| Input: : Parent population; : Offspring; : Reference vector; : Iteration number. |
| Output: : Selected population. |
| (1) Compare the ideal point of and the objective value of |
| (2) if was updated by the offspring then |
| (3) |
| (4) Cluster using and obtain subspaces in Algorithm 3 |
| (5) Calculate contribution of each individual in in Algorithm 4 |
| (6) Find the solutions which have the worst contribution |
| (7) if then |
| (8) if then |
| (9) Delete the solution as shown in Figure 3(a) |
| (10) else |
| (11) Find the most crowded subspaces and delete which has the maximum PBI value in as shown in |
| Figure 3(b) |
| (12) end if |
| (13) else |
| (14) if then |
| (15) Find and delete the solution with maximum PBI value as shown in Figure 4(a) |
| (16) else |
| (17) Find the most crowded subspaces and delete which has the maximum PBI value in as shown in |
| Figure 4(b) |
| (18) end if |
| (19) end if |
| (20) else |
| (21) |
| (22) Cluster using reference vectors |
| (23) Find the most crowded subspaces and delete which has the maximum PBI value in |
| (24) end if |
| (25) |
| (26) Adaptive reference vector |
| (27) if mod(,) = 0 then |
| (28) |
| (29) |
| (30) end if |
| (31) return |
