Optimization Technologies Center OPTEX (OPTEX Ltd.)
1.SOLVEX Fortran Libraries - numerical algorithms package which includes routines for unconctrained, nonlinear constrained, global minimization and multicriterial optimization (more than 20 various algorithms).
Optimization - Routines
R-algorithm of Shor URS
Modified Newton UMN
Minimization with simple Bounds
Direct search UBDS
Random search UBRS
Coordinates descent UBCD
Steepest descent UBSD
Conjugate gradient UBCG
Nonlinear Constrained Minimization
Possible directions NPD
Linearized function NLF
Parametrized function NPAF
Penalty function NPEF
Dual Lagrange NDUL
Direct Lagrange NDIL
Modified Newton NMN
Constrained Global Minimization
Nonuniform covering GNC
Cluster search GCS
Additive convolution MAC
Chebyshev convolution MCC
Goal programming MGP
Epsilon approximation MEA
Admissible point ADMP
Forward-difference gradient FDGRAD
SOLVEX for WINDOWS and SOLVEX OCX - optimization system for OCX (VBX) and WINDOWS applications.
SOLVEX is integrated application software intended for the following:
·solution of nonlinear constrained optimization problems;
Purposes: Solution of nonlinear optimization problems - constrained and unconstrained function minimization, search of global extremum of multi-variables nonlinear function, multicriterial Pareto-search and Pareto-approximation;
Applications: You can use SOLVEX for solving your problem directly, for including into application program, or as add-on tool for well-recognised mathematical systems (Mathematica, Matlab, MathCad etc.);
powerful algorithms library, module structure, special interface and VBX standard enables both a professional and nonprofessional users to solve difficult mathematical problems quickly and effectively.
Mode of operation: Symbolic expressions or DLL code for problems description;ODBC interface, open architecture for connection to external programs (VBX format).
Main system units
Optimization Algorithms library
Unconstrained minimization: Direct search method, conjugate gradients algorithm, conjugate directions Powell's and transform space gradients algorithms;
Constrained minimization: Penalty functions method, parametrization algorithm, modified Lagrange functions method;
Global (multi-extremal) optimization: non-uniform covering algorithms, cluster algorithm, genetic algorithms;
Multi-criterial (Pareto search) optimization: Convolution (including goal programming) and direct approximation algorithms.
Symbolic computation module
Calculus: Differentiation and integration, interval algebra;
Equation solving: Analytic solution of system of equations or unequalitys,Open Base of rules and trace of resolution.
Plots: superposition of elementary functions, parametrization functions, serial graphics, region of attractions, graphics for demonstration of the optimization process dynamic.
Microsoft Windows Version SOLVEX 3.1
System requirements: 80486 or higher, Windows 3.1 or later, 1 Mb HDD
User interface: Windows user interface and graphics
(MS-DOS Version SOLVEX 2.0 is available )