17 @@ namespace GridKit
{
assert(refframe_ || node_ref_->size() == 1);
assert(node_bus_->size() == 2);
// internals [\delta_i, Pi, Qi, phi_di, phi_qi, gamma_di, gamma_qi, il_di, il_qi, vo_di, vo_qi, io_di, io_qi]
// internals [Pi, Qi, phi_di, phi_qi, gamma_di, gamma_qi, il_di, il_qi, vo_di, vo_qi, io_di, io_qi, \delta_i]
// externals [\omega_ref, vba_out, vbb_out]
size_ = 16;
n_intern_ = refframe_ ? 12 : 13;
n_extern_ = refframe_ ? 4 : 3;
idc_ = id;
nnz_ = refframe_ ? 80 : 78;
nnz_ = refframe_ ? 77 : 78;

if (refframe_)
{
extern_indices_ = {0, 1, 2, 3};
extern_indices_ = {0, 1, 2, 15};
}
Comment thread
pelesh marked this conversation as resolved.
else
{
Expand Down Expand Up @@ -91,54 +91,56 @@ namespace GridKit
template <class ScalarT, typename IdxT>
int DistributedGenerator<ScalarT, IdxT>::evaluateInternalResidual()
{
ScalarT omega = wb_ - mp_ * y_[4];
ScalarT omega = wb_ - mp_ * y_[3];

// Take incoming voltages to current rotator reference frame
ScalarT vbd_in = std::cos(y_[3]) * y_[1] + std::sin(y_[3]) * y_[2];
ScalarT vbq_in = -std::sin(y_[3]) * y_[1] + std::cos(y_[3]) * y_[2];
ScalarT vbd_in = std::cos(y_[15]) * y_[1] + std::sin(y_[15]) * y_[2];
ScalarT vbq_in = -std::sin(y_[15]) * y_[1] + std::cos(y_[15]) * y_[2];

// ### Internal Componenets ##
// Rotor difference angle
f_[3] = -yp_[3] + omega - y_[0];

// P and Q equations
f_[4] = -yp_[4] + wc_ * (y_[12] * y_[14] + y_[13] * y_[15] - y_[4]);
f_[5] = -yp_[5] + wc_ * (-y_[12] * y_[15] + y_[13] * y_[14] - y_[5]);
f_[3] = -yp_[3] + wc_ * (y_[11] * y_[13] + y_[12] * y_[14] - y_[3]);
f_[4] = -yp_[4] + wc_ * (-y_[11] * y_[14] + y_[12] * y_[13] - y_[4]);

// Voltage control
ScalarT vod_star = Vn_ - nq_ * y_[5];
ScalarT vod_star = Vn_ - nq_ * y_[4];
ScalarT voq_star = static_cast<ScalarT>(0.0);

f_[6] = -yp_[6] + vod_star - y_[12];
f_[7] = -yp_[7] + voq_star - y_[13];
f_[5] = -yp_[5] + vod_star - y_[11];
f_[6] = -yp_[6] + voq_star - y_[12];

ScalarT ild_star = F_ * y_[14] - wb_ * Cf_ * y_[13] + Kpv_ * (vod_star - y_[12]) + Kiv_ * y_[6];
ScalarT ilq_star = F_ * y_[15] + wb_ * Cf_ * y_[12] + Kpv_ * (voq_star - y_[13]) + Kiv_ * y_[7];
ScalarT ild_star = F_ * y_[13] - wb_ * Cf_ * y_[12] + Kpv_ * (vod_star - y_[11]) + Kiv_ * y_[5];
ScalarT ilq_star = F_ * y_[14] + wb_ * Cf_ * y_[11] + Kpv_ * (voq_star - y_[12]) + Kiv_ * y_[6];

// Current control
f_[8] = -yp_[8] + ild_star - y_[10];
f_[9] = -yp_[9] + ilq_star - y_[11];
f_[7] = -yp_[7] + ild_star - y_[9];
f_[8] = -yp_[8] + ilq_star - y_[10];

ScalarT vid_star = -wb_ * Lf_ * y_[11] + Kpc_ * (ild_star - y_[10]) + Kic_ * y_[8];
ScalarT viq_star = wb_ * Lf_ * y_[10] + Kpc_ * (ilq_star - y_[11]) + Kic_ * y_[9];
ScalarT vid_star = -wb_ * Lf_ * y_[10] + Kpc_ * (ild_star - y_[9]) + Kic_ * y_[7];
ScalarT viq_star = wb_ * Lf_ * y_[9] + Kpc_ * (ilq_star - y_[10]) + Kic_ * y_[8];

// Output LC Filter
f_[10] = -yp_[10] - (rLf_ / Lf_) * y_[10] + omega * y_[11] + (vid_star - y_[12]) / Lf_;
f_[11] = -yp_[11] - (rLf_ / Lf_) * y_[11] - omega * y_[10] + (viq_star - y_[13]) / Lf_;
f_[9] = -yp_[9] - (rLf_ / Lf_) * y_[9] + omega * y_[10] + (vid_star - y_[11]) / Lf_;
f_[10] = -yp_[10] - (rLf_ / Lf_) * y_[10] - omega * y_[9] + (viq_star - y_[12]) / Lf_;

f_[12] = -yp_[12] + omega * y_[13] + (y_[10] - y_[14]) / Cf_;
f_[13] = -yp_[13] - omega * y_[12] + (y_[11] - y_[15]) / Cf_;
f_[11] = -yp_[11] + omega * y_[12] + (y_[9] - y_[13]) / Cf_;
f_[12] = -yp_[12] - omega * y_[11] + (y_[10] - y_[14]) / Cf_;

// Output Connector
f_[14] = -yp_[14] - (rLc_ / Lc_) * y_[14] + omega * y_[15] + (y_[12] - vbd_in) / Lc_;
f_[15] = -yp_[15] - (rLc_ / Lc_) * y_[15] - omega * y_[14] + (y_[13] - vbq_in) / Lc_;
f_[13] = -yp_[13] - (rLc_ / Lc_) * y_[13] + omega * y_[14] + (y_[11] - vbd_in) / Lc_;
f_[14] = -yp_[14] - (rLc_ / Lc_) * y_[14] - omega * y_[13] + (y_[12] - vbq_in) / Lc_;

// Rotor difference angle
if (!refframe_)
f_[15] = -yp_[15] + omega - y_[0];

return 0;
}

template <class ScalarT, typename IdxT>
int DistributedGenerator<ScalarT, IdxT>::evaluateExternalResidual()
{
ScalarT omega = wb_ - mp_ * y_[4];
ScalarT omega = wb_ - mp_ * y_[3];
// ref common ref motor angle
if (refframe_)
{
Expand All @@ -151,8 +153,8 @@ namespace GridKit

// output
// current transformed to common frame
f_[1] = std::cos(y_[3]) * y_[14] - std::sin(y_[3]) * y_[15];
f_[2] = std::sin(y_[3]) * y_[14] + std::cos(y_[3]) * y_[15];
f_[1] = std::cos(y_[15]) * y_[13] - std::sin(y_[15]) * y_[14];
f_[2] = std::sin(y_[15]) * y_[13] + std::cos(y_[15]) * y_[14];
return 0;
}

Expand Down Expand Up @@ -195,180 +197,190 @@ namespace GridKit
// Create dF/dy
// r = 1

ctemp = {3, 14, 15};
ctemp = {13, 14, 15};
rtemp.clear();
for (size_t i = 0; i < ctemp.size(); i++)
rtemp.push_back(1);
valtemp = {-sin(static_cast<RealT>(y_[3])) * static_cast<RealT>(y_[14]) - cos(static_cast<RealT>(y_[3])) * static_cast<RealT>(y_[15]),
cos(static_cast<RealT>(y_[3])),
-sin(static_cast<RealT>(y_[3]))};
valtemp = {
cos(static_cast<RealT>(y_[15])),
-sin(static_cast<RealT>(y_[15])),
-sin(static_cast<RealT>(y_[15])) * static_cast<RealT>(y_[13]) - cos(static_cast<RealT>(y_[15])) * static_cast<RealT>(y_[14]),
};
this->setJacValues(rtemp, ctemp, valtemp);

// r = 2

ctemp = {3, 14, 15};
ctemp = {13, 14, 15};
rtemp.clear();
for (size_t i = 0; i < ctemp.size(); i++)
rtemp.push_back(2);
valtemp = {cos(static_cast<RealT>(y_[3])) * static_cast<RealT>(y_[14]) - sin(static_cast<RealT>(y_[3])) * static_cast<RealT>(y_[15]),
sin(static_cast<RealT>(y_[3])),
cos(static_cast<RealT>(y_[3]))};
this->setJacValues(rtemp, ctemp, valtemp);

// r = 3

ctemp = {0, 3, 4};
rtemp.clear();
for (size_t i = 0; i < ctemp.size(); i++)
rtemp.push_back(3);
valtemp = {-1.0, -alpha_, -mp_};
valtemp = {
sin(static_cast<RealT>(y_[15])),
cos(static_cast<RealT>(y_[15])),
cos(static_cast<RealT>(y_[15])) * static_cast<RealT>(y_[13]) - sin(static_cast<RealT>(y_[15])) * static_cast<RealT>(y_[14]),
};
this->setJacValues(rtemp, ctemp, valtemp);

// r = 0
if (refframe_)
{
ctemp = {0, 4};
ctemp = {0, 3};
rtemp.clear();
for (size_t i = 0; i < ctemp.size(); i++)
rtemp.push_back(0);
valtemp = {-1.0, -mp_};
this->setJacValues(rtemp, ctemp, valtemp);
}

// r = 3
ctemp = {3, 11, 12, 13, 14};
rtemp.clear();
for (size_t i = 0; i < ctemp.size(); i++)
rtemp.push_back(3);
valtemp = {-wc_ - alpha_,
wc_ * static_cast<RealT>(y_[13]),
wc_ * static_cast<RealT>(y_[14]),
wc_ * static_cast<RealT>(y_[11]),
wc_ * static_cast<RealT>(y_[12])};
this->setJacValues(rtemp, ctemp, valtemp);

// r = 4
ctemp = {4, 12, 13, 14, 15};
ctemp = {4, 11, 12, 13, 14};
rtemp.clear();
for (size_t i = 0; i < ctemp.size(); i++)
rtemp.push_back(4);
valtemp = {-wc_ - alpha_,
wc_ * static_cast<RealT>(y_[14]),
wc_ * static_cast<RealT>(y_[15]),
-wc_ * static_cast<RealT>(y_[14]),
wc_ * static_cast<RealT>(y_[13]),
wc_ * static_cast<RealT>(y_[12]),
wc_ * static_cast<RealT>(y_[13])};
-wc_ * static_cast<RealT>(y_[11])};
this->setJacValues(rtemp, ctemp, valtemp);

// r = 5
ctemp = {5, 12, 13, 14, 15};
ctemp = {4, 5, 11};
rtemp.clear();
for (size_t i = 0; i < ctemp.size(); i++)
rtemp.push_back(5);
valtemp = {-wc_ - alpha_,
-wc_ * static_cast<RealT>(y_[15]),
wc_ * static_cast<RealT>(y_[14]),
wc_ * static_cast<RealT>(y_[13]),
-wc_ * static_cast<RealT>(y_[12])};
valtemp = {-nq_, -alpha_, -1.0};
this->setJacValues(rtemp, ctemp, valtemp);

// r = 6
ctemp = {5, 6, 12};
ctemp = {6, 12};
rtemp.clear();
for (size_t i = 0; i < ctemp.size(); i++)
rtemp.push_back(6);
valtemp = {-nq_, -alpha_, -1.0};
valtemp = {-alpha_, -1.0};
this->setJacValues(rtemp, ctemp, valtemp);

// r = 7
ctemp = {7, 13};
ctemp = {4, 5, 7, 9, 11, 12, 13};
rtemp.clear();
for (size_t i = 0; i < ctemp.size(); i++)
rtemp.push_back(7);
valtemp = {-alpha_, -1.0};
valtemp = {-Kpv_ * nq_, Kiv_, -alpha_, -1.0, -Kpv_, -Cf_ * wb_, F_};
this->setJacValues(rtemp, ctemp, valtemp);

// r = 8
ctemp = {5, 6, 8, 10, 12, 13, 14};
ctemp = {6, 8, 10, 11, 12, 14};
rtemp.clear();
for (size_t i = 0; i < ctemp.size(); i++)
rtemp.push_back(8);
valtemp = {-Kpv_ * nq_, Kiv_, -alpha_, -1.0, -Kpv_, -Cf_ * wb_, F_};
valtemp = {Kiv_, -alpha_, -1.0, Cf_ * wb_, -Kpv_, F_};
this->setJacValues(rtemp, ctemp, valtemp);

// r = 9
ctemp = {7, 9, 11, 12, 13, 15};
ctemp = {3, 4, 5, 7, 9, 10, 11, 12, 13};
rtemp.clear();
for (size_t i = 0; i < ctemp.size(); i++)
rtemp.push_back(9);
valtemp = {Kiv_, -alpha_, -1.0, Cf_ * wb_, -Kpv_, F_};
this->setJacValues(rtemp, ctemp, valtemp);

// r = 10
ctemp = {4, 5, 6, 8, 10, 11, 12, 13, 14};
rtemp.clear();
for (size_t i = 0; i < ctemp.size(); i++)
rtemp.push_back(10);
valtemp = {-mp_ * static_cast<RealT>(y_[11]),
valtemp = {-mp_ * static_cast<RealT>(y_[10]),
-(Kpc_ * Kpv_ * nq_) / Lf_,
(Kpc_ * Kiv_) / Lf_,
Kic_ / Lf_,
-(Kpc_ + rLf_) / Lf_ - alpha_,
-mp_ * static_cast<RealT>(y_[4]),
-mp_ * static_cast<RealT>(y_[3]),
-(Kpc_ * Kpv_ + 1.0) / Lf_,
-(Cf_ * Kpc_ * wb_) / Lf_,
(F_ * Kpc_) / Lf_};
this->setJacValues(rtemp, ctemp, valtemp);

// r = 11
ctemp = {4, 7, 9, 10, 11, 12, 13, 15};
// r = 10
ctemp = {3, 6, 8, 9, 10, 11, 12, 14};
rtemp.clear();
for (size_t i = 0; i < ctemp.size(); i++)
rtemp.push_back(11);
valtemp = {mp_ * static_cast<RealT>(y_[10]),
rtemp.push_back(10);
valtemp = {mp_ * static_cast<RealT>(y_[9]),
(Kiv_ * Kpc_) / Lf_,
Kic_ / Lf_,
mp_ * static_cast<RealT>(y_[4]),
mp_ * static_cast<RealT>(y_[3]),
-(Kpc_ + rLf_) / Lf_ - alpha_,
(Cf_ * Kpc_ * wb_) / Lf_,
-(Kpc_ * Kpv_ + 1.0) / Lf_,
(F_ * Kpc_) / Lf_};
this->setJacValues(rtemp, ctemp, valtemp);

// r = 12
ctemp = {4, 10, 12, 13, 14};
// r = 11
ctemp = {3, 9, 11, 12, 13};
rtemp.clear();
for (size_t i = 0; i < ctemp.size(); i++)
rtemp.push_back(12);
valtemp = {-mp_ * static_cast<RealT>(y_[13]),
rtemp.push_back(11);
valtemp = {-mp_ * static_cast<RealT>(y_[12]),
1.0 / Cf_,
-alpha_,
wb_ - mp_ * static_cast<RealT>(y_[4]),
wb_ - mp_ * static_cast<RealT>(y_[3]),
-1.0 / Cf_};
this->setJacValues(rtemp, ctemp, valtemp);

// r = 12
ctemp = {3, 10, 11, 12, 14};
rtemp.clear();
for (size_t i = 0; i < ctemp.size(); i++)
rtemp.push_back(12);
valtemp = {mp_ * static_cast<RealT>(y_[11]), 1.0 / Cf_, -wb_ + mp_ * static_cast<RealT>(y_[3]), -alpha_, -1.0 / Cf_};
this->setJacValues(rtemp, ctemp, valtemp);

// r = 13
ctemp = {4, 11, 12, 13, 15};
ctemp = {1, 2, 3, 11, 13, 14, 15};
rtemp.clear();
for (size_t i = 0; i < ctemp.size(); i++)
rtemp.push_back(13);
valtemp = {mp_ * static_cast<RealT>(y_[12]), 1.0 / Cf_, -wb_ + mp_ * static_cast<RealT>(y_[4]), -alpha_, -1.0 / Cf_};
valtemp = {
(1.0 / Lc_) * -cos(static_cast<RealT>(y_[15])),
(1.0 / Lc_) * -sin(static_cast<RealT>(y_[15])),
-mp_ * static_cast<RealT>(y_[14]),
1.0 / Lc_,
-rLc_ / Lc_ - alpha_,
wb_ - mp_ * static_cast<RealT>(y_[3]),
(1.0 / Lc_) * (sin(static_cast<RealT>(y_[15])) * static_cast<RealT>(y_[1]) - cos(static_cast<RealT>(y_[15])) * static_cast<RealT>(y_[2])),
};
this->setJacValues(rtemp, ctemp, valtemp);

// r = 14
ctemp = {1, 2, 3, 4, 12, 14, 15};
ctemp = {1, 2, 3, 12, 13, 14, 15};
rtemp.clear();
for (size_t i = 0; i < ctemp.size(); i++)
rtemp.push_back(14);
valtemp = {(1.0 / Lc_) * -cos(static_cast<RealT>(y_[3])),
(1.0 / Lc_) * -sin(static_cast<RealT>(y_[3])),
(1.0 / Lc_) * (sin(static_cast<RealT>(y_[3])) * static_cast<RealT>(y_[1]) - cos(static_cast<RealT>(y_[3])) * static_cast<RealT>(y_[2])),
-mp_ * static_cast<RealT>(y_[15]),
1.0 / Lc_,
-rLc_ / Lc_ - alpha_,
wb_ - mp_ * static_cast<RealT>(y_[4])};
valtemp = {
(1.0 / Lc_) * sin(static_cast<RealT>(y_[15])),
(1.0 / Lc_) * -cos(static_cast<RealT>(y_[15])),
mp_ * static_cast<RealT>(y_[13]),
1.0 / Lc_,
-wb_ + mp_ * static_cast<RealT>(y_[3]),
-rLc_ / Lc_ - alpha_,
(1.0 / Lc_) * (cos(static_cast<RealT>(y_[15])) * static_cast<RealT>(y_[1]) + sin(static_cast<RealT>(y_[15])) * static_cast<RealT>(y_[2])),
};
this->setJacValues(rtemp, ctemp, valtemp);

// r = 15
ctemp = {1, 2, 3, 4, 13, 14, 15};
rtemp.clear();
for (size_t i = 0; i < ctemp.size(); i++)
rtemp.push_back(15);
valtemp = {(1.0 / Lc_) * sin(static_cast<RealT>(y_[3])),
(1.0 / Lc_) * -cos(static_cast<RealT>(y_[3])),
(1.0 / Lc_) * (cos(static_cast<RealT>(y_[3])) * static_cast<RealT>(y_[1]) + sin(static_cast<RealT>(y_[3])) * static_cast<RealT>(y_[2])),
mp_ * static_cast<RealT>(y_[14]),
1.0 / Lc_,
-wb_ + mp_ * static_cast<RealT>(y_[4]),
-rLc_ / Lc_ - alpha_};
this->setJacValues(rtemp, ctemp, valtemp);
if (!refframe_)
{
// r = 15
ctemp = {0, 3, 15};
rtemp.clear();
for (size_t i = 0; i < ctemp.size(); i++)
rtemp.push_back(15);
valtemp = {-1.0, -mp_, -alpha_};
this->setJacValues(rtemp, ctemp, valtemp);
}

return 0;
}
Expand All @@ -384,7 +396,7 @@ namespace GridKit

if (refframe_)
{
this->setExternalConnectionNodes(3, INVALID_INDEX<IdxT>);
this->setExternalConnectionNodes(15, INVALID_INDEX<IdxT>);
}

return 0;
Expand Down
Loading
Loading