Note
Click here to download the full example code
IDF-based MDO on the Sobieski SSBJ test case¶
from __future__ import annotations
from gemseo.api import configure_logger
from gemseo.api import create_discipline
from gemseo.api import create_scenario
from gemseo.api import generate_n2_plot
from gemseo.problems.sobieski.core.problem import SobieskiProblem
configure_logger()
<RootLogger root (INFO)>
Instantiate the disciplines¶
First, we instantiate the four disciplines of the use case:
SobieskiPropulsion,
SobieskiAerodynamics,
SobieskiMission
and SobieskiStructure.
disciplines = create_discipline(
[
"SobieskiPropulsion",
"SobieskiAerodynamics",
"SobieskiMission",
"SobieskiStructure",
]
)
We can quickly access the most relevant information of any discipline (name, inputs,
and outputs) with Python’s print() function. Moreover, we can get the default
input values of a discipline with the attribute MDODiscipline.default_inputs
for discipline in disciplines:
print(discipline)
print(f"Default inputs: {discipline.default_inputs}")
SobieskiPropulsion
Default inputs: {'y_23': array([12562.01206488]), 'x_3': array([0.5]), 'x_shared': array([5.0e-02, 4.5e+04, 1.6e+00, 5.5e+00, 5.5e+01, 1.0e+03]), 'c_3': array([4360.])}
SobieskiAerodynamics
Default inputs: {'x_2': array([1.]), 'y_32': array([0.50279625]), 'x_shared': array([5.0e-02, 4.5e+04, 1.6e+00, 5.5e+00, 5.5e+01, 1.0e+03]), 'y_12': array([5.06069742e+04, 9.50000000e-01]), 'c_4': array([0.01375])}
SobieskiMission
Default inputs: {'y_14': array([50606.9741711 , 7306.20262124]), 'x_shared': array([5.0e-02, 4.5e+04, 1.6e+00, 5.5e+00, 5.5e+01, 1.0e+03]), 'y_24': array([4.15006276]), 'y_34': array([1.10754577])}
SobieskiStructure
Default inputs: {'y_21': array([50606.9741711]), 'y_31': array([6354.32430691]), 'x_1': array([0.25, 1. ]), 'x_shared': array([5.0e-02, 4.5e+04, 1.6e+00, 5.5e+00, 5.5e+01, 1.0e+03]), 'c_0': array([2000.]), 'c_1': array([25000.]), 'c_2': array([6.])}
You may also be interested in plotting the couplings of your disciplines.
A quick way of getting this information is the API function
generate_n2_plot(). A much more detailed explanation of coupling
visualization is available here.
generate_n2_plot(disciplines, save=False, show=True)
Build, execute and post-process the scenario¶
Then, we build the scenario which links the disciplines
with the formulation and the optimization algorithm. Here, we use the
IDF formulation. We tell the scenario to minimize -y_4 instead of
minimizing y_4 (range), which is the default option.
Instantiate the scenario¶
design_space = SobieskiProblem().design_space
print(design_space)
scenario = create_scenario(
disciplines,
"IDF",
objective_name="y_4",
design_space=design_space,
maximize_objective=True,
)
Design space:
+-------------+-------------+--------------------+-------------+-------+
| name | lower_bound | value | upper_bound | type |
+-------------+-------------+--------------------+-------------+-------+
| x_shared[0] | 0.01 | 0.05 | 0.09 | float |
| x_shared[1] | 30000 | 45000 | 60000 | float |
| x_shared[2] | 1.4 | 1.6 | 1.8 | float |
| x_shared[3] | 2.5 | 5.5 | 8.5 | float |
| x_shared[4] | 40 | 55 | 70 | float |
| x_shared[5] | 500 | 1000 | 1500 | float |
| x_1[0] | 0.1 | 0.25 | 0.4 | float |
| x_1[1] | 0.75 | 1 | 1.25 | float |
| x_2 | 0.75 | 1 | 1.25 | float |
| x_3 | 0.1 | 0.5 | 1 | float |
| y_14[0] | 24850 | 50606.9741711 | 77100 | float |
| y_14[1] | -7700 | 7306.20262124 | 45000 | float |
| y_32 | 0.235 | 0.5027962499999999 | 0.795 | float |
| y_31 | 2960 | 6354.32430691 | 10185 | float |
| y_24 | 0.44 | 4.15006276 | 11.13 | float |
| y_34 | 0.44 | 1.10754577 | 1.98 | float |
| y_23 | 3365 | 12194.2671934 | 26400 | float |
| y_21 | 24850 | 50606.9741711 | 77250 | float |
| y_12[0] | 24850 | 50606.9742 | 77250 | float |
| y_12[1] | 0.45 | 0.95 | 1.5 | float |
+-------------+-------------+--------------------+-------------+-------+
Set the design constraints¶
for c_name in ["g_1", "g_2", "g_3"]:
scenario.add_constraint(c_name, "ineq")
Define the algorithm inputs¶
We set the maximum number of iterations, the optimizer and the optimizer options
algo_options = {
"ftol_rel": 1e-10,
"ineq_tolerance": 1e-3,
"eq_tolerance": 1e-3,
"normalize_design_space": True,
}
scn_inputs = {"max_iter": 20, "algo": "SLSQP", "algo_options": algo_options}
Execute the scenario¶
scenario.execute(scn_inputs)
INFO - 17:23:48:
INFO - 17:23:48: *** Start MDOScenario execution ***
INFO - 17:23:48: MDOScenario
INFO - 17:23:48: Disciplines: SobieskiAerodynamics SobieskiMission SobieskiPropulsion SobieskiStructure
INFO - 17:23:48: MDO formulation: IDF
INFO - 17:23:48: Optimization problem:
INFO - 17:23:48: minimize -y_4(x_shared, y_14, y_24, y_34)
INFO - 17:23:48: with respect to x_1, x_2, x_3, x_shared, y_12, y_14, y_21, y_23, y_24, y_31, y_32, y_34
INFO - 17:23:48: subject to constraints:
INFO - 17:23:48: g_1(x_shared, x_1, y_31, y_21) <= 0.0
INFO - 17:23:48: g_2(x_shared, x_2, y_32, y_12) <= 0.0
INFO - 17:23:48: g_3(x_shared, x_3, y_23) <= 0.0
INFO - 17:23:48: y_31_y_32_y_34: y_31#y_32#y_34(x_shared, x_3, y_23): y_31(x_shared, x_3, y_23) - y_31 == 0.0
INFO - 17:23:48: y_32(x_shared, x_3, y_23) - y_32 == 0.0
INFO - 17:23:48: y_34(x_shared, x_3, y_23) - y_34 == 0.0
INFO - 17:23:48: y_21_y_23_y_24: y_21#y_23#y_24(x_shared, x_2, y_32, y_12): y_21(x_shared, x_2, y_32, y_12) - y_21 == 0.0
INFO - 17:23:48: y_23(x_shared, x_2, y_32, y_12) - y_23 == 0.0
INFO - 17:23:48: y_24(x_shared, x_2, y_32, y_12) - y_24 == 0.0
INFO - 17:23:48: y_12_y_14: y_12#y_14(x_shared, x_1, y_31, y_21): y_12(x_shared, x_1, y_31, y_21) - y_12 == 0.0
INFO - 17:23:48: y_14(x_shared, x_1, y_31, y_21) - y_14 == 0.0
INFO - 17:23:48: over the design space:
INFO - 17:23:48: +-------------+-------------+--------------------+-------------+-------+
INFO - 17:23:48: | name | lower_bound | value | upper_bound | type |
INFO - 17:23:48: +-------------+-------------+--------------------+-------------+-------+
INFO - 17:23:48: | x_shared[0] | 0.01 | 0.05 | 0.09 | float |
INFO - 17:23:48: | x_shared[1] | 30000 | 45000 | 60000 | float |
INFO - 17:23:48: | x_shared[2] | 1.4 | 1.6 | 1.8 | float |
INFO - 17:23:48: | x_shared[3] | 2.5 | 5.5 | 8.5 | float |
INFO - 17:23:48: | x_shared[4] | 40 | 55 | 70 | float |
INFO - 17:23:48: | x_shared[5] | 500 | 1000 | 1500 | float |
INFO - 17:23:48: | x_1[0] | 0.1 | 0.25 | 0.4 | float |
INFO - 17:23:48: | x_1[1] | 0.75 | 1 | 1.25 | float |
INFO - 17:23:48: | x_2 | 0.75 | 1 | 1.25 | float |
INFO - 17:23:48: | x_3 | 0.1 | 0.5 | 1 | float |
INFO - 17:23:48: | y_14[0] | 24850 | 50606.9741711 | 77100 | float |
INFO - 17:23:48: | y_14[1] | -7700 | 7306.20262124 | 45000 | float |
INFO - 17:23:48: | y_32 | 0.235 | 0.5027962499999999 | 0.795 | float |
INFO - 17:23:48: | y_31 | 2960 | 6354.32430691 | 10185 | float |
INFO - 17:23:48: | y_24 | 0.44 | 4.15006276 | 11.13 | float |
INFO - 17:23:48: | y_34 | 0.44 | 1.10754577 | 1.98 | float |
INFO - 17:23:48: | y_23 | 3365 | 12194.2671934 | 26400 | float |
INFO - 17:23:48: | y_21 | 24850 | 50606.9741711 | 77250 | float |
INFO - 17:23:48: | y_12[0] | 24850 | 50606.9742 | 77250 | float |
INFO - 17:23:48: | y_12[1] | 0.45 | 0.95 | 1.5 | float |
INFO - 17:23:48: +-------------+-------------+--------------------+-------------+-------+
INFO - 17:23:48: Solving optimization problem with algorithm SLSQP:
INFO - 17:23:48: ... 0%| | 0/20 [00:00<?, ?it]
INFO - 17:23:48: ... 5%|▌ | 1/20 [00:00<00:00, 143.88 it/sec, obj=-536]
INFO - 17:23:48: ... 10%|█ | 2/20 [00:00<00:00, 34.31 it/sec, obj=-1.49e+3]
INFO - 17:23:48: ... 15%|█▌ | 3/20 [00:00<00:00, 37.20 it/sec, obj=-3.83e+3]
INFO - 17:23:48: ... 20%|██ | 4/20 [00:00<00:00, 38.76 it/sec, obj=-3.87e+3]
INFO - 17:23:48: ... 25%|██▌ | 5/20 [00:00<00:00, 39.80 it/sec, obj=-3.96e+3]
INFO - 17:23:48: ... 30%|███ | 6/20 [00:00<00:00, 40.46 it/sec, obj=-3.96e+3]
INFO - 17:23:48: ... 35%|███▌ | 7/20 [00:00<00:00, 40.79 it/sec, obj=-3.96e+3]
INFO - 17:23:49: ... 40%|████ | 8/20 [00:00<00:00, 43.91 it/sec, obj=-3.96e+3]
INFO - 17:23:49: ... 45%|████▌ | 9/20 [00:00<00:00, 46.66 it/sec, obj=-3.96e+3]
INFO - 17:23:49: ... 50%|█████ | 10/20 [00:00<00:00, 49.15 it/sec, obj=-3.96e+3]
INFO - 17:23:49: ... 55%|█████▌ | 11/20 [00:00<00:00, 51.42 it/sec, obj=-3.96e+3]
INFO - 17:23:49: ... 60%|██████ | 12/20 [00:00<00:00, 53.45 it/sec, obj=-3.96e+3]
INFO - 17:23:49: ... 65%|██████▌ | 13/20 [00:00<00:00, 55.31 it/sec, obj=-3.96e+3]
INFO - 17:23:49: ... 70%|███████ | 14/20 [00:00<00:00, 57.01 it/sec, obj=-3.96e+3]
INFO - 17:23:49: ... 75%|███████▌ | 15/20 [00:00<00:00, 58.55 it/sec, obj=-3.96e+3]
INFO - 17:23:49: ... 80%|████████ | 16/20 [00:00<00:00, 60.03 it/sec, obj=-3.96e+3]
INFO - 17:23:49: ... 85%|████████▌ | 17/20 [00:00<00:00, 61.37 it/sec, obj=-3.96e+3]
INFO - 17:23:49: ... 90%|█████████ | 18/20 [00:00<00:00, 59.97 it/sec, obj=-3.97e+3]
INFO - 17:23:49: ... 95%|█████████▌| 19/20 [00:00<00:00, 61.16 it/sec, obj=-3.97e+3]
INFO - 17:23:49: ... 100%|██████████| 20/20 [00:00<00:00, 62.26 it/sec, obj=-3.96e+3]
INFO - 17:23:49: Optimization result:
INFO - 17:23:49: Optimizer info:
INFO - 17:23:49: Status: None
INFO - 17:23:49: Message: Maximum number of iterations reached. GEMSEO Stopped the driver
INFO - 17:23:49: Number of calls to the objective function by the optimizer: 22
INFO - 17:23:49: Solution:
INFO - 17:23:49: The solution is feasible.
INFO - 17:23:49: Objective: -3966.790848506069
INFO - 17:23:49: Standardized constraints:
INFO - 17:23:49: g_1 = [-0.01812782 -0.03339445 -0.0442868 -0.0518651 -0.05735184 -0.13720865
INFO - 17:23:49: -0.10279135]
INFO - 17:23:49: g_2 = 2.535682693616259e-05
INFO - 17:23:49: g_3 = [-7.67489138e-01 -2.32510862e-01 9.20908972e-05 -1.83255000e-01]
INFO - 17:23:49: y_12_y_14 = [ 3.34629122e-05 -1.85828412e-05 3.35589780e-05 -5.48378413e-05]
INFO - 17:23:49: y_21_y_23_y_24 = [ 0.00000000e+00 9.05477756e-05 -2.83060860e-04]
INFO - 17:23:49: y_31_y_32_y_34 = [6.90641576e-08 6.03563891e-08 1.07078689e-04]
INFO - 17:23:49: Design space:
INFO - 17:23:49: +-------------+-------------+---------------------+-------------+-------+
INFO - 17:23:49: | name | lower_bound | value | upper_bound | type |
INFO - 17:23:49: +-------------+-------------+---------------------+-------------+-------+
INFO - 17:23:49: | x_shared[0] | 0.01 | 0.06000633920673407 | 0.09 | float |
INFO - 17:23:49: | x_shared[1] | 30000 | 60000 | 60000 | float |
INFO - 17:23:49: | x_shared[2] | 1.4 | 1.4 | 1.8 | float |
INFO - 17:23:49: | x_shared[3] | 2.5 | 2.5 | 8.5 | float |
INFO - 17:23:49: | x_shared[4] | 40 | 70 | 70 | float |
INFO - 17:23:49: | x_shared[5] | 500 | 1500 | 1500 | float |
INFO - 17:23:49: | x_1[0] | 0.1 | 0.4 | 0.4 | float |
INFO - 17:23:49: | x_1[1] | 0.75 | 0.75 | 1.25 | float |
INFO - 17:23:49: | x_2 | 0.75 | 0.75 | 1.25 | float |
INFO - 17:23:49: | x_3 | 0.1 | 0.156259134365097 | 1 | float |
INFO - 17:23:49: | y_14[0] | 24850 | 44746.56244377587 | 77100 | float |
INFO - 17:23:49: | y_14[1] | -7700 | 19355.27544458607 | 45000 | float |
INFO - 17:23:49: | y_32 | 0.235 | 0.7325108278699097 | 0.795 | float |
INFO - 17:23:49: | y_31 | 2960 | 9433.274956925387 | 10185 | float |
INFO - 17:23:49: | y_24 | 0.44 | 8.059470045805423 | 11.13 | float |
INFO - 17:23:49: | y_34 | 0.44 | 0.9237614560895575 | 1.98 | float |
INFO - 17:23:49: | y_23 | 3365 | 5552.047264915329 | 26400 | float |
INFO - 17:23:49: | y_21 | 24850 | 44746.56244377587 | 77250 | float |
INFO - 17:23:49: | y_12[0] | 24850 | 44746.56244377587 | 77250 | float |
INFO - 17:23:49: | y_12[1] | 0.45 | 0.9028108618065331 | 1.5 | float |
INFO - 17:23:49: +-------------+-------------+---------------------+-------------+-------+
INFO - 17:23:49: *** End MDOScenario execution (time: 0:00:00.374632) ***
{'max_iter': 20, 'algo': 'SLSQP', 'algo_options': {'ftol_rel': 1e-10, 'ineq_tolerance': 0.001, 'eq_tolerance': 0.001, 'normalize_design_space': True}}
Save the optimization history¶
We can save the whole optimization problem and its history for further post processing:
scenario.save_optimization_history("idf_history.h5", file_format="hdf5")
INFO - 17:23:49: Export optimization problem to file: idf_history.h5
We can also save only calls to functions and design variables history:
scenario.save_optimization_history("idf_history.xml", file_format="ggobi")
INFO - 17:23:49: Export to ggobi for functions: ['-y_4', 'Iter', 'g_1', 'g_2', 'g_3', 'y_12_y_14', 'y_21_y_23_y_24', 'y_31_y_32_y_34']
INFO - 17:23:49: Export to ggobi file: idf_history.xml
Print optimization metrics¶
scenario.print_execution_metrics()
INFO - 17:23:49: Scenario Execution Statistics
INFO - 17:23:49: Discipline: SobieskiPropulsion
INFO - 17:23:49: Executions number: 20
INFO - 17:23:49: Execution time: 0.009780139996109938 s
INFO - 17:23:49: Linearizations number: 7
INFO - 17:23:49: Discipline: SobieskiAerodynamics
INFO - 17:23:49: Executions number: 20
INFO - 17:23:49: Execution time: 0.015109208000467333 s
INFO - 17:23:49: Linearizations number: 7
INFO - 17:23:49: Discipline: SobieskiMission
INFO - 17:23:49: Executions number: 20
INFO - 17:23:49: Execution time: 0.0015350510002463125 s
INFO - 17:23:49: Linearizations number: 7
INFO - 17:23:49: Discipline: SobieskiStructure
INFO - 17:23:49: Executions number: 20
INFO - 17:23:49: Execution time: 0.05731182700037607 s
INFO - 17:23:49: Linearizations number: 7
INFO - 17:23:49: Total number of executions calls: 80
INFO - 17:23:49: Total number of linearizations: 28
Plot the optimization history view¶
scenario.post_process("OptHistoryView", save=True, show=True)
<gemseo.post.opt_history_view.OptHistoryView object at 0x7fccd3e771f0>
Plot the quadratic approximation of the objective¶
scenario.post_process("QuadApprox", function="-y_4", save=False, show=True)
<gemseo.post.quad_approx.QuadApprox object at 0x7fccd2f7af70>
Total running time of the script: ( 0 minutes 7.295 seconds)