Chemical plant layout via graph partitioning—II. Multiple levels

Abstract Partitioning of units into multiple groups occurs naturally in multilevel plants where equipment must be distributed onto different floors, so as to minimize material flow costs. A significant complication introduced in multilevel layout is the dependence of the cost of material flows between units on the relative levels of the respective unit locations. This feature leads to three distinct costs for downward, upward and horizontal flow between any pair of units. In this paper, we formulate this problem as an Integer Non-Linear Program. A heuristic procedure proposed for the single floor graph partitioning problem in Part I of this work is extended to address this problem. Global optima were obtained for all examples for which the exact optimal solutions were obtainable by exhaustive enumeration. Solutions for problems with up to 70 units were obtained in reasonable computation times, suggesting applicability of the procedure to chemical plants of realistic scope. Additionally, a linearization scheme is used to convert the INLP to an MILP. Continuous relaxations of the MILP are solved by Lagrangean Relaxation and Subgradient Optimization to yield good lower bounds. It is demonstrated that the linearization is sufficiently tight that, for some problem instances, solutions to the continuous relaxation are optimal. In general, the combination of upper and lower bounding procedures is effective as a means of both obtaining good solutions and estimating the deviation from the optimal solution for problems of practical scope.

[1]  G. Reklaitis,et al.  Chemical plant layout via graph partitioning-1. Single level , 1994 .

[2]  R. McBride,et al.  An Implicit Enumeration Algorithm for Quadratic Integer Programming , 1980 .

[3]  Mokhtar S. Bazaraa,et al.  Nonlinear Programming: Theory and Algorithms , 1993 .

[4]  J. F. Benders Partitioning procedures for solving mixed-variables programming problems , 1962 .

[5]  A. M. Geoffrion,et al.  Integer Programming Algorithms: A Framework and State-of-the-Art Survey , 1972 .

[6]  Warren P. Adams,et al.  A Tight Linearization and an Algorithm for Zero-One Quadratic Programming Problems , 1986 .

[7]  Rafael Lazimy,et al.  Mixed-integer quadratic programming , 1982, Math. Program..

[8]  H. Sherali,et al.  On the choice of step size in subgradient optimization , 1981 .

[9]  Frank Körner,et al.  Integer quadratic optimization , 1985 .

[10]  Charles N. Eastman,et al.  Spatial Synthesis in Computer-Aided Building Design , 1975 .

[11]  G. Thompson,et al.  A heuristic method for the multi-story layout problem , 1988 .

[12]  C. E. Lemke,et al.  Equilibrium Points of Bimatrix Games , 1964 .

[13]  George L. Nemhauser,et al.  Some Facets for an Assignment Problem with Side Constraints , 1991, Oper. Res..

[14]  Philip Wolfe,et al.  Validation of subgradient optimization , 1974, Math. Program..

[15]  Alois X. Schmidt,et al.  Material and Energy Balances , 1962 .

[16]  Srinivasan Jayakumar Chemical plant layout via graph partitioning , 1992 .

[17]  Ignacio E. Grossmann,et al.  A modelling and decomposition strategy for the MINLP optimization of process flowsheets , 1989 .

[18]  Hanif D. Sherali,et al.  On the Use of Exact and Heuristic Cutting Plane Methods for the Quadratic Assignment Problem , 1982 .

[19]  Roger V. Johnson Spacecraft for Multi-Floor Layout Planning , 1982 .

[20]  A. Victor Cabot,et al.  A branch and bound algorithm for solving a class of nonlinear integer programming problems , 1986 .

[21]  Thomas L. Morin,et al.  Fixed Order Branch-and-Bound Methods for Mixed-Integer Programming: The ZOOM System , 1989, INFORMS J. Comput..

[22]  Hanif D. Sherali,et al.  Linearization Strategies for a Class of Zero-One Mixed Integer Programming Problems , 1990, Oper. Res..

[23]  Brian W. Kernighan,et al.  An efficient heuristic procedure for partitioning graphs , 1970, Bell Syst. Tech. J..