My restrictions does not working

[RafaHPSUnicamp]

Hi, I discovered that all my restrictions does not working. Like the first restriction:

S_total = S_comercio + S_insumo

The best results from my code does not follow that restriction, for example. I already checked with a corresponding sum to S_comercio and S_insumo calculated from Platypus. So, what I need to do to all my restrictions work in my code??

Here is the code:

#variaveis

S_preco = 1069.23
F_preco = 1071.09
OB_preco = 2006.66
OR_preco = 2669.21
B_preco = 2540.47
S_custo = 533.17
F_custo = 569.89
OB_custo = 1384.39
OR_custo = 1466.34
B_custo = 2389.89

S_total = 2329278

S_percentoleo = 0.2057
C_percentoleo = 0.0064

OBF_percentoleo = 0.22

OR_massamolar = 873*0.000001
M_massamolar = 32*0.000001
B_massamolar = 292*0.000001
G_massamolar = 92*0.000001

S_capacidade = 3600000
OR_capacidade = 367200
B_capacidade = 887760*(880/1000)

S_demanda = 80638
F_demanda = 398984
OB_demanda = 164700
OR_demanda = 164700
B_demanda = 77634

from platypus import NSGAII, Problem, Real, unique, nondominated

def belegundu(vars):
    S_comercio = vars[0]
    F_comercio = vars[1]
    OB_comercio = vars[2]
    OR_comercio = vars[3]
    B_total = vars[4]
    S_insumo = vars[5]
    C_insumo = vars[6]
    OB_total = vars[7]
    OR_total = vars[8]
    OR_biodiesel = vars[9]
    MOL = vars[10]
    M_insumo = vars[11]
    G_comercio = vars[12]

    objs = [S_comercio*S_preco - S_comercio*S_custo + F_comercio*F_preco - F_comercio*F_custo + OB_comercio*OB_preco - OB_comercio*OB_custo + OR_comercio*OR_preco - OR_comercio*OR_custo + B_total*B_preco - B_total*B_custo]
    constrs = [
        S_total - S_comercio - S_insumo,
        S_insumo - C_insumo - F_comercio - OB_total,
        C_insumo - 0.04*S_insumo,
        OB_total - (F_comercio*OBF_percentoleo)/(1 - OBF_percentoleo),
        OB_total - OB_comercio - OR_total,
        OR_total - OR_comercio - OR_biodiesel,
        OR_biodiesel - MOL*OR_massamolar,
        M_insumo - 3*MOL*M_massamolar,
        B_total - 3*MOL*B_massamolar,
        G_comercio - MOL*G_massamolar,
        S_insumo - S_capacidade,
        OR_total - OR_capacidade,
        B_total - B_capacidade,
        S_comercio - S_demanda,
        F_comercio - F_demanda,
        OB_comercio - OB_demanda,
        OR_comercio - OR_demanda,
        B_total - B_demanda
    ]
    return objs, constrs

problem = Problem(13, 1, 18)
problem.types[:] = [Real(0, 2329278), Real(0, 2329278), Real(0, 2329278), Real(0, 2329278), Real(0, 2329278), Real(0, 2329278), Real(0, 2329278), Real(0, 2329278), Real(0, 2329278), Real(0, 2329278), Real(0, 2329278), Real(0, 2329278), Real(0, 2329278)]
problem.constraints[0] = "==0"
problem.constraints[1] = "==0"
problem.constraints[2] = "==0"
problem.constraints[3] = "==0"
problem.constraints[4] = "==0"
problem.constraints[5] = "==0"
problem.constraints[6] = "==0"
problem.constraints[7] = "==0"
problem.constraints[8] = "==0"
problem.constraints[9] = "==0"
problem.constraints[10] = "<=0"
problem.constraints[11] = "<=0"
problem.constraints[12] = "<=0"
problem.constraints[13] = ">=0"
problem.constraints[14] = ">=0"
problem.constraints[15] = ">=0"
problem.constraints[16] = ">=0"
problem.constraints[17] = ">=0"
problem.function = belegundu
problem.directions[:] = Problem.MAXIMIZE

algorithm = NSGAII(problem)
algorithm.run(1000)

for solution in algorithm.result:
    print(solution.objectives)

for solution in algorithm.result:
    print(solution.variables)

for solution in unique(nondominated(algorithm.result)):
    print(solution.objectives)

for solution in unique(nondominated(algorithm.result)):
    print(solution.variables)

[dhadka]

Also print out solution.constraints and/or solution.feasible. I would guess that you would see the solutions you're getting are not feasible (violate one or more constraints). The first thing I would look at is increasing the NFE beyond 1000 as it's likely these algorithm's haven't yet converged.

[github-actions[bot]]

This issue is stale and will be closed soon. If you feel this issue is still relevant, please comment to keep it active. Please also consider working on a fix and submitting a PR.