Example: Synthetic module EL IV
This example shows how to build synthetic module EL IVs using single cell SDM parameters
[1]:
import os
from pathlib import Path
project_root = Path.cwd().parents[1]
os.chdir(project_root) # now cwd is .../pvcracks
from pvcracks.pvspice_lite.pvspice_tools import run_ngspice
from pvcracks.pvspice_lite.pvspice_tools import ngpsice_read_voltage_current_modules
from pvcracks.pvspice_lite.pvspice_tools import Create_Cell_NetCode
from pvcracks.pvspice_lite.pvspice_tools import MiniMod_Spice
from pvcracks.pvspice_lite.pvspice_helper import Read_IV
from pvcracks.pvspice_lite.pvspice_helper import Read_EL
from pvcracks.pvspice_lite.pvspice_helper import cells2Mod
from pvcracks.pvspice_lite.pvspice_helper import Extract_Params
import numpy as np
import pandas as pd
import cv2
import matplotlib.pyplot as plt
import pvlib
import scipy
Install ngspice
Install ngspice with conda: - conda install ngspice Find path (e.g. “/home//.conda/envs/pyhpc_spice/bin/ngspice”) Edit the spicepath below to be that path
[2]:
# spicepath = <your spice path>
# spicepath = '/home/bkbyfor/anaconda3/envs/Things/bin/ngspice'
spicepath = "/home/nrjost/.conda/envs/pyhpc_spice/bin/ngspice"
Read Data
[3]:
# Master DataFrame
Cell9Master = pd.read_csv('docs/data/ELdata_module_209.csv', index_col=0)
# 3x3 minimodule IVs
FM_Init = Read_IV('docs/data/Minimodule_IVs/209_s0_FM_1000.csv')
FM_Deg1 = Read_IV('docs/data/Minimodule_IVs/209_s1_FM_1000.csv')
FM_Deg2 = Read_IV('docs/data/Minimodule_IVs/209_s2_FM_1000.csv')
[4]:
Cell9Master.head()
[4]:
| ELPath | Module | Deg | Rs | Rsh | I | Is | N | Pmp | Vmp | Imp | |
|---|---|---|---|---|---|---|---|---|---|---|---|
| 0 | /docs/data/EL/209_A3/Init/209_A3_0005_2021_04_... | 209_A3 | Init | 0.008625 | 525.0 | 8.242323 | 6.545686e-08 | 1.292650 | 3.564256 | 0.465455 | 7.657580 |
| 1 | /docs/data/EL/209_C1/Init/209_C1_0005_2021_04_... | 209_C1 | Init | 0.009957 | 1543.5 | 8.201203 | 3.488834e-08 | 1.267408 | 3.547602 | 0.465455 | 7.621802 |
| 2 | /docs/data/EL/209_A2/Init/209_A2_0005_2021_04_... | 209_A2 | Init | 0.008988 | 1380.5 | 8.223689 | 6.476966e-08 | 1.293979 | 3.543461 | 0.465455 | 7.612904 |
| 3 | /docs/data/EL/209_A1/Init/209_A1_0005_2021_04_... | 209_A1 | Init | 0.009377 | 1789.5 | 8.185086 | 3.607070e-08 | 1.254117 | 3.524153 | 0.465455 | 7.571422 |
| 4 | /docs/data/EL/209_C2/Init/209_C2_0005_2021_04_... | 209_C2 | Init | 0.009753 | 1723.0 | 8.122330 | 2.261230e-08 | 1.224717 | 3.494077 | 0.465455 | 7.506806 |
Cell9Master description
ElPath : Path to the EL image of the cell
Module : Name of the cell
Deg : Degredation step of Mini Module
SDM params found through fitting of the measured cell IVs
Rs : Series Resistance
Rsh : Shunt Resistance
I : Current
Is : Dark Current
N : Diode Ideality Factor
Display Single Cell Fits
[5]:
# iterate over the dataframe to display each entry
for ind, row in Cell9Master.iterrows():
# File Path for writing Spice files
file_path = f"{os.getcwd()}/pvcracks/pvspice_lite/"
# Read image
img = Read_EL(f"{os.getcwd()}{row.ELPath}")
# Order SDM params in the passed DataFrame into a list of dictionary for MiniMod_Spice
Params = Extract_Params(Cell9Master.loc[ind:ind])
# Voltage points for Spice simulation
V = np.linspace(0, 0.64, 100)
# Spice simulation for series connected SDM cells
IV_Pred = MiniMod_Spice(Params, file_path, V, V_Step=0.001, file_name ='', spicepath=spicepath)
#Calculate maximum power point values
V_sim = IV_Pred[0]
I_sim = IV_Pred[1]
P = V_sim * I_sim
idx_mpp = np.argmax(P)
Pmp = P[idx_mpp]
# Plot Findings
fig, axs = plt.subplots(1, 2, figsize = (12,5))
fig.suptitle(f'Cell {row.Module} Deg {row.Deg}')
axs[0].imshow(img)
axs[0].set_title('EL image')
axs[1].plot(IV_Pred[0], IV_Pred[1], label = 'Spice IV')
axs[1].text(
0.75, 0.90, f"Pmp={Pmp:.3f}W",
transform=axs[1].transAxes,
fontsize=10,
verticalalignment='top',
bbox=dict(boxstyle='round', facecolor='white', alpha=0.7)
)
axs[1].set_xlabel('Voltage')
axs[1].set_ylabel('Current')
axs[1].set_title('IV Curves')
axs[1].legend()
plt.show()
Simulate 3x3 MiniMod
Initial
[6]:
# query the slice of initial readings sorted by the Module name
Init_Cells = Cell9Master.loc[Cell9Master.Deg == 'Init'].sort_values(by=['Module'])
[7]:
# File Path for writing Spice files
file_path = f"{os.getcwd()}/pvcracks/pvspice_lite/"
# Construct Image for cells into module
MMEL = cells2Mod(Init_Cells,(400,400),3,3)
# Order SDM params in the passed DataFrame into a list of dictionary for MiniMod_Spice
Params = Extract_Params(Init_Cells)
# Spice simulation for series connected SDM cells
IV_Pred = MiniMod_Spice(Params, file_path, FM_Init[0], V_Step=0.001, file_name ='', spicepath=spicepath)
# unpack measured vs. predicted
V_meas, I_meas = FM_Init
V_sp, I_sp = IV_Pred
# compute Power curves
P_meas = V_meas * I_meas
P_sp = V_sp * I_sp
# locate maximum‐power indices
idx_meas = np.argmax(P_meas)
idx_sp = np.argmax(P_sp)
# extract MPPs
Pmp_meas = P_meas[idx_meas]
Pmp_sp = P_sp[idx_sp]
# Plot Findigs
fig, axs = plt.subplots(1, 3, figsize = (17, 5))
axs[0].imshow(MMEL)
axs[0].set_title('EL image')
axs[1].plot(V_meas, I_meas, label = f'Measured IV (Pmpp={Pmp_meas:.3f} W)')
axs[1].plot(V_sp, I_sp, label = f'Spice IV (Pmpp={Pmp_sp:.3f} W)')
axs[1].set_xlabel('Voltage')
axs[1].set_ylabel('Current')
axs[1].set_title('IV Curves')
axs[1].legend()
axs[2].set_title('Residual')
axs[2].plot(FM_Init[0],FM_Init[1] - IV_Pred[1])
axs[2].set_ylabel('Current')
axs[2].set_xlabel('Sample')
[7]:
Text(0.5, 0, 'Sample')
Deg1
[8]:
# query the slice of Deg 1 readings sorted by the Module name
Deg1_Cells = Cell9Master.loc[Cell9Master.Deg == 'Deg1'].sort_values(by=['Module'])
[9]:
# File Path for writing Spice files
file_path = f"{os.getcwd()}/pvcracks/pvspice_lite/"
# Construct Image for cells into module
MMEL = cells2Mod(Deg1_Cells,(400,400),3,3)
# Order SDM params in the passed DataFrame into a list of dictionary for MiniMod_Spice
Params = Extract_Params(Deg1_Cells)
# Spice simulation for series connected SDM cells
IV_Pred = MiniMod_Spice(Params, file_path, FM_Deg1[0], V_Step=0.001, file_name ='', spicepath=spicepath)
# unpack measured vs. predicted
V_meas, I_meas = FM_Deg1
V_sp, I_sp = IV_Pred
# compute Power curves
P_meas = V_meas * I_meas
P_sp = V_sp * I_sp
# locate maximum‐power indices
idx_meas = np.argmax(P_meas)
idx_sp = np.argmax(P_sp)
# extract MPPs
Pmp_meas = P_meas[idx_meas]
Pmp_sp = P_sp[idx_sp]
# Plot Findigs
fig, axs = plt.subplots(1, 3, figsize = (17, 5))
axs[0].imshow(MMEL)
axs[0].set_title('EL image')
axs[1].plot(V_meas, I_meas, label = f'Measured IV (Pmpp={Pmp_meas:.3f} W)')
axs[1].plot(V_sp, I_sp, label = f'Spice IV (Pmpp={Pmp_sp:.3f} W)')
axs[1].set_xlabel('Voltage')
axs[1].set_ylabel('Current')
axs[1].set_title('IV Curves')
axs[1].legend()
axs[2].set_title('Residual')
axs[2].plot(FM_Deg1[0],FM_Deg1[1] - IV_Pred[1])
axs[2].set_ylabel('Current')
axs[2].set_xlabel('Sample')
[9]:
Text(0.5, 0, 'Sample')
Deg 2
[10]:
# query the slice of Deg 2 readings sorted by the Module name
Deg2_Cells = Cell9Master.loc[Cell9Master.Deg == 'Deg2'].sort_values(by=['Module'])
[11]:
# File Path for writing Spice files
file_path = f"{os.getcwd()}/pvcracks/pvspice_lite/"
# Construct Image for cells into module
MMEL = cells2Mod(Deg2_Cells,(400,400),3,3)
# Order SDM params in the passed DataFrame into a list of dictionary for MiniMod_Spice
Params = Extract_Params(Deg2_Cells)
# Spice simulation for series connected SDM cells
IV_Pred = MiniMod_Spice(Params, file_path, FM_Deg2[0], V_Step=0.001, file_name ='', spicepath=spicepath)
# unpack measured vs. predicted
V_meas, I_meas = FM_Deg2
V_sp, I_sp = IV_Pred
# compute Power curves
P_meas = V_meas * I_meas
P_sp = V_sp * I_sp
# locate maximum‐power indices
idx_meas = np.argmax(P_meas)
idx_sp = np.argmax(P_sp)
# extract MPPs
Pmp_meas = P_meas[idx_meas]
Pmp_sp = P_sp[idx_sp]
# Plot Findigs
fig, axs = plt.subplots(1, 3, figsize = (17, 5))
axs[0].imshow(MMEL)
axs[0].set_title('EL image')
axs[1].plot(V_meas, I_meas, label = f'Measured IV (Pmpp={Pmp_meas:.3f} W)')
axs[1].plot(V_sp, I_sp, label = f'Spice IV (Pmpp={Pmp_sp:.3f} W)')
axs[1].set_xlabel('Voltage')
axs[1].set_ylabel('Current')
axs[1].set_title('IV Curves')
axs[1].legend()
axs[2].set_title('Residual')
axs[2].plot(FM_Deg2[0],FM_Deg2[1] - IV_Pred[1])
axs[2].set_ylabel('Current')
axs[2].set_xlabel('Sample')
[11]:
Text(0.5, 0, 'Sample')
Constructing Synthetic MiniModule Data
[12]:
# Randomly select 4 init cells and combine them into a single dataframe
Synth_MiniMod_Init = Init_Cells.sample(n=4)
# Select the damaged cell (top left corner) in Deg1 and repeat it 4 times
Synth_MiniMod_Deg1 = Deg1_Cells.loc[[12,12,12,12]]
# Select the damaged cell (top left corner) in Deg2 and repeat it 4 times
Synth_MiniMod_Deg2 = Deg2_Cells.loc[[21,21,21,21]]
# Mix 2 initial cells and the most damaged cell in Deg2
Synth_MiniMod_Mix = pd.concat((Deg2_Cells.loc[[21,21]],Init_Cells.sample(n=2) ) )
[13]:
# File Path for writing Spice files
file_path = f"{os.getcwd()}/pvcracks/pvspice_lite/"
# Voltage over which to simulate
V = np.linspace(0, 0.64*4, 400)
# Construct Image for cells into module
MMEL_Mix = cells2Mod(Synth_MiniMod_Mix,(400, 400),2,2)
# Order SDM params in the passed DataFrame into a list of dictionary for MiniMod_Spice
Params = Extract_Params(Synth_MiniMod_Mix)
# Spice simulation for series connected SDM cells
IV_Pred_Mix = MiniMod_Spice(Params, file_path, V, V_Step=0.001, file_name ='', spicepath=spicepath)
# Construct Image for cells into module
MMEL_Init = cells2Mod(Synth_MiniMod_Init,(400, 400),2,2)
# Order SDM params in the passed DataFrame into a list of dictionary for MiniMod_Spice
Params = Extract_Params(Synth_MiniMod_Init)
# Spice simulation for series connected SDM cells
IV_Pred_Init = MiniMod_Spice(Params, file_path, V, V_Step=0.001, file_name ='', spicepath=spicepath)
# Construct Image for cells into module
MMEL_Deg1 = cells2Mod(Synth_MiniMod_Deg1,(400, 400),2,2)
# Order SDM params in the passed DataFrame into a list of dictionary for MiniMod_Spice
Params = Extract_Params(Synth_MiniMod_Deg1)
# Spice simulation for series connected SDM cells
IV_Pred_Deg1 = MiniMod_Spice(Params, file_path, V, V_Step=0.001, file_name ='', spicepath=spicepath)
# Construct Image for cells into module
MMEL_Deg2 = cells2Mod(Synth_MiniMod_Deg2,(400, 400),2,2)
# Order SDM params in the passed DataFrame into a list of dictionary for MiniMod_Spice
Params = Extract_Params(Synth_MiniMod_Deg2)
# Spice simulation for series connected SDM cells
IV_Pred_Deg2 = MiniMod_Spice(Params, file_path, V, V_Step=0.001, file_name ='', spicepath=spicepath)
def compute_Pmpp(iv_pred):
Varr, Iarr = iv_pred
Parr = Varr * Iarr
return Parr.max()
# compute all four Pmpp values
Pmp_mix = compute_Pmpp(IV_Pred_Mix)
Pmp_init = compute_Pmpp(IV_Pred_Init)
Pmp_deg1 = compute_Pmpp(IV_Pred_Deg1)
Pmp_deg2 = compute_Pmpp(IV_Pred_Deg2)
# 1×2 figures for each dataset
for MMEL, iv_pred, title, Pmp in [
(MMEL_Mix, IV_Pred_Mix, "Mix Synthetic Data", Pmp_mix),
(MMEL_Init, IV_Pred_Init, "Init Synthetic Data", Pmp_init),
(MMEL_Deg1, IV_Pred_Deg1, "Deg1 Synthetic Data", Pmp_deg1),
(MMEL_Deg2, IV_Pred_Deg2, "Deg2 Synthetic Data", Pmp_deg2),
]:
fig, axs = plt.subplots(1, 2, figsize=(10,5))
fig.suptitle(title)
axs[0].imshow(MMEL)
axs[0].axis('off')
axs[0].set_title('EL image')
axs[1].plot(iv_pred[0], iv_pred[1],
label=f"{title.split()[0]} IV (Pmp={Pmp:.3f} W)")
axs[1].set_xlabel('Voltage')
axs[1].set_ylabel('Current')
axs[1].set_title('IV Curves')
axs[1].legend()
plt.tight_layout(rect=[0,0,1,0.95])
plt.show()
# overlay with ΔPmpp relative to Init
plt.figure(figsize=(8,6))
plt.title("Overlay of synthetic cell IVs")
# plot Init first, with its absolute Pmpp
plt.plot(IV_Pred_Init[0], IV_Pred_Init[1],
label=f"Init (Pmp={Pmp_init:.3f} W)")
# plot the others, with ΔPmpp = Pmpp_x – Pmpp_init
for label, iv_pred, Pmp in [
('Mix', IV_Pred_Mix, Pmp_mix),
('Deg1', IV_Pred_Deg1, Pmp_deg1),
('Deg2', IV_Pred_Deg2, Pmp_deg2),
]:
delta = (Pmp - Pmp_init)/Pmp_init*100
plt.plot(iv_pred[0], iv_pred[1],
label=f"{label} (ΔPmp={delta:+.2f}%)")
plt.xlabel('Voltage')
plt.ylabel('Current')
plt.legend()
plt.tight_layout()
plt.show()
[14]:
# Randomly select 9 init cells and combine them into a single dataframe
Synth_MiniMod_Init = Init_Cells.sample(n=9)
# Select the damaged cell (top left corner) in Deg1 and repeat it 9 times
Synth_MiniMod_Deg1 = Deg1_Cells.loc[[12,12,12,12,12,12,12,12,12]]
# Select the damaged cell (top left corner) in Deg2 and repeat it 9 times
Synth_MiniMod_Deg2 = Deg2_Cells.loc[[21,21,21,21,21,21,21,21,21]]
# Mix 3 initial cells and the most damaged cell in Deg2
Synth_MiniMod_Mix = pd.concat((Deg2_Cells.loc[[21,21,21,21,21,21]], Init_Cells.sample(n=3)))
[15]:
# File Path for writing Spice files
file_path = f"{os.getcwd()}/pvcracks/pvspice_lite/"
# Voltage over which to simulate
V = np.linspace(0, 0.64*9, 400)
# Construct Image for cells into module
MMEL_Mix = cells2Mod(Synth_MiniMod_Mix, (400, 400), 3, 3)
# Order SDM params in the passed DataFrame into a list of dictionary for MiniMod_Spice
Params = Extract_Params(Synth_MiniMod_Mix)
# Spice simulation for series connected SDM cells
IV_Pred_Mix = MiniMod_Spice(Params, file_path, V, V_Step=0.001, file_name ='', spicepath=spicepath)
# Construct Image for cells into module
MMEL_Init = cells2Mod(Synth_MiniMod_Init, (400, 400), 3, 3)
# Order SDM params in the passed DataFrame into a list of dictionary for MiniMod_Spice
Params = Extract_Params(Synth_MiniMod_Init)
# Spice simulation for series connected SDM cells
IV_Pred_Init = MiniMod_Spice(Params, file_path, V, V_Step=0.001, file_name ='', spicepath=spicepath)
# Construct Image for cells into module
MMEL_Deg1 = cells2Mod(Synth_MiniMod_Deg1, (400, 400), 3, 3)
# Order SDM params in the passed DataFrame into a list of dictionary for MiniMod_Spice
Params = Extract_Params(Synth_MiniMod_Deg1)
# Spice simulation for series connected SDM cells
IV_Pred_Deg1 = MiniMod_Spice(Params, file_path, V, V_Step=0.001, file_name ='', spicepath=spicepath)
# Construct Image for cells into module
MMEL_Deg2 = cells2Mod(Synth_MiniMod_Deg2, (400, 400), 3, 3)
# Order SDM params in the passed DataFrame into a list of dictionary for MiniMod_Spice
Params = Extract_Params(Synth_MiniMod_Deg2)
# Spice simulation for series connected SDM cells
IV_Pred_Deg2 = MiniMod_Spice(Params, file_path, V, V_Step=0.001, file_name ='', spicepath=spicepath)
def compute_Pmpp(iv_pred):
Varr, Iarr = iv_pred
Parr = Varr * Iarr
return Parr.max()
# compute all four Pmpp values
Pmp_mix = compute_Pmpp(IV_Pred_Mix)
Pmp_init = compute_Pmpp(IV_Pred_Init)
Pmp_deg1 = compute_Pmpp(IV_Pred_Deg1)
Pmp_deg2 = compute_Pmpp(IV_Pred_Deg2)
# 1×2 figures for each dataset
for MMEL, iv_pred, title, Pmp in [
(MMEL_Mix, IV_Pred_Mix, "Mix Synthetic Data", Pmp_mix),
(MMEL_Init, IV_Pred_Init, "Init Synthetic Data", Pmp_init),
(MMEL_Deg1, IV_Pred_Deg1, "Deg1 Synthetic Data", Pmp_deg1),
(MMEL_Deg2, IV_Pred_Deg2, "Deg2 Synthetic Data", Pmp_deg2),
]:
fig, axs = plt.subplots(1, 2, figsize=(10,5))
fig.suptitle(title)
axs[0].imshow(MMEL)
axs[0].axis('off')
axs[0].set_title('EL image')
axs[1].plot(iv_pred[0], iv_pred[1],
label=f"{title.split()[0]} IV (Pmp={Pmp:.3f} W)")
axs[1].set_xlabel('Voltage')
axs[1].set_ylabel('Current')
axs[1].set_title('IV Curves')
axs[1].legend()
plt.tight_layout(rect=[0,0,1,0.95])
plt.show()
# overlay with ΔPmpp relative to Init
plt.figure(figsize=(8,6))
plt.title("Overlay of synthetic cell IVs")
# plot Init first, with its absolute Pmpp
plt.plot(IV_Pred_Init[0], IV_Pred_Init[1],
label=f"Init (Pmp={Pmp_init:.3f} W)")
# plot the others, with ΔPmpp = Pmpp_x – Pmpp_init
for label, iv_pred, Pmp in [
('Mix', IV_Pred_Mix, Pmp_mix),
('Deg1', IV_Pred_Deg1, Pmp_deg1),
('Deg2', IV_Pred_Deg2, Pmp_deg2),
]:
delta = (Pmp - Pmp_init)/Pmp_init*100
plt.plot(iv_pred[0], iv_pred[1],
label=f"{label} (ΔPmp={delta:+.2f}%)")
plt.xlabel('Voltage')
plt.ylabel('Current')
plt.legend()
plt.tight_layout()
plt.show()
