how to apply bounds restriction in PSO

Hello,

i need more clarification about the different ways to ensure that the end results in PSO are still in bounds

in pso, bound controls should be applied on velocity, however doing so still gives me value out of bound in the particle position

because in pso the new position in iteration i, X(i)= X(i-1) + Velocity , that means the position could still receive a value out of bounds when we add the value of previous position.

my question is how to i fix this without changing the fundamentals of PSO.

i appreciate your help,

[Walter Roberson]

Are you using https://www.mathworks.com/help/gads/particleswarm.html ?

[chaker ben salem]

No,am using a simple self made pso to better understand how it work

Answer 1

[Walter Roberson]

If you were using particleswarm() and if each position element was paired with a velocity element, then the way it would be handled would be to create a matrix of linear constraints

for k = 1:N
  A(k, k) = 1;
  A(k, k+N) = 1
  b(k) = position_upper_bound(k);
  A(k+N, k) = -1;
  A(k+N, k+N) = -1;
  b(k+N) = -position_lower_bound(k);
end

and then it would just be a normal linear constraints check.