Analytical Target Cascading Example With Coupling Variables

A simple optimization problem is used to illustrate ATC with coupling variables.

Analytical Target Cascading (ATC) is a method for solving large-scale distributed optimization problems. It can be applied to multidisciplinary design optimization (MDO) problems. A very simple example problem was chosen so that the ATC solution process could be illustrated without requiring much effort to understand the underlying optimization problem. The benefits of ATC are not realized when solving this or other small problems; ATC is helpful for large optimization problems with sparse dependence structures that are suitable for decomposition.
This example is similar to ATC_Demo1:

http://www.mathworks.com/matlabcentral/fileexchange/23260

but instead of demonstrating how to use the ATC coordination process to bring shared design variables into a consistent state, this newer example illustrates how to use the same process to coordinate coupling variables.

To run the demo type ATC_Demo2 at the MATLAB® command prompt. The file contains a full description of the example. For more information on ATC (and ALC, a related method) see:

http://ode.engin.umich.edu/publications/PapalambrosPapers/2000/126.pdf
http://ode.engin.umich.edu/publications/PapalambrosPapers/2005/216SMO.pdf
http://ode.engin.umich.edu/publications/PapalambrosPapers/2008/254.pdf