public void TestCalcFitness_NoCharsInCorrectPositions()
{
CodeBreaker cb = new CodeBreaker("ABCD", new Random());
Chromosome c = new Chromosome("DCBA");
Utility.CalcFitness(c, "ABCD");
Assert.AreEqual(0, c.Fitness);
}
public void TestCalcFitness_NoCommonChars()
{
CodeBreaker cb = new CodeBreaker("ABCD", new Random());
Chromosome c = new Chromosome("WXYZ");
Utility.CalcFitness(c, "ABCD");
Assert.AreEqual(0, c.Fitness);
}
public void TestCalcFitness_AllCorrect()
{
CodeBreaker cb = new CodeBreaker("ABCD", new Random());
Chromosome c = new Chromosome("ABCD");
Utility.CalcFitness(c, "ABCD");
Assert.AreEqual(4, c.Fitness);
}
public void TestCalcFitness_OneCharInCorrectPosition()
{
CodeBreaker cb = new CodeBreaker("ABCD", new Random());
Chromosome c = new Chromosome("AXYZ");
Utility.CalcFitness(c, "ABCD");
Assert.AreEqual(1, c.Fitness);
}
public void TestOnePointCrossover()
{
Random r = new Random(0);
Chromosome c1 = new Chromosome("AAAA");
Chromosome c2 = new Chromosome("BBBB");
Utility.OnePointCrossover(c1, c2, r);
Assert.AreEqual("AABB", new string(c1.Genes));
Assert.AreEqual("BBAA", new string(c2.Genes));
}
private Tuple<Chromosome, Chromosome> Crossover(Chromosome c1, Chromosome c2)
{
switch (Properties.crossoverMethod)
{
case Properties.Crossover.OnePoint:
return Utility.OnePointCrossover(c1, c2, r);
break;
case Properties.Crossover.TwoPoint:
throw new NotImplementedException();
case Properties.Crossover.Uniform:
return Utility.UniformCrossover(c1, c2, r);
break;
default:
throw new ArgumentException();
}
}
public void ChromosomeConstructorTest()
{
var elementProviderMock = new Mock<IElementProvider>();
elementProviderMock.Setup(provider => provider.Elements).Returns(new[] {
new Element() { Name = "Element 1", Value = 1, Weight = 1, Active = false },
new Element() { Name = "Element 2", Value = 2, Weight = 1, Active = false },
});
elementProviderMock.Setup(provider => provider.Size).Returns(2);
var randomNumberGeneratorMock = new Mock<IRandomNumberGenerator>();
randomNumberGeneratorMock.Setup(provider => provider.NextDouble()).Returns(1);
Chromosome chromosome = new Chromosome(elementProviderMock.Object, randomNumberGeneratorMock.Object);
Assert.AreEqual(chromosome.Elements.Length, 2);
Assert.IsTrue(chromosome.Elements[0] == elementProviderMock.Object.Elements[0]);
Assert.IsTrue(chromosome.Elements[1] == elementProviderMock.Object.Elements[1]);
}
public static Solution <T>[] PMXCrossover(Solution <T> a, Solution <T> b, FitnessFunction <T> fitnessFunction)
{
Random r = new Random();
var ae = a.Chromosomes.GetEnumerator();
var be = b.Chromosomes.GetEnumerator();
List <Chromosome <T> > cl1 = new List <Chromosome <T> >();
List <Chromosome <T> > cl2 = new List <Chromosome <T> >();
// For each chromosome
while (ae.MoveNext() && be.MoveNext())
{
Chromosome <T> ac = ae.Current;
Chromosome <T> bc = be.Current;
int point = r.Next(Math.Min(ac.GenesCount, bc.GenesCount));
// Split chromosomes into parts
var aFrontGenes = ac.Genes.GetRange(0, point);
var bFrontGenes = bc.Genes.GetRange(0, point);
var aTailGenes = ac.Genes.GetRange(point, ac.GenesCount - point);
var bTailGenes = bc.Genes.GetRange(point, bc.GenesCount - point);
// Find not genes that are unique in aFrontGenes and bFrontGenes
Stack <T> uniqueAGenes = new Stack <T>();
Stack <T> uniqueBGenes = new Stack <T>();
foreach (T aGene in aFrontGenes)
{
int i = bFrontGenes.FindIndex(x => x.Equals(aGene));
if (i < 0)
{
uniqueAGenes.Push(aGene);
}
}
foreach (T bGene in bFrontGenes)
{
int i = aFrontGenes.FindIndex(x => x.Equals(bGene));
if (i < 0)
{
uniqueBGenes.Push(bGene);
}
}
// Swap elements in Tail parts
while (uniqueAGenes.Count > 0 && uniqueBGenes.Count > 0)
{
T aEl = uniqueAGenes.Pop();
T bEl = uniqueBGenes.Pop();
int pos;
pos = aTailGenes.FindIndex(x => x.Equals(bEl));
aTailGenes[pos] = aEl;
pos = bTailGenes.FindIndex(x => x.Equals(aEl));
bTailGenes[pos] = bEl;
}
// Create new, crossed chromosomes
List <T> gl1 = new List <T>();
gl1.AddRange(aFrontGenes);
gl1.AddRange(bTailGenes);
cl1.Add(new Chromosome <T>(gl1));
List <T> gl2 = new List <T>();
gl2.AddRange(bFrontGenes);
gl2.AddRange(aTailGenes);
cl2.Add(new Chromosome <T>(gl2));
}
Solution <T> s1 = new Solution <T>(cl1);
s1.Fitness = fitnessFunction(s1);
Solution <T> s2 = new Solution <T>(cl2);
s2.Fitness = fitnessFunction(s2);
return(new Solution <T>[] { s1, s2 });
}
public void TestUniformCrossover()
{
Random r = new Random(1);
Chromosome c1 = new Chromosome("AAAA");
Chromosome c2 = new Chromosome("BBBB");
Utility.UniformCrossover(c1, c2, r);
Assert.AreEqual("BBBA", new string(c1.Genes));
Assert.AreEqual("AAAB", new string(c2.Genes));
}
/// <summary>
/// Sets up the population to a state before generation 1
/// </summary>
public void InitializePopulation()
{
Console.WriteLine("Population initialized");
//make lists
currentGen = new List<Organism>();
//setup ideal
//TODO make this accessible to the topmost level. i.e. population constructor takes in the ideal and allow population to externally set a new ideal
Chromosome idealGeneticInfo = new Chromosome();
ideal = new Organism(idealGeneticInfo);
//fill the first list with _popSize elements
for (int i = 0; i < _populationSize; i++)
{
Organism organism = new Organism();
organism.testFitness(ideal);
currentGen.Add(organism);
}
}
private static void Algorithm3()
{
List <Chromosome> generation;
Stopwatch sw;
TimeSpan ts3;
generation = Chromosome.Copy(initialChromosome);
//算法GA
//开始计时
sw = new Stopwatch();
sw.Start();
double lastMin = 0;
for (int i = 0, noIm = 0; i < G; i++)
{
Console.WriteLine("代数：" + i);
generation.Sort(); //从小到大排列染色体
// for (int jj = 0; jj < N; jj++)
// {
// Console.WriteLine("No."+jj+":"+ string.Join(",", generation[jj].encoded));
// Console.WriteLine("No."+jj+":"+ string.Join(",",generation[jj].GetDecoded()));
// }
//Console.WriteLine("方法GA最优染色体编码为: " + string.Join(",", generation.First().encoded));
// Console.WriteLine("方法GA最优适应函数值为: " + lastMin);
//Console.WriteLine("代数1：" + i);
if (i != 0) //至少执行一次
{
if (lastMin - generation.First().GetFitness() <= NoImprovedCriticalCondition) //如果没有改进,计数加一
{
noIm++;
}
else
{
Console.WriteLine("未改进代数：" + noIm);
noIm = 0;
}
}
if (noIm >= MaxGenerationNoImproved)//
{
Console.WriteLine("代数：" + i);
break;
}
lastMin = generation.First().GetFitness();//更新 lastmin
//generation.Sort(); //从小到大排列染色体
var elite = generation.GetRange(0, N);
//精英 进行 选择 ,交叉,变异
elite = Chromosome.Select(elite);
//Console.WriteLine("代数2：" + i);
Chromosome.Crossover(elite);
//Console.WriteLine("代数3：" + i);
Chromosome.Mutation(elite);
//Console.WriteLine("代数4：" + i);
generation.Clear();
generation.AddRange(elite);
//Chromosome.Intensify(generation.GetRange(N / 2, N / 2), generation.Min());
}
generation.Sort();
// Console.WriteLine("方法GA最优染色体编码为: " + string.Join(",", generation.First().encoded));
Console.WriteLine("方法GA最优染色体序列为: " + string.Join(",", generation.First().GetDecoded()));
Console.WriteLine("方法GA最优适应函数值为: " + generation.First().GetFitness());
Console.WriteLine("最优适应函数值为: " + HistoryRecords.Values.Min());
//结束计时
sw.Stop();
ts3 = sw.Elapsed;
Console.WriteLine("方法GA总共花费{0}ms.", ts3.TotalMilliseconds);
Console.WriteLine(" ");
}
static void Main(string[] args)
{
int better = 0;
int worse = 0;
double grandAvgSum = 0;
int runs = 0;
for( ; runs < 1; runs++)
{
int populationSize = 200;
int currentGeneration = 1;
Chromosome[] population = new Chromosome[populationSize];
int maxScore = 0;
int minScore = Int32.MaxValue;
int total = 0;
//initial setup of the population
for (int i = 0; i < populationSize; i++)
{
population[i] = new Chromosome(4, .5, .001);
if (population[i].getFitnessScore() > maxScore)
maxScore = population[i].getFitnessScore();
if (population[i].getFitnessScore() < minScore)
minScore = population[i].getFitnessScore();
total += population[i].getFitnessScore();
}
int firstMax = maxScore;
int lastMax = 0;
double lastAvg = 0;
double firstAvg = (double)(total) / (double)(populationSize);
for (; currentGeneration <= 10000; currentGeneration++)
{
Chromosome[] selected = Chromosome.rouletteSelect(population, populationSize / 2);
Chromosome[] nextGen = new Chromosome[populationSize];
maxScore = 0;
minScore = Int32.MaxValue;
total = 0;
//first half are selected and mated from previous set
for (int j = 0; j < populationSize / 2; j += 2)
{
Chromosome[] offspring = Chromosome.mate(selected[j], selected[j + 1]);
nextGen[j] = offspring[0];
nextGen[j + 1] = offspring[1];
}
//second half are just the 'parents' that live on
for (int j = populationSize / 2; j < populationSize; j++)
{
nextGen[j] = selected[j - (populationSize / 2)];
}
for (int j = 0; j < populationSize; j++)
{
if (nextGen[j].getFitnessScore() > maxScore)
maxScore = nextGen[j].getFitnessScore();
if (nextGen[j].getFitnessScore() < minScore)
minScore = nextGen[j].getFitnessScore();
total += nextGen[j].getFitnessScore();
}
population = nextGen;
lastMax = maxScore;
lastAvg = ((double)(total) / (double)(populationSize));
}
if (firstAvg < lastAvg)
better++;
else
worse++;
grandAvgSum += lastAvg;
}
Console.Write("Better performance: ");
Console.WriteLine(better);
Console.Write("Worse performance: ");
Console.WriteLine(worse);
Console.Write("Grand Average: ");
Console.WriteLine(grandAvgSum/runs);
Console.ReadLine();
}
private static Chromosome SelectAnswerCandidate(Chromosome source, Chromosome target)
{
return(IsCandidateAnswer(source) ? source : target);
}
public static Chromosome[] rouletteSelect(Chromosome[] population, int numToSelect)
{
Chromosome[] toReturn = new Chromosome[numToSelect];
int[] popIndex = new int[population.Length];
bool[] hasBeenSelected = new bool[population.Length];
// sort and compute total fitness
int fitnessSum = 0;
population = population.OrderByDescending(x => x.getFitnessScore()).ToArray();
int previousFitness = 0;
int counter = 0;
foreach (Chromosome chrome in population)
{
popIndex[counter] = chrome.getFitnessScore() + previousFitness;
previousFitness = popIndex[counter];
counter++;
fitnessSum += chrome.getFitnessScore();
}
int selected;
for(int i = 0; i < numToSelect; i++)
{
selected = Program.rng.Next(fitnessSum);
int index = 0;
while (selected >= popIndex[index])
{
index++;
}
// chrome @ current index was selected from the roulette
// if the chromosome has already been selected, we just randomly
// generate another number. This is an easy way to do this but
// can potentially take a long time to finish the selection if
// the proportion of selected to not selected is high.
if (!hasBeenSelected[index])
{
toReturn[i] = population[index];
hasBeenSelected[index] = true;
}
else
{
// we don't want to go to the next index since we tried
// to select something already selected
i--;
}
}
return toReturn;
}
/// <summary>
/// Compares the current chromosome with another chromosome.
/// </summary>
/// <param name="otherChromosome">A chromosome to compare with this chromosome.</param>
/// <returns>
/// Less than zero if this chromosome is less than the <c>otherChromosome</c> (this chromosome is less fit than the <c>otherChromosome</c>),
/// zero if this chromosome equals the <c>otherChromosome</c> (this chromosome and the <c>otherChromosome</c> are equally fit), and
/// greater than zero if this chromosome is more than the <c>otherChromosome</c> (this chromosome is fitter than the <c>otherChromosome</c>).
/// </returns>
public int CompareTo(Chromosome <TGene> otherChromosome)
{
return(evaluation.CompareTo(otherChromosome.Evaluation));
}
// Methods public abstract Chromosome PerformCrossover(Chromosome firstParent, Chromosome secondParent);
public override void RunTest()
{
Console.WriteLine("Chromosome Tests");
Console.WriteLine("~~~~~~~~~~~~~~~~");
Console.WriteLine("Created a Chromosome (random alphanumeric string)");
Chromosome chr1 = new Chromosome();
Console.WriteLine(chr1.TestGenetics.First<string>());
Console.WriteLine();
Console.WriteLine("randomize its genetic information");
chr1.RandomizeGenome();
Console.WriteLine(chr1.TestGenetics.First<string>());
Console.WriteLine();
//TODO: set this up so you any added string truncates after the number of chars in the string before you set it
Console.WriteLine("change to specific genetic info: \"Abnormality\"");
string geneticInfo = "Abnormality";
chr1.TestGenetics.RemoveRange(0, 1);
chr1.TestGenetics.Add(geneticInfo);
Console.WriteLine(chr1.TestGenetics.First<string>());
Console.WriteLine();
}
private void GenerateNextPopulation()
{
Random rnd = new Random();
Chromosome oldBestSolution = new Chromosome(ProblemParameters.Dimension);
Chromosome newBestSolution, newWrostSolution;
// Store best found solution
_bestOfCurrentPopulation.CopyTo(oldBestSolution);
//
#region Selection (Algorithm: Tournament)
int tourSize; // Tournament size
int bestChromosomeIndex = 0;
double bestFitness;
if (PopulationSize < 32)
{
tourSize = 3;
}
else
{
tourSize = (int)Math.Log(PopulationSize, 2) - 1;
}
for (int i = 0; i < PopulationSize; i++)
{
bestFitness = Double.NegativeInfinity;
for (int j = 0; j < tourSize; j++)
{
int rndIndex;
rndIndex = rnd.Next(PopulationSize);
if (_currentPopulation[rndIndex].FitnessValue > bestFitness)
{
bestFitness = _currentPopulation[rndIndex].FitnessValue;
bestChromosomeIndex = rndIndex;
}
}
// Copy the "winner of tournament" to Intermediate Population
_currentPopulation[bestChromosomeIndex].CopyTo(_intermediatePopulation[i]);
}
#endregion
#region Crossover (Algorithm: Arithmetic Crossover)
for (int i = 0; i < PopulationSize / 2; i++)
{
if (rnd.NextDouble() < ProbabilityOfCrossover)
{
// Do Crossover
Chromosome.DoCrossover(_intermediatePopulation[i], _intermediatePopulation[i + 1],
_currentPopulation[i], _currentPopulation[i + 1]);
}
else
{
// Do NOT Crossover (Copy parents to the new population without any change)
_intermediatePopulation[i].CopyTo(_currentPopulation[i]);
_intermediatePopulation[i + 1].CopyTo(_currentPopulation[i + 1]);
}
}
#endregion
#region Mutation
newBestSolution = newWrostSolution = _currentPopulation[0];
for (int i = 0; i < PopulationSize; i++)
{
if (rnd.NextDouble() < ProbabilityOfMutation)
{
_currentPopulation[i].DoMutation();
}
// Find wrost of this (new) population
if (_currentPopulation[i].FitnessValue < newWrostSolution.FitnessValue)
{
newWrostSolution = _currentPopulation[i];
}
// Find best of this (new) population
if (_currentPopulation[i].FitnessValue > newBestSolution.FitnessValue)
{
newBestSolution = _currentPopulation[i];
}
}
#region Elitism
oldBestSolution.CopyTo(newWrostSolution); // After this statement, "newWrostSolution" will point to "Old Best Solution"!
// Now we must find best of current and old populations
_bestOfCurrentPopulation = (oldBestSolution.FitnessValue < newBestSolution.FitnessValue)
? newBestSolution : newWrostSolution;
#endregion
#endregion
}
private static void Algorithm1()
{
//算法一
//首先复制初始染色体,注意,因为list是一个引用类型,不能简单的赋值,又因为lsit中存放的也是引用类型,也不能使用浅克隆.
var generation = Chromosome.Copy(initialChromosome);
//开始计时
Stopwatch sw = new Stopwatch();
sw.Start();
double lastMin = 0;
for (int i = 0, noIm = 0; i < G; i++)
{
generation.Sort(); //从小到大排列染色体
if (i != 0) //至少执行一次
{
if (lastMin - generation.First().GetFitness() <= NoImprovedCriticalCondition) //如果没有改进,计数加一
{
noIm++;
}
else
{
Console.WriteLine("未改进代数：" + noIm);
noIm = 0;
}
}
if (noIm >= MaxGenerationNoImproved)//
{
Console.WriteLine("代数：" + i);
break;
}
lastMin = generation.First().GetFitness();//更新 lastmain
//舍弃适应度函数值较大的后2/1染色体
generation.RemoveRange(N / 2, N / 2);
//强化前二分之一
Chromosome.Intensify(generation, generation.Min());
var elite = Chromosome.Copy(generation);
//选择操作
elite = Chromosome.Select(elite);
//交叉
Chromosome.Crossover1(elite);
//变异
Chromosome.Mutation(elite);
generation.AddRange(elite);
}
generation.Sort();
//Console.WriteLine("方法1最优染色体编码为: " + string.Join(",", generation.First().encoded));
Console.WriteLine("方法1最优染色体序列为: " + string.Join(",", generation.First().GetDecoded()));
Console.WriteLine("方法1最优适应函数值为: " + generation.First().GetFitness());
Console.WriteLine("最优适应函数值为: " + HistoryRecords.Values.Min());
//结束计时
sw.Stop();
TimeSpan ts2 = sw.Elapsed;
Console.WriteLine("方法1总共花费{0}ms.", ts2.TotalMilliseconds);
Console.WriteLine(" ");
}
// crossing genes of 2 mates
public static Chromosome[] mate(Chromosome mate1, Chromosome mate2)
{
//crossover points = 2
//For 2 crossover points, just do the process twice
uint[] childGenes1 = new uint[mate1.getNumGenes()];
uint[] childGenes2 = new uint[mate2.getNumGenes()];
Chromosome[] toReturn = new Chromosome[2];
//chose which change to split on
int geneToSplit = Program.rng.Next(mate1.getNumGenes());
//calculate how many bits we are using
int bitPos = 0;
int bitsCalc = maxGeneVal;
while (bitsCalc != 1)
{
bitsCalc = bitsCalc >> 1;
bitPos++;
}
int pivot = Program.rng.Next(bitPos + 1);
//no mid-gene swap
if (pivot == 0)
{
childGenes1[geneToSplit] = mate1.getGenes()[geneToSplit];
childGenes2[geneToSplit] = mate2.getGenes()[geneToSplit];
}//swap all bits of this gene
else if (pivot == bitPos)
{
childGenes1[geneToSplit] = mate2.getGenes()[geneToSplit];
childGenes2[geneToSplit] = mate1.getGenes()[geneToSplit];
}//mid bitstring swap, actually have to do stuff here
else
{
uint saveMask = uint.MaxValue << pivot;
uint swapMask = uint.MaxValue >> 32 - pivot;
// half of the chunk that is swapping
uint swapChunk1 = mate1.getGenes()[geneToSplit] & swapMask;
uint swapChunk2 = mate2.getGenes()[geneToSplit] & swapMask;
//put two havles together with bitwise 'or'
childGenes1[geneToSplit] = (mate1.getGenes()[geneToSplit] & saveMask) | swapChunk2;
childGenes2[geneToSplit] = (mate2.getGenes()[geneToSplit] & saveMask) | swapChunk1;
}
//then, copy all genes before the pivot, and swap all genes after pivot
for (int i = 0; i < mate1.getNumGenes(); i++)
{
if (i < geneToSplit)
{
childGenes1[i] = mate1.getGenes()[i];
childGenes2[i] = mate2.getGenes()[i];
}
else if (i > geneToSplit)
{
childGenes1[i] = mate2.getGenes()[i];
childGenes2[i] = mate1.getGenes()[i];
}
}
Chromosome newChild1 = new Chromosome(childGenes1, mate1.getCrossoverRate(), mate1.getMutationRate());
Chromosome newChild2 = new Chromosome(childGenes2, mate2.getCrossoverRate(), mate2.getMutationRate());
toReturn[0] = newChild1;
toReturn[1] = newChild2;
if (hasMated)
{
hasMated = false;
toReturn[0] = mutateGenes(toReturn[0], mate1.getMutationRate());
toReturn[1] = mutateGenes(toReturn[1], mate2.getMutationRate());
return toReturn;
}
else
{
hasMated = true;
return mate(newChild1, newChild2);
}
}
private static bool IsCandidateAnswer(Chromosome candidate)
{
return(candidate.Score == 0);
}
// random chance that a bit is flipped
private static Chromosome mutateGenes(Chromosome toMutate, double mutationRate)
{
double target = 1.0 / mutationRate;
if (target == Program.rng.Next(1000) + 1)
{
// mutation! now we randomly select bit to change
int geneToMutate = Program.rng.Next(toMutate.getNumGenes());
int bitsCalc = maxGeneVal;
int bitPos = 0;
while (bitsCalc != 1)
{
bitsCalc = bitsCalc >> 1;
bitPos++;
}
int bitToFlip = Program.rng.Next(bitPos);
uint newGeneVal = toMutate.getGenes()[geneToMutate];
newGeneVal ^= (uint)(1 << bitToFlip);
toMutate.setGene(geneToMutate, newGeneVal);
}
return toMutate;
}
private static void SetCandidates()
{
source = myHabitat.ExtractCandidate(pool);
target = myHabitat.ExtractCandidate(pool);
}
public Tuple <bool, double, double[], bool[]> Execute(Point goal)
{
Goal = goal;
/*Chs = new Chromosome<double>[GenSize];
* Fit = new double[GenSize];
*
* for (int i = 0; i < GenSize; i++)
* {
* Chs[i] = new Chromosome<double>(ParamNum);
* for (int j = 0; j < ParamNum; j++)
* Chs[i].Genes[j] = Agent.StepRanges[j, 0] + (Agent.StepRanges[j, 1] - Agent.StepRanges[j, 0]) * Rng.NextDouble();
* }
*
* Chromosome<double> dominant = null;
* double Min = double.PositiveInfinity;
* bool Converged = false;
* Stopwatch sw = new Stopwatch();
* sw.Start();
* while (Time++ < MaxTime)
* {
* //fitting
* FitnessFunction();
* Array.ConvertAll(Fit, (t) => { return Math.Abs(t); });
*
* //qualification
* Min = Fit.Min();
* if (Min < Precision)
* {
* dominant = Chs[Array.IndexOf(Fit, Min)];
* Converged = true;
* break;
* }
*
* //selection
* RouletteWheelSelection(Chs, Fit);
*
* //mutation
* Mutation(Chs, Pm, t => t + -0.1 + 0.2 * Rng.NextDouble());
*
* //crossover
* Crossover(Chs);
* }
* sw.Stop();
* TotalRealTime += (int)sw.ElapsedTicks / 10;
* Console.WriteLine("IKP Time: {0}; Real time: {1}", Time, sw.ElapsedTicks / 10);
*
* //if dominant chromosome wasn't found, consider last result the best result
* if (dominant == null)
* {
* dominant = Chs[Array.IndexOf(Fit, Min)];
* }
*
* //decoding chromosome
* double[] dq = dominant.GetParamAll(Decode);
*
* //detect all collisions
* Manipulator AgentNext = new Manipulator(Agent)
* {
* q = Agent.q.Zip(dq, (t, s) => { return t + s; }).ToArray()
* };
* bool[] Collisions = DetectCollisions(AgentNext);
*
* Time = 0;
* return new Tuple<bool, double, double[], bool[]>(Converged, Min, dq, Collisions);*/
double range = 0;
for (int i = 0; i < ParamNum; i++)
{
range = -120 * Math.PI / 180 - Agent.q[i];
Agent.StepRanges[i, 0] = range / 5; // <= -1 ? -1 : range;
range = 120 * Math.PI / 180 - Agent.q[i];
Agent.StepRanges[i, 1] = range / 5; // >= 1 ? 1 : range;
}
Manipulator Contestant = new Manipulator(Agent);
double[] q_init = new double[ParamNum];
Array.Copy(Agent.q, q_init, ParamNum);
double dist = Agent.DistanceTo(Goal), init_dist = dist, coeff = 1;
double Min = double.PositiveInfinity;
bool Converged = false;
Stopwatch sw = new Stopwatch();
sw.Start();
while (Time++ < 1000)
{
Chromosome <double> ch = new Chromosome <double>(ParamNum);
for (int i = 0; i < ParamNum; i++)
{
ch.Genes[i] = Agent.StepRanges[i, 0] + (Agent.StepRanges[i, 1] - Agent.StepRanges[i, 0]) * Rng.NextDouble();
ch.Genes[i] *= coeff;
}
double[] dq = ch.GetParamAll(Decode);
Contestant.q = Agent.q.Zip(dq, (t, s) => { return(t + s); }).ToArray();
double dist_new = Contestant.DistanceTo(Goal);
if (dist_new < dist)
{
Array.Copy(Contestant.q, Agent.q, ParamNum);
Min = dist = dist_new;
coeff = dist / init_dist;
}
if (dist < 0.002)
{
Converged = true;
break;
}
}
sw.Stop();
Console.WriteLine("IKP Time: {0}; Real time: {1}", Time, sw.ElapsedTicks / 10);
bool[] Collisions = DetectCollisions(Agent);
Time = 0;
return(new Tuple <bool, double, double[], bool[]>(Converged, Min, Agent.q.Zip(q_init, (t, s) => { return t - s; }).ToArray(), Collisions));
}
private static void Algorithm2()
{
List <Chromosome> generation;
Stopwatch sw;
TimeSpan ts2;
generation = Chromosome.Copy(initialChromosome);
//算法二
//开始计时
sw = new Stopwatch();
sw.Start();
double lastMin = 0;
for (int i = 0, noIm = 0; i < G; i++)
{
generation.Sort(); //从小到大排列染色体
// Console.WriteLine("方法2最优染色体编码为: " + string.Join(",", generation.First().encoded));
// Console.WriteLine("方法2最优适应函数值为: " + generation.First().GetFitness());
// Console.WriteLine("方法2最差适应函数值为: " + generation.Last().GetFitness());
if (i != 0) //至少执行一次
{
if (lastMin - generation.First().GetFitness() <= NoImprovedCriticalCondition) //如果没有改进,计数加一
{
noIm++;
}
else
{
noIm = 0;
}
}
//Console.WriteLine("未改进代数：" + noIm);
if (noIm >= MaxGenerationNoImproved)//
{
Console.WriteLine("代数：" + i);
break;
}
lastMin = generation.First().GetFitness();//更新 lastmain
var elite = generation.GetRange(0, N / 2);
var ordinary = generation.GetRange(N / 2, N / 2);
//Console.WriteLine("代数：" + i);
//Console.WriteLine("目前代最优适应函数值为: " + generation.Min().GetFitness());
//Console.WriteLine("目前代第一适应函数值为: " + generation.First().GetFitness());
//Console.WriteLine("目前代最差适应函数值为: " + generation.Last().GetFitness());
//Console.WriteLine("方法2最优染色体序列为: " + string.Join(",", generation.Last().GetDecoded()));
//Console.WriteLine("目前精英最优适应函数值为: " + elite.Min().GetFitness());
//精英 进行 选择 ,交叉,变异
elite = Chromosome.Select(elite);
// Console.WriteLine("目前代最优适应函数值为: " + generation.Min().GetFitness());
//Console.WriteLine("目前精英最优适应函数值为: " + elite.Min().GetFitness());
Chromosome.Crossover(elite);
// Console.WriteLine("目前代最优适应函数值为: " + generation.Min().GetFitness());
// Console.WriteLine("目前精英最优适应函数值为: " + elite.Min().GetFitness());
Chromosome.Mutation(elite);
//Console.WriteLine("目前代最优适应函数值为: " + generation.Min().GetFitness());
//Console.WriteLine("目前精英最优适应函数值为: " + elite.Min().GetFitness());
Chromosome.Intensify(ordinary, elite.Min());
generation.Clear();
generation.AddRange(elite);
generation.AddRange(ordinary);
// Console.WriteLine("目前代最优适应函数值为: " + generation.Min().GetFitness());
//Console.WriteLine("目前精英最优适应函数值为: " + elite.Min().GetFitness());
//Console.WriteLine(" ");
}
generation.Sort();
// Console.WriteLine("方法2最优染色体编码为: " + string.Join(",", generation.First().encoded));
Console.WriteLine("方法2最优染色体序列为: " + string.Join(",", generation.First().GetDecoded()));
Console.WriteLine("方法2最优适应函数值为: " + generation.First().GetFitness());
Console.WriteLine("最优适应函数值为: " + HistoryRecords.Values.Min());
//结束计时
sw.Stop();
ts2 = sw.Elapsed;
Console.WriteLine("方法2总共花费{0}ms.", ts2.TotalMilliseconds);
Console.WriteLine(" ");
}