Ooot by DesignEngrLab

Project Description

A library (.dll) of various linear, nonlinear, and stochastic numerical optimization techniques. While some of these are older than 40 years, many have yet to take advantage of an object-oriented programming model. This toolbox was originally created to aid in the automated design of various engineering artifacts.

Here is a table of content of wiki files I have created so far for OOOT:

Overview presents a motivation and introduction for the toolbox.
- The Main Classes Diagram attempts to show the logic in the organization of some of the files. It lies that there are six main optimization routines. There are currently 9 and more are on the way (Branch & Bound, Particle Swarm, etc.).
Problem Formulation offers basic thinking in how problems are provided to an optimization routine.
Convergence Criteria lists the 10 different convergence criteria and how they can be combined per optimization technique.
Examples
- Example 1: Simple 2D Function is a simple Nelder-Mead optimization performed on Rosenbrock's Banana Function.
- Example 2: Setting up a Genetic Algorithm shows how to run a GA and how to store the objective function and constraints in files that inherit from OOOT interfaces, which by the way are shown in more details in Problem Formulation.
  - The A Study of C# implementations of GAs surveys other GA's on the internet and presents some motivation for rolling my own.
- Example 3: Using XML and Comparison illustrates how saving and opening problem defintions can be accomplished as well as how multiple optimization methods can be invoked in the same problem formulation.
- Example 4 Using Dependent Analysis introduces a useful technique for combining the determination of dependent paramters in one analysis that may be used by multiple constraints and objectives.

Finally, I provide a rough To Do List to offer to fortune-telling in where all of this is going: