Comprehensive Learning Fungal Growth Optimizer for Numerical Optimization and Reservoir Production Optimization

View full-text article in PMC

. 2026 May 27;11(6):370. doi: 10.3390/biomimetics11060370

Algorithm 2 Comprehensive Learning Fungal Growth Optimizer (CLFGO)

1:
Input: Population size N, maximum function evaluations $MaxFEs$ , dimension D, bounds $\vec{S_{L}}, \vec{S_{U}}$ , stagnation threshold m, learning bounds $a, b$
2:
Initialize N hyphae $\vec{S_{i}}$ using Equation (1); evaluate fitness; identify $\vec{S^{*}}$
3:
Initialize personal best positions $\vec{p b e s t_{i}} \leftarrow \vec{S_{i}}$ , stagnation counters $s c_{i} \leftarrow 0$ , $\forall i$
4:
Compute learning probabilities $P_{c_{i}}$ using Equation (6), $\forall i$
5:
$F E s \leftarrow N$ , $t \leftarrow 1$
6:
while $F E s < MaxFEs$ do
7:
for $i = 1$ to N do ▹ CLS Module
8:
if $F E s \geq MaxFEs$ then
9:
break
10:
end if
11:
if $s c_{i} \geq m$ then ▹ Individual i is stagnant
12:
for $j = 1$ to D do ▹ Build exemplar vector $e_{i}$
13:
Generate $rand \sim rand (0, 1)$
14:
if $rand < P_{c_{i}}$ then
15:
Randomly select $k_{1}, k_{2} \neq i$
16:
$e_{i} (j) \leftarrow arg {min}_{k \in {k_{1}, k_{2}}} p f i t (k)$ ▹ Tournament selection
17:
else
18:
$e_{i} (j) \leftarrow i$ ▹ Learn from own best
19:
end if
20:
end for
21:
if $e_{i} (j) = i, \forall j$ then ▹ Safeguard check
22:
Randomly select dimension $j^{'}$ and individual $k_{3} \neq i$
23:
$e_{i} (j^{'}) \leftarrow k_{3}$
24:
end if
25:
Construct candidate $\vec{S_{i}^{CL}}$ : for each dimension j, set $S_{i, j}^{CL} \leftarrow p b e s t_{e_{i} (j), j}$
26:
Evaluate $\vec{S_{i}^{CL}}$ and set $F E s \leftarrow F E s + 1$
27:
if $f (\vec{S_{i}^{CL}}) < f (\vec{S_{i}^{t}})$ then $\vec{S_{i}^{t}} \leftarrow \vec{S_{i}^{CL}}$ ; $s c_{i} \leftarrow 0$
28:
end if
29:
end for
30:
▹ Standard FGO Update (Algorithm 1, Lines 6–16)
31:
Execute Hyphal Tip Growth or Branching-Germination loop on all individuals while respecting $MaxFEs$
32:
Perform greedy selection for each individual; update $\vec{p b e s t_{i}}$ and $s c_{i}$ accordingly
33:
Update global best $\vec{S^{*}}$ if a better solution is found
34:
$t \leftarrow t + 1$
35:
end while
36:
Output: Global best solution $\vec{S^{*}}$