Land Use Change Simulation Based On Particle Swarm Opyimization Grey-Markov Model

Land use change is complicated,and the construction of land use prediction model is helpful to understand its change process and provide reference for urban planning.The existing quantitative prediction models mostly ignore the impact of social development variability on land use.In order to solve this problem,this paper constructed an improved gry-Markov model based on particle swarm optimization algorithm.The model can effectively combine the influence of social factors and improve the accuracy of land use quantity prediction.The prediction results can provide reference for land planning in the study area.The main research results are as follows:(1)Through the screening of common quantitative prediction models(exponential smoothing method,trend extrapolation method,grey prediction model,Markov model),it is found that grey prediction model is more suitable for the study of land use quantity prediction.In order to fit the quantifiable social factors with the quantity of each land use type,the average simulation accuracy was improved by 14.6% with the help of principal component analysis and multiple regression model.The effective combination of land use change simulation and quantifiable factors was realized,and the grey prediction results after integrating social factors were closer to the real data.(2)The simulated data of grey prediction model are divided into state intervals and modified by Markov model,which can effectively make up for the shortcomings of grey prediction model in the prediction of random data.Verified by example,Markov correction on the division of state interval,will directly affect the precision of prediction results,therefore,by introducing the particle swarm algorithm to find the best interval,reduce the error produced by state division is not accurate,simulation precision is improved by 16.6% on average,the optimized result is more conform to the data characteristics dividing state interval,get the best predictive value.(3)Based on the difference coefficient of land use structure and intensity analysis method,the land use pattern change of Shuangyashan city was studied.It was found that the change of cultivated land and construction land was obvious,and the active period of land change was from2005 to 2010 and from 2015 to 2020 due to the influence of coal industry.On this basis,the optimized model was applied and relevant parameters were set,and the feasibility of the model was verified by simulation of the study area in 2020,and the land use pattern of Shuangyashan city in 2025 was further predicted.The evaluation results show that compared with the single Markov model,the simulation accuracy of the improved model in 2020 is improved by 27.5%,indicating that the overall accuracy of the simulation of land use change can be significantly improved through the improvement of quantitative prediction,and the reliability is high.