Multi-objective evolutionary algorithm based on decomposition and improved epsilon constraint-handling mechanism. MOEA/D-IEpsilon is a constrained multi-objective evolutionary algorithm (CMOEA).
Python 100.00%
moea_d-iepsilon's Introduction
MOEA/D-IEpsilon: Multi-objective evolutionary algorithm based on decomposition and improved epsilon constraint-handling mechanism
Reference: Fan Z, Li W, Cai X, et al. An improved epsilon constraint-handling method in MOEA/D for CMOPs with large infeasible regions[J]. Soft Computing, 2019, 23: 12491-12510.
MOEA/D-IEpsilon is a constrained multi-objective evolutionary algorithm (CMOEA). The epsilon level is adjust according to the ratio of feasible to total solutions in the current population.
Variables
Meaning
npop
Population size
iter
Iteration number
lb
Lower bound
ub
Upper bound
T
Neighborhood size (default = 30)
delta
The probability of selecting individuals in the neighborhood (default = 0.9)
nr
The maximal number of solutions replaced by a child (default = 2)
tau
Control the scale factor multiplied by the maximum overall constraint violation (default = 0.1)
alpha
Control the searching preference between the feasible and infeasible regions (default = 0.95)
Tc
Control generation (default = 0.8 * iter)
CR
Crossover rate (default = 1)
F
Mutation scalar number (default = 0.5)
pm
Mutation probability (default = 1)
eta_m
Spread factor distribution index (default = 20)
nvar
The dimension of decision space
nobj
The dimension of objective space
V
Weight vectors
B
The T closet weight vectors
pop
Population
objs
Objectives
phi
Constraint violations
phi_max
The maximum overall constraint violation found so far