E-dyNSGA-III: A Multi-Objective Algorithm for Handling Pareto Optimality over Time
Algorithm 1
Input: reference points on normalized hyperplane Pr with specific location 1…… , parent population npop,t, offspring population npop,o,t, elite population Pe, distance of elite population to the ideal point Pi, max_gen = 500
Output: elite population Pe,t, updated reference point location with respect to ideal point lup ref,t
2009
while
(1)
Apply adaptive mutation strategy specified in (1) and (2)
(2)
Obtain the set of elites Pe and current location of updated reference points
else if
(3)
Increase the number of reference points Pr by and arrange in the normalized hyperplane by associating each member of Pe,t with a reference point
(4)
Execute step 1
(5)
Obtain set of elites Pe and current location of updated reference points according to: [Pe,t + 1,lup ref, t + 1] = Update Elite(lup ref, t, Pe,t,npop,t,npop,o,t)
(6)
Compare current location of solutions with ideal Pareto front using Minkowski distance
