#!/usr/bin/env python # -*- coding: utf-8 -*- """ *********************************************************************************** opt_tutorial1.py DAE Tools: pyDAE module, www.daetools.com Copyright (C) Dragan Nikolic *********************************************************************************** DAE Tools is free software; you can redistribute it and/or modify it under the terms of the GNU General Public License version 3 as published by the Free Software Foundation. DAE Tools is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for more details. You should have received a copy of the GNU General Public License along with the DAE Tools software; if not, see <http://www.gnu.org/licenses/>. ************************************************************************************ """ __doc__=""" This tutorial introduces IPOPT NLP solver, its setup and options. """ import sys from time import localtime, strftime from daetools.pyDAE import*from daetools.solvers.ipopt import pyIPOPTclassmodTutorial(daeModel):def__init__(self, Name, Parent=None, Description=""): daeModel.__init__(self, Name, Parent, Description) self.x1=daeVariable("x1", no_t, self) self.x2=daeVariable("x2", no_t, self) self.x3=daeVariable("x3", no_t, self) self.x4=daeVariable("x4", no_t, self) self.dummy=daeVariable("dummy", no_t, self, "A dummy variable to satisfy the condition that there should be at least one-state variable and one equation in a model")defDeclareEquations(self): daeModel.DeclareEquations(self) eq=self.CreateEquation("Dummy") eq.Residual=self.dummy()classsimTutorial(daeSimulation):def__init__(self): daeSimulation.__init__(self) self.m=modTutorial("opt_tutorial1") self.m.Description=__doc__defSetUpParametersAndDomains(self):passdefSetUpVariables(self): self.m.x1.AssignValue(1) self.m.x2.AssignValue(5) self.m.x3.AssignValue(5) self.m.x4.AssignValue(1)defSetUpOptimization(self): # Set the objective function (minimization). # The objective function can be accessed by using ObjectiveFunction property which always returns the 1st obj. function, # for in general case more than one obj. function. can be defined, so ObjectiveFunctions[0] can be used as well: # self.ObjectiveFunctions[0].Residual = ... # Obviously defining more than one obj. function has no meaning when using opt. software such as Ipopt, Bonmin or Nlopt # which cannot do the multi-objective optimization. The number of obj. functions can be defined in the function # optimization.Initialize by using the named argument NumberOfObjectiveFunctions (the default is 1). # Other obj. functions can be obtained by using ObjectiveFunctions[i] property. self.ObjectiveFunction.Residual=self.m.x1()*self.m.x4()*(self.m.x1()+self.m.x2()+self.m.x3())+self.m.x3() # Set the constraints (inequality, equality) # Constraints are in the following form: # - Inequality: g(i) <= 0 # - Equality: h(i) = 0 self.c1=self.CreateInequalityConstraint("Constraint 1") # g(x) >= 25: 25 - x1*x2*x3*x4 <= 0 self.c1.Residual=25-self.m.x1()*self.m.x2()*self.m.x3()*self.m.x4() self.c2=self.CreateEqualityConstraint("Constraint 2") # h(x) == 40 self.c2.Residual=self.m.x1()*self.m.x1()+self.m.x2()*self.m.x2()+self.m.x3()*self.m.x3()+self.m.x4()*self.m.x4()-40 # Set the optimization variables, their lower/upper bounds and the starting point # The optimization variables can be stored and used later to get the optimization results or # to interact with some 3rd party software. self.x1=self.SetContinuousOptimizationVariable(self.m.x1, 1, 5, 2);self.x2=self.SetContinuousOptimizationVariable(self.m.x2, 1, 5, 2);self.x3=self.SetContinuousOptimizationVariable(self.m.x3, 1, 5, 2);self.x4=self.SetContinuousOptimizationVariable(self.m.x4, 1, 5, 2);defsetOptions(nlpsolver): # 1) Set the options manuallytry: nlpsolver.SetOption('print_level', 0) nlpsolver.SetOption('tol', 1e-7) nlpsolver.SetOption('mu_strategy', 'adaptive') # Print options loaded at pyIPOPT startup and the user set options: #nlpsolver.PrintOptions() #nlpsolver.PrintUserOptions() # ClearOptions can clear all options: #nlpsolver.ClearOptions()exceptExceptionas e: print(str(e))defrun(**kwargs): simulation=simTutorial() # Achtung! Achtung! NLP solver options can only be set after optimization.Initialize() # Otherwise seg. fault occurs for some reasons. nlpsolver=pyIPOPT.daeIPOPT()returndaeActivity.optimize(simulation, reportingInterval=1, timeHorizon=1, nlpsolver=nlpsolver, nlpsolver_setoptions_fn=setOptions, reportSensitivities=True,**kwargs)if__name__=="__main__": guiRun=Falseif(len(sys.argv)>1andsys.argv[1]=='console')elseTrue run(guiRun=guiRun)