evomo.problems.constrained.lircmop

Classes

LIRCMOP	Constrained benchmark MOP with large infeasible regions.
LIRCMOP1	Constrained benchmark MOP with large infeasible regions.
LIRCMOP2	Constrained benchmark MOP with large infeasible regions.
LIRCMOP3	Constrained benchmark MOP with large infeasible regions.
LIRCMOP4	Constrained benchmark MOP with large infeasible regions.
LIRCMOP5	Constrained benchmark MOP with large infeasible regions.
LIRCMOP6	Constrained benchmark MOP with large infeasible regions.
LIRCMOP7	Constrained benchmark MOP with large infeasible regions.
LIRCMOP8	Constrained benchmark MOP with large infeasible regions.
LIRCMOP9	Constrained benchmark MOP with large infeasible regions.
LIRCMOP10	Constrained benchmark MOP with large infeasible regions.
LIRCMOP11	Constrained benchmark MOP with large infeasible regions.
LIRCMOP12	Constrained benchmark MOP with large infeasible regions.
LIRCMOP13	Constrained benchmark MOP with large infeasible regions.
LIRCMOP14	Constrained benchmark MOP with large infeasible regions.

Module Contents

class evomo.problems.constrained.lircmop.LIRCMOP(d: int = None, m: int = None, n_iq: int = None, ref_num: int = 1000, **kwargs)

Bases: evomo.problems.constrained.base.CMOP

Constrained benchmark MOP with large infeasible regions.

Z. Fan, W. Li, X. Cai, H. Huang, Y. Fang, Y. You, J. Mo, C. Wei, and E. Goodman, An improved epsilon constraint-handling method in MOEA/D for CMOPs with large infeasible regions, Soft Computing, 2019, 23: 12491-12510.

Parameters:

• d – Number of decision variables.

• m – Number of objectives.

• n_iq – Number of inequality constraints.

• n_eq – Number of equality constraints.

• ref_num – Number of reference points used in the problem.

device

ref_num = 1000

class evomo.problems.constrained.lircmop.LIRCMOP1(d: int = 30, m: int = 2, n_iq: int = 2, ref_num: int = 1000, **kwargs)

Bases: LIRCMOP

Constrained benchmark MOP with large infeasible regions.

Z. Fan, W. Li, X. Cai, H. Huang, Y. Fang, Y. You, J. Mo, C. Wei, and E. Goodman, An improved epsilon constraint-handling method in MOEA/D for CMOPs with large infeasible regions, Soft Computing, 2019, 23: 12491-12510.

Parameters:

• d – Number of decision variables.

• m – Number of objectives.

• n_iq – Number of inequality constraints.

• n_eq – Number of equality constraints.

• ref_num – Number of reference points used in the problem.

g1(X: torch.Tensor) → torch.Tensor

g2(X: torch.Tensor) → torch.Tensor

f1(X: torch.Tensor, g: torch.Tensor) → torch.Tensor

f2(X: torch.Tensor, g: torch.Tensor) → torch.Tensor

c(g: torch.Tensor) → torch.Tensor

fn(X: torch.Tensor) → torch.Tensor

pf()

class evomo.problems.constrained.lircmop.LIRCMOP2(d: int = 30, m: int = 2, n_iq: int = 2, ref_num: int = 1000, **kwargs)

Bases: LIRCMOP1

Constrained benchmark MOP with large infeasible regions.

Z. Fan, W. Li, X. Cai, H. Huang, Y. Fang, Y. You, J. Mo, C. Wei, and E. Goodman, An improved epsilon constraint-handling method in MOEA/D for CMOPs with large infeasible regions, Soft Computing, 2019, 23: 12491-12510.

Parameters:

• d – Number of decision variables.

• m – Number of objectives.

• n_iq – Number of inequality constraints.

• n_eq – Number of equality constraints.

• ref_num – Number of reference points used in the problem.

f2(X: torch.Tensor, g: torch.Tensor) → torch.Tensor

pf()

class evomo.problems.constrained.lircmop.LIRCMOP3(d: int = 30, m: int = 2, n_iq: int = 3, ref_num: int = 1000, **kwargs)

Bases: LIRCMOP1

Constrained benchmark MOP with large infeasible regions.

Z. Fan, W. Li, X. Cai, H. Huang, Y. Fang, Y. You, J. Mo, C. Wei, and E. Goodman, An improved epsilon constraint-handling method in MOEA/D for CMOPs with large infeasible regions, Soft Computing, 2019, 23: 12491-12510.

Parameters:

• d – Number of decision variables.

• m – Number of objectives.

• n_iq – Number of inequality constraints.

• n_eq – Number of equality constraints.

• ref_num – Number of reference points used in the problem.

c3(X: torch.Tensor) → torch.Tensor

fn(X: torch.Tensor) → torch.Tensor

pf()

class evomo.problems.constrained.lircmop.LIRCMOP4(d: int = 30, m: int = 2, n_iq: int = 3, ref_num: int = 1000, **kwargs)

Bases: LIRCMOP2

Constrained benchmark MOP with large infeasible regions.

Z. Fan, W. Li, X. Cai, H. Huang, Y. Fang, Y. You, J. Mo, C. Wei, and E. Goodman, An improved epsilon constraint-handling method in MOEA/D for CMOPs with large infeasible regions, Soft Computing, 2019, 23: 12491-12510.

Parameters:

• d – Number of decision variables.

• m – Number of objectives.

• n_iq – Number of inequality constraints.

• n_eq – Number of equality constraints.

• ref_num – Number of reference points used in the problem.

c3(X: torch.Tensor) → torch.Tensor

fn(X: torch.Tensor) → torch.Tensor

pf()

class evomo.problems.constrained.lircmop.LIRCMOP5(d: int = 30, m: int = 2, n_iq: int = 2, ref_num: int = 1000, cons_params: torch.Tensor | None = None, **kwargs)

Bases: LIRCMOP

Constrained benchmark MOP with large infeasible regions.

Z. Fan, W. Li, X. Cai, H. Huang, Y. Fang, Y. You, J. Mo, C. Wei, and E. Goodman, An improved epsilon constraint-handling method in MOEA/D for CMOPs with large infeasible regions, Soft Computing, 2019, 23: 12491-12510.

Parameters:

• d – Number of decision variables.

• m – Number of objectives.

• n_iq – Number of inequality constraints.

• n_eq – Number of equality constraints.

• ref_num – Number of reference points used in the problem.

cons_params = None

g1(X: torch.Tensor) → torch.Tensor

g2(X: torch.Tensor) → torch.Tensor

f1(X: torch.Tensor, g: torch.Tensor) → torch.Tensor

f2(X: torch.Tensor, g: torch.Tensor) → torch.Tensor

c(f1: torch.Tensor, f2: torch.Tensor, params: torch.Tensor) → torch.Tensor

fn(X: torch.Tensor) → torch.Tensor

pf()

class evomo.problems.constrained.lircmop.LIRCMOP6(d: int = 30, m: int = 2, n_iq: int = 2, ref_num: int = 1000, cons_params: torch.Tensor | None = None, **kwargs)

Bases: LIRCMOP5

Constrained benchmark MOP with large infeasible regions.

Z. Fan, W. Li, X. Cai, H. Huang, Y. Fang, Y. You, J. Mo, C. Wei, and E. Goodman, An improved epsilon constraint-handling method in MOEA/D for CMOPs with large infeasible regions, Soft Computing, 2019, 23: 12491-12510.

Parameters:

• d – Number of decision variables.

• m – Number of objectives.

• n_iq – Number of inequality constraints.

• n_eq – Number of equality constraints.

• ref_num – Number of reference points used in the problem.

cons_params = None

f2(X, g)

pf()

class evomo.problems.constrained.lircmop.LIRCMOP7(d: int = 30, m: int = 2, n_iq: int = 3, ref_num: int = 1000, **kwargs)

Bases: LIRCMOP5

Constrained benchmark MOP with large infeasible regions.

Z. Fan, W. Li, X. Cai, H. Huang, Y. Fang, Y. You, J. Mo, C. Wei, and E. Goodman, An improved epsilon constraint-handling method in MOEA/D for CMOPs with large infeasible regions, Soft Computing, 2019, 23: 12491-12510.

Parameters:

• d – Number of decision variables.

• m – Number of objectives.

• n_iq – Number of inequality constraints.

• n_eq – Number of equality constraints.

• ref_num – Number of reference points used in the problem.

pf()

class evomo.problems.constrained.lircmop.LIRCMOP8(d: int = 30, m: int = 2, n_iq: int = 3, ref_num: int = 1000, **kwargs)

Bases: LIRCMOP6

Constrained benchmark MOP with large infeasible regions.

Z. Fan, W. Li, X. Cai, H. Huang, Y. Fang, Y. You, J. Mo, C. Wei, and E. Goodman, An improved epsilon constraint-handling method in MOEA/D for CMOPs with large infeasible regions, Soft Computing, 2019, 23: 12491-12510.

Parameters:

• d – Number of decision variables.

• m – Number of objectives.

• n_iq – Number of inequality constraints.

• n_eq – Number of equality constraints.

• ref_num – Number of reference points used in the problem.

pf()

class evomo.problems.constrained.lircmop.LIRCMOP9(d: int = 30, m: int = 2, n_iq: int = 2, ref_num: int = 1000, c1_params: torch.Tensor | None = None, c2_params: torch.Tensor | None = None, **kwargs)

Bases: LIRCMOP5

Constrained benchmark MOP with large infeasible regions.

Z. Fan, W. Li, X. Cai, H. Huang, Y. Fang, Y. You, J. Mo, C. Wei, and E. Goodman, An improved epsilon constraint-handling method in MOEA/D for CMOPs with large infeasible regions, Soft Computing, 2019, 23: 12491-12510.

Parameters:

• d – Number of decision variables.

• m – Number of objectives.

• n_iq – Number of inequality constraints.

• n_eq – Number of equality constraints.

• ref_num – Number of reference points used in the problem.

c1_params = None

c2_params = None

f1(X: torch.Tensor, g: torch.Tensor) → torch.Tensor

f2(X: torch.Tensor, g: torch.Tensor) → torch.Tensor

c1(f1: torch.Tensor, f2: torch.Tensor, params: torch.Tensor) → torch.Tensor

c2(f1: torch.Tensor, f2: torch.Tensor, params: torch.Tensor) → torch.Tensor

fn(X: torch.Tensor) → torch.Tensor

pf()

class evomo.problems.constrained.lircmop.LIRCMOP10(d: int = 30, m: int = 2, n_iq: int = 2, ref_num: int = 1000, c1_params: torch.Tensor | None = None, c2_params: torch.Tensor | None = None, **kwargs)

Bases: LIRCMOP9

Constrained benchmark MOP with large infeasible regions.

Z. Fan, W. Li, X. Cai, H. Huang, Y. Fang, Y. You, J. Mo, C. Wei, and E. Goodman, An improved epsilon constraint-handling method in MOEA/D for CMOPs with large infeasible regions, Soft Computing, 2019, 23: 12491-12510.

Parameters:

• d – Number of decision variables.

• m – Number of objectives.

• n_iq – Number of inequality constraints.

• n_eq – Number of equality constraints.

• ref_num – Number of reference points used in the problem.

c1_params = None

c2_params = None

f2(X: torch.Tensor, g: torch.Tensor) → torch.Tensor

pf()

class evomo.problems.constrained.lircmop.LIRCMOP11(d: int = 30, m: int = 2, n_iq: int = 2, ref_num: int = 1000, c1_params: torch.Tensor | None = None, c2_params: torch.Tensor | None = None, **kwargs)

Bases: LIRCMOP10

Constrained benchmark MOP with large infeasible regions.

Z. Fan, W. Li, X. Cai, H. Huang, Y. Fang, Y. You, J. Mo, C. Wei, and E. Goodman, An improved epsilon constraint-handling method in MOEA/D for CMOPs with large infeasible regions, Soft Computing, 2019, 23: 12491-12510.

Parameters:

• d – Number of decision variables.

• m – Number of objectives.

• n_iq – Number of inequality constraints.

• n_eq – Number of equality constraints.

• ref_num – Number of reference points used in the problem.

pf()

class evomo.problems.constrained.lircmop.LIRCMOP12(d: int = 30, m: int = 2, n_iq: int = 2, ref_num: int = 1000, c1_params: torch.Tensor | None = None, c2_params: torch.Tensor | None = None, **kwargs)

Bases: LIRCMOP9

Constrained benchmark MOP with large infeasible regions.

Z. Fan, W. Li, X. Cai, H. Huang, Y. Fang, Y. You, J. Mo, C. Wei, and E. Goodman, An improved epsilon constraint-handling method in MOEA/D for CMOPs with large infeasible regions, Soft Computing, 2019, 23: 12491-12510.

Parameters:

• d – Number of decision variables.

• m – Number of objectives.

• n_iq – Number of inequality constraints.

• n_eq – Number of equality constraints.

• ref_num – Number of reference points used in the problem.

pf()

class evomo.problems.constrained.lircmop.LIRCMOP13(d: int = 30, m: int = 3, n_iq: int = 2, ref_num: int = 1000, cons_params: torch.Tensor | None = None, **kwargs)

Bases: LIRCMOP

Constrained benchmark MOP with large infeasible regions.

Z. Fan, W. Li, X. Cai, H. Huang, Y. Fang, Y. You, J. Mo, C. Wei, and E. Goodman, An improved epsilon constraint-handling method in MOEA/D for CMOPs with large infeasible regions, Soft Computing, 2019, 23: 12491-12510.

Parameters:

• d – Number of decision variables.

• m – Number of objectives.

• n_iq – Number of inequality constraints.

• n_eq – Number of equality constraints.

• ref_num – Number of reference points used in the problem.

cons_params = None

g(X: torch.Tensor) → torch.Tensor

f123(X: torch.Tensor, g: torch.Tensor) → torch.Tensor

c(f1: torch.Tensor, f2: torch.Tensor, f3: torch.Tensor, params: torch.Tensor) → torch.Tensor

fn(X: torch.Tensor) → torch.Tensor

pf()

class evomo.problems.constrained.lircmop.LIRCMOP14(d: int = 30, m: int = 2, n_iq: int = 2, ref_num: int = 1000, **kwargs)

Bases: LIRCMOP13

Constrained benchmark MOP with large infeasible regions.

Z. Fan, W. Li, X. Cai, H. Huang, Y. Fang, Y. You, J. Mo, C. Wei, and E. Goodman, An improved epsilon constraint-handling method in MOEA/D for CMOPs with large infeasible regions, Soft Computing, 2019, 23: 12491-12510.

Parameters:

• d – Number of decision variables.

• m – Number of objectives.

• n_iq – Number of inequality constraints.

• n_eq – Number of equality constraints.

• ref_num – Number of reference points used in the problem.

pf()