FBCA: Flexible Besiege and Conquer Algorithm for Multi-Layer Perceptron Optimization Problems

. 2025 Nov 19;10(11):787. doi: 10.3390/biomimetics10110787

Algorithm 2 The pseudocode of the FBCA

1:
Input: Population Size: N, Problem Dimension: D, Max Iteration: Max_Gen, The Number Of Soldiers: nSoldiers
2:
Output: Obtained best solution
3:
Initialize the solutions’ positions randomly
4:
Initialize the velocities and BCB parameters
5:
while t ← 1 to Max_Gen do
6:
for i ← 1 to nArmies do
7:
for j ← 1 to nSoldiers do
8:
for $d \leftarrow 1$ to D do
9:
if rand < BCB then
10:
if rand < $p_{1}$ then
11:
Creation1: Sine-Guided Soft Asymmetric Gaussian Perturbation
12:
initial Gaussian perturbation item $N$
13:
Update the soldier position by Equation (9)
14:
End Creation1
15:
else
16:
Creation2: Exponentially Modulated Spiral Perturbation
17:
Update $S p i r a l_F a c t o r$ by Equation (10)
18:
Update the soldier position by Equation (11)
19:
End Creation2
20:
end if
21:
Update the soldier position that exceeds the search boundaries By Equation (3)
22:
else
23:
Update $p_{2}$ by Equation (16)
24:
if rand < $p_{2}$ then
25:
Creation3: Nonlinear Cognitive Velocity Update
26:
Update $v_{i, d}$ by Equation (14)
27:
Update the position by Equation (13)
28:
End Creation3
29:
else
30:
Update the position by Equation (2)
31:
end if
32:
Update the soldier position that exceeds the search boundaries By Equation (4)
33:
end if
34:
end for
35:
end for
36:
end for
37:
Update gBest and gBestPos.
38:
end while