This data set is the weight data formed by the model solution and collation of the prediction factors of river flow, river closure and river opening of Bayan Gaole hydrological station in the Ningmeng section of the Yellow River from 1986 to 2020. This data set contains three Excel files, which are composed of the weight elements of the prediction factors in different periods of Bayan Gaole hydrological station, namely: the weight data of the prediction factors on the date of river flow, the weight data of the prediction factors on the date of river closure and the weight data of the prediction factors on the date of river opening.
The source data was obtained through consultation with the staff of the hydrology bureau of the Ningmeng reach of the Yellow River.
The researchers of relevant projects of China Academy of water resources and hydropower research obtained the data by solving the model. The solution is as follows:
(1) Taking the influence factor matrix as the input vector of BP network and the ice condition result as the target output vector, BP neural network is trained by L-M algorithm to obtain the weight matrix w of input layer and hidden layer and the weight vector w of hidden layer and output layer.
(2) Calculate the overall weight vector ω=| W | * | w |, where ω= （ ω_ 1， ω_ 2，…， ω_ n）。
(3) Calculate the direct relationship matrix B of each influencing factor index
(4) Normalized direct correlation matrix X
(5) Calculate the full incidence matrix t
(6) Establish a cause and effect diagram. Define D as the sum of the rows of T, and R as the sum of the columns of T. D+r is defined as the centrality of index I. the larger it is, the greater the importance and role of this index in the influencing factor system. D − R is defined as the cause degree of indicator I, which can be used to distinguish the cause group and the result group. If the D − r of indicator I is greater than 0, this indicator belongs to the cause group. If the indicator D − R is less than 0, this indicator belongs to the result group. Among many influencing factors, the influencing factor in the result group is the influence result of the influencing factor in the cause group. Define the comprehensive importance degree, which can reflect the center degree and cause degree at the same time.
Source of model: Cui Qiang, Wu Chunyou, Kuang Haibo Application of bp-dematel in the identification of influencing factors of airport competitiveness [j] System engineering theory and practice, 2013,33 (6):1471-1478
The solver is a professional. It has been checked many times and the data quality is good.
This work is licensed under a
Creative
Commons Attribution 4.0 International License.
CSTR:11738.11.ncdc.iceflood.db2371.2022
10.12072/ncdc.iceflood.db2371.2022