A construction method of urban ca model based on the principle of maximum entropy

Abstract

The invention discloses a method for constructing an urban CA model based on the principle of maximum entropy, including obtaining two phases of land use grid data and reclassifying them; superimposing the classified data to obtain the growth range of urban land, and performing random sampling and coordinates within this range Calculate and obtain sample point data; process spatial variables that affect urban land expansion, and use the results as classification model constraints; use sample points and constraints for model training to obtain the classification model with the largest conditional entropy; establish a land use matrix, and matrix elements correspond to elements Input the spatial coordinates of the cells into the model, and obtain the conditional probability of each cell classified as urban land under the influence of constraints, as the conversion probability of the CA model cells, and on this basis, combine the neighborhood constraints to construct the urban expansion CA model. The present invention constructs the maximum entropy CA model, applies it to the simulation of urban land expansion, considers the problem of randomness, and helps to more accurately simulate urban development.

Description

technical field [0001] The invention belongs to the technical field of geographic simulation, and relates to a method for constructing an urban CA model, in particular to a method for constructing an urban CA model based on the principle of maximum entropy. Background technique [0002] Cellular automaton is a dynamic system with discrete time, space and state, and powerful spatial modeling and computing capabilities. It has been widely used in the simulation research of systems with complex spatio-temporal characteristics, such as urban expansion. With the rapid development of 3S (GPS, RS, GIS) technology, the development of CA model has achieved great progress in spatial data and software, and the research on CA model has gradually deepened. The conversion rules are the core components of the CA model. Scholars have adopted a variety of methods to mine the conversion rules of the CA model in order to improve the simulation accuracy of the CA model. However, for the study ...

AI Technical Summary

Problems solved by technology

However, for the study of conversion rules, many scholars have started to mine rules from spatial variables and land use data changes. For the randomness of the model, only a simple random function is used to generate a random disturbance matrix to solve the problem. This method considers random has nothing to do with the variables in the model space, and a simple random perturbation matrix often cannot accurately solve the randomness problem of the model, which limits the improvement of the simulation accuracy. new way of sex

Images