دانلود رایگان ترجمه مقاله بهینه سازی مکانی کاربری ارضی مبنی بر الگوریتم PSO (نشریه تیلور و فرانسیس 2011)

بهینه سازی مکانی کاربری ارضی مبنی بر الگوریتم بهینه سازی ازدحام ذرات

Land-use spatial optimization based on PSO algorithm

Abstract

The optimization of land-use spatio-structure is one of the most important areas of land use management; constructing a spatial optimization model that is based on the micro spatial unit in a bottom-up mode plays an important role in coupling the quantity structure and spatial structure effectively. The objective of this research is to develop a land use spatial optimization model based on particle swarm optimization to make spatial decision in land use management. The model is implemented using real datasets to emulate the process of spatial structure optimization in order to get the best landscape pattern under the control of decision environments. Simulation results revealed that the particle swarm optimization model has the ability to utilize the quantity and spatial structure. Furthermore, the result demonstrated that it can be used to stimulate the landscape pattern in designing the appropriate optimization environment, which could land quantity target to the basic spatial units effectively and provide appropriate spatio-structure for regional land use space layout decision making.

1 Introduction

Realizing the sustainable utilization of land resources is a very important issue of land resources management with the accelerated process of industrialization and urbanization, which has weakened agricultural sustainable development of resources because of nonagriculturalization and ecological environment deterioration. Meanwhile, the land resource structure optimization allocation also gets further study as the important research content of sustainable development. According to the characteristics of land resource and its suitability assessment and based on certain scientific technology and management, the land resource configuration optimization can get land resources within the area more reasonable arrangement and spatial distribution to achieve a certain economic, social, and ecological targets to improve the efficiency of land use, maintain the relative equilibrium of land ecosystem, and realize the sustainable utilization of land resources.[1] It is very important to use this model to optimize the land use structure. Land resource optimization alloca-tion is a complicated engineering system, which is a multiple-target multilevel continuous decision-making process; land resource configuration optimization model is also in the constant development and improvement. It has formed a number of models, such as linear programming, multiobjective and multicriteria optimization decision system dynamics, landscape ecology, spatial logistic regression, genetic algorithm (GA), and cellular automata (CA) model.[2-5] Land resources optimization allocation also breaks the traditional simple quantity structure optimization, turning to study the spatial structure, and tend to coupling them. However, it appears very difficult to simulate effectively due to the complexity and the space characteristics of the geographical landscape system, where it is difficult to get the ideal effect by some simple model; thus, land resources optimization allocation methods are also steering to geographic information science and intelligent information processing technology from the simple mathematical model in order to optimize regional land use space structure in bottom-up model at microcosmic space level of land use. However, how to match the land use goal to the corresponding space unit effectively in micro level remains to be a difficult technical problem, and the conventional models in quantity processing or treatment of space have certain limitation. The development of intelligent geographic information technology provides important technical support for spatial decision making in land resources optimizing process.[6-14] Thus, combining the intelligent geographic information technology with land optimization model and constructing intelligent land optimization model to realize the reasonable allocation of land resources both in quantity and space has become a hotspot to these researchers concerned, and it also promotes the development of scientific research about the land use optimization. The two typical kinds of space optimization models, namely, cellular automata and genetic algorithm,[15-21] also have their limitations though they are widely used at present. Though genetic algorithm can make global optimization with the coding of each land use map spot, its efficiency of managing high dimensional space information is slightly inadequate, and this disadvantage reflects more obvious especially when thousands of land use decision space is concerned. Cellular automata have the advantage in spatial evolution timing simulation, but the cell is susceptible to neighborhood and conversion rules. Meanwhile, the model is very weak in the cytoplasm optimization ability, and the spatial evolution results may not achieve global optimization, so new cell model that is under the control of multiobjective constraints or meta cell based on agent has become a new development direction.[22-24] Therefore, a model that can effectively couple quantity structure and spatial structure in microlevel in terms of land use from bottom-up approach becomes a research focus, and the intelligent processing of spatial information provides the solving tool for this optimization pattern undoubtedly.[25,26] Particle swarm optimization (PSO) is a kind of evolutionary algorithm[27] that can optimize decision-making by using the information sharing mechanism between particles. At present, some scholars have taken particle swarm into space optimization fields, such as Du, and other researchers used the spatial optimal decision research with particle swarm algorithm.[28-30] This paper brings a land use spatial optimization model based on the study of mechanism of particle swarm algorithm in order to couple the land resources optimization allocation quantity target and spatial structure effectively.[31-34] A detailed explanation is given about the design ideas and key technology when using this model. A county area was selected as a case study area to test the optimizing ability of this model thru experiments.