City CA model establishment method based on maximum entropy principle

Abstract

The invention discloses a city CA model establishment method based on a maximum entropy principle. The method comprises the steps that two-stage land utilization grid data is acquired and reclassified; the classified data is superimposed to obtain a city land utilization increase range, and random sampling and coordinate calculation are conducted in the range to obtain sample point data; a space variable influencing city land utilization extension is processed, and a result is used as a classification model constraint condition; a sample point and the constraint condition are used for model training to obtain a classification model with the maximum conditional entropy; a land utilization matrix and a cell corresponding to a matrix element are established, a space coordinate of the cell isinput into the model, the conditional probability of each cell classified as urban land under influence of the constraint condition is obtained and used as the CA model cell transition probability, and a city extension CA model is established on the basis of the CA model cell transition probability and in combination with neighborhood constraint. According to the method, the CA model with the maximum entropy is established and applied to urban land extension simulation, the randomness problem is considered, and the method helps to more precisely simulate city 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