A k-means-based storm surge disaster key defensive area selection method

Abstract

The invention discloses a k-means-based storm surge disaster key defensive area selection method. The method specifically comprises the following steps: S01, obtaining a storm surge assessment research result of a to-be-selected area as a data set; s02, selecting four groups of different element indexes including one group of independent judgment indexes and three groups of parallel judgment indexes in the data set as an initial class cluster center; s03, respectively carrying out single-attribute K-Means clustering analysis on the four indexes serving as the initial class cluster center; s04, performing spatial superposition according to single-attribute K-Means clustering analysis results of three groups of parallel judgment indexes in the four indexes of the initial class cluster center to obtain a potential key defensive area range; and S05, performing multi-attribute K-Means clustering on the data set according to the four indexes of the initial class cluster center to determine a key defensive area. According to the invention, K-Means clustering is effectively utilized to determine the key defensive area, and the technical problems that in the prior art, subjectivity is too high, scientificity is low, the method is affected by factors such as scientificity of analogy project index standard value setting and similarity between two projects, and the universality of the method is poor are solved.

Description

technical field [0001] The invention is applied to the field of selection and planning of natural disaster defense areas, and specifically relates to a k-means-based method for selection and planning of key defense areas for storm surge disasters. Background technique [0002] Storm surge disasters are one of the major marine disasters that our country is facing, almost all over the coast of China, with high disaster frequency and high disaster intensity, causing heavy human and economic losses, and have been included in my country's catastrophe category. Although my country has promulgated the "Technical Guidelines for the Delineation of Key Defense Areas for Storm Surge Disasters" (draft for approval), which has played a guiding role in the delineation of key defense areas for storm surge disasters, but the guidelines did not propose key defense areas for storm surges ( The criteria for the selection of land area) only proposed the selection criteria of "combining the chara...

AI Technical Summary

Problems solved by technology

At this stage, the commonly used index standard assignment methods in China mainly include: 1) Experience judgment method, that is, according to the project leader’s own work experience, subjectively set the index standard value. This method is the easiest to operate, but also the most subjective , depends heavily on the project leader's own experience and level, and is suitable for situations where the dependence on objective data is not high and the scientificity of the assignment results is not required; The standard value of the index is determined by means of consultation with experts. This method has the advantages of simplicity, strong intuition, and simple calculation method, but it also reflects subjective consciousness in general, and has strict requirements on the representativeness and professionalism of experts; 3 ) analogy analysis method, that is to refer to the index standard value of similar projects with strong correlation as the index standard value of this project. Influenced by factors such as sex and similarity between two items, this method cannot be used effectively if there is no similar item or the similarity between the two items is insufficient; Sort out, find certain mathematical laws, and select statistically significant eigenvalues ​​as index standard values. This method is based on objective data and has a good scientific nature. However, different projects involve different characteristics of data samples, and different data need to be used. The eigenvalues ​​can only be found by the processing method. Even if the same data sample is obtained by different data processing methods, the eigenvalues ​​may be different. At this time, it is necessary to verify the rationality of the different results and select the best value. Therefore, this method is suitable for data samples. There are many situations, and the specific data analysis method adopted may have a strong correlation with the specific project, and there are few universally applicable features

Images