MultiDimensionalSolver.GlobalAnnealing Method

Finds the location of the global extremum (i.e. minimum or maximum) of a general multidimensional function using the technique known as simulated annealing applied to a modified version of the Needler and Mead's downhill simplex algorithm.

Overload List

Finds the location of the global extremum (i.e. minimum or maximum) of a general multidimensional function using the technique known as simulated annealing applied to a modified version of the Needler and Mead's downhill simplex algorithm.

public double[] GlobalAnnealing(ExtremumTypes,double[],int,double,double,double,double,int);

Seeks a global extremum (minimum or maximum) of a differentiable multidimensional object function using simulated annealing applied in conjunction with one of the following algorithms for finding the local extremum of differentiable object functions:

1. Steepest-Descent (i.e. DerivSteepestDescent)
2. Fletcher-Powell (i.e. DerivFletcherPowell)
3. Broyden-Fletcher-Goldfarb-Shanno (BFGS) (i.e. DerivBFGS)
4. Polak-Riviere (i.e. DerivPolakRiviere)
5. Fletcher-Reeves (i.e. DerivFletcherReeves)

Overview of Approach

The general idea used here is that after each step within the selected algorithm for finding the local extremum, we perturb the point by a random jump corresponding to "temperature of the surface".

public double[] GlobalAnnealing(ExtremumTypes,double[],int,double,double,BracketingAlgorithm,LocateAlgorithm,double,double,double,int,double,AnnealingAlgorithmTypes);

See Also

MultiDimensionalSolver Class | WebCab.COM.Math.Optimization.MultiDimensional Namespace

---