부산대학교 도서관

The phrase "Planning and Scheduling" represents a class of problems where a resource in limited supply is to be optimally applied. Planning and Scheduling problems are accommodated by a branch of mathematics known as Operations Research. The goal of Operations Research technology is to find a solution to a known set of conditions such that some minimum "cost" or maximum "profit" is achieved, within some defined constraints. Several algorithms, such as Linear Programming, Dynamic Programming, Constraint Logic Programming, have arisen in the past three to four decades to solve particular classes of problems, but real-world problems continue to be extremely difficult to characterize and solve robustly. Therefore, good Operations Research is still dominated by the ability to transform the real problem into one or more simplified - and solvable - problems without sacrificing too much of the real problem's objectives and constraints. The subject technique evolved from the environment of multi-level resource management, where the decision variables represent requests for use of a given resource at a given level. Like a job-shop scheduling problem, resource management problem requests have inherent multiplicities (the part can be manufactured on any of n machines) that cause combinatorial explosions in the search for optimal resource assignments. The multi-level nature arises when the use of a resource at one level in turn requires or precludes the use of resource(s) at other levels, thus creating derived decision variables associated with the other levels.