Algorithm 3. RPSO

Input:

N: population size

T: Matrix of population information

Maxfes: Maximum number of iterations

Dim: Dimension of the problem

Lb: Lower bound of the problem

Ub: Upper bound of the problem

f: objective function

Output:

xgbest: Matrix of newly generated stock information

Fmin: optimal solution

1.      The position matrix x is obtained by random initialization based on the set parameters.

2.      Calculate the objective function value based on f to obtain the optimal position xgbest, the optimal solution fmin

3.       Fes = N;

4.       T = 1;

5.       While (t <= T)&&(Fes <= Maxfes)

6.           Calculate the new population position x according to Equations (1) and (7)

7.       Calculate the objective function value according to f to get the current optimal position xl its subscripts q, update the optimal position xgbest, the optimal solution fmin

8.       Fes = Fes + N;

9.           Nk = 0

10.         For i = 1:N

11.       If i ≠ q

12.       The cosine similarity between the individual and xl was calculated to obtain nc according to Equation (2)

13.       End

14.       If nc < 0

15.       Nk = Nk + 1;

16.       end

17.       end

18.         If Nk > N/2

19.       Perform according to Algorithm 2.

20.       Calculate the objective function value based on f to obtain the optimal position xgbest, the optimal solution fmin

21.       Fes = Fes + TS;

22.       else

23.       Perform according to Algorithm 1.

24.       Calculate the objective function value based on f to obtain the optimal position xgbest, the optimal solution fmin

25.       Fes = Fes + N;

26.         End

27.       t = t + 1;

28.       End

Return xgbest, fmin