关键词: 城市布局, 城市土地利用规划, 优化技术, 模拟退火算法

Abstract:

Urban land-use planning is a complex, time-consuming process and involves complex decision-making problems. Generally, urban land-use planning is a multi-objective and nonlinear programming problem with high dimensionality, but it is difficult to run into numerical and high time cost problems with conventional optimisation tools. Currently, planners have to use some qualitative methods to generate urban land-use plans, this process has long been viewed as a ‘black box’ inside which planners are more or less subjective and the resulting land-use spatial scheme may be non-optimal. This paper develops a method called simulated annealing algorithms for handling such a conflict. The simulated annealing algorithm is quite effective and efficient for dealing with many complicated mathematical models. The mathematical optimisation model for generating optimal urban land-use plans was established, in which two main objectives, the minimum development cost of land and maximum spatial harmony were considered. Since spatial constraint considerably slows down the process of searching feasible solutions, some spatial constraints are straightforwardly integrated into the objective function in our SA, including the contiguity and compact form constraints, while the direction and distance constraints are combined into the objective function by means of adding accessional penalty cost to the objective function. The SA method is successfully applied to generate optimal land-use plans for Muyun industrial district in Changsha city, but a final realistic plan requires more detailed modeling. SA presented in this paper should therefore be regarded as a fast and simple technique useful in an early stage of the decision-making process. However, in order to maintain the transparency of urban land-use planning, robust and simple techniques as simulated annealing are promising.

Key words: urban layout, urban land-use planning, optimisation techniques, simulated annealing algorithm