摘要
针对月平均降水量时间序列存在模糊性和随机性的特点,给出一种结合云推理和模糊逻辑关系的干旱等级预测方法。采用徐州站1951-2014年逐月降水量数据,通过计算标准化降水指数（SPI）,得到实测干旱等级序列,以1952-2013年SPI指数数据作为样本数据,并提取样本模糊时间序列的52条模糊逻辑推理规则,将某月份的SPI指数数据作为输入变量,利用云发生器进行云推理,得到未来相应月份的干旱等级预测结果。结果表明,研究方法对干旱发生具有一定预测能力,尤其是对无旱的预测比较准确,但是对于干旱状态突变的预测能力较弱,主要是由于发生严重旱灾的可能性较少,导致模糊推理规则较少。因此,对于江苏省以轻旱为主的苏北地区,可以作为早期干旱预警的参考。
For the monthly average precipitation time series having the features of the fuzziness and randomness, a drought class prediction combination method based on the cloud reasoning and fuzzy logic relation are given. By calculating the standardized precipitation index（SPI） of Xuzhou station monthly precipitation data from 1951 to 2014, we get the measured drought grade sequence. We use SPI data from1952 to 2013 as the sample data, and extract 52 fuzzy logic inference rules of the fuzzy time series. The SPI data of a month is used as the input variable, and the cloud generator is used to obtain the forecast result of the drought level in the future. The results show that this method has certain prediction ability for drought occurrence, especially for no drought prediction, which is more accurate. But it is weak for the drought state mutations prediction, mainly due to the possibility of a severe drought is less, resulting in a less fuzzy inference rules.Therefore, it can be used as a reference for early warning of drought in northern Jiangsu Province, which is given priority to the light drought.
出处
《长江流域资源与环境》
CAS
CSSCI
CSCD
北大核心
2016年第8期1273-1278,共6页
Resources and Environment in the Yangtze Basin
基金
国家自然科学基金项目(71303074)
江苏省社会科学基金项目(14GLC004)
江苏省自然科学基金项目(BK20130242)
中央高校基本科研业务费项目(2015B23914
2015B28014)~~
