CN-121544048-B - Water regime abnormal risk prediction method and system based on AI algorithm

CN121544048BCN 121544048 BCN121544048 BCN 121544048BCN-121544048-B

Abstract

The invention relates to the technical field of water regime prediction, in particular to a water regime abnormal risk prediction method and a water regime abnormal risk prediction system based on an AI algorithm, wherein the method comprises the steps of screening abnormal monitoring points from all monitoring points according to water level abnormal conditions of all the monitoring points; according to the water level change intensity and danger degree of each upstream monitoring point of the abnormal monitoring points in the abnormal period, the influence degree of upstream rainfall on the abnormal monitoring points in the abnormal period is determined by combining the distance between the upstream monitoring points and the abnormal monitoring points, and the water condition abnormal factors of the abnormal monitoring points are obtained by combining the lengths of the abnormal river segments where the abnormal monitoring points are located, and the current water levels of the monitoring points are fused by the water condition abnormal factors of the monitoring points to obtain the current water level of the river. Considering that different monitoring points are affected differently, the abnormal water condition conditions of the monitoring points are fused, the data reliability of the river water level can be ensured, the accuracy of the river water level is improved, and the accuracy of water condition risk early warning is improved.

Inventors

LI YAQIANG
SHANG WEI
LIU TAIXIN
ZHANG JI

Assignees

特变电工国际工程有限公司

Dates

Publication Date: 20260508
Application Date: 20260116

Claims (9)

1. The water condition abnormal risk prediction method based on the AI algorithm is characterized by comprising the following steps of: according to the abnormal water level conditions of all monitoring points, abnormal monitoring points are obtained through screening from all monitoring points; Determining the influence degree of upstream rainfall on the abnormal monitoring points in the abnormal period by combining the distance between the abnormal monitoring points and the influence of the water level change intensity and the danger degree of each upstream monitoring point of the abnormal monitoring points on the abnormal monitoring points in the abnormal period; Fusing the influence degree of the abnormal monitoring point, and combining the length of the abnormal river section where the abnormal monitoring point is positioned to obtain a water condition abnormal factor of the abnormal monitoring point; fusing the current water levels of the monitoring points by the water condition abnormal factors of the monitoring points to obtain the current water level of the river; The screening process of the abnormal monitoring points comprises the steps of obtaining the water choking degree of each monitoring point in each abnormal period according to the water level abnormal degree of each monitoring point in each abnormal period and the time length of each abnormal period, wherein the water choking degree is positively correlated with the water level abnormal degree and the time length, combining the water choking degree of each monitoring point in all abnormal periods and the gradient and the river channel width of each monitoring point to obtain the dangerous degree of each monitoring point, wherein the dangerous degree is positively correlated with the water choking degree and the gradient and inversely correlated with the river channel width, and screening the abnormal monitoring points from the monitoring points according to the dangerous degree of each monitoring point.
2. The water condition abnormal risk prediction method based on the AI algorithm as claimed in claim 1, wherein the acquiring process of the abnormal period comprises the following steps: acquiring the water level of the monitoring point at each moment in the historical time period, taking the moment corresponding to the water level which is larger than the water level safety threshold value as an abnormal moment, and forming the continuous abnormal moment into an abnormal time period of the monitoring point; the process for acquiring the water level abnormality degree in the abnormality period comprises the following steps: Acquiring the ultra-safe water level at each abnormal moment in the abnormal period, wherein the ultra-safe water level is the difference value between the water level and the water level safety threshold value; And performing straight line fitting on the ultra-safe water levels at different abnormal moments in the abnormal period, and taking the slope of the obtained fitting straight line as the abnormal degree of the water level in the abnormal period.
3. The water condition abnormal risk prediction method based on the AI algorithm as claimed in claim 1, wherein the water level change severity obtaining process comprises the following steps: The method comprises the steps of obtaining the peak quantity ratio, the average value of peak-to-peak values and the average value of peak time intervals in a water level sequence of a target rainfall period, wherein the target rainfall period is a rainfall period associated with an abnormal period, the water level sequence is composed of water levels at all moments in the target rainfall period, the peak-to-peak value is a water level difference value between a peak and a trough closest to the peak in the water level sequence, and the peak time interval is a time interval between two adjacent peaks; And obtaining the intensity of the water level change in the abnormal period according to the peak quantity ratio, the average value of the peak and the peak time interval and the average value of the peak and the peak time interval, wherein the intensity of the water level change is positively correlated with the peak quantity ratio and the average value of the peak and the peak time interval and inversely correlated with the average value of the peak time interval.
4. The water condition abnormal risk prediction method based on the AI algorithm as claimed in claim 1, wherein the obtaining process of the influence degree comprises the following steps: obtaining an influence index corresponding to each upstream monitoring point according to the water level change intensity and the danger degree of each upstream monitoring point of the abnormal monitoring point in the abnormal period and the distance between the upstream monitoring point and the abnormal monitoring point, wherein the influence index is positively correlated with the water level change intensity and the danger degree and inversely correlated with the distance; and fusing the influence indexes corresponding to all upstream monitoring points of the abnormal monitoring points to obtain the influence degree of the abnormal monitoring points by the upstream rainfall in the abnormal period.
5. The water condition abnormality risk prediction method based on the AI algorithm as claimed in claim 1, wherein the water condition abnormality factor obtaining process includes: Obtaining hydrologic bearing capacity of the abnormal monitoring point according to the influence degree of the abnormal monitoring point in each abnormal period and the time interval between adjacent abnormal periods, wherein the hydrologic bearing capacity is inversely related to the influence degree and positively related to the time interval between adjacent abnormal periods; Obtaining a water condition abnormal factor of the abnormal monitoring point according to the difference of the water condition bearing capacity of the abnormal monitoring point and the length of the abnormal river section where the abnormal monitoring point is located, wherein the water condition abnormal factor is positively related to the difference and the length of the abnormal river section where the abnormal monitoring point is located, the difference is the difference value between the overall water condition bearing capacity level and the water condition bearing capacity of the abnormal monitoring point, and the overall water condition bearing capacity level is the water condition bearing capacity average value of all the abnormal monitoring points in the abnormal river section where the abnormal monitoring point is located.
6. The water condition abnormal risk prediction method based on the AI algorithm as claimed in claim 1, wherein the obtaining process of the current water level of the river comprises the following steps: obtaining the water level influence weight of each monitoring point according to the water condition abnormal factors of each monitoring point; and carrying out weighted summation on the current water level of each monitoring point according to the water level influence weight of each monitoring point to obtain the current water level of the river.
7. The AI algorithm-based water condition abnormality risk prediction method of claim 6, wherein among the water condition abnormality factors of the monitoring points, the water condition abnormality factors of the other monitoring points except for the abnormality monitoring point are set to a preset value, and the preset value is smaller than the water condition abnormality factor of the abnormality monitoring point.
8. The AI algorithm-based water condition abnormal risk prediction method of claim 1, wherein the abnormal river segment is composed of continuous abnormal monitoring points.
9. The water condition abnormal risk prediction system based on the AI algorithm is characterized by comprising a memory and a processor, wherein the memory is connected with the processor and is used for storing program instructions, and the processor is used for realizing the water condition abnormal risk prediction method based on the AI algorithm according to any one of claims 1-8 when the program instructions are executed.

Description

Water regime abnormal risk prediction method and system based on AI algorithm Technical Field The invention relates to the technical field of water regime prediction, in particular to a water regime abnormal risk prediction method and system based on an AI algorithm. Background Through river water level prediction, the flood disasters can be judged in advance, and instructive data is provided for disaster early warning and disaster prevention and control. The abnormal water condition risk prediction refers to predicting the water level at the future moment by analyzing and predicting the abnormal condition of the water body change (such as the water level, etc.), so as to predict the risk according to the predicted water level. When the current method utilizes the real-time water level of the river to predict the water level, the water level of the river is obtained by directly taking the average value of the water levels of all monitoring points in the river as the water level of the river. However, because the positions of the different monitoring points in the river are different (such as upstream, midstream, downstream, etc.), the influence on the different monitoring points is different, such as the time delay of the water level change trend among the different monitoring points and the influence degree of rainfall in the upstream area on each monitoring point are different, the hydrologic bearing capacity of each monitoring point is different, if the average value of the water level of each monitoring point in the river is directly used as the water level of the river, the influence on the water level of the river is not considered, the data reliability of the river is influenced, the water level of the river is inaccurate, and the accuracy of the water condition risk early warning is reduced. Disclosure of Invention In order to solve the technical problem of low data reliability of the current river water level, the invention aims to provide a water condition abnormal risk prediction method and system based on an AI algorithm, and the adopted technical scheme is as follows: in a first aspect of the present invention, there is provided a water condition abnormal risk prediction method based on an AI algorithm, including: according to the abnormal water level conditions of all monitoring points, abnormal monitoring points are obtained through screening from all monitoring points; Determining the influence degree of upstream rainfall on the abnormal monitoring points in the abnormal period by combining the distance between the abnormal monitoring points and the influence of the water level change intensity and the danger degree of each upstream monitoring point of the abnormal monitoring points on the abnormal monitoring points in the abnormal period; Fusing the influence degree of the abnormal monitoring point, and combining the length of the abnormal river section where the abnormal monitoring point is positioned to obtain a water condition abnormal factor of the abnormal monitoring point; And fusing the current water levels of the monitoring points by the water condition abnormal factors of the monitoring points to obtain the current water level of the river. In an exemplary embodiment, the screening process of the abnormal monitoring points includes: According to the water level abnormality degree of each monitoring point in each abnormal period, combining the time length of each abnormal period to obtain the water choking degree of each monitoring point in each abnormal period, wherein the water choking degree is positively correlated with the water level abnormality degree and the time length; Fusing the water choking degree of each monitoring point in all abnormal time periods, and combining the gradient of each monitoring point and the width of the river channel to obtain the dangerous degree of each monitoring point, wherein the dangerous degree is positively correlated with the water choking degree and the gradient and inversely correlated with the width of the river channel; and screening abnormal monitoring points from the monitoring points according to the dangerous degree of the monitoring points. In an exemplary embodiment, the acquiring of the abnormal period includes: acquiring the water level of the monitoring point at each moment in the historical time period, taking the moment corresponding to the water level which is larger than the water level safety threshold value as an abnormal moment, and forming the continuous abnormal moment into an abnormal time period of the monitoring point; the process for acquiring the water level abnormality degree in the abnormality period comprises the following steps: Acquiring the ultra-safe water level at each abnormal moment in the abnormal period, wherein the ultra-safe water level is the difference value between the water level and the water level safety threshold value; And performing straight line fitting on the ultra-safe water levels at diff