A matheuristic algorithm for the one-dimensional cutting stock and scheduling problem with heterogeneous orders (Q2307325)

The article proposes an upper bound for the planning horizon and an iterative algorithm for solving the quite large integer programming problem. Cutting stock deals with the orders of a number of customers (each has a due date). The customers order a number of items (with certain lengths). The problem is to optimize the ordering of items to fulfill all production constraints (length limit per period) and hold all (or most of) the due dates. A precomputation of an upper limit of the planning horizon is done. Using this horizon an iterative algorithm (they call it ``Matheuristic'') is designed which trys to improve the solution with local search techniques (most of the values are fixes, some are changed). The algorithm is tested numerically and from the performance side (quality and computational effort) against a standard solver for integer programming problems (CPLEX). Most of the instances show the better performance of their proposed algorithm compared with the standard solver. Here it must be stated that due to the computational complexity of the problem the instances are of medium size (up to 30 orders)