An n‐step, 2‐variable search algorithm for the component placement problem

Abstract

The component placement problem is a specialization of the quadratic assignment problem that has been extensively studied for a decade and which is of considerable practical value. Recently, interest in component placement algorithms has risen primarily as a result of increased activity in the field of computer-aided design automation. This paper deals with the methodology of component placement and is based on the results of considerable operational experience. A tutorial presentation of tree search placement algorithms is provided, and an improved placement procedure is described which is demonstrated to be effective in generating near optimal solutions to the component placement problem. These solutions are completely reproducible and are obtained at an acceptable expenditure of computational resources. An additional objective is an assessment of performance of the class of near optimal algorithms. In particular, the question-how close to optimal are the near optimal solutionsis examined.

Demand Point Aggregation Analaysis for Location Models

“…White (1993) proposed a new approach, where the actual data are relaxed by embedding them in a data space that satisfies an extension of the metric triangle property. Arora et al (2002) proposed a randomized procedure for rounding fractional perfect assignments to integral assignments and a also Heider (1973), Mirchandani and Obata (1979), Bruijs (1984), Pardalos et al (1993), Burkard and Cela (1995), Li and Smith (1995), Anderson (1996), Talbi et al (1998a), Deineko and Woeginger (2000), Misevicius (2000a), Mills et al (2003.…”

The single-finger keyboard layout problem

“…One can see that the contour locations are O(n), so the neighborhood size is O(n 2 ). To evaluate the entire neighborhood we use a technique adapted from those proposed by Heider [24] and Burkard and Rendl [8] for QAP. Let be the current solution and the solution obtained from by moving symbol i to location j.…”

A directional decomposition heuristic for one-dimensional, non-equidistant machine-cell location problems

“…Many heuristics for MLPs have been developed as a result of research in this area since the early 1960s. The major contributors to this problem are, amongst others, Hillier and Connors [7], Nugent et al [8], Heider [9], Burkard and Stratmann [10], Bazaraa and Kirca [11], Aneke and Carrie [12], Co and Arrar [13], Houshyar and McGinnis [14], Kouvelis et al [15], Sarker et al [16], and Diponegoro and Sarker [17].…”