CN-121981512-A - River basin ecological flow early warning analysis method based on geographic knowledge graph

CN121981512ACN 121981512 ACN121981512 ACN 121981512ACN-121981512-A

Abstract

The invention relates to the technical field of drainage basin ecological flow early warning, in particular to a drainage basin ecological flow early warning analysis method based on a geographic knowledge graph. According to the invention, through integrating multi-source data collection and adopting an advanced data grabbing technology, comprehensive and multidimensional data viewing angles are provided, the comprehensiveness and relativity of the data are enhanced, the characteristics are extracted by utilizing deep learning technologies such as a convolutional neural network, a recurrent neural network and the like, the accuracy and the efficiency of pattern recognition are improved, accurate anomaly detection and flow prediction are carried out, scientific risk assessment is carried out through cluster analysis and pattern recognition, geographic knowledge maps and decision support algorithm deep analysis data are constructed, comprehensive hole finding is provided, an early warning notice is issued by an integrated communication platform, information is rapidly transmitted, and powerful technical support is provided for ecological management of a river basin.

Inventors

ZHANG ZHIPENG
LI LONG
NIU LE
LI ZHAORUI
YUAN HE
XUE XIANGKUN
WU YAN

Assignees

山东乾元工程集团有限公司垦利分公司

Dates

Publication Date: 20260505
Application Date: 20231207

Claims (10)

1. The river basin ecological flow early warning analysis method based on the geographic knowledge graph is characterized by comprising the following steps of: based on multi-source data collection, adopting a data grabbing technology and an API integration method to acquire and integrate data, and generating an original data set; Based on the original data set, adopting a data cleaning algorithm and a data standardization technology to perform data preprocessing to generate a preprocessed data set; based on the preprocessing data set, adopting a deep learning technology comprising a convolutional neural network and a recurrent neural network to perform feature extraction to generate a feature data set; Based on the characteristic data set, performing abnormal mode analysis by adopting an abnormal detection method comprising an isolated forest algorithm to generate an abnormal detection report; Based on the characteristic data set, a long-term and short-term memory network is adopted to conduct flow change prediction, and a drainage basin flow prediction report is generated; Based on the anomaly detection report and the drainage basin flow prediction report, performing risk assessment by adopting a cluster analysis and pattern recognition technology to generate a risk assessment report; based on the characteristic data set and the risk assessment report, constructing a geographic knowledge graph by adopting a graph database technology, and generating the geographic knowledge graph; Based on the geographic knowledge graph and the risk assessment report, determining an early warning level and a response strategy by adopting a decision support algorithm, and generating an early warning response strategy; Based on the early warning response strategy, an integrated communication platform is adopted to issue early warning notification, and final early warning notification is generated; The original data set comprises meteorological data, hydrological data, geographic data and ecological data, the preprocessing data set is specifically data with invalid values, missing values and abnormal values removed and in a unified format, the characteristic data set comprises an identified mode and a trend, the abnormal detection report is specifically an identified abnormal mode, the river basin flow prediction report is specifically a future river basin ecological flow trend, the risk assessment report comprises a risk level of river basin ecological flow, the geographic knowledge graph integrates geographic, environmental and ecological data and provides data relation management and query, the early warning response strategy is specifically an action scheme aiming at multiple early warning levels, and the final early warning notification comprises early warning information and response measures.
2. The method for early warning and analyzing the ecological flow of the river basin based on the geographic knowledge graph according to claim 1, wherein the steps of acquiring and integrating data and generating an original data set are specifically as follows, based on multi-source data collection, by adopting a data grabbing technology and an API (application program interface) integration method: Based on multi-source data collection, adopting a web crawler technology to collect meteorological and hydrological data and generate a preliminary meteorological hydrological data set; based on a geographic information system and an ecological database, collecting geographic and ecological data by adopting an API (application program interface) integration method, and generating a preliminary geographic ecological data set; Based on the preliminary meteorological hydrological dataset and the preliminary geographic ecological dataset, carrying out data matching and time sequence synchronization by adopting a data fusion technology to generate a combined dataset; Based on the merged data set, data integrity and consistency are verified using a data verification technique, generating an original data set.
3. The method for early warning and analyzing the ecological flow of the river basin based on the geographical knowledge graph according to claim 1, wherein the step of preprocessing data based on the original data set by adopting a data cleaning algorithm and a data standardization technology to generate a preprocessed data set is specifically as follows: Based on the original data set, performing null value processing and abnormal value identification by using a data cleaning algorithm, and removing invalid and missing data to generate a cleaned data set; based on the cleaned data set, performing linear interpolation and time sequence interpolation by adopting a data interpolation method to generate a filled data set; Based on the filled data set, a data conversion technology is applied to perform data normalization and standardization processing to generate a standardized data set; Based on the standardized dataset, feature selection and dimension reduction are performed using principal component analysis, generating a preprocessed dataset.
4. The method for early warning and analyzing the ecological flow of the river basin based on the geographical knowledge graph according to claim 1, wherein based on the preprocessing data set, a deep learning technology comprising a convolutional neural network and a recurrent neural network is adopted to perform feature extraction, and the step of generating a feature data set is specifically as follows: based on the preprocessing data set, extracting spatial features by using a convolutional neural network, and generating a spatial feature data set; Based on the space feature data set, a recurrent neural network is applied to extract time sequence features, and a time sequence feature data set is generated; Based on the spatial feature data set and the time sequence feature data set, integrating the differentiated type features by adopting a feature fusion technology to generate a fusion feature data set; And optimizing feature selection by using an automatic feature selection algorithm based on the fused feature data set to generate a feature data set.
5. The method for forewarning and analyzing the ecological flow of the river basin based on the geographical knowledge graph according to claim 1, wherein based on the characteristic data set, an anomaly detection method comprising an isolated forest algorithm is adopted to analyze an anomaly mode, and the step of generating an anomaly detection report is specifically as follows: based on the characteristic data set, performing preliminary outlier detection by using an isolated forest algorithm, dividing a characteristic space to identify outliers, and generating a preliminary outlier detection result; based on the preliminary abnormal point detection result, performing abnormal point visual analysis by applying principal component analysis and t distribution random neighborhood embedding, and generating an abnormal point depth visual analysis chart; Based on the abnormal point depth visual analysis chart, carrying out association rule mining by using an Apriori algorithm and an FP-Growth algorithm, analyzing association modes of abnormal points and data, and generating an abnormal point association mode analysis report; And based on the abnormal point association mode analysis report, adopting a data aggregation technology and logistic regression analysis to perform data aggregate interpretation and generate an abnormal detection report.
6. The method for early warning and analyzing the ecological flow of the river basin based on the geographical knowledge graph according to claim 1, wherein the step of predicting the flow change and generating a flow prediction report of the river basin by adopting a long-term memory network based on the characteristic data set is specifically as follows: Based on the characteristic data set, applying a long-term and short-term memory network to perform flow time sequence analysis to generate a flow time sequence analysis result; Based on the flow time sequence analysis result, carrying out trend analysis by adopting seasonal decomposition and autoregressive moving average model to generate a trend analysis report; based on the trend analysis report, performing uncertainty analysis by using Monte Carlo simulation and a Bayesian network, and generating an analysis report; and integrating analysis results by using a data fusion and multiple linear regression technology based on the analysis report to generate a drainage basin flow prediction report.
7. The method for early warning and analyzing the ecological flow of the river basin based on the geographical knowledge graph as set forth in claim 1, wherein the step of performing risk assessment by using cluster analysis and pattern recognition technology based on the anomaly detection report and the river basin flow prediction report to generate a risk assessment report is specifically as follows: Based on the anomaly detection report and the drainage basin flow prediction report, performing risk pattern recognition by using a K-means and hierarchical clustering technology, and generating a risk pattern clustering result; Based on the risk pattern clustering result, performing risk feature analysis by using a support vector machine and a random forest algorithm, and generating a risk feature analysis report; based on the risk characteristic analysis report, performing risk level assessment by using a decision tree analysis and a logistic regression model to generate a risk level assessment report; and based on the risk level assessment report, applying decision matrix analysis and Monte Carlo simulation technology, integrating a risk analysis result, and generating a risk assessment report.
8. The method for early warning and analyzing the ecological flow of the river basin based on the geographical knowledge map according to claim 1, wherein the step of constructing the geographical knowledge map and generating the geographical knowledge map by adopting a map database technology based on the characteristic data set and the risk assessment report is specifically as follows: Based on the characteristic data set and the risk assessment report, carrying out data alignment and entity identification by using a data integration technology, integrating the river basin geography and environment data, and generating an integrated geography environment data set; Based on the integrated geographic environment data set, a graph data modeling method is applied to conduct graph structure design and node relation mapping, a data structure of a geographic knowledge graph is designed, and a geographic knowledge graph data model is generated; based on the geographic knowledge graph data model, constructing and storing a graph by using a graph database technology, and establishing nodes and relationships to generate a geographic knowledge graph database; And carrying out data query and map optimization based on the geographic knowledge map database, and carrying out data indexing and map updating to generate a geographic knowledge map.
9. The method for early warning and analyzing the ecological flow of the river basin based on the geographical knowledge graph according to claim 1, wherein the step of determining the early warning level and the response strategy by adopting a decision support algorithm based on the geographical knowledge graph and the risk assessment report and generating the early warning response strategy is specifically as follows: Based on the geographical knowledge graph and the risk assessment report, performing risk analysis and classification by using a decision tree algorithm, constructing a decision tree model and performing branch judgment to generate a preliminary risk classification result; Based on the preliminary risk classification result, adopting a Bayesian network, and carrying out risk probability assessment and prediction through network structure design and probability calculation to generate a risk probability assessment report; Based on the risk probability evaluation report, determining an early warning level by using multi-standard decision analysis, and generating an early warning level decision result; Based on the early warning level decision result, a specific response strategy is designed by using a scene planning and risk management model, emergency measure planning and long-term coping strategy formulation are carried out, and an early warning response strategy is generated.
10. The method for early warning and analyzing the ecological flow of the river basin based on the geographical knowledge graph according to claim 1, wherein the step of issuing the early warning notification by adopting an integrated communication platform based on the early warning response strategy to generate the final early warning notification is specifically as follows: Integrating early warning information by using an information arrangement technology based on the early warning response strategy, and carrying out message formatting and content arrangement to generate integrated early warning information; Based on the integrated early warning information, an information release system is used for preparing release of the early warning information, release channel selection and message typesetting are carried out, and an early warning information release preparation state is generated; based on the early warning information release preparation state, auditing the early warning information by adopting a content management system and an automatic verification algorithm to generate the audited early warning information; And based on the checked early warning information, issuing an early warning notice by using an integrated communication platform to generate a final early warning notice.

Description

River basin ecological flow early warning analysis method based on geographic knowledge graph Technical Field The invention relates to the technical field of river basin ecological flow early warning, in particular to a river basin ecological flow early warning analysis method based on a geographic knowledge graph. Background The core of the technical field of river basin ecological flow early warning is to ensure the health and sustainability of a aquatic ecological system in a river basin, wherein water flow is one of the most important factors in the ecological system of the river basin, not only affects the survival of aquatic organisms, but also affects the sediment movement, water quality and growth of aquatic plants of a river, and the technical application of the river basin ecological flow early warning comprises the steps of monitoring the change of water flow, predicting ecological risk and giving early warning when necessary so as to take measures to protect the ecological system. The method for early warning and analyzing the ecological flow of the river basin based on the geographic knowledge graph is a method for monitoring and analyzing the ecological flow of the river basin by using a geographic information system and a knowledge graph technology, and aims to predict and identify the flow change threatening the ecological system of the river basin by integrating and analyzing a large amount of geographic and environmental data about the river basin, and aims to realize early detection and effective early warning of the change of the ecological system of the river basin, in order to take timely protection measures and reduce negative effects on an ecological system, the method is usually combined with advanced data collection technology and data processing technology, can predict future flow change trend through deep analysis of data, identify potential risk areas and timely send out early warning, and provides scientific basis for decision makers and managers to take corresponding environmental protection measures. The traditional river basin ecological flow early warning method depends on a single and limited data source, lacks comprehensive and multidimensional view angles, lacks efficient algorithm support in the aspects of feature extraction and pattern recognition, is difficult to accurately and quickly recognize complex data patterns, lacks systematicness and scientificity in the aspect of risk assessment, and is difficult to comprehensively evaluate and predict potential risks. Disclosure of Invention The invention aims to solve the defects in the prior art, and provides a river basin ecological flow early warning analysis method based on a geographic knowledge graph. In order to achieve the purpose, the invention adopts the following technical scheme that the drainage basin ecological flow early warning analysis method based on the geographic knowledge graph comprises the following steps: s1, based on multi-source data collection, adopting a data grabbing technology and an API integration method to acquire and integrate data to generate an original data set; S2, based on the original data set, performing data preprocessing by adopting a data cleaning algorithm and a data standardization technology to generate a preprocessed data set; s3, based on the preprocessing data set, adopting a deep learning technology comprising a convolutional neural network and a recurrent neural network to perform feature extraction to generate a feature data set; s4, based on the characteristic data set, performing abnormal mode analysis by adopting an abnormal detection method comprising an isolated forest algorithm, and generating an abnormal detection report; s5, based on the characteristic data set, adopting a long-period memory network to conduct flow change prediction, and generating a drainage basin flow prediction report; S6, based on the anomaly detection report and the drainage basin flow prediction report, performing risk assessment by adopting a cluster analysis and pattern recognition technology to generate a risk assessment report; s7, constructing a geographic knowledge graph by adopting a graph database technology based on the characteristic data set and the risk assessment report, and generating the geographic knowledge graph; s8, determining an early warning level and a response strategy by adopting a decision support algorithm based on the geographic knowledge graph and the risk assessment report, and generating an early warning response strategy; S9, based on the early warning response strategy, issuing an early warning notice by adopting an integrated communication platform to generate a final early warning notice; The original data set comprises meteorological data, hydrological data, geographic data and ecological data, the preprocessing data set is specifically data with invalid values, missing values and abnormal values removed and in a unified format, the characteristic data set