Algorithm 3 Evolution Algorithm

Input:  Population size $n_p$, parent solutions $\mathcal{P}\left(q\right)$, crossover probability $p_{crossover}$, and mutation probability $p_{mutation}$. Output:  Offspring solutions: $\mathcal{P}{\left(q\right)}_{EA}$.    1:Randomly divide $\mathcal{P}\left(q\right)$ into two populations: $\mathcal{P}{\left(q\right)}_C$ and $\mathcal{P}{\left(q\right)}_M$;    2:$\widehat{\mathcal{V}},\Phi \left(\widehat{\mathcal{V}}\right)\leftarrow \mathbf{MSS}(\mathcal{P}{\left(q\right)}_C,\mathcal{K},H)$;    3:Sort the solutions $\mathcal{P}{\left(q\right)}_c$ in an ascending order based on their $\Phi \left(\widehat{\mathcal{V}}\right)$;    4:Compose $\{{\mathcal{P}}_i^b\left(q\right),{\mathcal{P}}_j^w\left(q\right)\}$ of the first 50% of the populations $\left\{{\mathcal{P}}_i^b\left(q\right)\right\}$ and the last 50% of the populations $\left\{{\mathcal{P}}_j^w\left(q\right)\right\}$;    5:for each pair of solutions $({\mathcal{P}}_i^b\left(q\right),{\mathcal{P}}_j^w\left(q\right))$ do    6:   if A randomly generated value $p_1\le p_{crossover}$ then    7:     Each pair of genes $(x_i,x_j)$ in ${\mathcal{P}}_i^b\left(q\right)$ and ${\mathcal{P}}_j^w\left(q\right)$ is randomly assigned a value of 0 or 1, thus constituting the sequence M;    8:     for each $(x_i,x_j)$ in $\{{\mathcal{P}}_i^b\left(q\right),{\mathcal{P}}_j^w\left(q\right)\}$ do    9:        if $(x_i,x_j)$ corresponds to a value of 1 in the sequence M then  10:          Replace the values of $x_i$ and $x_j$ to obtain a new ${\mathcal{P}}_i{\left(q\right)}^C$ and ${\mathcal{P}}_j{\left(q\right)}^C$;  11:        end if  12:     end for  13:   end if  14:end for  15:Set $\left\{\mathcal{P}{\left(q\right)}^C\right\}=\left\{{\mathcal{P}}_i{\left(q\right)}^C\right\}\cup \left\{{\mathcal{P}}_j{\left(q\right)}^C\right\}$;  16:for each solution ${\mathcal{P}}_i{\left(q\right)}_M$ in $\mathcal{P}{\left(q\right)}_M$ do  17:   for each weight parameter y in ${\mathcal{P}}_i{\left(q\right)}_M$ do  18:     if A randomly generated value $p_2\le p_{mutation}$ then  19:        Update y value to a value N(−1,1),get ${\mathcal{P}}_i{\left(q\right)}^M$;  20:     end if  21:   end for  22:end for  23:$\widehat{\mathcal{V}},\Phi \left(\widehat{\mathcal{V}}\right)\leftarrow \mathbf{MSS}(\mathcal{P}{\left(q\right)}^C\cup \mathcal{P}{\left(q\right)}^M,\mathcal{K},H)$;  24:The $n_p$ populations from $\mathcal{P}{\left(q\right)}^C\cup \mathcal{P}{\left(q\right)}^M$ with the highest $\Phi \left(\widehat{\mathcal{V}}\right)$ are selected as $\mathcal{P}{\left(q\right)}_{EA}$;