CN-121997280-A - Multi-source data coupling system and method

CN121997280ACN 121997280 ACN121997280 ACN 121997280ACN-121997280-A

Abstract

The invention relates to the technical field of multi-source data processing, in particular to a multi-source data coupling system and method, comprising the steps of acquiring original monitoring data of a target area from a plurality of data sources, carrying out quality evaluation on the data of each data source and generating quality scores; the method comprises the steps of establishing a data source contribution degree dynamic weight distribution model by combining with geological activity stage parameters, calculating real-time dynamic contribution weights of all data sources, weighting and fusing original data, introducing geological priori knowledge to conduct self-adaptive feature screening on fused data, reserving key features and carrying out collaborative analysis with the geological priori knowledge to obtain a geological state coupling result, constructing a closed loop optimization mechanism, calibrating data quality scores by using the coupling result, updating the weight model, and fusing the next period. The method can adapt to dynamic change of geological activity, improves reliability of fusion data, realizes adaptive coupling optimization of multi-source monitoring data, and adapts to complex geological monitoring scenes.

Inventors

WANG ZHENGCHAO
WANG JUNJIE
ZHANG GUANGXING
ZHANG XIAOCHEN
HUANG QINGKANG
PENG LEIXIANG
ZHAO GUORUI
SUN SHUO
ZHANG LEI
GAO QIUYE
PENG PENG
JU WEIJUN
LI ZHONGQIAN

Assignees

通用技术集团工程设计有限公司

Dates

Publication Date: 20260508
Application Date: 20260409

Claims (10)

1. A method of multi-source data coupling, the method comprising: Acquiring original monitoring data of a target area from a plurality of data sources, constructing an original monitoring data quality evaluation unit, performing quality evaluation on the original monitoring data of each data source, and generating a data quality score of each data source; based on the data quality scores of the data sources, combining with the geological activity stage parameters of the target area, establishing a data source contribution degree dynamic weight distribution model, and calculating to obtain the real-time dynamic contribution weight of each data source; According to the real-time dynamic contribution weight, carrying out weighted fusion on the original monitoring data of each data source to generate a weighted fusion monitoring data set; introducing geological priori knowledge of a target area, including lithology distribution characteristics and historical deformation rules, and executing self-adaptive characteristic screening processing on the weighted fusion monitoring data set; The key feature monitoring data is reserved through self-adaptive feature screening processing, and a screened feature monitoring data set is formed; carrying out collaborative analysis on the screened feature monitoring dataset and the geological priori knowledge to generate a geological state coupling analysis result of a target area; Constructing a closed-loop optimization mechanism, recalibrating the data quality scores of all the data sources by using the geological state coupling analysis result, and updating the data source contribution dynamic weight distribution model; and using the updated data source contribution dynamic weight distribution model for the weighted fusion of the original monitoring data of the next period, thereby realizing the self-adaptive coupling optimization of the multi-source monitoring data.
2. The method for coupling multi-source data according to claim 1, wherein the acquiring the original monitoring data of the target area from the plurality of data sources specifically comprises: The original monitoring data comprise terrain variation data, geological deformation time sequence data, lithology distribution data and historical activity record data; Obtaining terrain variation data of a target area from a first type of data source, wherein the terrain variation data comprises digital elevation model difference data acquired at different time points; obtaining geological deformation time sequence data of a target area from a second type of data source, wherein the geological deformation time sequence data is a surface displacement time sequence continuously acquired through a sensor network; Acquiring lithology distribution data of a target area from a third type of data source, wherein the lithology distribution data are rock type spatial distribution diagrams obtained through geological exploration; Acquiring historical activity record data of a target area from a fourth type of data source, wherein the historical activity record data comprises time-space information of historical earthquake events, landslide events and fault activity records; Registering and aligning the topographic variation data, the geological deformation time sequence data, the lithology distribution data and the historical activity record data according to the unified space-time reference to form an initial original monitoring data set.
3. The method for coupling multi-source data according to claim 2, wherein the constructing an original monitoring data quality evaluation unit performs quality evaluation on the original monitoring data of each data source to generate a data quality score of each data source, specifically: for the topographic variation data, evaluating the spatial resolution, the time span coverage integrity and the elevation measurement precision of the topographic variation data to generate topographic data quality item scores; Aiming at the geological deformation time sequence data, evaluating the continuity, the missing value proportion and the sensor acquisition noise level of the time sequence, and generating deformation data quality item scores; aiming at lithology distribution data, evaluating the exploration point density, lithology classification and spatial interpolation uncertainty of the lithology distribution data, and generating lithology data quality score; Aiming at the historical activity record data, evaluating the completeness, space-time positioning and reliability of the event intensity description of the record event to generate a historical data quality score; And integrating the quality item scores corresponding to each data source, and calculating to obtain the integrated data quality score of each data source through a preset score aggregation algorithm.
4. The multi-source data coupling method according to claim 3, wherein the step of establishing a data source contribution degree dynamic weight distribution model based on the data quality scores of the data sources and in combination with the geological activity stage parameters of the target area, and calculating to obtain the real-time dynamic contribution weight of each data source is specifically as follows: Obtaining a geological activity stage parameter representing the current geological activity intensity of a target area; Setting a basic weight factor of a data source contribution dynamic weight distribution model, wherein the basic weight factor and the comprehensive data quality score of each data source form a positive correlation; Introducing an adjusting coefficient of a geological activity stage parameter into the data source contribution dynamic weight distribution model, and increasing a basic weight factor of a data source sensitive to time sequence change when the geological activity stage parameter indicates that the activity is aggravated, otherwise, reducing; And (3) performing simultaneous calculation on the comprehensive data quality scores, the basic weight factors and the adjustment coefficients of the geological activity stage parameters to obtain the real-time dynamic contribution weights of each data source in the coupling analysis process, and ensuring that the sum of the real-time dynamic contribution weights of all the data sources is one.
5. The method for coupling multi-source data according to claim 4, wherein the raw monitoring data of each data source is weighted and fused according to the real-time dynamic contribution weight to generate a weighted and fused monitoring data set, specifically: reading real-time dynamic contribution weights corresponding to the topographic variation data, the geological deformation time sequence data, the lithology distribution data and the historical activity record data; Multiplying the numerical value of each space grid unit in the terrain change data by the corresponding real-time dynamic contribution weight to obtain weighted terrain change data; multiplying the numerical value of each time sequence data point in the geological deformation time sequence data by the corresponding real-time dynamic contribution weight to obtain weighted geological deformation time sequence data; Multiplying each lithology type distribution probability in lithology distribution data by the corresponding real-time dynamic contribution weight to obtain weighted lithology distribution data; multiplying the intensity and the influence factor of each historical event in the historical activity record data by the corresponding real-time dynamic contribution weight to obtain weighted historical activity record data; And superposing and spatially fusing all weighted terrain change data, geological deformation time sequence data, lithology distribution data and historical activity record data to generate a unified weighted fusion monitoring data set.
6. The multi-source data coupling method according to claim 5, wherein the introducing of the geological priori knowledge of the target region includes lithology distribution characteristics and historical deformation rules, and the performing of the adaptive feature screening process on the weighted fusion monitoring dataset is specifically: Calling lithology distribution characteristics of a target area from a geological priori knowledge base, wherein the lithology distribution characteristics identify spatial distribution of different rock types and physical and mechanical properties thereof; Invoking a historical deformation rule of the target region from a geological priori knowledge base, wherein the historical deformation rule describes a typical deformation mode and spatial distribution characteristics of the target region in past geological activities; Setting an initial global threshold value of the adaptive feature screening; combining the lithology distribution characteristics, dynamically adjusting the initial global threshold value by weighting and fusing monitoring data of different lithology subareas, adopting a low characteristic screening threshold value in a lithology unstable region and adopting a high characteristic screening threshold value in a stable region; combining the historical deformation rule, and in the area with active deformation in the history, carrying out weighted fusion on the monitoring characteristics conforming to the historical deformation mode in the monitoring data, and further reducing the characteristic screening threshold value to reserve the key points; And screening each monitoring characteristic data in the weighted fusion monitoring data set based on the dynamically adjusted characteristic screening threshold value, filtering out the data lower than the threshold value as noise or secondary characteristics, and reserving the data higher than the threshold value as key characteristic monitoring data to form a screened characteristic monitoring data set.
7. The multi-source data coupling method according to claim 6, wherein the screened feature monitoring dataset and the geological priori knowledge are cooperatively analyzed to generate a geological state coupling analysis result of the target area, specifically: carrying out space association analysis on the screened characteristic monitoring dataset and lithology distribution characteristics in geological priori knowledge, and identifying the correlation of the monitoring characteristics and specific lithology in space; Performing space-time pattern matching on the screened feature monitoring dataset and a historical deformation rule in geological priori knowledge, and judging whether the current monitoring feature accords with a known historical deformation pattern or not; the result of space correlation analysis and space-time pattern matching is synthesized, and the potential intensities of geological activities of different space positions in the target area are estimated; Based on the spatial distribution of the potential intensity of the geological activity, the spatial-temporal evolution trend of the characteristic values in the characteristic monitoring dataset is combined after screening, and a geological state coupling analysis result comprising the position of a potential geological risk area, the risk level and the activity mode prediction is generated.
8. The method for coupling multi-source data according to claim 7, wherein the constructing a closed-loop optimization mechanism uses the geologic state coupling analysis result to recalibrate the data quality scores of the data sources, and updates the data source contribution dynamic weight distribution model specifically as follows: in the geological state coupling analysis result, identifying a high-confidence analysis area which is confirmed to be accurate by a follow-up field check or independent monitoring means; backtracking the role played by the original monitoring data provided by each data source in the coupling analysis process of the high-confidence analysis area; correcting the initial comprehensive data quality score according to the analysis contribution of each data source in the high-confidence analysis area, wherein the data source with large contribution improves the data quality score, and the data source with small contribution or providing interference information reduces the data quality score; Recalculating a basic weight factor in the dynamic weight distribution model of the contribution degree of the data sources by using the data quality scores of the corrected data sources; substituting the recalculated basic weight factors into the data source contribution degree dynamic weight distribution model to generate an updated data source contribution degree dynamic weight distribution model for carrying out weighted fusion on the original monitoring data which are newly acquired subsequently.
9. The method for coupling multi-source data according to claim 8, further comprising the step of uncertainty quantization and transmission of the original monitoring data prior to the weighted fusion, in particular: for the original monitoring data of each data source, based on the data acquisition principle, the processing flow and the data quality score obtained by evaluation, quantifying the data uncertainty of the original monitoring data, and generating an uncertainty spatial distribution diagram of each data source; When the original monitoring data of each data source are subjected to weighted fusion, the uncertainty spatial distribution diagram corresponding to each data source is synchronously subjected to weighted fusion according to the respective real-time dynamic contribution weight; generating a post-fusion uncertainty distribution map spatially corresponding to the weighted fusion monitoring dataset; and when the self-adaptive feature screening processing is executed, taking the fused uncertainty distribution diagram as an auxiliary criterion, adopting a stricter feature screening threshold value for the monitoring features from the high uncertainty region and adopting a relatively loose feature screening threshold value for the monitoring features from the low uncertainty region.
10. A multi-source data coupling system comprising a memory, a processor and a computer program stored in the memory and running on the processor, characterized in that the processor, when executing the computer program, realizes the steps of a multi-source data coupling method according to any of the preceding claims 1 to 9.

Description

Multi-source data coupling system and method Technical Field The invention relates to the technical field of multi-source data processing, in particular to a multi-source data coupling system and a multi-source data coupling method. Background In the field of geological monitoring, multi-source data coupling is an important means for realizing accurate analysis of geological states of a target area, a technical scheme of multi-source monitoring data fusion is commonly adopted in the industry at present, original monitoring data are obtained from a plurality of data sources, and after simple weighted fusion, analysis is performed by combining with relevant geological information, so that a geological state evaluation result is obtained. In the prior art, the weight distribution mostly adopts a mode of presetting fixed weights, or determines the contribution weights of all data sources according to a single data quality index, and the influence of dynamic changes of geological activity stages of a target area on the effectiveness of the data sources is not considered. The prior art scheme has the defects that the fixed weight or single index weight distribution mode cannot adapt to monitoring requirements of different stages of geological activities, so that the effect of part of high-value data sources is not fully exerted, the interference of low-quality data sources is difficult to effectively inhibit, the reliability of fusion data is influenced, meanwhile, the data coupling process is mostly a unidirectional processing flow, the result after fusion analysis cannot be reversely acted on data quality evaluation and weight distribution, the coupling effect cannot be optimized in an iterative manner along with a monitoring period, complex and changeable geological monitoring scenes are difficult to adapt, and the self-adaptive coupling of multi-source data cannot be realized. Disclosure of Invention The invention aims to solve the defects in the prior art and provides a multi-source data coupling system and a multi-source data coupling method. In order to achieve the above purpose, the invention adopts the following technical scheme that the multi-source data coupling method comprises the following steps: Acquiring original monitoring data of a target area from a plurality of data sources, constructing an original monitoring data quality evaluation unit, performing quality evaluation on the original monitoring data of each data source, and generating a data quality score of each data source; based on the data quality scores of the data sources, combining with the geological activity stage parameters of the target area, establishing a data source contribution degree dynamic weight distribution model, and calculating to obtain the real-time dynamic contribution weight of each data source; According to the real-time dynamic contribution weight, carrying out weighted fusion on the original monitoring data of each data source to generate a weighted fusion monitoring data set; introducing geological priori knowledge of a target area, including lithology distribution characteristics and historical deformation rules, and executing self-adaptive characteristic screening processing on the weighted fusion monitoring data set; The key feature monitoring data is reserved through self-adaptive feature screening processing, and a screened feature monitoring data set is formed; carrying out collaborative analysis on the screened feature monitoring dataset and the geological priori knowledge to generate a geological state coupling analysis result of a target area; Constructing a closed-loop optimization mechanism, recalibrating the data quality scores of all the data sources by using the geological state coupling analysis result, and updating the data source contribution dynamic weight distribution model; and using the updated data source contribution dynamic weight distribution model for the weighted fusion of the original monitoring data of the next period, thereby realizing the self-adaptive coupling optimization of the multi-source monitoring data. As a further aspect of the present invention, the acquiring the original monitoring data of the target area from the plurality of data sources specifically includes: The original monitoring data comprise terrain variation data, geological deformation time sequence data, lithology distribution data and historical activity record data; Obtaining terrain variation data of a target area from a first type of data source, wherein the terrain variation data comprises digital elevation model difference data acquired at different time points; obtaining geological deformation time sequence data of a target area from a second type of data source, wherein the geological deformation time sequence data is a surface displacement time sequence continuously acquired through a sensor network; Acquiring lithology distribution data of a target area from a third type of data source, wherein the li