public static double GetLoadRate(this Bee bee, double demandPower)
{
var totalPower = bee.GetTotalPower();
if (demandPower > totalPower)
{
return(0);
}
return(demandPower / totalPower);
}
public static double Fitness(this Bee bee, double demandPower)
{
if (bee.GetTotalPower() < demandPower)
{
return(0);
}
var loadRate = bee.GetLoadRate(demandPower);
var averageCost = bee.GetAverageCost(demandPower);
var averageCo2 = bee.GetAverageCO2(demandPower);
var fitness = loadRate + 1 / averageCost + 1 / averageCo2;
return(fitness);
}
public static double GetAverageCO2(this Bee bee, double demandPower)
{
var totalPower = bee.GetTotalPower();
if (demandPower > totalPower)
{
return(double.PositiveInfinity);
}
var totalCo2 = 0.0;
foreach (var gene in bee)
{
totalCo2 += gene.CO2 * gene.Power * gene.Count;
}
return(totalCo2 / totalPower);
}
static void Main(string[] args)
{
var stopWatch = new Stopwatch();
stopWatch.Start();
var maxGeneration = 2000;
var generation = 0;
var bestFitness = 0.0;
Bee bestBee = null;
var populationSizeEmployee = 800;
var populationSizeOnlooker = 200;
var demandPower = 10000;
//Init employee bee population
var populationEmployeeBee = PopulationGenerator.Generate(populationSizeEmployee).ToList();
//Onlooker bee collection
var populationOnlookerBee = new List <Bee>();
while (generation < maxGeneration)
{
//For each employee bee
foreach (var b in populationEmployeeBee)
{
//Grab a random other employee and perform a close search
var k = Rand.Next(0, populationSizeEmployee);
var beeK = populationEmployeeBee[k];
//Check if this bee has exceed search limit
//If so this function will turn this bee into a scout bee,
//which is, to generate a new bee.
if (!b.IsReSearchLimitExceed())
{
//If not, perform a close search
b.ReSearch(beeK, demandPower);
}
}
//Clear onlooker bee's collection
populationOnlookerBee.Clear();
//Get bees from employee collcetion using Roulette selection
foreach (var _ in Enumerable.Range(0, populationSizeOnlooker))
{
populationOnlookerBee.Add(populationEmployeeBee.RouletteSelect().Copy());
}
//For each onlooker bee, perform a close search
foreach (var b in populationOnlookerBee)
{
//Grab a random bee from onlooker bee's collection
var k = Rand.Next(0, populationSizeOnlooker);
var beeK = populationOnlookerBee[k];
//Perform close search
b.ReSearch(beeK, demandPower);
}
//Get the best one, for now
populationOnlookerBee = populationOnlookerBee.OrderBy(x => x.FitnessLast).ToList();
var lastBee = populationOnlookerBee.Last();
//Save the best one
if (lastBee.GetTotalPower() >= demandPower)
{
var fitness = lastBee.FitnessLast;
if (fitness > bestFitness)
{
bestFitness = fitness;
bestBee = lastBee.Copy();
}
}
Console.WriteLine(bestBee);
Console.WriteLine($"Best fitness: {bestBee.Fitness(demandPower)}");
Console.WriteLine($"Best Total Power: {bestBee.GetTotalPower()}");
Console.WriteLine($"Best LoadRate: {bestBee.GetLoadRate(demandPower)}");
Console.WriteLine($"Best Cost: {bestBee.GetAverageCost(demandPower)}");
Console.WriteLine($"Best CO2: {bestBee.GetAverageCO2(demandPower)}");
Console.WriteLine($"Gen: {generation}");
Console.WriteLine($"=========================");
//This the known best solution.
if (Math.Abs(bestFitness - 1.1054208443867) < 0.0001)
{
break;
}
generation++;
}
stopWatch.Stop();
Console.WriteLine($"Execution Time:{stopWatch.Elapsed.TotalSeconds}");
Console.ReadKey();
}
