From DDL Wiki
In the area of engineering design, decomposition may refer to problem decomposition, product decomposition, or process decomposition. Kusiak (1995) discusses these in detail.
In the field of optimization, a decomposition-based design optimization method is an approach that solves optimization problems by replacing the original problem with a number of smaller sub-problems and then, in a coordinated manner, solving the sub-problems iteratively until it reaches a complete solution to the original problem. Often, decomposition approaches will exploit some special structure that exists in the optimization problem. Decomposition may be used to to improve efficiency, robustness, or convenience in solving the problem.
There are many decomposition-based design optimization methods, including Dantzig-Wolfe decomposition, model coordination, goal coordination, bi-level integrated system synthesis (BLISS), analytical target cascading, and collaborative optimization.
Kusiak, A., 1995, Decomposition and Representation Methods in Mechanical Design, Journal of Mechanical Design, 117, pp. 17-24.