Solving Land-use Suitability Analysis And Planning Problem By Hybrid Metaheuristic Algorithms

The aim of Land-use Suitability Analysis and Planning Problem is the basic basis of scientific land compilation,which is of great significance to land use structure adjustment and land partition.The aim of Land-use Suitability Analysis and Planning Problem is the most suitable land use types in one piece of the future after considering the influence factors of multiple conflicting.At present,the aim of Land-use Suitability Analysis and Planning is mainly based on the single objective optimization model.The commonly used solutions are mainly limited to a single heuristic algorithm.However,the single heuristic algorithm is difficult to adapt to the multi-objective and complexity of Land-use Suitability Analysis and Planning model,which leads to easy to fall into the local optimum and can not obtain the optimal solution.In this paper,the comprehensive weight of land use types was obtained by combining AHP with DEMATET.Then,the hybrid heuristic algorithm was constructed by combining the tabu search algorithm(TS),variable neighborhood search algorithm(VNS),genetic algorithm(GA)and particle swarm flipimization algorithm(PSO)to solve the problems of land use suiteness analysis and planning.The research work of this paper mainly includes:(1)the research of Land-use Suitability Analysis and Planning Problem and its corresponding mathematical model,the mathematical model includes three objective functions:maximization of land use suitability,maximization of tightness and minimization of traffic flow cost,and three constraints: the number of land type units,the uniqueness of land type units,and the binary decision variable.Because the influencing factors of land use suitability are compensatory,the weight assignment method of AHP combined with DEMATEL is adfliped.Firstly,AHP is used to determine the initial weight value of each influencing factor,and then DEMATEL method is used to further flipimize the weight obtained by using AHP to improve the accuracy of weight.(2)A hybrid heuristic algorithm for land use suitability analysis and planning is constructed.Based on the idea of hybrid heuristic algorithm,the algorithm structure,basic process and advantages and disadvantages of tabu search algorithm(TS),variable neighborhood search algorithm(VNS),genetic algorithm(GA)and particle swarm flipimization algorithm(PSO)are studied.They are combined to construct three hybrid heuristic algorithms,namely TS-VNS algorithm,TS-VNS-GA algorithm and TS-VNS-PSO algorithm.In tabu search algorithm(TS),the solution space is increased to increase the diversity of solutions,that is,the probability of global optimal solution is increased.In the VNS algorithm,the search of disturbance shaking operator,swap operator,Flip-first operator and Flip-two operator are used.Improved search speed to solve the problem of too large area leading to slow search speed.A constraint on the compactness of the target was added into the crossover operator of GA algorithm to improve the accuracy of the objective function of land use suitability.In PSO,the spatial distribution of particle swarm is used to simulate the target in combination with the spatial searching property of the algorithm,spatial pattern to optimize the multi-objective model.(3)Taking the land use of Zhengzhou airport area as an example,three hybrid heuristic algorithms were constructed to verify the results.The results show that the structure of three kinds of hybrid heuristic algorithm to generate the goal of the flipimal solution in the suitability of maximization,cost minimization maximum compactness and traffic flow on the three aspects are better than the existing MOSVNS algorithm,TS-VNS-GA algorithm is the best effect.The search ability of the three hybrid heuristic algorithms is improved obviously,which can effectively avoid the trap of local flipimum and search for feasible solutions to a greater extent,so that the complexity of the algorithm is balanced with the effectiveness of the solution.