Skip to content

Commit 3f5daac

Browse files
AbyAbraham21AbyAbraham21
authored
Merge pull request #45 from simvue-io/main
Merge in changes from main
2 parents 739b8b8 + 96bb96d commit 3f5daac

File tree

2 files changed

+166
-0
lines changed

2 files changed

+166
-0
lines changed

‎examples/GeometryOptimisation/bluemira_simvue_geometry_optimisation.py

Lines changed: 159 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,159 @@
1+
# bluemira is an integrated inter-disciplinary design tool for future fusion
2+
# reactors. It incorporates several modules, some of which rely on other
3+
# codes, to carry out a range of typical conceptual fusion reactor design
4+
# activities.
5+
#
6+
# Copyright (C) 2021 M. Coleman, J. Cook, F. Franza, I.A. Maione, S. McIntosh, J. Morris,
7+
# D. Short
8+
#
9+
# bluemira is free software; you can redistribute it and/or
10+
# modify it under the terms of the GNU Lesser General Public
11+
# License as published by the Free Software Foundation; either
12+
# version 2.1 of the License, or (at your option) any later version.
13+
#
14+
# bluemira is distributed in the hope that it will be useful,
15+
# but WITHOUT ANY WARRANTY; without even the implied warranty of
16+
# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
17+
# Lesser General Public License for more details.
18+
#
19+
# You should have received a copy of the GNU Lesser General Public
20+
# License along with bluemira; if not, see <https://www.gnu.org/licenses/>.
21+
"""
22+
A quick tutorial on the optimisation of geometry in bluemira
23+
"""
24+
25+
# We're going to set up some geometry optimisation problems and solve them with different
26+
# optimisastion algorithms.
27+
import numpy as np
28+
29+
from bluemira.geometry.optimisation import GeometryOptimisationProblem, minimise_length
30+
from bluemira.geometry.parameterisations import PrincetonD
31+
from bluemira.utilities.opt_problems import OptimisationConstraint, OptimisationObjective
32+
from bluemira.utilities.optimiser import Optimiser, approx_derivative
33+
from bluemira.utilities.tools import set_random_seed
34+
35+
from simvue import Run,Handler
36+
import logging
37+
38+
# Let's set up a simple GeometryOptimisationProblem, where we minimise the length of
39+
# parameterised geometry.
40+
41+
# First, we set up the GeometryParameterisation, with some bounds on the variables.
42+
x1_lower = 2
43+
x1_value = 2.05
44+
x1_upper = 6
45+
x2_lower = 80
46+
x2_value = 198.5
47+
x2_upper = 260
48+
dz_lower = -0.5
49+
dz_upper = 0.5
50+
max_eval = 500
51+
ftol_abs = 1e-12
52+
ftol_rel = 1e-12
53+
54+
run = Run()
55+
56+
logger = logging.getLogger(__name__)
57+
logger.setLevel(logging.DEBUG)
58+
sth = Handler(run)
59+
logger.addHandler(sth)
60+
61+
run.init(metadata={'dataset.x1_lower': x1_lower,
62+
'dataset.x1_upper': x1_upper,
63+
'dataset.x1_value': x1_value,
64+
'dataset.x2_lower': x2_lower,
65+
'dataset.x2_upper': x2_upper,
66+
'dataset.x2_value': x2_value,
67+
'dataset.dz_lower': dz_lower,
68+
'dataset.dz_upper': dz_upper,
69+
'optimiser.max_eval': max_eval,
70+
'optimiser.ftol_abs': ftol_abs,
71+
'optimiser.ftol_rel': ftol_rel,
72+
},
73+
description="A simple GeometryOptimisationProblem, where we minimise the length of parameterised geometry using gradient-based optimisation algorithm."
74+
)
75+
76+
logger.info("Initialised run")
77+
logger.info("Create parameterisation")
78+
parameterisation_1 = PrincetonD(
79+
{
80+
"x1": {"lower_bound": x1_lower, "value": x1_value, "upper_bound": x1_upper},
81+
"x2": {"lower_bound": x2_lower, "value": x2_value, "upper_bound": x2_upper},
82+
"dz": {"lower_bound": dz_lower, "value": 0, "upper_bound": dz_upper},
83+
}
84+
)
85+
86+
# Here we're minimising the length, and we can work out that the dz variable will not
87+
# affect the optimisation, so let's just fix at some value and remove it from the problem
88+
parameterisation_1.fix_variable("dz", value=0)
89+
90+
# Now, we set up our optimiser. We'll start with a gradient-based optimisation algorithm
91+
logger.info("Define Optimiser")
92+
93+
slsqp_optimiser = Optimiser(
94+
"SLSQP", opt_conditions={"max_eval": max_eval, "ftol_abs": ftol_abs, "ftol_rel": ftol_rel}
95+
)
96+
97+
98+
#Define the call back function
99+
def calculate_length(vector, parameterisation):
100+
"""
101+
Calculate the length of the parameterised shape for a given state vector.
102+
"""
103+
104+
parameterisation.variables.set_values_from_norm(vector)
105+
print('logging metrics', float(parameterisation.variables['x1'].value))
106+
run.log_metrics({'x1_value': float(parameterisation.variables['x1'].value), 'x1_lower': float(parameterisation.variables['x1'].lower_bound), 'x1_upper': float(parameterisation.variables['x1'].upper_bound)})
107+
run.log_metrics({'x2_value': float(parameterisation.variables['x2'].value), 'x2_lower': float(parameterisation.variables['x2'].lower_bound), 'x2_upper': float(parameterisation.variables['x2'].upper_bound)})
108+
109+
return parameterisation.create_shape().length
110+
111+
112+
def my_minimise_length(vector, grad, parameterisation, ad_args=None):
113+
"""
114+
Objective function for nlopt optimisation (minimisation) of length.
115+
116+
Parameters
117+
----------
118+
vector: np.ndarray
119+
State vector of the array of coil currents.
120+
grad: np.ndarray
121+
Local gradient of objective function used by LD NLOPT algorithms.
122+
Updated in-place.
123+
ad_args: Dict
124+
Additional arguments to pass to the `approx_derivative` function.
125+
126+
Returns
127+
-------
128+
fom: Value of objective function (figure of merit).
129+
"""
130+
ad_args = ad_args if ad_args is not None else {}
131+
print(vector)
132+
length = calculate_length(vector, parameterisation)
133+
if grad.size > 0:
134+
grad[:] = approx_derivative(
135+
calculate_length, vector, f0=length, args=(parameterisation,), **ad_args
136+
)
137+
run.update_metadata({'x1_value': float(parameterisation.variables['x1'].value), 'x2_value': float(parameterisation.variables['x2'].value) })
138+
return length
139+
140+
141+
# Next, we make our objective function, using in this case one of the ready-made ones.
142+
# NOTE: This `minimise_length` function includes automatic numerical calculation of the
143+
# objective function gradient, and expects a certain signature.
144+
objective = OptimisationObjective(
145+
my_minimise_length,
146+
f_objective_args={"parameterisation": parameterisation_1},
147+
)
148+
149+
150+
# Finally, we initialise our `GeometryOptimisationProblem` and run it.
151+
logger.info("Call optimiser")
152+
my_problem = GeometryOptimisationProblem(parameterisation_1, slsqp_optimiser, objective)
153+
my_problem.optimise()
154+
155+
156+
# Here we're minimising the length, within the bounds of our PrincetonD parameterisation,
157+
# so we'd expect that x1 goes to its upper bound, and x2 goes to its lower bound.
158+
run.save('bluemira_simvue_geometry_optimisation.py', 'code')
159+
run.close()

‎examples/GeometryOptimisation/readme.md

Lines changed: 7 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,7 @@
1+
# Geometry optimisation using Bluemira
2+
3+
Bluemira is an integrated inter-disciplinary design tool for future fusion reactors. It incorporates several modules, some of which rely on other codes,
4+
to carry out a range of typical conceptual fusion reactor design activities. (https://github.com/Fusion-Power-Plant-Framework/bluemira)
5+
6+
7+
For details of installation of Bluemira please refer to https://bluemira.readthedocs.io/en/develop/installation.html

0 commit comments

Comments
 (0)