Solving land-use suitability analysis and planning problem by a hybrid meta-heuristic algorithm |
| |
| Authors: | Kaveh Khalili-Damghani Bahram Aminzadeh-Goharrizi Saeed Rastegar Babak Aminzadeh-Goharrizi |
| |
| Institution: | 1. Department of Industrial Engineering, Azad University, Tehran, Iran;2. Department of Urban Planning, Imam Khomeini International University, Qazvin, Iran;3. Department of Industrial Engineering, Industrial Management Institute, Tehran, Iran |
| |
| Abstract: | The aim of Land-use Suitability Analysis and Planning Problem (LSAPP) is to identify the most suitable parcels of land for future land-uses considering several conflicting criteria. LSAPP can be modeled using a variant of a well-known combinatorial optimization problem called Quadratic Assignment Problem (QAP). In this paper, a multi-objective mathematical model is developed for LSAPP based on QAP modeling. The large-size instances of the proposed multi-objective mathematical model are difficult to solve in a reasonable CPU time using exact algorithms. So, an efficient three-phase hybrid solution procedure is proposed. In the first phase, the compensatory objectives are integrated using Analytic Hierarchy Process (AHP) and Decision-Making Trial and Evaluation Laboratory. Then, based on the aforementioned suitability objective function and other spatial objectives and constraints, a multi-objective LSAPP is constructed. Finally, a hybrid multiple objective meta-heuristic algorithm is proposed to solve the LSAPP. The core of the proposed algorithm is based on Scatter Search while Tabu Search and Variable Neighborhood Search are also utilized. The proposed algorithm is equipped with the concepts of Pareto optimality and Veto Threshold, which improve its efficacy. The proposed algorithm is applied on a real LSAPP case study, in ‘Persian Gulf Knowledge Village’, wherein its performance is compared with a well-known evolutionary computation algorithm called Vector Evaluated Genetic Algorithm (VEGA) using comprehensive statistical analysis. A survey on time complexity of the proposed algorithm is also accomplished. The results show that MOSVNS is significantly superior to VEGA both in single and in multi-objective modes. Furthermore, analysis of time complexity of the proposed algorithm shows that it is of polynomial time and can be applied to significantly larger problems with multiple compensatory and non-compensatory objectives. |
| |
| Keywords: | land-use suitability analysis and planning scatter search variable neighborhood search tabu search analytic hierarchy process DEMATEL |
|
|
