基于遗传算法的地震短期综合预报分类系统研究

简要介绍了基于遗传算法的分类器和在此基础上建立的地震预报分类法系的基本原理。设计出基于遗传算法的地震短期综合预报分类系统的研究思路和方法,即以异常持续时间为编码依据。以遗传算法为优选预报规则的工具,以预报3个月地震为目标,综合测震和前兆多手段进行地震预测,对北天山西段进行多参数初步试算后表明该方法具有很高的学习正确率和较高检验和预测能力。

STUDY ON THE CLASSIFING SYSTEM OF SHORT-TERM COMPREHENSIVE EARTHQUAKE PREDICTION BASED ON THE GENETIC ALGORITHMS

The genetic algorithms classifier and principle of classifing system of earthquake prediction based on the genetic algorithms is briefly introduced. The method to design the classifing system of shortterm comprehensive earthquake prediction based on the genetic algorithms is developed. In the method the genetic algorithms is used as the tool to optimize the prediction rules, the lasting times of abnormity is assigned as the codes; the prediction period is three months; and the seismicity is combined with many precursor methods to predict earthquake. As an example, the method is used in North Tianshan seismic zone with multiple parameters. The preliminary calculating result shows that the method has higher self-study correctness and higher ability of testing and predicting.