Urban dynamic expansion simulation method based on partition and improved CNN-CA model

Abstract

The invention relates to an urban dynamic expansion simulation method based on partitions and an improved CNN-CA model, and belongs to the field of geographic information. The method comprises the following steps: S1, carrying out region division on a research region; s2, establishing an improved CNN model; and S3, coupling the improved CNN with the CA model. According to the method, an improved deep learning method is combined with a traditional CA model, the idea of ensemble learning is introduced, the neighborhood effect is fully considered, the city expansion simulation precision is high,and data support and reference are provided for decision makers in related fields.

Description

technical field [0001] The invention belongs to the field of geographic information, and relates to a method for simulating urban dynamic expansion based on partitions and an improved CNN-CA model. Background technique [0002] Cellular Automata (Cellular Automata, CA) is a grid dynamics model in which time, space, and state are all discrete, and spatial interaction and temporal causality are localized. It has the ability to simulate the spatiotemporal evolution process of complex systems. The core problem of CA is to obtain effective transformation rules. Nowadays, there are many urban expansion simulation methods and models, such as logistic regression, artificial neural network and random forest algorithm, etc. coupled with CA model. Logistic regression can explain the importance of driving factors for urban sprawl simulation, but the simulation accuracy is low when dealing with nonlinear and complex expansion simulation changes. Artificial neural network is better than...

AI Technical Summary

Problems solved by technology

These algorithms only evaluate the suitability of urban development through the evaluation of the driving factors of a single cell. Due to the existence of spatial heterogeneity, the simulation accuracy is often low.

On the other hand, these methods usually consider the impact of various driving factors (geographical conditions, socioeconomic factors, etc.) on urban sprawl, and have certain shortcomings in terms of neighborhood effects

Images