Driving Force Analysis And Dynamic Simulation Of Mining Cities’ Spatial Pattern Based On Improved Case-Based Reasoning Model

Mining cities are important strategic bases for energy and resources in China and important support for the sustainable and healthy development of the national economy.However,the excessive exploitation of mining resources and the disorderly expansion of urban layout lead to the problems of land destruction and cultivated land erosion,which seriously affect the regional environmental quality,food security,ecological security and social security.It is an effective way to guarantee the sustainable development of mining cities to provide decision support for the development of mining cities by researching temporal and spatial evolution law and trend prediction.Taking Jixi City,a typical mining city in Heilongjiang Province as the research area,this paper adopts the process of“evolution law analysis-simulation analysis and model improvement-model verification-trend prediction” by combing qualitative and quantitative analysis methods.The concrete work of this paper includes:(1)Based on the land use data of Jixi City in 2005,2010 and 2015,the spatial and temporal distribution characteristics and evolution characteristics of urban pattern in Jixi City were analyzed comprehensively.On this basis,the driving forces of the mining city expansion are analyzed from the aspects of terrain factors,mineral resources distribution,economic development and population,traffic and road,and policy planning.(2)An improved UGSCBR(Urban Growth Simulation Case Based Reasoning)model for mining urban growth is proposed.With the urban growth CBR model inspected,it is improved from the following two aspects:(1)In order to solve the problem that the traditional CBR model only retrieves a single case as the reasoning basis in the simulation process,the study proposes a comprehensive retrieval strategy that classifies the cases according to their evolution results,and then retrieves multiple most similar cases for each case.(2)considering the increasing housing vacancy rate in mining cities,based on nighttime light data,the study introduces the housing vacancy index which can reflect the human factors in mining cities for the first time to control the unreasonable expansion phenomenon in the simulation of urban spatial pattern.(3)The UGSCBR model was used to simulate the urban pattern of Jixi City in 2015,and the simulation results were tested and analyzed.Then CA model was used to simulated the urban pattern of Jixi City in 2015,and compared with UGSCBR model.(4)The UGSCBR model of urban growth in Jixi City was used to simulate the urban growth pattern from 2020 to 2035,and the spatial and temporal evolution of the urban growth pattern in the future was analyzed.The main research conclusions of this paper include:(1)From 2005 to 2015,the urban pattern of Jixi City was mainly distributed in flat areas,and the urban growth mainly occupied cultivated land in low-lying areas.Due to the depletion of coal resources and the destruction of ecological environment,the scale of urban expansion gradually slowed down.Although the overall pattern of the city was loose,there was a trend of spatial agglomeration and development during the study period,and this trend was more obvious in the central urban area.(2)Topographic factors,mineral resources distribution,economic development,transportation network and policy planning are the main driving forces affecting the evolution of mining city pattern.Among them,the distribution of mineral resources,as a special driving factor of mining cities,has both attracted and repelled the spatial development of urban land.The spatial layout planning of "one core,multi-center and point-axis development" has established the development strategy of group urban structure in the central urban area.In Jixi City,as a coal and grain composite area,the concentration of basic farmland has become the most important limiting factor of urban land expansion.(3)Through the authenticity evaluation of the simulation results,it is found that the simulation accuracy reaches 97.02%,Kappa is 85.51%,the Fo M is 0.1700,the simulation results are relatively ideal,which reflects the effectiveness of the model when the retrieval quantity parameter x=10.By comparing with CA simulation result,it is found that the UGSCBR model can better simulate the trend of urban growth,and the construction of the model is simpler and easier to understand.(4)Under the current development trend,the urban land area of Jixi City will continue to increase from 2020 to 2035,but its expansion scale will continue to decrease.The trend of urban growth expanding outwards from the center of each urban area and the central urban area further gathering will make the urban pattern more compact gradually.