A river and lake health monitoring system based on multi-dimensional data fusion

Abstract

This invention discloses a river and lake health monitoring system based on multi-dimensional data fusion, comprising: a monitoring data preprocessing module for periodically collecting multi-dimensional monitoring data from river and lake sub-regions to generate standardized regional health monitoring data; a river and lake sub-region health assessment module for conducting health assessments based on the standardized regional health monitoring data to generate periodic health assessment results for the river and lake sub-regions; a calculation and analysis module for calculating the health self-repair index of the river and lake sub-regions to achieve regional health change trend analysis and generate health change monitoring data; a health status prediction module for obtaining health status change prediction results for the river and lake sub-regions; and a river and lake health monitoring and restoration module for determining and adjusting health restoration priority schemes for the river and lake sub-regions based on the health status change prediction results. According to the above technical solution, effective technical support can be provided for river and lake management, ensuring the stability and sustainable development of river and lake ecosystems.

Description

Technical Field

[0001] This invention relates to the field of ecological monitoring technology, and more specifically, to a river and lake health monitoring system based on multi-dimensional data fusion. Background Technology

[0002] River and lake ecosystems are vital ecological support systems in nature, playing a crucial role in maintaining biodiversity, ensuring water resource security, and regulating regional climate. With increasing human activity and changes in the natural environment, river and lake ecosystems face multiple pressures, including water pollution, biodiversity loss, and habitat destruction. Therefore, conducting scientific and effective health monitoring and management has become an urgent issue.

[0003] Existing methods for monitoring river and lake health often rely on single types of data (such as focusing only on water pollution or biodiversity), making it difficult to comprehensively reflect the health status of rivers and lakes. Furthermore, fixed data collection cycles lack flexibility and cannot adapt to the dynamic changes of different indicators. Health assessment models rarely consider key influencing factors such as season, hydrology, and pollution type, resulting in insufficient accuracy in self-repair capacity analysis and regional difference assessment. Therefore, restoration priority determination based on existing river and lake health monitoring methods relies solely on self-repair time, failing to consider ecological importance, socio-economic impact, and restoration cost-effectiveness, thus limiting its practicality. It also struggles to cope with complex scenarios such as sudden pollution and extreme weather. Moreover, the high technical implementation threshold of existing river and lake health monitoring methods makes routine implementation difficult for grassroots management departments. In summary, traditional river and lake health monitoring and restoration management suffers from core pain points: single data dimensions, biased and subjective assessments, distorted and unfounded predictions, blind priority determination, and poor dynamic adaptability. Summary of the Invention

[0004] To address the aforementioned issues, this application provides a river and lake health monitoring system based on multi-dimensional data fusion, comprising: The monitoring data preprocessing module is used to periodically collect multi-dimensional monitoring data from river and lake sub-regions and generate standardized regional health monitoring data. The standardized regional health monitoring data reflects river and lake health assessment indicators, including: biodiversity data, water pollutant concentration data, key hydrological data, habitat quality data, and human activity disturbance data. The river and lake sub-region health assessment module is used to conduct health assessments based on standardized regional health monitoring data and generate periodic health assessment results for the river and lake sub-regions. The periodic health assessment results include the periodic comprehensive health score, health level, and key weakness indicators for the river and lake sub-regions. Among them, the health level includes: excellent, good, average, and poor. The calculation and analysis module is used to calculate the health self-repair index of river and lake sub-regions to realize regional health change trend analysis. Regional health change trend analysis refers to: calculating the health difference characteristics of adjacent river and lake sub-regions based on periodic health assessment results and health self-repair index, generating health difference curves, and obtaining the health change trend of the target river and lake sub-region; the health difference characteristics, health difference curves, and health change trends constitute health change monitoring data. The health status prediction module is used to acquire health change monitoring data of representative river and lake sub-regions, load the health change trend prediction model, and obtain the health status change prediction results of the river and lake sub-regions. The health status change prediction results include the Simpson diversity index, pollutant concentration prediction results, comprehensive health prediction score, health level prediction value, and predicted deterioration index for each river and lake sub-region in the future preset time period. The river and lake health monitoring and restoration module is used to determine and adjust the health restoration priority plan for river and lake sub-regions based on the prediction results of changes in health status; the content of the health restoration priority plan includes the key points of governance, the time sequence, and the expected goals.

[0005] The health assessment based on standardized regional health monitoring data includes the following steps: Set the influence weight of each indicator according to the preset ratio; Load the regional health threshold library for rivers and lakes, and determine the scores of individual indicators for sub-regions of rivers and lakes based on standardized regional health monitoring data; Based on a single parameter index score, a comprehensive health score is calculated, and the health level evaluation result is determined by combining the health level evaluation rules. The health level of river and lake sub-regions is determined based on the single-parameter index scoring rule, and periodic health assessment results are generated. The periodic health assessment results include the periodic comprehensive health score, health level, and key weakness indicators of river and lake sub-regions.

[0006] The calculation and analysis module includes a self-repair index calculation unit and a regional health change trend analysis unit. The self-repair index calculation unit is used to analyze the health self-repair capacity of river and lake sub-regions based on the results of periodic health assessments, and generates a health self-repair index. The regional health change trend analysis unit is used to analyze the health differences between adjacent river and lake sub-regions based on the periodic health assessment results and the comprehensive similarity coefficient, to draw health difference curves and generate regional health change trends of target rivers and lakes, thereby realizing regional health change trend analysis.

[0007] Furthermore, generating a health self-repair index includes the following steps: Based on the results of periodic health assessments, river and lake sub-regions with health levels of "medium" and "poor" are selected and uniformly marked to form marked river and lake sub-regions; The periodic health assessment results of the marked river and lake sub-regions are sorted in chronological order of monitoring time to form a time-series health monitoring table for each sub-region, which clarifies the biodiversity data and water pollutant concentration data for each monitoring period; Based on the time-series health monitoring table, biodiversity change maps and water pollutant concentration change maps were plotted for each marked river and lake sub-region, and core parameters were calculated. These core parameters include: the average periodic change in biodiversity, the average periodic change in pollutant concentration in a single water body, and the comprehensive periodic change in pollutant concentration in the overall water body. The calculation methods for these core parameters include: Average value of biodiversity cyclical variation ; Average value of pollutant concentration in a single water body during a period ; Overall water body pollutant concentration periodic variation comprehensive value ,in, This represents the secondary weighting value for pollutants; Based on the actual environmental characteristics of the target rivers and lakes, multi-dimensional correction coefficients are introduced to adapt to the impact of different scenarios on restoration capabilities; these multi-dimensional correction coefficients include seasonal coefficients. Hydrological condition coefficient Pollution type coefficient ; The health self-repair index HI of each marked river / lake sub-region is determined; the calculation method of the health self-repair index is as follows: ,in As the weight of biodiversity indicators, As the weight of water pollutant concentration index, and + =1; The health self-repair index HI is divided into numerical values ​​to determine the repair capability level.

[0008] Further analysis of regional health change trends includes the following steps: Based on the results of periodic health assessments, biological and abiotic data are extracted for all river and lake sub-regions. The biological data refers to the list of fish / aquatic plant species and the number of individuals of each species, while the abiotic data refers to the concentration of water pollutants and key hydrological parameters. Calculate the comprehensive similarity coefficient of adjacent sub-regions The calculation method is as follows: ,in: Let be the Jaccard class similarity coefficient, and: ; Let be the coefficient of species diversity, and: ; The similarity coefficient for non-biological indicators is expressed as: ; α, β, and γ are weights, and α+β+γ=1. Usually, α=0.4, β=0.3, and γ=0.3 are set, but can be adjusted according to different river and lake conditions. Based on the comprehensive similarity coefficient Identify health differences between adjacent sub-regions; Health difference curves are plotted in spatial order for the health differences of all consecutive adjacent sub-regions. Key influencing factors are labeled on the health difference curves to form a trend analysis chart. The trend analysis chart is used to reflect the macro trend of the target river and lake and output the conclusion of the regional health change trend.

[0009] Furthermore, before loading the health change trend prediction model, construct the health change trend prediction model; The steps involved in building a predictive model for health trends are as follows: Multiple sets of health change monitoring data from representative river and lake sub-regions were selected as training samples. Determine the model type, including: first, use Moran's I index to test the spatial correlation of river and lake health; if the spatial correlation is strong, choose the Kriging model; if the spatial correlation is weak, choose the random forest model. The trend analysis chart and the regional health change trend conclusions output by the trend analysis chart are used as model fitting constraints; the training samples are input into the model, spatial interpolation is introduced, the health change status of all non-sample sub-regions is predicted, and the preliminary health status change prediction results are output, thus realizing the construction of the health change trend prediction model. Error correction is performed on the health change trend prediction model.

[0010] The river and lake health monitoring and restoration module includes a restoration plan determination unit and a dynamic update unit. The restoration plan determination unit is used to correct the predicted results of changes in health status based on the health self-repair index, and to construct a multi-dimensional priority judgment system by combining ecological importance, socio-economic impact, and restoration cost-effectiveness to determine the health restoration priority plan for each river and lake sub-region. The dynamic update unit is used to dynamically adjust the health assessment results and repair priority plan in a regular manner based on the assessment results, update the trigger conditions and update cycle.

[0011] The process of determining the priority plan for the health restoration of river and lake sub-regions includes the following steps: Based on the health self-repair index HI and the predicted health level, the self-repair time period T required for each river and lake sub-region to recover from the state in the predicted health level to the health level is calculated. Constructing a multi-dimensional priority determination system includes: setting determination dimensions for scoring standards, including: ecological importance dimension E, socio-economic impact dimension S, and restoration cost-effectiveness dimension C; and determining the scores of each determination dimension for river and lake sub-regions. The comprehensive score is calculated based on a multi-dimensional priority judgment system. This includes: calculating the basic comprehensive score, expressed as: Determine future risk level based on comprehensive health prediction score and Calculate the final comprehensive score , is represented as: The future risk level is determined based on the comprehensive health prediction score. Determine the final overall score Then, repair priorities are assigned and a healthy repair priority scheme is generated.

[0012] Furthermore, the dynamic update unit supports regular updates and triggered updates; regular updates refer to: obtaining the restoration priority of river and lake sub-regions according to the restoration priority determination criteria, and updating the health assessment results and restoration priority plan at a specified cycle; Triggered updates refer to updating health assessment results and remediation priority plans when environmental information meets specified conditions.

[0013] Furthermore, triggered updates also support emergency response; The conditions for emergency response include sudden pollution incidents and ecological crises caused by extreme weather. The emergency response-related update and repair priority scheme for health assessment results is implemented through a rapid assessment process; the rapid assessment process includes the following steps: The monitoring data preprocessing module is invoked to collect real-time monitoring data, supplement the collection of emergency sample data from multiple representative sub-regions along the pollution diffusion path, and update standardized regional health monitoring data. Adjust the health change trend prediction model, temporarily adjust the adjustment correction coefficient, and recalculate the health self-repair index; Load a health change trend prediction model to predict the pollution diffusion path and impact range, simplify the error verification to RMSE≤15%, and ensure rapid output of results; An emergency remediation priority is generated by adding a weight based on the urgency of pollution spread to the remediation priority.

[0014] According to this invention, the limitations of traditional single-dimensional data collection in monitoring can be overcome, ensuring comprehensive, reliable, and unified input data, laying a high-quality data foundation for subsequent assessment and prediction, and solving the problem of data fragmentation in traditional river and lake health monitoring and restoration management. It outputs the health level of river and lake sub-regions, accurately locating key weakness indicators, and addressing the shortcomings of traditional assessments that are "vague in qualitative analysis and lack clear problem identification." It can quantify the self-restoration potential of each sub-region and embed it as a core parameter in health prediction and priority determination, solving the problem of traditional predictions being "one-size-fits-all and unrealistic." Furthermore, it overcomes the one-sidedness of traditional methods that "only determine priority based on restoration time," constructing a multi-dimensional judgment system that considers prediction results, self-restoration timeliness, ecological importance, socio-economic impact, and cost-effectiveness. Through weight allocation and future risk correction coefficients, it achieves a comprehensive balance between current state, future trends, and multiple values. This invention also supports a closed-loop update and emergency response mechanism, supporting a triple dynamic mechanism of regular updates based on risk level, triggered updates due to data mutations or external intervention, and emergency responses to sudden scenarios, forming a closed-loop process of assessment-prediction-decision-adjustment-reassessment, ensuring "dynamic adaptation of management." Attached Figure Description

[0015] Figure 1 This is a schematic diagram of the structure of a river and lake health monitoring system based on multi-dimensional data fusion, provided in an embodiment of the present invention. Detailed Implementation

[0016] Addressing the core pain points of traditional river and lake health monitoring and restoration management, such as single data dimensions, biased and subjective assessments, distorted and unfounded predictions, blind prioritization, and poor dynamic adaptability, this invention constructs a full-chain technical system of "multi-dimensional data fusion - precise health assessment - scientific trend analysis - dynamic priority decision-making." This system upgrades river and lake health management from "passively responding to problems" to "proactively predicting risks" and from "one-size-fits-all management" to "personalized and precise governance." It provides practical, reusable, and dynamically optimizable technical support for grassroots river and lake governance, ultimately ensuring the stability and sustainable development of river and lake ecosystems.

[0017] The specific implementation of the present invention will now be described in detail with reference to the accompanying drawings.

[0018] The structure of the river and lake health monitoring system provided by this invention is as follows: Figure 1 As shown, it includes the following parts: The P100 monitoring data preprocessing module is used to periodically collect multi-dimensional monitoring data from river and lake sub-regions and generate standardized regional health monitoring data. The river and lake sub-region health assessment module first divides the target river and lake into multiple river and lake sub-regions, and collects or gathers multi-dimensional raw monitoring data of each river and lake sub-region from monitoring equipment according to a preset cycle.

[0019] The multi-dimensional raw monitoring data includes the following indicators: 1) Biodiversity data: including the species and abundance of fish and aquatic plants, collected on a monthly basis; 2) Water pollutant concentration data: including organic pollutants, heavy metals, and nutrient concentrations, collected weekly as a regular cycle, and collected in real time during sudden pollution events, using automatic sensing equipment for monitoring; 3) Key hydrological data: including flow rate, flow velocity, water level, water temperature, dissolved oxygen, pH value, and transparency, collected on a daily basis and acquired in real time using automatic hydrological monitoring stations; 4) Habitat quality data: including substrate type, shoreline stability, aquatic vegetation coverage, and shoreline buffer zone width, collected quarterly, using a combination of remote sensing inversion and on-site verification. 5) Human activity disturbance data: including the location and discharge volume of sewage outlets, shipping frequency, land use types along the coast, and distribution of water intakes, collected at specified intervals; Furthermore, the multi-dimensional raw monitoring data undergoes quality control preprocessing (e.g., outlier removal and missing value imputation) and is integrated into standardized regional health monitoring data. The standardized regional health monitoring data reflects river and lake health assessment indicators, including: biodiversity data, water pollutant concentration data, key hydrological data, habitat quality data, and human activity disturbance data.

[0020] The P110 River and Lake Sub-region Health Assessment Module is used to conduct health assessments based on standardized regional health monitoring data and generate periodic health assessment results for river and lake sub-regions.

[0021] During the health assessment process, based on standardized regional health monitoring data, the influence weights of indicators are set, and the calculation of diversity index and threshold comparison are introduced to generate periodic health assessment results for each river and lake sub-region. The health assessment specifically includes the following steps: Set the influence weight of each indicator according to the preset ratio. For example, set biodiversity data to 0.25, water pollutant concentration data to 0.25, key hydrological data to 0.2, habitat quality data to 0.15, and human activity disturbance data to 0.15. Load the regional health threshold library for rivers and lakes, and determine the scores of individual indicators for sub-regions of rivers and lakes based on standardized regional health monitoring data; Among them, the regionalized health threshold database for rivers and lakes is a pre-set standard for judging the health of rivers and lakes, the contents of which are shown in Table 1: Table 1: Regionalized Health Threshold Database for Rivers and Lakes

[0022] When obtaining the score for a single indicator, the monitoring indicator data of each sub-region in the standardized regional health monitoring data are compared with the corresponding thresholds in the river and lake regional health threshold database. The score for each single indicator is obtained based on the single-parameter indicator scoring rules. The single-parameter indicator scoring rules are shown in Table 2. Table 2: Scoring Rules for Single Parameter Indicators

[0023] Based on a single parameter index score, a comprehensive health score is calculated, and the health level evaluation result is determined by combining the health level evaluation rules. The calculation method for the overall health score is: Overall Health Score = Σ (Score of individual indicator × corresponding weight).

[0024] Furthermore, the health level of river and lake sub-regions is determined based on the single-parameter index scoring rule, and periodic health assessment results are generated. The periodic health assessment results include the periodic comprehensive health score, health level, and key weakness indicators of the river and lake sub-regions; among which, the health level includes: excellent, good, average, and poor.

[0025] The health level evaluation rules are shown in Table 3: Table 3: Rules for Health Level Evaluation

[0026] The P120 calculation and analysis module is used to calculate the health self-repair index of the river and lake sub-regions and realize the regional health change trend analysis. The regional health change trend analysis refers to: calculating the health difference characteristics of adjacent river and lake sub-regions based on periodic health assessment results and health self-repair index, generating health difference curves, and obtaining the health change trend of the target river and lake sub-region; the health difference characteristics, health difference curves, and health change trends constitute health change monitoring data; Specifically, the calculation and analysis module includes a self-repair index calculation unit and a regional health change trend analysis unit; The P121 self-repair index calculation unit is used to analyze the health self-repair capacity of river and lake sub-regions based on periodic health assessment results, taking into account influencing factors such as season, hydrology, and pollution type, and to generate a health self-repair index. Generating a health self-repair index involves the following steps: 1) Based on the results of periodic health assessments, river and lake sub-regions with health levels of "medium" and "poor" are selected and uniformly marked to form marked river and lake sub-regions; 2) Sort the periodic health assessment results of the marked river and lake sub-regions in chronological order of monitoring time to form a time series health monitoring table for each sub-region, and clarify the biodiversity data and water pollutant concentration data for each monitoring period; 3) Based on the time-series health monitoring table, plot biodiversity change maps and water pollutant concentration change maps for each marked river / lake sub-region, and calculate core parameters. Core parameters include: the average periodic change in biodiversity, the average periodic change in water pollutant concentration, and the comprehensive periodic change in water pollutant concentration. The calculation method is as follows: Average value of biodiversity cyclical variation (ΔB): Average periodic variation of pollutant concentration in a single water body (ΔP) i ): The overall value of the periodic changes in pollutant concentration in the water body is ,in, This represents the secondary weighting value for pollutants; Specifically, the common pollutant types and their secondary weight values ​​are shown in Table 3: Table 3: Common Pollutant Types and Secondary Weights

[0027] 4) Based on the actual environmental characteristics of the target rivers and lakes, multi-dimensional correction coefficients are introduced to adapt to the impact of different scenarios on restoration capacity; these multi-dimensional correction coefficients include seasonal coefficients, hydrological condition coefficients, and pollution type coefficients. Seasonal coefficient The value of is usually 0.75 in winter, 1.0 in summer, and 0.9 in spring and autumn. Hydrological condition coefficient The value selection rule is as follows: 1.0 is taken when the flow rate is within the historical suitable range (determined based on hydrological data of the past 3 years), and 0.6-0.8 is taken when the flow rate exceeds the upper limit of the suitable range or is below the lower limit; Pollution type coefficient The value selection rules are as follows: 1.0 for organic pollution, 1.5 for heavy metal pollution, and 1.2 for nutrient pollution.

[0028] 5) Determine the Health Self-Repair Index (HI) and Repair Capacity Level for each marked river / lake sub-region; The calculation method for the health self-repair index is as follows: ,in As the weight of biodiversity indicators, As the weight of water pollutant concentration index, and + =1, such as =0.6, =0.4, which can be adjusted according to actual conditions.

[0029] The health self-repair index can reflect the restoration capacity of the river and lake sub-area. In this invention, the health self-repair index is divided into restoration capacity levels according to the value (for example: HI>0.01 is "strong", 0.005~0.01 is "medium" and <0.005 is "weak").

[0030] The P122 Regional Health Change Trend Analysis Unit is used to analyze the health differences between adjacent river and lake sub-regions based on the results of periodic health assessments, using a comprehensive similarity coefficient, to plot health difference curves and generate regional health change trends for target rivers and lakes, thus achieving regional health change trend analysis.

[0031] Specifically, the analysis of regional health change trends includes the following steps: 1) Based on the results of periodic health assessments, biological and abiotic data are extracted for all river and lake sub-regions. Biological data refers to the list of fish / aquatic plant species and the number of individuals of each species, while abiotic data refers to the concentration of water pollutants and key hydrological parameters (flow variation coefficient, dissolved oxygen, etc.). 2) Calculate the comprehensive similarity coefficient of adjacent sub-regions. The calculation method is as follows: ,in: Let be the Jaccard class similarity coefficient, and: ; Let be the coefficient of species diversity, and: ; The similarity coefficient for abiotic indicators is obtained by normalizing the Euclidean distance between pollutant concentration and hydrological parameters, and then normalizing it to 0-1, and is expressed as follows: Furthermore, the smaller the Euclidean distance, the closer the similarity coefficient is to 1, and the smaller the difference. α, β, and γ are weights, and α+β+γ=1. Usually, α=0.4, β=0.3, and γ=0.3 are set, but can be adjusted according to different river and lake conditions. 3) Based on the comprehensive similarity coefficient The health differences between adjacent sub-regions are determined, with S≥0.7 indicating "low difference", S ranging from 0.3 to 0.7 indicating "medium difference", and S<0.3 indicating "high difference". In this invention, health differences can reflect the differences between adjacent regions in biodiversity, water pollutant concentration, and hydrological conditions. 4) Draw a health difference curve diagram of all consecutive adjacent sub-regions in spatial order, and mark key influencing factors such as sewage outlet location, shoreline type, and water conservancy project on the health difference curve diagram to form a trend analysis diagram; the trend analysis diagram can reflect the macro trend of the target river and lake, and can output the conclusion of regional health change trend. The conclusions on regional health change trends include trends in river and lake location, trends in health difference coefficients, fluctuation trends, core influencing factors, and overall health status. For example, from upstream to downstream, the health difference coefficient decreased from 0.8 to 0.2, showing a fluctuation of "low difference → high difference". The core influencing factor is the concentrated discharge of sewage outlets along the downstream banks, and the overall health status gradually deteriorated.

[0032] The P130 health status prediction module is used to acquire health change monitoring data of representative river and lake sub-regions, load the health change trend prediction model, and obtain the health status change prediction results of the river and lake sub-regions. Before loading the health change trend prediction model, construct the health change trend prediction model; First, multiple sets of health change monitoring data from representative river and lake sub-regions are selected as training samples. For example, health change monitoring data from more than 10% of the target river and lake sub-regions are selected as prediction samples. The samples need to cover different health levels (including: excellent, good, medium and poor) and different regional characteristics (including: upstream, midstream, downstream, shore and lake center). The time span of the sample data is the health change monitoring data within a preset time period (such as 6 months or 1 year), which can reflect the core indicator data such as biodiversity and pollutant concentration during the monitoring period.

[0033] When determining the model type, Moran's I index is first used to test the spatial correlation of river and lake health. If the spatial correlation is strong (Moran's I ≥ 0.3), the Kriging model is selected, and if the spatial correlation is weak (Moran's I < 0.3), the random forest model is selected.

[0034] After determining the model type, the trend analysis chart and its output conclusions on regional health change trends are used as constraints for model fitting. Health change monitoring data from sample river and lake sub-regions are input into the model, and spatial interpolation is introduced to predict the health changes in all non-sample sub-regions, outputting preliminary health status prediction results, thus constructing a health change trend prediction model. The constraints ensure that the prediction results align with the overall health status; for example, if the downstream health trend is known to be deteriorating, the model prediction must follow this pattern to avoid unreasonable results such as "the downstream suddenly improving." Spatial interpolation supports Kriging interpolation and random forest interpolation.

[0035] Furthermore, error correction is performed on the health change trend prediction model: the mean absolute error (MAE) and root mean square error (RMSE) of the prediction results are calculated. If RMSE ≤ 10%, the prediction is deemed valid and the "final health status change prediction result" is directly output; if RMSE > 10%, the model parameters are re-optimized or sample data is supplemented, and the prediction is repeated until RMSE ≤ 10%.

[0036] The final health status change prediction results output by the error-corrected health change trend prediction model include the Simpson index, pollutant concentration prediction results, comprehensive health prediction score, health level prediction value, and predicted deterioration indicators for each sub-region over a preset time period.

[0037] The P140 River and Lake Health Monitoring and Restoration Module is used to determine and adjust the priority scheme for health restoration of river and lake sub-regions based on the prediction results of changes in health status.

[0038] The river and lake health monitoring and restoration module includes a restoration plan determination unit and a dynamic update unit; The P141 restoration scheme determination unit is used to correct the predicted results of changes in health status based on the health self-repair index, and to construct a multi-dimensional priority judgment system by combining ecological importance, socio-economic impact, and restoration cost-effectiveness to determine the health restoration priority scheme for each river and lake sub-region. The process of determining the priority plan for the health restoration of river and lake sub-regions includes the following steps: 1) Based on the health self-repair index HI and the health level prediction value output by the health change trend prediction model, calculate the self-repair time period T required for each sub-region to recover from the state in the health level prediction value to the health level (excellent or good). The self-healing time period T is calculated as follows: ,in The minimum standard score corresponding to the health level. A comprehensive health prediction score for a sub-region over a predetermined future time period; 2) Construct a multi-dimensional priority determination system: First, we establish the criteria for scoring to cover the core considerations of governance. The criteria include: ecological importance (E), socio-economic impact (S), and cost-effectiveness of restoration (C). Then determine the scores for each judgment dimension of the river and lake sub-region: When determining the ecological importance dimension E, if the river and lake sub-region is a core habitat, water source protection area, or biodiversity hotspot, the ecological importance dimension E is scored 3 points; if the river and lake sub-region is a general ecological protection area, the ecological importance dimension E is scored 1 point. When determining the socio-economic impact dimension S, if the river and lake sub-region is close to residential areas, industrial and agricultural water use areas, or important shipping channels, the socio-economic impact dimension S is scored 3 points; if the river and lake sub-region is close to remote areas with no important production and living needs, the socio-economic impact dimension S is scored 1 point. When determining the cost-effectiveness dimension C of restoration, the benefit ratio is calculated as follows: ecological benefits after restoration ÷ total restoration cost. Ecological benefits after restoration include water quality improvement and biodiversity restoration benefits, while total restoration cost includes engineering costs and operation and maintenance costs. If the benefit ratio is ≥3, the cost-effectiveness dimension C of restoration receives 3 points; if the benefit ratio is between 1 and 3, the cost-effectiveness dimension C of restoration receives 2 points; and if the benefit ratio is <1, the cost-effectiveness dimension C of restoration receives 1 point.

[0039] 3) Calculate the overall priority score based on the multi-dimensional priority determination system: First, calculate the basic comprehensive score using the weighted formula, as follows: ; Then, based on the comprehensive health prediction score, the future risk level is determined. And calculate the final comprehensive score. , is represented as: ; The future risk level is determined based on the comprehensive health prediction score predicted by the health change trend prediction model. When the comprehensive health prediction score is less than 60 points, the future risk level is considered high. The score is 1.2; when the comprehensive health prediction score is 60-74, the future risk level is medium risk. A score of 1.0 indicates a low risk level if the overall health prediction score is ≥75. It is 0.8.

[0040] Determine the final overall score Then, prioritize repairs and generate a health repair priority plan; First, combine the self-healing time period T and the final comprehensive score. The repair priorities are divided into three categories: high priority, medium priority, and low priority; the specific rules for this division are as follows: High priority: ≥2.5 points, and T>the preset intervention period; or the future risk is high, and it is located in the core ecological area; Medium priority: The score is 1 point, and T ≤ the preset intervention period, but T > 1 year; or the future risk is medium risk. Low priority: The score is 1 point, and T≤1 year; or the future risk is low risk. Output the health restoration priority plan for each sub-region, including the key points of governance, timeline, and expected goals. The key points of governance are the critical weakness indicators and the predicted deterioration indicators. The timeline for high-priority regions is to start restoration in the near future (e.g., within 3 months), the timeline for medium-priority regions is to start restoration in the medium term (e.g., within 6 months), and the timeline for low-priority regions is to start restoration in the long term (e.g., within 1 year) and focus on monitoring. The expected goals are the health level improvement targets within the preset time period in the future.

[0041] The health repair priority scheme can also be adjusted through a dynamic update unit.

[0042] In the P142 dynamic update unit, the trigger conditions and update cycle for updating assessment results are set to enable routine dynamic adjustment of health assessment results and remediation priority plans; it also supports the initiation of rapid response procedures for sudden pollution, extreme weather and other situations to adjust health assessment results and remediation priority plans in a timely manner.

[0043] The dynamic update unit supports both regular updates and triggered updates.

[0044] 1) Regular updates refer to: obtaining the restoration priority of river and lake sub-regions according to the restoration priority determination criteria, and updating the health assessment results and restoration priority plan at a specified cycle.

[0045] In this step, the criteria for determining the repair priority include: sub-regions with an overall health level of "poor" or ≥30% are "high repair priority"; sub-regions with an overall health level of "medium" or 10%~30% are "medium repair priority"; and sub-regions with an overall health level of "excellent" or "good" and <10% are "low repair priority".

[0046] Set the specified periods for low repair priority, medium repair priority, and high-low repair priority, such as quarterly, monthly, and bi-weekly, respectively.

[0047] Set different update methods with different repair priorities: When updating the health assessment results and restoration priority schemes for river and lake sub-regions with low restoration priority, the newly added monitoring data for the current quarter is supplemented into the standardized regional health monitoring data. The health assessment module and calculation and analysis module for river and lake sub-regions are called to recalculate the comprehensive health score and health self-restoration index. Based on the updated score, the health status prediction module is called to fine-tune the restoration priority scheme. If there is no significant change, the original scheme is maintained. When updating the health assessment results and restoration priority plans for river and lake sub-regions with medium restoration priority, the monitoring data preprocessing module, river and lake sub-region health assessment module, calculation and analysis module, and health status prediction module are fully invoked to recalculate the health status change prediction results; the prediction results before and after the update are compared to adjust the timing of the health restoration priority plan; if the predicted deterioration of a certain sub-region is accelerated, restoration is initiated in advance. When updating the health assessment results and restoration priority plans for river and lake sub-regions with high restoration priority, the monitoring data preprocessing module, river and lake sub-region health assessment module, calculation and analysis module, and health status prediction module are fully invoked throughout the process to recalculate the health status change prediction results; the regional health change trend and pollutant concentration prediction results are verified in a key manner, and the governance priorities of high-priority sub-regions are dynamically adjusted.

[0048] 2) Triggered updates refer to updating health assessment results and remediation priority plans when environmental information meets specified conditions. Specified conditions include: changes in monitoring data reaching a specified threshold (e.g., quantity changes ≥ 20%), new pollution sources, extreme weather, water conservancy project scheduling, or completion of remediation projects.

[0049] During triggered updates, based on the changed core data, the monitoring data preprocessing module, river and lake sub-region health assessment module, calculation and analysis module, and health status prediction module are invoked. The focus is on updating the comprehensive health score, adjusting the correction coefficient to recalculate the health self-repair index, and updating the parameters of the health change trend prediction model. On this basis, the repair priority is dynamically adjusted, specifically including: comparing the priority scores before and after the update; if the health status of a sub-region improves (e.g., score increases by ≥10 points), the priority is reduced; if it deteriorates (e.g., score decreases by ≥10 points), the priority is increased.

[0050] After each triggered update, the update result information is output, including the reason for the change, a comparison of core data changes, and an explanation of priority adjustments.

[0051] Triggered updates also support emergency response, with conditions including sudden pollution incidents and ecological crises caused by extreme weather. Sudden pollution incidents include situations such as chemical leaks and illegal discharges. These incidents can be identified through real-time monitoring of water pollutant concentrations using the monitoring data preprocessing module. For example, a sudden pollution incident is triggered when the concentration exceeds the unhealthy threshold by more than twice. Ecological crises caused by extreme weather include situations such as pollutant diffusion due to heavy rains and water flow interruptions due to droughts. These can be detected through real-time monitoring of key hydrological data using the monitoring data preprocessing module.

[0052] The updating of health assessment results and the remediation of priority plans in response to emergency situations are achieved through a rapid assessment process, which includes the following steps: Prioritize the use of the monitoring data preprocessing module to collect real-time monitoring data, supplement the collection with emergency sample data from multiple (e.g., 3) representative sub-regions along the pollution diffusion path, and update standardized regional health monitoring data; Adjust the health change trend prediction model and temporarily adjust the correction coefficients (e.g., for sudden heavy metal pollution, K3 is increased from 1.5 to 2.0; for abnormal flow caused by rainstorms, K2 is decreased to 0.6), and recalculate the health self-repair index; Load the health change trend prediction model to predict the pollution diffusion path and impact range. In this step, the error verification is simplified to RMSE≤15% to ensure rapid output of results. In this step, an emergency priority is generated by superimposing a pollution spread urgency weight on the remediation priority. This generates an emergency remediation priority (including: emergency response, rapid prevention and control, and routine monitoring). The pollution spread urgency weight is applied to core ecological areas and residential areas along the spread path. When calculating the priority score, the original priority score is multiplied by a coefficient greater than 1 (e.g., 1.5).

[0053] Generate a repair priority scheme corresponding to the emergency repair priority, including: Emergency Response (Highest Priority): Specific measures will be implemented for the core sub-areas along the pollution spread path (such as releasing pollutant purification agents, constructing temporary interception dams, and relocating sensitive aquatic organisms). Rapid prevention and control (medium priority): For the spread buffer zone, it is recommended to increase the monitoring frequency (once every 2 hours) and close water intakes / shipping channels; Routine monitoring (low priority): In unaffected areas, maintain the original monitoring cycle and track the dynamics of pollution spread in real time.

[0054] Furthermore, within 72 hours after the emergency response, a trigger-based update is initiated to verify the emergency response effectiveness throughout the entire process (recalculating health scores, health self-repair index, and regular priorities). If the health status returns to a safe range after emergency response, the regular update cycle will resume; if risks still exist, the river or lake will be upgraded to "high risk" and continuously tracked with updates every half month.

[0055] This invention employs a multi-dimensional data fusion and standardization system, breaking through the limitations of traditional single-dimensional data collection in monitoring. It systematically integrates multiple types of core data, setting differentiated collection cycles and diverse collection methods according to the dynamic characteristics of indicators. Simultaneously, it establishes a standardized mechanism for outlier removal and missing value filling, ensuring comprehensive, reliable, and consistent input data. This lays a high-quality data foundation for subsequent assessment and prediction, solving the problem of data fragmentation in traditional river and lake health monitoring and restoration management. Furthermore, this invention adopts a regionalized hierarchical assessment system, abandoning the traditional one-size-fits-all assessment model with uniform thresholds. It adapts to different river and lake ecological characteristics based on a health threshold database. It introduces the Simpson diversity index to quantify biodiversity and establishes a standardized assessment logic for health levels. This not only outputs health levels but also accurately identifies key weakness indicators, overcoming the shortcomings of traditional assessments that are "vague in qualitative analysis and lack clear problem identification," achieving a breakthrough in "precise assessment."

[0056] In the method for predicting the trend of river and lake health changes, this invention adopts a personalized correction mechanism of the health self-repair index, introduces multi-dimensional correction coefficients to calculate the health self-repair index, quantifies the self-repair potential of each sub-region, and embeds it as a core parameter into health prediction and priority determination. This solves the problem of traditional prediction being "one-size-fits-all and not in line with reality" and makes up for the shortcoming of "prediction ignoring individual differences". In terms of predicting changes in river and lake health status, it combines a collaborative prediction system that combines regional health change trend analysis, spatial correlation adaptation model and error correction verification, integrates the health change laws of upstream and downstream, and provides spatial constraints for prediction. Among them, the adaptation of Kriging and random forest models can avoid the limitations of a single model.

[0057] At the level of application of river and lake health monitoring results, this invention adopts a multi-dimensional dynamic priority judgment system, which breaks through the one-sidedness of the traditional "priority judgment based solely on restoration time" and constructs a multi-dimensional judgment system that considers prediction results, self-restoration timeliness, ecological importance, socio-economic impact, and cost-effectiveness. Through weight allocation and future risk correction coefficients, it achieves a comprehensive balance between the current state and future trends and multiple values.

[0058] This invention also supports a closed-loop update and emergency response mechanism, supporting a triple dynamic mechanism of regular updates with cycles determined by risk level, trigger updates caused by data mutations or external interventions, and emergency responses to sudden scenarios, forming a closed-loop process of assessment-prediction-decision-adjustment-reassessment to ensure "dynamic management adaptation".

[0059] The above-disclosed embodiments are merely a few specific examples of the present invention. However, the present invention is not limited thereto, and any variations that can be conceived by those skilled in the art should fall within the protection scope of the present invention.

Claims

1. A river and lake health monitoring system based on multi-dimensional data fusion, characterized in that, include: The monitoring data preprocessing module is used to periodically collect multi-dimensional monitoring data from river and lake sub-regions and generate standardized regional health monitoring data. The standardized regional health monitoring data reflects river and lake health assessment indicators, including: biodiversity data, water pollutant concentration data, key hydrological data, habitat quality data, and human activity disturbance data; The river and lake sub-region health assessment module is used to conduct health assessments based on standardized regional health monitoring data and generate periodic health assessment results for the river and lake sub-regions. The periodic health assessment results include the periodic comprehensive health score, health level, and key weakness indicators for the river and lake sub-regions. Among them, the health level includes: excellent, good, average, and poor. The calculation and analysis module is used to calculate the health self-repair index of the river and lake sub-region to realize regional health change trend analysis. The regional health change trend analysis refers to: calculating the health difference characteristics of adjacent river and lake sub-regions based on the periodic health assessment results and the health self-repair index, generating health difference curves, and obtaining the health change trend of the target river and lake sub-region. The health difference characteristics, health difference curves, and health change trends constitute health change monitoring data. The health status prediction module is used to acquire health change monitoring data of representative river and lake sub-regions, load the health change trend prediction model, and obtain the health status change prediction results of the river and lake sub-regions. The health status change prediction results include the Simpson diversity index, pollutant concentration prediction results, comprehensive health prediction score, health level prediction value, and predicted deterioration index for each river and lake sub-region in the future preset time period. The river and lake health monitoring and restoration module is used to determine and adjust the health restoration priority plan for river and lake sub-regions based on the prediction results of changes in health status; the content of the health restoration priority plan includes the key points of governance, the time sequence, and the expected goals.

2. The river and lake health monitoring system based on multi-dimensional data fusion according to claim 1, characterized in that, The health assessment based on standardized regional health surveillance data includes the following steps: Set the influence weight of each indicator according to the preset ratio; Load the regional health threshold library for rivers and lakes, and determine the scores of individual indicators for sub-regions of rivers and lakes based on standardized regional health monitoring data; Based on a single parameter score, a comprehensive health score is calculated, and the health level evaluation result is determined by combining the health level evaluation rules. The health level of river and lake sub-regions is determined based on a single indicator scoring rule, and periodic health assessment results are generated. The periodic health assessment results include the periodic comprehensive health score, health level, and key weakness indicators of the river and lake sub-regions.

3. The river and lake health monitoring system based on multi-dimensional data fusion according to claim 1, characterized in that, The calculation and analysis module includes a self-repair index calculation unit and a regional health change trend analysis unit; The self-repair index calculation unit is used to combine influencing factors and analyze the health self-repair capacity of river and lake sub-regions based on periodic health assessment results, and generate a health self-repair index. The regional health change trend analysis unit is used to analyze the health differences between adjacent river and lake sub-regions based on the periodic health assessment results and the comprehensive similarity coefficient, to draw health difference curves and generate regional health change trends of target rivers and lakes, thereby realizing regional health change trend analysis.

4. The river and lake health monitoring system based on multi-dimensional data fusion according to claim 3, characterized in that, The generation of the health self-repair index includes the following steps: Based on the results of periodic health assessments, river and lake sub-regions with health levels of "medium" and "poor" are selected and uniformly marked to form marked river and lake sub-regions; The periodic health assessment results of the marked river and lake sub-regions are sorted in chronological order of monitoring time to form a time-series health monitoring table for each sub-region, which clarifies the biodiversity data and water pollutant concentration data for each monitoring period; Based on the time-series health monitoring table, biodiversity change maps and water pollutant concentration change maps were plotted for each marked river and lake sub-region, and core parameters were calculated. These core parameters include: the average periodic change in biodiversity, the average periodic change in pollutant concentration in a single water body, and the comprehensive periodic change in pollutant concentration in the overall water body. The calculation methods for these core parameters include: Average value of biodiversity cyclical variation ; Average value of pollutant concentration in a single water body during a period ; Overall water body pollutant concentration periodic variation comprehensive value ,in, These are the secondary weight values ​​for pollutants; Based on the actual environmental characteristics of the target rivers and lakes, multi-dimensional correction coefficients are introduced to adapt to the impact of different scenarios on restoration capabilities; these multi-dimensional correction coefficients include seasonal coefficients. Hydrological condition coefficient Pollution type coefficient ; The health self-repair index HI of each marked river / lake sub-region is determined; the calculation method of the health self-repair index is as follows: ,in As the weight of biodiversity indicators, As the weight of water pollutant concentration index, and + =1; The health self-repair index HI is divided into numerical values ​​to determine the repair capability level.

5. The river and lake health monitoring system based on multi-dimensional data fusion according to claim 3, characterized in that, Regional health trend analysis includes the following steps: Based on the results of periodic health assessments, biological and abiotic data are extracted for all river and lake sub-regions. The biological data refers to the list of fish / aquatic plant species and the number of individuals of each species, while the abiotic data refers to the concentration of water pollutants and key hydrological parameters. Calculate the comprehensive similarity coefficient of adjacent sub-regions The calculation method is as follows: ,in: Let be the Jaccard class similarity coefficient, and: ; Let be the coefficient of species diversity, and: ; The similarity coefficient for non-biological indicators is expressed as: ; α, β, and γ are weights, and α+β+γ=1. Usually, α=0.4, β=0.3, and γ=0.3 are set, but can be adjusted according to different river and lake conditions. Based on the comprehensive similarity coefficient Identify health differences between adjacent sub-regions; Health difference curves are plotted in spatial order for the health differences of all consecutive adjacent sub-regions. Key influencing factors are labeled on the health difference curves to form a trend analysis diagram. The trend analysis diagram is used to reflect the macro trend of the target river and lake and output the conclusion of the regional health change trend.

6. The river and lake health monitoring system based on multi-dimensional data fusion according to claim 1, characterized in that, Before loading the health change trend prediction model, a health change trend prediction model is constructed. The construction of the health change trend prediction model includes the following steps: Multiple sets of health change monitoring data from representative river and lake sub-regions were selected as training samples. Determine the model type, including: first, use Moran's I index to test the spatial correlation of river and lake health; if the spatial correlation is strong, choose the Kriging model; if the spatial correlation is weak, choose the random forest model. The trend analysis chart and the regional health change trend conclusions output by the trend analysis chart are used as model fitting constraints; the training samples are input into the model, spatial interpolation is introduced, the health change status of all non-sample sub-regions is predicted, and the preliminary health status change prediction results are output, thus realizing the construction of the health change trend prediction model. Error correction is performed on the health change trend prediction model.

7. The river and lake health monitoring system based on multi-dimensional data fusion according to claim 1, characterized in that, The river and lake health monitoring and restoration module includes a restoration plan determination unit and a dynamic update unit; The restoration plan determination unit is used to correct the predicted results of changes in health status based on the health self-repair index, and to construct a multi-dimensional priority judgment system by combining ecological importance, socio-economic impact, and restoration cost-effectiveness to determine the health restoration priority plan for each river and lake sub-region. The dynamic update unit is used to dynamically adjust the health assessment results and repair priority schemes in a regular manner based on the assessment results, updating the trigger conditions and update cycle.

8. The river and lake health monitoring system based on multi-dimensional data fusion according to claim 7, characterized in that, The process of determining the priority scheme for health restoration of river and lake sub-regions includes the following steps: Based on the health self-repair index HI and the predicted health level, the self-repair time period T required for each river and lake sub-region to recover from the state in the predicted health level to the health level is calculated. Constructing a multi-dimensional priority determination system includes: setting determination dimensions for scoring standards, including: ecological importance dimension E, socio-economic impact dimension S, and restoration cost-effectiveness dimension C; and determining the scores of each determination dimension for river and lake sub-regions. The comprehensive score is calculated based on a multi-dimensional priority judgment system. This includes: calculating the basic comprehensive score, expressed as: Determining future risk levels based on comprehensive health prediction scores and Calculate the final overall score , represented as: The future risk level is determined based on the comprehensive health prediction score. Determine the final overall score Then, repair priorities are assigned and a healthy repair priority scheme is generated.

9. The river and lake health monitoring system based on multi-dimensional data fusion according to claim 7, characterized in that, The dynamic update unit supports regular updates and triggered updates; the regular update refers to: obtaining the restoration priority of river and lake sub-regions according to the restoration priority determination criteria, and updating the health assessment results and restoration priority schemes at a specified cycle; The triggered update refers to updating the health assessment results and repair priority scheme when the environmental information meets the specified conditions.

10. The river and lake health monitoring system based on multi-dimensional data fusion according to claim 9, characterized in that, The triggered update also supports emergency response; The conditions for emergency response include sudden pollution incidents and ecological crises caused by extreme weather. The updating of health assessment results and the repair of priority plans corresponding to the emergency response are achieved through a rapid assessment process; the rapid assessment process includes the following steps: The monitoring data preprocessing module is invoked to collect real-time monitoring data, supplement the collection of emergency sample data from multiple representative sub-regions along the pollution diffusion path, and update standardized regional health monitoring data. Adjust the health change trend prediction model, temporarily adjust the correction coefficient, and recalculate the health self-repair index; Load a health change trend prediction model to predict the pollution diffusion path and impact range, simplify the error verification to RMSE≤15%, and ensure rapid output of results; An emergency remediation priority is generated by adding a weight based on the urgency of pollution spread to the remediation priority.

Images