Algorithm 1: Pseudocode of ADVCSO algorithm
Initialize population using good point set and apply perturbation to top 10% of elites;
Define parameters: epoch, Gmax, pop_size, N, G, and jixiaoliang;
1:	While t < Gmax:
2:	If (t % Gcurrent == 0):
3:	Sort population by fitness and dynamically assign roles:
4:	-Roosters (NR): top-ranked individuals
5:	-Hens (NH): middle-ranked individuals
6:	-Chicks (NC): remaining individuals
7:	Update mating pairs and mother–chick relationships;
8:	Adaptively adjust Gcurrent via cosine annealing;
9:	End if
10:	For i = 1 to NR:
11:	Update roosters’ positions with adaptive learning factor;
12:	Apply hybrid mutation (Cauchy/Gaussian based on diversity threshold);
13:	End for
14:	For i = 1 to NH:
15:	Update hens’ positions using rank-based following strategy;
16:	Replace partial dimensions with global best values;
17:	End for
18:	For i = 1 to NC:
19:	Update chicks’ positions via spiral learning with FL factor;
20:	Inherit elite dimensions from global best;
21:	End for
22:	Evaluate new solutions and update local/global best;
23:	Adjust mutation threshold based on population diversity;
24:	End while
25:	Return global best solution;