摘要
目的为了推求出精度更高的城市暴雨强度参数.方法依据遗传算法全局寻优的原理与特点,提出采用遗传算法对皮尔逊-Ⅲ(P-Ⅲ)型分布的统计参数进行估计的方法.并在编制某城市暴雨强度公式的过程中,利用该城市52年的雨量资料进行了实际验证和与其他方法对比分析.结果计算结果表明,总的绝对误差仅为0.02,远远小于其他推求方法,也远远小于国家标准规定;在相同拟合准则的条件下,由该方法得到的P-Ⅲ型分布曲线与实测暴雨强度资料拟合的精度最高.结论遗传算法用于P-Ⅲ型分布统计参数的估计具有原理直观、方法简便、精度高、通用性强和便于计算机实现等特点,可以作为编制暴雨强度公式的P-Ⅲ型分布理论频率曲线拟合的新方法.
According to the principle and character of genetic algorithm, which asks for best in this arithmetic, the method adopting genetic algorithm to estimate statistic parameter of Pearson Ⅲ distribution was put forward. And practices testifying and contrasting analysis with other methods were carried out, utilizing rainfall data of a certain city for fifty-two years, when rainstorm intensity formula of a certain city was compiled. The result demonstrated that the general absolute error was only 0.02, far less than that by other methods and far less than that by national standard by similar fit rule. The fit precision between the P Ⅲ distributing curve gained from this method and real rainfall data, was more exact than others. The estimation has many characters, such as a simple principle, brief methods, high precision, better currency and easiness to control by computer. It can be regarded as a new method to compile PⅢ distributing theory frequency curve of rainstorm intensity formula.
出处
《沈阳建筑大学学报（自然科学版）》
CAS
2006年第2期299-302,共4页
Journal of Shenyang Jianzhu University：Natural Science
基金
辽宁省自然基金项目(20072005)
沈阳市建设委员会基金项目
关键词
遗传算法
暴雨强度公式
皮尔逊-Ⅲ型分布
统计参数
拟合准则
genetic algorithm
rainstorm intensity formula
pearson Ⅲ distribution
statistic parameter
simi-lar fit rule
