基于GOA-RCMSE模型区域降水复杂性测度分析  

Analysis of regional precipitation complexity measurement based on GOARCMSE model

作　　者：刘东 白镜筱 张亮亮 李雪松 LIU Dong;BAI Jingxiao;ZHANG Liangliang;LI Xuesong(School of Water Conservancy&Civil Engineering,Northeast Agricultural University,Harbin 150030,China;Laboratory of Effective Utilization of Agricultural Water Resources of Ministry of Agriculture and Rural Affairs,Harbin 150030,China;Heilongjiang Provincial Key Laboratory of Water Resources and Water Conservancy Engineering in Cold Region,Harbin 150030,China)

机构地区：[1]东北农业大学水利与土木工程学院,哈尔滨150030 [2]农业农村部农业水资源高效利用重点实验室,哈尔滨150030 [3]黑龙江省寒区水资源与水利工程重点实验室,哈尔滨150030

出　　处：《东北农业大学学报》2023年第9期80-89,共10页Journal of Northeast Agricultural University

基　　金：国家自然科学基金项目(52179008,51579044,41071053);国家杰出青年科学基金项目(51825901);国家自然科学基金联合基金项目(U20A20318);黑龙江省自然科学基金联合引导项目(LH2023E003)。

摘　　要：文章以黑龙江省13个地区1967~2016年(50年)旬降水量为例,构建基于蝗虫优化算法改进精细复合多尺度熵模型(The improved refined composite multi-scale entropy based on grasshopper optimization algorithm,GOARCMSE),在此基础上采用信息贡献率方法对不同尺度熵值作加权,全面、准确、可靠地评估区域降水复杂性。此外,基于黑龙江省旬降水复杂性测度结果,探索影响黑龙江省降水复杂性潜在因素。结果表明,黑龙江省旬降水复杂性呈现西部低东部高的显著空间分布特征。此外,水域面积和城建面积与降水复杂性测度结果相关系数分别为-0.629和0.451,存在显著相关关系。为分析模型性能,引入蝗虫优化算法改进多尺度熵模型(The multiscale entropy based on grasshopper optimization algorithm,GOA-MSE),可知GOA-RCMSE区分度和Spearman等级相关系数分别为1.1141和0.995,而GOA-MSE区分度和Spearman等级相关系数分别为1.0935和0.973,表明GOARCMSE具备更高的可靠性和稳定性。综上,GOA-RCMSE可全面合理评价区域降水复杂性,同时为不同区域解决降水复杂性测度问题提供新思路。The precipitation in 13 regions of Heilongjiang Province was taken from 1967 to 2016(50 years)as an example,and the improved refined Composite multi-scale entropy based on grasshopper optimization algorithm(GOA-RCMSE)was constructed.On this basis,the information contribution rate method was used to weight the entropy values of different scales,so as to comprehensively,accurately and reliably evaluate the regional precipitation complexity.In addition,based on the measurement results of the ten-day precipitation complexity in Heilongjiang Province,the potential influencing factors of the precipitation complexity in Heilongjiang Province were explored.The results showed that the ten-day precipitation complexity in Heilongjiang Province presented a significant spatial distribution characteristic of low in the west and high in the east.At the same time,it was found that the correlation coefficients between the water area and urban construction area and the results of the precipitation complexity measurement reached-0.629 and 0.451,respectively,and there was a significant correlation.To analyze the model performance,the multi-scale entropy based on grasshopper optimization algorithm(GOA-MSE)was introduced.The results showed that the discriminative and Spearman rank correlation coefficients of GOA-RCMSE were 1.1141 and 0.995,respectively,the discriminative and Spearman rank correlation coefficients of GOA-MSE were 1.0935 and 0.973,respectively,indicating that GOA-RCMSE had higher reliability and stability.In conclusion,GOA-RCMSE could make a comprehensive and reasonable evaluation of regional precipitation complexity,and at the same time,it could provide a relatively novel idea for solving the problem of precipitation complexity measurement in different regions.

关 键 词：旬降水 复杂性测度 蝗虫优化算法 多尺度熵 

分 类 号：S271[农业科学—农业水土工程]