So far, mathematical models used in urban planning were descriptive or predictive. It is the purpose of this paper to describe a planning model, Joe. a model whose purpose is to produce a plan rather than to predict the consequences of a given plan. It produces a land use plan as well as a transportation plan and takes into account the interactions between land use and transportation.
In part I, I take the transportation network for given and formalize the land use planning problem as a quadratic assignment problem as formulated by Koopmans and Beckmann (ref. 4). I describe the principles for a branch and bound algorithm yielding an exact optimal solution, first stated by Lawler (ref. 5). This algorithm has been progranmned and computational efficiency is examined here.
In part II, I present the problem of optimizing jointly land use and transportation. The formalization chosen is an extension of the quadratic assignment problem. Principles for a branch and bound algorithm are statedï¾° This algorithm has been progranmned and its efficiency is examined.
In part III, heuristic techniques for solving the joint problem are examined briefly.
[1]
Britton Harris,et al.
Planning as a branch and bound process
,
1971
.
[2]
Hoang Hai Hoc.
A Computational Approach to the Selection of an Optimal Network
,
1973
.
[3]
E. L. Lawler,et al.
Branch-and-Bound Methods: A Survey
,
1966,
Oper. Res..
[4]
R. Weischedel,et al.
Optimal Network Problem: A Branch-and-Bound Algorithm
,
1973
.
[5]
T. Koopmans,et al.
Assignment Problems and the Location of Economic Activities
,
1957
.
[6]
E. Lawler.
The Quadratic Assignment Problem
,
1963
.
[7]
Britton Harris,et al.
The city of the future: The problem of optimal design
,
1967
.