COnstrained Two-Dimensional Cutting Stock Problems a Best-First Branch-and-Bound Exact Algorithm

In this paper we develop a new version of the algortihm proposed in [17] for solving exactly some variants of (un)weighted constrained two-dimensional cutting stock problems. We introduce one-dimensional bounded knapsacks in order to obtain an improved initial lower bound for limitating initially the space search, an improved upper bound at each internal node, a symmetric-strategy for neglecting some unnecessary duplicate patterns, a new data structure of the set representing the best subproblems and we use some characteristics of the BOB-library.