Particle swarm optimisation

In summary, the conversation is about adding position with velocity in the simple PSO algorithm. The question is raised about how to handle the different units of measurement for position and velocity. It is clarified that units are not normally specified in the equations for the PSO algorithm, but it is important to consider when adding the two components together. The suggestion is made to use a velocity vector in meters per second multiplied by a small time increment in seconds to calculate the new position in meters. The use of delta-time in the original equation is further discussed and its representation in coding the PSO algorithm is questioned.
  • #1

Mke

5
0
Hello everyone,I have a short question about the PSO since I am a new comer to this field. how can we "add" position with velocity in the simple PSO algorithm, when they are of different units?
 
Mathematics news on Phys.org
  • #2
Welcome to PF!

Can you elaborate more on your question?

What do you mean by units?

What do you mean by "add" position to velocity?

Is the posititon in inches and feet (ie english units) but the velocity is in meters/sec (metric units) ?
 
  • #3
Hello thank you for your reply.
When the velocity is determined in the PSo algorithm, one needs to add the weighted velocity to the difference between the ( current positions and the best experience of that position ) . The former is velocity which is normally measured in different units than the position. In PSO it's not common to include units for the position and velocity that is what got me confused.
 
  • #5
the units of velocity and position in the PSo are normally specified but they are still added together.
 
  • #6
What are the units for position?

What are the units for velocity?
 
  • #7
they are not normally specified in the equations that produce the PSO. However, this is not important since regardless of the units, we are adding two different components that have different units. its like adding Kg and Km
 
  • #8
The best I can say here is that you compute the new position by using a velocity vector in say m/s times a small time increment in seconds to a change in position in meters:

NewPositionmeters = OldPositionmeters + Velocitymeters/sec * delta-timeseconds
 
  • #9
Thank you but where does this delta-time come in the original equation that produces the new NewPosition which is merely : NewPositionmeters = OldPositionmeters + Velocitymeters/sec . Could you please further explain this ?

Also, how the delta-time is normally represented when the PSO algorithm is coded?
 
  • #10
The delta time is implied in your equation to be 1 second.
 

1. What is Particle Swarm Optimisation (PSO)?

Particle Swarm Optimisation (PSO) is a computational method inspired by the social behavior of bird flocking and fish schooling. It is a population-based stochastic optimization algorithm that aims to find the optimal solution by simulating the collective movement and cooperation of particles in a multidimensional search space.

2. How does PSO work?

In PSO, a group of particles, each representing a potential solution, move around the search space and adjust their positions based on their own experience and the experience of their neighbors. The movement of particles is guided by two main components: personal best (pbest) and global best (gbest). The pbest is the best solution found by each particle, while the gbest is the best solution found by any particle in the entire swarm. These two components help the particles to converge towards the optimal solution.

3. What are the advantages of PSO?

PSO has several advantages, such as its simplicity, efficiency, and ability to handle large-scale optimization problems. It also does not require any gradient information, making it suitable for non-differentiable and noisy objective functions. PSO is also relatively easy to implement and can find near-optimal solutions even in complex search spaces.

4. What are the applications of PSO?

PSO has been widely used in various fields, including engineering, economics, finance, and machine learning. It has been successfully applied to solve problems such as parameter estimation, pattern recognition, and data clustering. PSO has also been used to train neural networks and optimize the parameters of deep learning models.

5. What are the limitations of PSO?

Despite its advantages, PSO also has some limitations. One of the main limitations is that it can get stuck in local optima and may fail to find the global optimal solution. In addition, PSO may require a large number of iterations to converge, and its performance can be affected by the choice of parameters. It may also struggle with high-dimensional and multimodal optimization problems.

Suggested for: Particle swarm optimisation

Replies
2
Views
801
Replies
2
Views
562
Replies
4
Views
1K
Replies
3
Views
1K
Replies
1
Views
2K
Replies
1
Views
2K
Replies
2
Views
1K
Replies
1
Views
813
Replies
1
Views
834
Back
Top