基于信息熵和统计参数的径流随机模拟序列检验方法
|
摘要
传统的径流模拟序列检验方法是基于统计参数开展的,但由于径流具有高度的非线性、耗散性和复杂性,传统的检验方法虽能反映径流模拟序列和原始径流序列在各截口统计参数的吻合程度,但却忽视了径流作为一种时间序列的变化特征。因此,引入样本熵、小波熵概念,提出基于信息熵和统计参数的径流随机模拟序列检验方法,作为对传统检验方法的补充。以淮河支流沙河上漯河水文站月径流模拟系列为例进行应用研究,结果表明该方法可发现径流模拟序列中的较大月径流量、较大月径流位置的异常情况,筛选出精度高的模拟序列,为检验模拟序列精度提供了一条新途径。
The traditional runoff simulation sequence test method is based on the statistical parameters.Due to the high nonlinearity,dissipation and complexity of runoff,the traditional test method can reflect the coincidence degree between the runoff simulation sequence and the original runoff sequence at each time,but it neglects the characteristics of changes as a time series for the runoff.In this paper,the concept of sample entropy and wavelet entropy were introduced,and a runoff stochastic simulation sequence test method based on information entropy and statistical parameters was proposed as a supplement to the traditional inspection methods.The monthly runoff simulation series of the Luohe Hydrological Station in the Shahe River,one branch of the Huaihe River,were used as an example.The results show that the proposed method can detect the larger monthly flow in the runoff simulation sequence and the abnormality of the location of the larger monthly runoff and select a high-precision simulation sequence,which provides a new way for verifying the accuracy of the simulation sequence.
The traditional runoff simulation sequence test method is based on the statistical parameters.Due to the high nonlinearity,dissipation and complexity of runoff,the traditional test method can reflect the coincidence degree between the runoff simulation sequence and the original runoff sequence at each time,but it neglects the characteristics of changes as a time series for the runoff.In this paper,the concept of sample entropy and wavelet entropy were introduced,and a runoff stochastic simulation sequence test method based on information entropy and statistical parameters was proposed as a supplement to the traditional inspection methods.The monthly runoff simulation series of the Luohe Hydrological Station in the Shahe River,one branch of the Huaihe River,were used as an example.The results show that the proposed method can detect the larger monthly flow in the runoff simulation sequence and the abnormality of the location of the larger monthly runoff and select a high-precision simulation sequence,which provides a new way for verifying the accuracy of the simulation sequence.
引文
[1]丁晶,邓育仁.随机水文学[M].四川:成都科技大学出版社,1988.
[2]胡庆芳,王银堂,刘克琳,等.基于改进的两参数月水量平衡模型的月径流模拟[J].河海大学学报:自然科学版,2007,35(6):638-642.
[3]张勃,王海青,张华.基于人工神经网络的莺落峡月径流模拟预测[J].自然资源学报,2009,24(12):2 169-2 177.
[4]谢萍萍,宋松柏.非参数解集模型在月径流随机模拟中的应用[J].人民黄河,2010,32(8):36-38,42.
[5]闫宝伟,郭生练,刘攀,等.基于Copula函数的径流随机模拟[J].四川大学学报:工程科学版,2010,42(1):5-9.
[6]夏军.水问题的复杂性与不确定性研究与进展[M].北京:中国水利水电出版社,2004.
[7]王磊,颜威利,杨硕,等.基于样本熵的精神任务脑电信号分类研究[J].微计算机信息,2008,24(30):199-201.
[8]郑旭,盛立辉,崔宵语.基于小波熵的时间序列分段聚合近似表示[J].计算机仿真,2015,32(1):411-415.
[9]王浩,严登华,杨大文,等.水文学方法研究[M].北京:科学出版社,2012.
[10]邹阳,苗夺谦,王登.醉酒者和正常人脑电的样本熵研究[J].中国生物医学工程学报,2010,29(6):939-942.
[11]马细霞,穆浩泽,郭慧芳.基于小波-ANFIS的水库月径流预报模型[J].水电能源科学,2008,26(1):26-29,74.
[2]胡庆芳,王银堂,刘克琳,等.基于改进的两参数月水量平衡模型的月径流模拟[J].河海大学学报:自然科学版,2007,35(6):638-642.
[3]张勃,王海青,张华.基于人工神经网络的莺落峡月径流模拟预测[J].自然资源学报,2009,24(12):2 169-2 177.
[4]谢萍萍,宋松柏.非参数解集模型在月径流随机模拟中的应用[J].人民黄河,2010,32(8):36-38,42.
[5]闫宝伟,郭生练,刘攀,等.基于Copula函数的径流随机模拟[J].四川大学学报:工程科学版,2010,42(1):5-9.
[6]夏军.水问题的复杂性与不确定性研究与进展[M].北京:中国水利水电出版社,2004.
[7]王磊,颜威利,杨硕,等.基于样本熵的精神任务脑电信号分类研究[J].微计算机信息,2008,24(30):199-201.
[8]郑旭,盛立辉,崔宵语.基于小波熵的时间序列分段聚合近似表示[J].计算机仿真,2015,32(1):411-415.
[9]王浩,严登华,杨大文,等.水文学方法研究[M].北京:科学出版社,2012.
[10]邹阳,苗夺谦,王登.醉酒者和正常人脑电的样本熵研究[J].中国生物医学工程学报,2010,29(6):939-942.
[11]马细霞,穆浩泽,郭慧芳.基于小波-ANFIS的水库月径流预报模型[J].水电能源科学,2008,26(1):26-29,74.