[Pyomo]How to generate a constraints via a loop - pyomo

Question

all
i want to create a set of constraints, as following:
Pg[0] <= Pmax[0]
Pg[1] <= Pmax[1]
Pg[3] <= Pmax[3]
Pg[4] <= Pmax[4]
Pg[5] <= Pmax[5]
and, my code is as following:
import pyomo.environ as pyo
model = pyo.ConcreteModel()
Ngen = 5
Pmax = [40, 170, 520, 200, 100]
model.Ngen = pyo.Set(dimen=Ngen)
model.Pg = pyo.Var(model.Ngen)
def Co1(model):
return ((model.Pg[ii] for ii in model.Ngen) <= (Pmax[jj] for jj in range(Ngen)))
model.Co1 = pyo.Constraint(rule=Co1)
but, the Python Console tells me:
" TypeError: '<=' not supported between instances of 'generator' and 'generator' "
HOW DO I RECORRECT this?
and the other quesiton, if the Pmax is not a list, but a numpy-ndarray. somethings will different?
thanks a lot!

Answer 1

The clue is in the fact that your constraint is defined over a set. Pyomo allows you to pass args to the Constraint constructor that define the set(s) of that constraint.
Therefore your final line should look something like:
model.Co1 = pyo.Constraint(model.Ngen, rule=Co1)
The second issue is the constraint expression itself. Now that this is defined with respect to a Set, you need only express the constraint in terms of a given index (say i), because the rule also expects arguments pertaining to each index of each Set of the Constraint:
def Co1(model, i):
return model.Pg[i] <= Pmax[i]
Finally, the Set you're defining is of size=5, not dimen=5. Sets are usually dimen=1 unless you're working with say arc links of a network in which case they would be of dimension 2. Since your set is defined over integers, the easiest way to add this is via RangeSet which defines all the integers from 1 to N:
model.Ngen = pyo.RangeSet(Ngen)
Given this, the only thing to change is that the list Pmax is 0-indexed and so our constraint will need to account for i being offset by one:
import pyomo.environ as pyo
model = pyo.ConcreteModel()
Ngen = 5
Pmax = [40, 170, 520, 200, 100]
model.Ngen = pyo.RangeSet(Ngen)
model.Pg = pyo.Var(model.Ngen)
def Co1(model, i):
return model.Pg[i] <= Pmax[i - 1]
model.Co1 = pyo.Constraint(model.Ngen, rule = Co1)
model.pprint()
# 1 RangeSet Declarations
# Ngen : Dim=0, Dimen=1, Size=5, Domain=Integers, Ordered=True, Bounds=(1, 5)
# Virtual
#
# 1 Var Declarations
# Pg : Size=5, Index=Ngen
# Key : Lower : Value : Upper : Fixed : Stale : Domain
# 1 : None : None : None : False : True : Reals
# 2 : None : None : None : False : True : Reals
# 3 : None : None : None : False : True : Reals
# 4 : None : None : None : False : True : Reals
# 5 : None : None : None : False : True : Reals
#
# 1 Constraint Declarations
# Co1 : Size=5, Index=Ngen, Active=True
# Key : Lower : Body : Upper : Active
# 1 : -Inf : Pg[1] : 40.0 : True
# 2 : -Inf : Pg[2] : 170.0 : True
# 3 : -Inf : Pg[3] : 520.0 : True
# 4 : -Inf : Pg[4] : 200.0 : True
# 5 : -Inf : Pg[5] : 100.0 : True
#
# 3 Declarations: Ngen Pg Co1
NumPy arrays can be sliced in exactly the same way as lists so the above constraint will still work fine if you change Pmax = np.array([40, 170, 520, 200, 100])