PARID - Parameter identification

PARID is designed to solve nonlinear optimization problems for determining multidimensional parameters from measured or simulated data.

These problems are characterized by a mathematical model J, which depends on parameters x and known quantities t, and the task, called multi-parameter-fitting, is to determine the unknowns x in such a way that the computed model J(t,x) fits in an optimal way the measured (or simulated) data g.

Use PARID to study the influence of parameters without performing experiments - or use it to develop appropriate models.

PARID provides the following features:

- Rapid model implementation
- efficient model evaluation (compiler used instead of interpreter)
- simulation as well as inverse problem
- arbitrary dimensions of data, model function and parameter
- user defined weighted norm
- "brute force" method to find good initial guess

A Microsoft Visual C++ or Borland C++ compiler has to be installed on your computer to generate your own model implementation.

Parid is free for eduacational purposes

THIS VERSION IS BETA; feel free to send comments and questions to doerr@isca-louis.com

Download PARID 0.4 beta Win 95/98/NT Version