PSO (Particle Swarm Optimization)

Installation

bash go get github.com/umitanilkilic/pso

Usage

Usage

To use the PSO library, you can import it into your Go project and start optimizing your problems. Here's an example of how to use it:

```go package main

import ( "fmt" "math" "math/rand"

"github.com/umitanilkilic/pso"

)

var spaceSize = 100.0 var psOptimizer pso.PSO

func main() { //easom function

f := func(p pso.Position) float64 {
    return -1 * (math.Cos(p.GetCoordinates()[0]) * math.Cos(p.GetCoordinates()[1]) * math.Exp(-1*(math.Pow(p.GetCoordinates()[0]-math.Pi, 2)+math.Pow(p.GetCoordinates()[1]-math.Pi, 2))))
}
//initialize particles
particles := createParticles()
//initialize swarm
swarm := pso.NewSwarm(0.5, 0.5, 0.5, particles, f, constraintFunc)
psOptimizer = pso.NewPSO(&swarm)
psOptimizer.SetIterationCount(10)
//run pso
psOptimizer.Optimize(printParticleInfo)

fmt.Printf("Global best: %v\n", swarm.GetGlobalBest())
fmt.Printf("Fitness: %v\n", f(swarm.GetGlobalBest()))

}

func constraintFunc(position *pso.Position) { for i := 0; i < position.GetDimension(); i++ { if position.GetCoordinates()[i] < -spaceSize { position.GetCoordinates()[i] = -spaceSize } else if position.GetCoordinates()[i] > spaceSize { position.GetCoordinates()[i] = spaceSize } } }

func createParticles() []pso.Particle { particles := make([]pso.Particle, 0) for i := 0; i < 100; i++ { pos := pso.NewPosition(float64(rand.Float64()spaceSize), float64(rand.Float64()spaceSize)) vel := pso.NewPosition(float64(rand.Float64()spaceSize), float64(rand.Float64()spaceSize)) particles = append(particles, pso.NewParticle(i, pos, vel)) } return particles }

func printParticleInfo(particles *pso.Swarm) { iteration := psOptimizer.GetCurrentIteration() fmt.Printf("Iteration: %v, Global Best Position: %v\n", iteration, particles.GetGlobalBest()) }

```

Make sure to replace the objective function logic and the bounds with your own problem-specific implementation. You can also adjust the number of dimensions, particles, and iterations according to your needs.

For more information, please refer to the PSO documentation.

Citation

If you use this code in your research, please cite the following paper:

bibtex @software{KILIC_Particle_Swarm_Optimization, author = {KILIC, UMIT ANIL}, license = {MIT}, title = {{Particle Swarm Optimization Library}}, url = {https://github.com/umitanilkilic/pso} }

License

This project is licensed under the MIT License - see the LICENSE file for details.