Adjustment for Simulated Annealing in CSharp (DRAFT)

28.3K views

Simulated Annealing in C#

The simulated annealing is a metaheuristic, a random search algorithm inspired from physics sciences. In this example, we will doing a simple thing : adjusting one coefficent for having a better results for the algorithm to found the global minimum of the function :

f(X)=0.11(0.5 X² + Cos[Pi X] - 2 Sin[2 Pi X] + Cos[3 Pi X]*Sin[Pi X] in [-5; 5].

// {
using System;
using System.Collections.Generic;
using System.Linq;
using System.Text;
using System.Threading.Tasks;
namespace SimulatedAnnealing
{
    class Program
    {
        static Random rand = new Random();
        static int ct0, ct1, ct2, ct3;
        static double random_x(double xmin, double xmax)
        {
            return xmin + (xmax - xmin) * rand.NextDouble();
        }
        static double random_variate(double x, double xmin, double xmax, double delta)
        {
            var x0 = x - delta < xmin ? xmin : x - delta;
            var x1 = x + delta > xmax ? xmax : x + delta;
            return random_x(x0, x1);
        }
        static double energy(double x)
        {
            return COEF * myFunction(x);
        }
        static double acceptanceProbability(double energy, double new_energy, double temp)
        {
            if (new_energy < energy)
            {
                ct1++;
                return 1.0;
            }
            return Math.Exp((energy - new_energy) / temp);
        }
        static int COEF = 100;
        static double simulated_annealing()
        {
            double XMIN = -5;
            double XMAX = 5;
            double Temperature = 1000;
            double minTemp = 1;