In this work, we use the constraint dominance definition
developed by Deb et al. [5] to modify EMOCA for handling the
constraint on generation capacity. A solution vector x is said to
constraint dominate a solution vector y if any of the following
conditions is true.
x is feasible by y is not
x dominates y
Both x and y are infeasible but x has an
overall lower constraint violation
EMOCA employs a binary tournament selection to generate the
mating population where the fitness of each individual is
computed as the sum of its non-domination rank and diversity
rank. NSGA-II primarily uses non-domination for selection. In
EMOCA, each offspring is compared with one of the parents to
form the new pool considering both domination and crowding
density. There are three possible cases:
Case 1: If the offspring dominates the parent, then the offspring
is added to the new pool.
Case 2: If dominated by the parent and if it has a higher
crowding distance than the parent, then the offspring is added
to the new pool with probability 1-exp(ψ(parent)-ψ(offspring))
where ψ denotes the crowding distance of a solution as defined
in NSGA-II.
Case 3: Otherwise, if the offspring has a higher crowding
distance than the parent, then it is added to the new pool, else
the parent is added to the new pool.
We use two point crossover and the mutation operator results in
randomly swapping the energy schedules of two users. We have
chosen a probability of 0.8 for crossover and 0.2 for mutation.
In our simulations, small variations in these probabilities did
not have a significant impact on the performance of the
algorithm.