Table 5 Summary of parameter settings for NCRBMO and competitive algorithms on the CEC2017(continue).
Algorithm | Parameter Name | Value/Range | Parameter Description | Function and Impact |
|---|---|---|---|---|
BKA | p_BKA | 0.9 | Attack behavior selection threshold | Controls selection between two attack behavior strategies |
r_BKA | [0,1] uniform distribution | Random coefficient | Used multiple times in attack and migration behaviors, increasing algorithm randomness | |
n_BKA | 0.05*exp(−2*(t/T)^2) | Attack behavior step factor | Gaussian decay function controlling attack behavior step size, decreasing with iteration | |
m_BKA | 2*sin(r + π/2) | Migration behavior coefficient | Sine function based on random number r, controlling migration intensity | |
s_BKA | Random integer [1, pop] | Random individual index | Randomly selects an individual for comparison in migration behavior | |
ori_value | [0,1] uniform distribution | Original random value | Base random value for generating Cauchy distribution random numbers | |
cauchy_value | tan((ori_value-0.5)*π) | Cauchy distribution random number | Based on Cauchy distribution with heavy-tailed characteristics, enhancing exploration capability | |
APO | Epsilon | unifrnd(0,1) | Uniform distribution random coefficient | [0,1] uniform random number for underwater foraging stage step control |
R | rand(1,dim) | [0,1] uniform distribution random vector | Used as input angle for tangent function calculation | |
step2 | X_rand1 - X_rand2 | Random direction vector | Difference vector between two random individuals for underwater foraging stage | |
S_APO | tan((R-0.5)*π) | Tangent function value | Maps random angles to tangent values, producing large step changes | |
SCSO | S_SCSO | 2 | Maximum sensitivity range | Controls sand cat perception range, determining exploration capability upper limit |
rg | S - ((S)*t/(Max_iter)) | Guiding parameter | Decreases linearly with iteration, controlling exploration to exploitation transition | |
r_SCSO | rand * rg | Random distance factor | Random scaling based on rg, determining movement step size | |
R_SCSO | ((2rg)*rand) - rg | Phase transition control parameter | Range [-rg, rg], controlling search strategy switching | |
p_SCSO | [1:360] | Roulette wheel selection probability array | Contains integers 1 to 360 for roulette wheel angle selection | |
DBO | P_percent | 0.2 | Producer proportion | Proportion of dung beetle (producer) population |
pNum | round(pop * P_percent) | Number of producers | Number of individuals executing rolling behavior | |
r2_DBO | [0,1] uniform distribution | Producer behavior selection probability | Controls which rolling strategy producer uses | |
r1_DBO | [0,1] uniform distribution | Random coefficient | Increases random disturbance | |
a_DBO | Random value 1 or −1 | Direction coefficient | 90% probability of 1, 10% probability of −1, controlling movement direction | |
θ_DBO | aaa * π/180 | Radian angle | Converts angle value to radians for tangent function | |
R_DBO | 1 - t/M | Contraction factor | Decreases linearly with iteration, controlling boundary contraction | |
WOA | a_WOA | 2-t*((2)/Max_iter) | Linear decreasing parameter | Controls encircling prey behavior step size, linearly decreasing from 2 to 0 |
a2 | −1 + t*((−1)/ Max_iter) | Spiral update parameter | Linearly decreases from − 1 to −2, calculating parameter l for spiral update | |
r1, r2_WOA | [0,1] uniform distribution | Random coefficients | Affect search agent movement | |
b_WOA | 1 | Spiral shape constant | Defines logarithmic spiral shape | |
l | (a2-1)*rand + 1 | Spiral update random parameter | Randomly varies in [−1,1], used in spiral update formula |
