CN-121706031-B - Water quality analysis method and system for water receiving lake in diversion engineering by combining multi-source data fusion

Abstract

The invention provides a water quality analysis method and a system for a water receiving lake in a diversion project by combining multi-source data fusion, which relate to the technical field of water resource management and analysis, and are characterized in that field basic data are collected and a water quality association field is constructed; the method comprises the steps of carrying out interaction relation modeling operation on a field unit to obtain a field unit interaction model, carrying out dynamic coupling fusion processing on field basic data based on the field unit interaction model to generate coupling water quality data, combining water quality associated field change trend data with the coupling water quality data to deduce a water quality trend evolution path of a water receiving lake, and finally identifying a core interaction unit and combining the trend evolution path to generate a diversion engineering water quality dynamic regulation scheme. The invention can comprehensively analyze the water quality of the water-bearing lake, realize accurate dynamic regulation and control and ensure stable water quality.

Inventors

• GUO DANDAN
• LUO CHI
• YUAN MIN
• LI JIAPING
• DU PANPAN
• YIN MING
• YU FANGQUAN
• CHEN JIAXIN

Assignees

• 西华大学
• 四川凯运工程勘测设计有限公司

Dates

Publication Date

20260508

Application Date

20260213

Claims (10)

1. 1. A water quality analysis method of a water receiving lake in water diversion engineering combined with multi-source data fusion is characterized by comprising the following steps: collecting field basic data, defining a dynamic action relation between data in the field basic data based on the field basic data, and constructing a water diversion project and a water quality association field of a water lake, wherein the water quality association field presents a field effect influenced by water quality through the dynamic action relation between data, the field basic data comprises water diversion source dynamic output data, water delivery process real-time interaction data, water lake in-situ perception data and surrounding environment dynamic association data, the water quality association field divides the field units by classifying the data with association strength higher than a preset threshold value into the same field unit, and each field unit corresponds to a core water quality influence factor and an association data set; Performing interaction relation modeling operation on each field unit in the water quality association field to obtain a field unit interaction model, wherein the field unit interaction model is used for quantitatively describing the transmission mode, transmission strength and dynamic change characteristics of water quality influence among different field units; based on the interaction relation described by the field unit interaction model, performing dynamic coupling fusion processing on the field basic data of the water quality association field to generate coupling water quality data updated in real time; Combining the field change trend data of the water quality associated field with the coupling water quality data, performing deduction operation of a water quality trend evolution path of the receiving lake, and generating a trend evolution path containing continuous change characteristics of water quality states and key turning point information, wherein the field change trend data is extracted from the water quality associated field and comprises expansion and contraction trends, interaction strength change trends and data flow rate change trends of all field units; and identifying a core interaction unit from the water quality association field, combining the trend evolution path with the core interaction unit, and generating a diversion project water quality dynamic regulation scheme, wherein the diversion project water quality dynamic regulation scheme comprises regulation instructions acting on core field interaction nodes influencing the water quality trend.
2. 2. The method for analyzing the water quality of a water receiving lake in a diversion project combined with multi-source data fusion according to claim 1, wherein the collecting field basic data, based on the field basic data, defines a dynamic action relation between data in the field basic data, and constructs a water quality association field of the diversion project and the water receiving lake, and the method comprises the following steps: collecting water diversion source dynamic output data according to a fixed time period, and recording the collecting time information collected each time, wherein the water diversion source dynamic output data comprises water quality component data in the water source supplying process, flow change data in the water source supplying process and output state data in the water source supplying process; The method comprises the steps of sectionally acquiring real-time interaction data of a water delivery process according to a water delivery flow, and marking acquisition position information acquired each time, wherein the real-time interaction data of the water delivery process comprises interaction data of a water delivery pipeline and a water body, water quality exchange data along the water delivery line and water delivery facility operation state related data; the method comprises the steps of synchronously collecting in-situ sensing data of a water receiving lake through distributed sensing equipment in a multipoint mode, wherein the in-situ sensing data of the water receiving lake comprise water quality parameter data of different areas of the lake, water body flow data of different areas of the lake and lake ecological association data of different areas of the lake; surrounding environment dynamic association data are collected in a classified mode according to the environment influence type, and the surrounding environment dynamic association data comprise atmospheric environment data, surface runoff data, vegetation coverage data and human activity association data; Classifying and mapping the diversion source dynamic output data, the water delivery process real-time interaction data, the in-situ sensing data of the watershed and the surrounding environment dynamic association data according to data dimensions, establishing a corresponding relation between the data dimensions and water quality influence factors, and forming a field basic data dimension mapping table; based on the standardized field basic data, calculating the water quality influence correlation strength among the different source data to form a data correlation strength comparison table, wherein the calculation of the correlation strength is based on the contribution degree of the standardized data to the same water quality influence factor, and the calculation of the correlation strength is based on the contribution degree of the data to the same water quality influence factor; Based on the data association strength comparison table, performing establishment operation according to field unit division, classifying the data with association strength higher than a preset threshold into the same field unit, and enabling each field unit to correspond to a core water quality influence factor and an association data set; defining a space boundary of each field unit based on a physical position range of data acquisition, and defining a time boundary of each field unit based on a time synchronization range of data acquisition to form a field unit boundary characteristic; Performing space layout and time synchronization on each field unit according to the association relation in the data association strength comparison table, and constructing an initial water quality association field frame, wherein the initial water quality association field frame comprises field unit distribution, inter-unit association channels and data flow paths; Dividing the field basic data into units of the field unit boundary characteristics and the initial water quality association field frame, filling the units into corresponding field units and association channels, and supplementing dynamic rules of data flow among the units into the association channels to form the water quality association field.
3. 3. The method for analyzing water quality of a water receiving lake in water diversion engineering combined with multi-source data fusion according to claim 2, wherein the step of performing an interaction relation modeling operation on each field unit in the water quality association field to obtain a field unit interaction model comprises the steps of: Extracting core data characteristics of each field unit from the water quality association field, wherein the core data characteristics are key data expression forms capable of representing water quality influence attributes of the field units, and one field unit corresponds to a group of core data characteristic sets; After the core data feature extraction is completed, analyzing the overlapping features among the core data feature sets of the adjacent field units, and determining the water quality influence factors coexisting among the adjacent units; Identifying interaction types among adjacent field units based on the coexisting water quality influence factors, wherein the interaction types are enhanced interaction, suppressed interaction and neutral interaction, and the identification of the interaction types is based on a change response relation of core data characteristics; Calculating interaction strength parameters corresponding to each interaction type, wherein the interaction strength parameters are determined based on the difference of contribution degrees of the standardized common influence factors and corresponding values in the data association strength comparison table, and an initial interaction strength comparison table is formed; Monitoring the change of the core data characteristics of each field unit in the water quality association field, capturing the real-time fluctuation condition of the core data characteristics, and establishing a data characteristic fluctuation sequence; based on the data characteristic fluctuation sequence, executing the dynamic adjustment operation of the interaction intensity parameters in the initial interaction intensity comparison table to enable the interaction intensity parameters to be matched with the amplitude and the frequency of the data characteristic fluctuation to form a dynamic updated interaction intensity comparison table; analyzing indirect interaction relations transmitted by non-directly adjacent field units through an intermediate field unit, deriving indirect interaction strength parameters based on the direct interaction strength parameters, and supplementing the indirect interaction strength parameters into the dynamically updated interaction strength comparison table; after the indirect interaction relationship is supplemented, determining a transfer delay characteristic of interaction of each field unit, wherein the transfer delay characteristic is determined based on the space distance of the field unit, the data flow path length and the characteristics of a transfer medium, and a transfer delay parameter set is formed; integrating the interaction type, the dynamically updated interaction strength comparison table and the transfer delay parameter set to construct a field unit interaction model; After the field unit interaction model is constructed, simulating the interaction process of each field unit through the field unit interaction model, verifying the fit degree of the interaction result output by the field unit interaction model and the actual data change in the water quality association field, and optimizing model parameters based on the fit degree to form a stable interaction relation modeling result.
4. 4. The method for analyzing water quality of a water receiving lake in a diversion project combined with multi-source data fusion according to claim 1, wherein the performing dynamic coupling fusion processing on field base data of the water quality associated field based on the interaction relation described by the field unit interaction model to generate real-time updated coupling water quality data comprises: Extracting core interaction factors from the interaction relation, wherein the core interaction factors are key data items for determining interaction strength and interaction direction among field units, and each interaction type corresponds to a group of core interaction factors; Integrating rules related to the core interaction factors to form a coupling fusion operator, wherein the coupling fusion operator comprises an interaction intensity weight distribution rule, a data complementation fusion rule and a dynamic adjustment rule; Classifying and extracting field basic data of each field unit in the water quality association field according to the core interaction factors to form factor association data subsets, wherein one factor association data subset corresponds to one core interaction factor and associated field unit data; The method comprises the steps of inputting a factor-associated data subset into a corresponding coupling fusion operator, carrying out standardized preprocessing on original data in the data subset, distributing fusion weights for the standardized data according to an interaction strength weight distribution rule to enable the distributed weight values to be consistent with corresponding interaction strength parameters in a dynamically updated interaction strength comparison table; performing fusion calculation on the data distributed with the fusion weights according to a data complementation fusion rule, and integrating complementation information about the same core interaction factor in different field unit data to generate factor-level fusion data; extracting delay information in a transfer delay parameter set, and performing time synchronization calibration on factor-level fusion data to ensure that the fusion data of different field units keep consistency in time dimension and generate time synchronization fusion data; Monitoring the dynamic change of the interaction relation, and triggering the parameter update of the coupling fusion operator when the parameter in the dynamically updated interaction strength comparison table is adjusted, so that the weight distribution rule and the fusion rule of the coupling fusion operator follow the change of interaction; Based on the updated coupling fusion operator, carrying out real-time fusion processing on field basic data acquired subsequently to generate new factor-level fusion data and new time synchronization fusion data; Performing factor-crossing integration on new time synchronization fusion data corresponding to all core interaction factors, eliminating redundant information and conflict data among factors, and forming initial global coupling fusion data; And (3) performing field adaptability adjustment on the initial global coupling fusion data, and optimizing the spatial distribution and time sequence characteristics of the data by combining the field boundary characteristics and the data flow rules of the water quality associated field to generate coupling water quality data updated in real time.
5. 5. The method for analyzing the water quality of a water receiving lake in a diversion project combined with multi-source data fusion according to claim 2, wherein the method for analyzing the water quality of the water receiving lake in the diversion project combined with multi-source data fusion is characterized by comprising the steps of carrying out standardized pretreatment on different source data mapped to the same water quality influence factor based on the field basic data dimension mapping table to generate standardized field basic data, calculating the water quality influence correlation strength among the different source data based on the standardized field basic data to form a data correlation strength comparison table, and comprising the following steps: extracting all water quality influence factors from the field basic data dimension mapping table to form a water quality influence factor set, wherein each water quality influence factor corresponds to at least two data dimensions of different sources; Extracting original data associated with corresponding data dimensions aiming at each water quality influence factor in the water quality influence factor set, and carrying out standardized pretreatment on the original data to form a standardized exclusive data set of the water quality influence factors, wherein the standardized exclusive data set comprises related standardized data records of different sources; Analyzing the characterization sensitivity of each standardized data record in the standardized exclusive data set to the water quality influence factor, wherein the characterization sensitivity is determined based on the response relation between the variation of the standardized data record and the variation of the standardized value of the water quality influence factor; According to the characterization sensitivity of each standardized data record, sequencing all standardized data records in the standardized exclusive data set in priority to form a data priority sequence; Based on the data priority sequence, distributing basic contribution weight to each data source, setting the basic contribution weight value of the data source with the priority being ranked in front to be higher than that of the data source with the priority being ranked in back, and reserving dynamic adjustment space for the basic contribution weight to adapt to data change; Calculating the co-contribution coefficient between different data sources under the same water quality influence factor, wherein the co-contribution coefficient is determined based on the synchronicity and complementarity of standardized data changes between the data sources, and the value of the co-contribution coefficient of the data source with outstanding synchronicity and complementarity is set to be higher than that of the data source with unobtrusiveness in synchronicity and complementarity; combining the basic contribution weight with the cooperative contribution coefficient, and calculating the comprehensive contribution degree of each data source to the corresponding water quality influence factor; taking water quality influence factors as rows, taking data sources as columns, constructing an initial comparison table frame, and filling the comprehensive contribution degree of each data source to the corresponding water quality influence factors into the corresponding position of the initial comparison table frame; And executing normalization processing on all comprehensive contribution degrees in the initial comparison table frame, binding the normalized comparison table with source identification of field basic data and data dimension identification of the field basic data, and supplementing metadata information of the normalized comparison table to form a data association strength comparison table.
6. 6. The method for analyzing water quality of a water receiving lake in a diversion project combined with multi-source data fusion according to claim 3, wherein the executing the operation of dynamically adjusting the interaction intensity parameter in the initial interaction intensity comparison table based on the data characteristic fluctuation sequence to match the interaction intensity parameter with the amplitude and the frequency of the data characteristic fluctuation to form a dynamically updated interaction intensity comparison table comprises: Extracting fluctuation characteristics of the standardized data characteristic fluctuation sequence, and extracting key fluctuation parameters of fluctuation amplitude, fluctuation frequency, fluctuation period and fluctuation trend, wherein the core data characteristics of each field unit correspond to a group of fluctuation parameters based on standardized data; integrating a physical rule and a historical data verification result of water quality influence transfer to form a mapping relation model of fluctuation parameters and interaction strength parameters, wherein the mapping relation model records the adjustment direction and adjustment amplitude of the interaction strength parameters corresponding to different fluctuation parameter combinations; inputting the fluctuation parameters of each field unit into the mapping relation model to obtain interaction strength parameter adjustment suggestions corresponding to each field unit, wherein the adjustment suggestions comprise parameter adjustment amounts and adjusted target interaction strength parameters; Extracting current interaction strength parameters between corresponding field units in an initial interaction strength comparison table, comparing the current parameters with target interaction strength parameters in an adjustment suggestion, and determining an actual adjustment difference value; The corresponding parameters in the initial interaction strength comparison table are adjusted point by point according to the actual adjustment difference value, the direct interaction strength parameters between adjacent field units are adjusted firstly, and then the adjustment quantity of the indirect interaction strength parameters is deduced based on the direct parameter adjustment result; Performing internal consistency verification on the adjusted interaction strength comparison table to ensure that interaction strength parameters between the same field unit and different adjacent units are logically unified, so that interaction strength parameters of different levels are mutually adapted; Setting an adjustment threshold of the interaction strength parameter based on the overall stability requirement of a field of the water quality association field, and executing adjustment operation step by step in stages when the single adjustment quantity exceeds the adjustment threshold; recording the time node, the adjustment reason and the adjustment amplitude of each parameter adjustment to form a parameter adjustment log; And aligning the adjusted interaction strength comparison table with the time stamp of the data characteristic fluctuation sequence, and integrating the adjusted interaction strength comparison tables of all the time nodes to form a dynamically updated interaction strength comparison table sequence.
7. 7. The method for analyzing the water quality of a water receiving lake in a diversion project combined with multi-source data fusion according to claim 4, wherein the real-time fusion processing is performed on the field basic data collected subsequently based on the updated coupling fusion operator, and new factor-level fusion data and new time synchronization fusion data are generated, and the method comprises the following steps: A real-time interactive link of the field basic data is built, so that the field basic data which is acquired later can be transmitted to a fusion processing module within a preset time limit, a data receiving buffer area is arranged in the fusion processing module, the field basic data which is transmitted in real time is temporarily stored and is regular in format, weight distribution rules and fusion rules are extracted from an updated coupling fusion operator, and the weight distribution rules and the fusion rules are loaded to a real-time fusion processing engine; Real-time data are extracted from the data receiving buffer area according to the core interaction factor classification, standardized pretreatment is carried out on the real-time data to form a standardized real-time factor association data subset, the standardized real-time factor association data subset is input into a real-time fusion processing engine, real-time fusion weights are distributed to each piece of standardized data in the standardized real-time factor association data subset according to the updated weight distribution rule, and the distributed weight values are synchronous with the latest dynamic interaction strength comparison table parameters; Performing complementary fusion calculation on the data distributed with the real-time fusion weights according to the updated fusion rule, and integrating effective information in the data to generate real-time factor-level fusion data; the latest delay information in the transfer delay parameter set is called, time synchronization calibration is carried out on the real-time factor-level fusion data, and time sequence splicing is carried out on the real-time factor-level fusion data after the time synchronization calibration and the historical factor-level fusion data to form a continuous factor-level fusion data sequence; and carrying out outlier rejection processing on the continuous factor-level fusion data sequence, rejecting abnormal data points caused by data acquisition errors or transmission interference, storing the processed factor-level fusion data sequence and corresponding time synchronization fusion data into a fusion data warehouse, and updating real-time data identifiers in the data warehouse.
8. 8. The method for analyzing water quality of a water receiving lake in a diversion project combined with multi-source data fusion according to claim 1, wherein the step of combining the field variation trend data of the water quality related field with the coupling water quality data, performing a deduction operation of a water quality trend evolution path of the water receiving lake, and generating a trend evolution path including continuous variation characteristics of a water quality state and key turning point information comprises the steps of: extracting a core water quality index sequence in the coupled water quality data, and carrying out standardized treatment on the core water quality index sequence, wherein the core water quality index sequence corresponds to core interaction factors in a water quality association field one by one; Analyzing time variation characteristics of a standardized core water quality index sequence, extracting time sequence characteristic parameters of variation rate, variation direction and variation amplitude in the standardized core water quality index sequence, and forming a time sequence characteristic parameter set; extracting field change trend data from a water quality association field, wherein the field change trend data comprises expansion and shrinkage trend, interaction strength change trend and data flow rate change trend of each field unit; Performing association analysis on the time sequence characteristic parameter set and the field change trend data to determine causal correspondence between the core water quality index change and the field change; Integrating causal related map related data to form a water quality trend evolution model, wherein the water quality trend evolution model takes field change trend data as an input variable, takes change characteristics of core water quality indexes as an output variable, and model parameters are calibrated based on historical related data; Inputting field change trend data at the current moment into a water quality trend evolution model, and calculating to obtain a recent prediction result of a core water quality index, wherein the recent prediction result comprises an index change track within a preset recent prediction time length in the future; Based on a recent prediction result, a long-term change trend of the core water quality index is deduced by combining with a future development trend characteristic of the field change trend, wherein the long-term change trend comprises a stepwise characteristic and a development direction of index change within a preset long-term trend deduction time length in the future, and the long-term trend deduction time is longer than the recent prediction time length; determining potential key turning points of water quality change by analyzing extreme points in a core water quality index sequence and mutation points in a field change trend, wherein the potential key turning points correspond to time nodes of significant change of water quality state; Based on a recent prediction result, a long-term change trend and key turning point information, constructing a trend evolution path of the water quality of the water-receiving lake, wherein the trend evolution path takes time as an axis, and records continuous change characteristics of the water quality state and specific information of the key turning points; Combining the trend evolution path with the water quality protection target of the water-receiving lake and the operation constraint condition of the diversion project, and marking a section exceeding the protection target range or violating the operation constraint condition in the trend evolution path.
9. 9. The method for analyzing the water quality of a water receiving lake in a diversion project by combining multi-source data fusion according to claim 2, wherein the step of performing spatial layout and time synchronization on each field unit according to the association relation in the data association strength comparison table to construct an initial water quality association field frame comprises the following steps: extracting a correlation strength value corresponding to each field unit in a data correlation strength comparison table, wherein the correlation strength value is obtained based on standardized field basic data calculation; Based on the association relation priority sequence, establishing a space layout basis of the field units, wherein the field units with high association strength values are arranged in a space priority adjacent mode; Determining the actual space coordinate of each field unit according to the physical layout of the diversion project, the geographical partition of the water receiving lake and the distribution characteristics of the surrounding environment; Placing each field unit in the virtual field space according to the space layout basis and the actual space coordinates, and building a correlation channel between the field units in adjacent layout based on the correlation relationship in the correlation strength comparison table, wherein the width of the correlation channel is in direct proportion to the correlation strength value; Configuring a data transmission rule for each association channel, wherein the data transmission rule is determined based on the type of the association relationship and the association strength value, and determines the transmission direction, the transmission rate and the transmission priority of the data in the association channel; Extracting time stamp information in field basic data, analyzing acquisition time synchronicity of different source data, determining a time synchronicity reference, adjusting time marks of the basic data in each field unit based on the time synchronicity reference, enabling the data of the same period of time of different field units to be aligned in a time dimension, and planning a data flow path among the field units according to the data flow requirement of the field units after time synchronization and the transmission rule of the associated channels; and integrating the spatial distribution of the field units, the associated channels among the field units and the planned data flow paths to form an initial water quality associated field frame.
10. 10. The utility model provides a water quality analysis system of receiving lake in diversion engineering that combines multisource data to fuse which characterized in that includes: A processor; A machine-readable storage medium storing machine-executable instructions for the processor; Wherein the processor is configured to perform the method of water quality analysis of a water-receiving lake in a diversion project incorporating multi-source data fusion of any one of claims 1 to 9 via execution of the machine-executable instructions.

Description

Water quality analysis method and system for water receiving lake in diversion engineering by combining multi-source data fusion Technical Field The invention relates to the technical field of water resource management and analysis, in particular to a water quality analysis method and system for a water receiving lake in water diversion engineering by combining multi-source data fusion. Background In the water diversion project, the water quality condition of the water receiving lake is directly related to the surrounding ecological environment, the safety of water for residents and the whole benefit of the project. At present, the analysis of the water quality of the water-receiving lake in the diversion engineering mainly depends on a single data source or a mode of simply superposing multi-source data. The single data source analysis method is used for evaluating the water quality based on in-situ sensing data of the water receiving lake, and is difficult to comprehensively reflect complex causes of water quality change due to lack of consideration of link information such as diversion sources, water delivery processes and the like. The method for simply superposing the multi-source data integrates partial data of different links, but does not deeply excavate the internal dynamic action relation among the data, and cannot accurately present the field effect of water quality influence. For example, in the face of complex situations such as diversion flow change and sudden pollution of surrounding environment, the existing method is difficult to accurately capture the change trend of water quality, and an effective water quality regulation scheme cannot be formulated in time, so that serious problems such as serious fluctuation of water quality of a water receiving lake and even water quality deterioration can be caused. Disclosure of Invention In view of the above-mentioned problems, in combination with the first aspect of the present invention, the present invention provides a method for analyzing water quality in a water-receiving lake in a diversion project in combination with multi-source data fusion, the method comprising: Collecting field basic data, defining a dynamic action relation between data in the field basic data based on the field basic data, and constructing a water diversion project and a water quality association field influenced by a water lake, wherein the water quality association field presents a field effect influenced by water quality through the dynamic action relation between the data, and the field basic data comprises dynamic output data of a water diversion source, real-time interaction data of a water delivery process, in-situ perception data of the water lake and dynamic association data of surrounding environments; Performing interaction relation modeling operation on each field unit in the water quality association field to obtain a field unit interaction model, wherein the field unit interaction model is used for quantitatively describing the transmission mode, transmission strength and dynamic change characteristics of water quality influence among different field units; Based on the interaction relation described by the interaction model of the field unit, performing dynamic coupling fusion processing on the field basic data of the water quality association field to generate coupling water quality data updated in real time, wherein the dynamic coupling fusion processing performs adjustment operation of a fusion strategy when the interaction relation changes so that the fusion strategy follows the change of the field interaction; Combining the field change trend data of the water quality association field with the coupling water quality data, executing deduction operation of a water quality trend evolution path of the water receiving lake, and generating a trend evolution path containing water quality state continuous change characteristics and key turning point information; and identifying a core interaction unit from the water quality association field, combining the trend evolution path with the core interaction unit, and generating a diversion project water quality dynamic regulation scheme, wherein the diversion project water quality dynamic regulation scheme comprises regulation instructions acting on core field interaction nodes influencing the water quality trend. In still another aspect, the present invention also provides a system for analyzing water quality of a water receiving lake in a diversion project in combination with multi-source data fusion, including: The system comprises a processor, a machine-readable storage medium, and a machine-executable instruction of the processor, wherein the processor is configured to execute the water quality analysis method of the water receiving lake in the diversion engineering combined with the multi-source data fusion by executing the machine-executable instruction. In still another aspect, the present invention furth