/// <summary>
/// 开始优化
/// </summary>
/// <param name="ConfigFile">算法参数配置文件路径</param>
/// <param name="X_Init">优化变量初值</param>
/// <param name="X_MaxStep">优化变量最大步长</param>
/// <param name="X_Lb">优化变量下限</param>
/// <param name="X_Ub">优化变量上限</param>
/// <returns>使适应度函数最小的变量值</returns>
/// <returns>最优目标函数</returns>
public double[] StartOpt(string ConfigFile, double[] X_Init, double[] X_MaxStep, double[] X_Lb, double[] X_Ub, out double Fun, out double Delta)
{
if (X_Init.Length != X_MaxStep.Length || X_Init.Length != X_Lb.Length || X_Init.Length != X_Ub.Length)
{
throw new Exception("Variable number are not set correctly!");
}
this.G = new General();
List <string[]> Config = new List <string[]>();
try
{
Config = G.ReadCSV(ConfigFile);
this.Max_FEs = int.Parse(Config[0][1]); // 最大评价次数
this.Min_TolFun = double.Parse(Config[1][1]); // 结束条件
}
catch (System.Exception ex)
{
throw new Exception("ConfigFile are not existed!");
}
var M = Matrix <double> .Build; // 生成矩阵
//var V = Vector<double>.Build; // 生成向量
this.X_Dim = X_Init.Length;
this.X = new double[X_Dim];
this.X_Lb = new double[X_Dim];
this.X_Ub = new double[X_Dim];
double[] X_Diff = new double[X_Dim];
for (int i = 0; i < X_Dim; i++)
{
this.X_Lb[i] = Math.Max(X_Init[i] - X_MaxStep[i], X_Lb[i]);
this.X_Ub[i] = Math.Min(X_Init[i] + X_MaxStep[i], X_Ub[i]);
X_Diff[i] = this.X_Ub[i] - this.X_Lb[i];
}
this.X_Opt = new double[X_Dim];
this.xmeanw = M.Dense(X_Dim, 1);
this.xold = M.Dense(X_Dim, 1);
this.zmeanw = M.Dense(X_Dim, 1);
this.flginiphase = true;
this.sigma = 0.25; // Initial step size
this.Min_sigma = 1.0e-15; // Minimal step size
this.Max_sigma = X_Diff.Max() / Math.Sqrt((double)X_Dim);
for (int i = 0; i < X_Dim; i++)
{
this.xmeanw[i, 0] = X_Init[i];
}
this.lambda = 4 + (int)Math.Floor(3.0 * Math.Log((double)X_Dim));
this.mu = (int)Math.Floor((double)lambda / 2.0);
this.arweights = M.Dense(mu, 1);
for (int i = 0; i < mu; i++)
{
this.arweights[i, 0] = Math.Log((lambda + 1.0) / 2.0) - Math.Log(i + 1.0);
}
this.cc = 4.0 / (X_Dim + 4.0);
this.ccov = 2.0 / Math.Pow((X_Dim + Math.Pow(2.0, 0.5)), 2.0);
this.cs = 4.0 / (X_Dim + 4.0);
this.damp = (1.0 - Math.Min(0.7, X_Dim * lambda / Max_FEs)) / cs + 1.0;
this.B = M.DenseIdentity(X_Dim);
this.D = M.DenseIdentity(X_Dim);
this.BD = M.Dense(X_Dim, X_Dim);
this.C = M.Dense(X_Dim, X_Dim);
this.CC = M.Dense(X_Dim, X_Dim);
this.BD = B * D;
this.C = BD * BD.Transpose();
this.pc = M.Dense(X_Dim, 1);
this.ps = M.Dense(X_Dim, 1);
this.cw = arweights.RowSums().Sum() / arweights.L2Norm();
this.chiN = Math.Pow(X_Dim, 0.5) * (1.0 - 1.0 / (4.0 * X_Dim) + 1.0 / (21.0 * Math.Pow(X_Dim, 2.0)));
this.arz = M.Dense(X_Dim, lambda);
this.arx = M.Dense(X_Dim, lambda);
this.arz_mu = M.Dense(X_Dim, mu);
this.arx_mu = M.Dense(X_Dim, mu);
this.Fun = new double[lambda]; // 各个个体的适应度
this.Delta = new double[lambda]; // 各个个体的约束违反度
this.Index = new int[lambda]; // 各个个体的序号
this.rd = new Random();
Opt(); // 优化
Fun = Best_Fun;
Delta = Best_Delta;
return(X_Opt);
}
