Algorithm 3.
Evolution with Knowledge Transfer
| Input: POP1, POP2, …, POPM, A1, A2, …, AM, δ (transfer interval) |
| Output: New Population and New Archives |
| /* Preparation */ |
| t ← 1 |
| Generate initial M population randomly. |
| Initialize M archive with corresponding population. |
| /* Evolution */ |
| While ending condition not satisfied Do |
| /* Searching */ |
| For each subpopulation POPi Do |
| If t % δ = = 0 Then |
| /* Transfer */ |
| k ← rand(1, M) |
| POPi ← Transfer(POPi, Ak) |
| Else |
| Self Evolution |
| End For |
| /* Updating */ |
| For each archive Ai Do |
| Ai ← Update(POPi, Ai) as Algorithm 2 |
| End For |
| t ← t + 1 |
| End While |