Seawall ecologicalization real-time monitoring method based on multi-source data fusion

Abstract

The invention relates to the technical field of environment monitoring, in particular to a seawall ecologicalization real-time monitoring method based on multi-source data fusion. The method comprises the following steps: acquiring structure monitoring data, ecological monitoring data and water quality monitoring data of a seawall through an Internet of Things sensor, and acquiring remote sensing data of a seawall area in combination with satellite remote sensing and an unmanned aerial vehicle image; performing space-time registration and fusion on different data sources to generate a unified multi-source fusion data set; identifying a red tide abnormal event based on the fused data set, and constructing a red tide influence conduction model for predicting the damage degree of the red tide to the ecological system and the influence of the red tide to the stability of the embankment body; establishing a coupling evaluation model of ecological indexes and structural indexes, calculating a coupling coefficient by using a fuzzy comprehensive evaluation method, dynamically adjusting ecological and structural weights according to a model output result, and calculating a seawall comprehensive health index; graded early warning information is automatically generated according to a preset threshold value, and collaborative monitoring and intelligent early warning of seawall ecology and structure safety are achieved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of environmental monitoring, and in particular to a real-time monitoring method for ecological seawall based on multi-source data fusion. BACKGROUND

[0002] In recent years, the concept of ecological seawall construction has gradually been valued. By introducing ecological revetment, salt-tolerant vegetation, and water-permeable substrate into the structure of the seawall, the ecological environment can be effectively improved, and the collaborative stability of the seawall and the ecological system can be enhanced. However, the ecological system of the seawall is long-term under the influence of complex marine dynamics and human activities, and its ecological and structural state has obvious spatiotemporal dynamics and multi-factor coupling characteristics. For example, the action of tides, waves, and runoff can cause settlement of the seawall foundation and erosion of the slope surface; changes in water quality and red tide outbreaks can lead to the death of vegetation and the reduction of benthic organisms, thereby indirectly weakening the soil-retaining and protective capacity of the seawall.

[0003] The existing seawall monitoring system mainly relies on a single data source, such as structural monitoring instruments or remote sensing images, and lacks dynamic quantitative means for the interaction between ecological processes and structural safety. In particular, when sudden ecological abnormal events such as red tide occur, traditional monitoring methods can only detect water quality abnormalities and cannot timely assess the damage of red tide to the ecological system of the seawall and its potential impact on the stability of the seawall, making it difficult to achieve early warning and precise management. SUMMARY

[0004] In order to make up for the above shortcomings, the present application provides a real-time monitoring method for ecological seawall based on multi-source data fusion, which aims to improve the problem that traditional monitoring methods cannot timely assess the damage of red tide to the ecological system of the seawall and its potential impact on the stability of the seawall, making it difficult to achieve early warning and precise management.

[0005] In the first aspect, the present application provides a real-time monitoring method for ecological seawall based on multi-source data fusion, comprising: Collecting structural monitoring data, ecological monitoring data, and water quality monitoring data of the seawall through Internet of Things sensors, and obtaining remote sensing data of the seawall area through remote sensing equipment; Performing spatiotemporal registration on the structural monitoring data, ecological monitoring data, water quality monitoring data, and remote sensing data, unifying the spatiotemporal resolution of different data sources, and generating a fusion data set; Based on the water quality characteristic parameters in the fusion data set, identifying red tide abnormal events; Constructing a red tide influence transmission model, which is used to predict the damage of red tide to the ecological system; A coupled evaluation model of ecological and structural indicators is established, and the weights of ecological and structural indicators in the model are dynamically adjusted based on the prediction results of the red tide anomaly event and the red tide impact transmission model. The comprehensive health index of the seawall is calculated based on the coupled evaluation model, and graded early warning information is generated according to the preset threshold.

[0006] Preferably, the structural monitoring data includes tilt angle, strain value, and settlement displacement; the ecological monitoring data includes vegetation coverage, soil moisture content, and soil salinity; the water quality monitoring data includes chlorophyll a concentration, dissolved oxygen concentration, water temperature, and nutrient content; and the remote sensing data includes the vegetation index NDVI, water chlorophyll concentration distribution data, and high-resolution remote sensing images.

[0007] Preferably, the step of generating the fused dataset includes: Spatial coordinate unification and timestamp standardization were performed on structural monitoring data, ecological monitoring data, water quality monitoring data, and remote sensing data, respectively. The Kriging interpolation algorithm is used to perform interpolation calculations on data with inconsistent spatial resolutions to generate unified spatial grid data. Synchronize data with different sampling frequencies based on time series alignment algorithm to eliminate time delay error; After spatial and temporal registration, the multi-source data is normalized and features are extracted to form a fusion dataset containing unified spatiotemporal features.

[0008] Preferably, the steps for identifying red tide anomalies include: Based on the time series mutation detection algorithm, abnormal fluctuation ranges of chlorophyll a concentration and dissolved oxygen concentration are identified; By combining a multi-parameter threshold model of water temperature and nutrient content, a secondary determination is made on the abnormal fluctuation range; When the determination result meets the characteristics of red tide formation, the corresponding spatial area is marked as a suspected red tide area; The spectral characteristics of the water body in the suspected area are compared and verified with remote sensing image data. If the matching degree is higher than the preset threshold, it is confirmed as a red tide anomaly.

[0009] Preferably, the steps for constructing a red tide impact transmission model include: Based on the spatiotemporal distribution information of red tide anomalies in the fused dataset, red tide intensity, water quality degradation indicators, and ecosystem response parameters are extracted. A correlation model between red tide intensity and ecosystem damage was established. The correlation model used the rate of change of chlorophyll a concentration, the rate of decrease of dissolved oxygen concentration, and the magnitude of change of nutrient concentration as input features, and the vegetation mortality rate and the rate of decrease of benthic organism density as output indicators. The mapping relationship between red tide impact factors and the degree of ecosystem damage was obtained by training on historical red tide data based on multiple regression analysis. A red tide impact transmission model is constructed using the aforementioned mapping relationship.

[0010] Preferably, the steps for establishing the coupled evaluation model include: The initial weight relationship between ecological and structural indicators was determined based on the analytic hierarchy process, forming a multi-factor comprehensive evaluation framework. The degree of ecological damage output by the red tide impact transmission model is used as a dynamic adjustment factor to adjust the weights of ecological indicators in real time. The coupling coefficient between ecological and structural indicators is calculated using the fuzzy comprehensive evaluation method to reflect the coupling effect of ecosystem degradation on structural stability. The parameters of the coupling evaluation model are updated based on the dynamic changes in the coupling coefficient.

[0011] Preferably, the implementation process of the fuzzy comprehensive evaluation method includes: Construct a fuzzy evaluation matrix of ecological and structural indicators, where the elements in the fuzzy evaluation matrix represent the degree of membership of each indicator to the health status of the seawall. The membership values ​​of ecological and structural indicators are calculated using triangular membership functions. Based on the dynamic adjustment factor and the initial weight relationship, calculate the fuzzy weight vector of the ecological index and the structural index; By performing fuzzy synthesis operations, the fuzzy evaluation matrix and the fuzzy weight vector are synthesized to obtain the fuzzy comprehensive evaluation result of the coupling coefficient.

[0012] Preferably, the tiered early warning information includes: When the comprehensive health index of the seawall is greater than 0.8, a low-risk warning is issued; when the comprehensive health index of the seawall is less than 0.5, a high-risk warning is issued; otherwise, a medium-risk warning is issued.

[0013] Secondly, the present invention provides the following technical solution: a real-time monitoring system for the ecological protection of seawalls based on multi-source data fusion, the system comprising: The data acquisition module is used to collect structural monitoring data, ecological monitoring data and water quality monitoring data of the seawall through IoT sensors, and to acquire remote sensing data of the seawall area through remote sensing equipment. The data fusion module is used to perform spatiotemporal registration on the structural monitoring data, ecological monitoring data, water quality monitoring data and remote sensing data, unify the spatiotemporal resolution of different data sources, and generate a fused dataset. The red tide event identification module is used to identify abnormal red tide events based on the water quality characteristic parameters in the fused dataset; An ecological damage assessment module is used to construct a red tide impact transmission model, which is used to predict the degree of damage that red tides cause to the ecosystem. The coupled evaluation module is used to establish a coupled evaluation model of ecological indicators and structural indicators, and dynamically adjust the weights of ecological indicators and structural indicators in the model based on the prediction results of the red tide anomaly event and the red tide impact transmission model. The graded early warning module is used to calculate the comprehensive health index of the seawall based on the coupled evaluation model and generate graded early warning information according to preset thresholds.

[0014] The present invention has the following beneficial effects: 1. This invention achieves multi-layered collaboration between structural monitoring, ecological monitoring, water quality monitoring, and remote sensing observation by constructing a multi-source heterogeneous data fusion architecture. It employs Kriging interpolation and time series alignment algorithms to uniformly register data from different sources and at different resolutions, forming a high-precision spatiotemporal fusion dataset. This effectively solves the problems of data dispersion, inconsistent accuracy, and insufficient timeliness in traditional seawall monitoring.

[0015] 2. This invention introduces a red tide identification and ecological damage transmission model, combined with an ecology-structure coupling evaluation mechanism, enabling early identification, damage assessment, and tiered early warning of sudden ecological events such as red tides. By dynamically adjusting the weights of ecological and structural indicators, and comprehensively calculating the seawall health index, it achieves synergistic analysis of ecological degradation and structural stability. Attached Figure Description

[0016] Figure 1 This is a flowchart of a real-time monitoring method for the ecological protection of seawalls based on multi-source data fusion proposed in this invention. Figure 2 This is a structural diagram of a real-time monitoring method for the ecological protection of seawalls based on multi-source data fusion proposed in this invention. Detailed Implementation

[0017] The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0018] Example 1 In the first embodiment of the present invention, the present invention provides a method for real-time monitoring of seawall ecology based on multi-source data fusion, such as... Figure 1 As shown, it includes the following steps: The structure monitoring data, ecological monitoring data and water quality monitoring data of the seawall are collected by IoT sensors, and remote sensing data of the seawall area is obtained by remote sensing equipment. Preferably, the structural monitoring data includes tilt angle, strain value, and settlement displacement; the ecological monitoring data includes vegetation coverage, soil moisture content, and soil salinity; the water quality monitoring data includes chlorophyll a concentration, dissolved oxygen concentration, water temperature, and nutrient content; and the remote sensing data includes the vegetation index NDVI, water chlorophyll concentration distribution data, and high-resolution remote sensing images.

[0019] Specifically, tilt sensors, strain gauges, and settlement monitoring points are deployed at key locations on the seawall (such as the wave-facing side, the crest, and the backwater side). Tilt sensors are used to monitor changes in the tilt angle of the seawall structure in real time, reflecting the stress state and deformation trend. Strain gauges are used to monitor the strain values ​​of the concrete or reinforced concrete structure to determine if the stress distribution is abnormal. Settlement monitoring points record the settlement displacement of the seawall foundation using high-precision displacement sensors, identifying potential areas of structural instability. The collected data is uploaded to the monitoring center server in real time via a low-power wireless communication module and updated at a fixed sampling period.

[0020] Ecological sensor nodes were deployed on the surface of the seawall and in the near-seawall ecological zone to collect data on vegetation cover, soil moisture content, and soil salinity. Vegetation cover was calculated by measuring the reflectance ratio using a multispectral photosensitive element; soil moisture content was measured in real time using a capacitive humidity sensor; and soil salinity was obtained using a conductivity sensor to assess the impact of salinization on vegetation growth.

[0021] Multi-parameter water quality monitoring buoys were deployed in nearshore waters and at the toe of dikes to collect real-time data on chlorophyll a concentration, dissolved oxygen concentration, water temperature, and nutrient content. Chlorophyll a concentration was measured by a fluorescence sensor to reflect phytoplankton biomass; dissolved oxygen concentration was determined by polarography or an optical dissolved oxygen probe to characterize the ecological activity of the water body; water temperature was monitored in real-time using a thermistor as a key factor influencing red tide formation; and nutrient content (such as nitrate and phosphate) was periodically sampled and detected using ion-selective electrodes or colorimetric sensors.

[0022] The remote sensing data comprises two parts: satellite remote sensing and UAV remote sensing. Satellite remote sensing periodically acquires data on the NDVI (NDV) and chlorophyll concentration distribution in the water of the seawall area via remote sensing satellites, with a spatial resolution of 10–30 meters and a temporal resolution of 5–10 days. UAV remote sensing utilizes UAVs equipped with high-resolution optical and multispectral cameras to conduct low-altitude inspections, capturing images of vegetation growth, surface cracks, and erosion marks on the seawall surface. The preferred flight altitude for UAVs is 50–100 meters, and the inspection cycle is preferably once a week. After image preprocessing (including radiometric correction, geometric correction, and orthophoto projection), the remote sensing data is coordinate-registered with the ground-based sensing data.

[0023] Through the above data collection and transmission steps, integrated, continuous, and high spatiotemporal resolution monitoring of seawall structural safety, ecological environment status, and water quality conditions can be achieved, ensuring the diversity and reliability of data sources and providing sufficient data support for subsequent red tide identification, ecological damage assessment, and comprehensive health index calculation.

[0024] The structural monitoring data, ecological monitoring data, water quality monitoring data, and remote sensing data are spatiotemporally registered to unify the spatiotemporal resolution of different data sources and generate a fused dataset. Preferably, the step of generating the fused dataset includes: Spatial coordinate unification and timestamp standardization were performed on structural monitoring data, ecological monitoring data, water quality monitoring data, and remote sensing data, respectively. The Kriging interpolation algorithm is used to perform interpolation calculations on data with inconsistent spatial resolutions to generate unified spatial grid data. Synchronize data with different sampling frequencies based on time series alignment algorithm to eliminate time delay error; After spatial and temporal registration, the multi-source data is normalized and features are extracted to form a fusion dataset containing unified spatiotemporal features.

[0025] Specifically, the spatial coordinate system of data from different monitoring devices is first unified. Ground point data collected by each sensor is based on the WGS-84 coordinate system, with latitude and longitude information automatically recorded via GPS modules. Remote sensing data is unified to the same coordinate system through geometric correction and orthophoto projection. Sensor points and remote sensing images are spatially overlaid to establish a geographic reference grid, ensuring that all data can be located within the same spatial framework. Different data sources have different sampling frequencies; for example, the sampling interval for structural sensors is 5 minutes, while the sampling interval for ecological and water quality monitoring nodes is 10 minutes, and the update cycle for remote sensing data is several days. To ensure time alignment, the system adds a standardized timestamp to each data record and uses linear interpolation or moving average to fill in missing time points.

[0026] The spatial resolution of different monitoring data varies greatly. For example, ground sensors provide discrete point data, while remote sensing images provide continuous raster data.

[0027] To achieve spatial consistency of data, this invention employs the Kriging interpolation algorithm to perform spatial interpolation on discrete point data. First, a semivariance function model is constructed based on the location and measurement values ​​of each sensor sampling point. Then, the spatial correlation coefficient and optimal weight matrix are calculated. Finally, equally spaced raster data are generated within a preset grid range to form a spatial continuous field.

[0028] Based on spatial registration, the system performs time series alignment on various types of data to address the issue of inconsistent sampling frequencies among multi-source data. This invention preferably uses a dynamic time warping algorithm to match time series data, achieving non-linear alignment of multi-source time series by calculating the similarity matrix between adjacent time periods. For missing or anomalous sampling points, an adaptive interpolation method is used for reconstruction to ensure the continuity and comparability of the time series.

[0029] Multi-source data after spatial and temporal registration were subjected to dimensional unification and normalization. Continuous parameters such as tilt angle, vegetation cover, and chlorophyll a concentration were standardized using Z-score; discrete parameters such as structural crack grade and vegetation type were converted using one-heat encoding. In addition, representative feature variables such as ecological characteristics (NDVI, soil moisture content change rate), structural characteristics (strain gradient, sedimentation trend), and water quality characteristics (chlorophyll a change rate, dissolved oxygen fluctuation amplitude) were extracted. The above data were integrated into a unified high-dimensional spatiotemporal feature matrix to form a fusion dataset that can be used for subsequent red tide identification and coupled analysis.

[0030] The spatiotemporal registration and fusion dataset generation method implemented through the above steps can effectively unify data from different sources, with different resolutions and different sampling frequencies, and significantly improve the consistency and fusion quality of multi-source data in the spatiotemporal dimension.

[0031] Based on the water quality characteristic parameters in the fused dataset, red tide anomalies are identified. Preferably, the steps for identifying red tide anomalies include: Based on the time series mutation detection algorithm, abnormal fluctuation ranges of chlorophyll a concentration and dissolved oxygen concentration are identified; By combining a multi-parameter threshold model of water temperature and nutrient content, a secondary determination is made on the abnormal fluctuation range; When the determination result meets the characteristics of red tide formation, the corresponding spatial area is marked as a suspected red tide area; The spectral characteristics of the water body in the suspected area are compared and verified with remote sensing image data. If the matching degree is higher than the preset threshold, it is confirmed as a red tide anomaly.

[0032] Specifically, key water quality parameters closely related to marine biological activity are extracted from the fused dataset, including chlorophyll a concentration, dissolved oxygen concentration, water temperature, and nutrient content (such as phosphate and nitrate concentrations). Noise filtering is applied to the time series of each parameter, preferably using a moving median filter algorithm to remove abnormal peak values ​​while preserving trend information. A time series mutation detection algorithm is used to identify abnormal fluctuation ranges in chlorophyll a concentration and dissolved oxygen concentration. This invention preferably employs a Bayesian change point detection algorithm to statistically analyze the mean changes in the time series. When the chlorophyll a concentration experiences a sudden increase exceeding a set threshold (preferably 30%–50% of the average value) within a continuous sampling period (e.g., 6 hours), and the dissolved oxygen concentration decreases simultaneously, the system marks this time period as a "preliminary abnormal interval." Simultaneously, the first and second differences of the rate of change are calculated to identify the stability of the mutation trend and prevent misjudging short-term disturbances as abnormal events.

[0033] The data within the initial anomaly range were further filtered to confirm whether conditions for red tide formation were met. A multi-parameter threshold model incorporating variables such as water temperature, nutrient content, and light intensity was constructed. ; in, Water temperature This refers to the concentration of nutrients. This refers to the concentration of chlorophyll a. Here, represents the dissolved oxygen concentration, and represents the weights of each parameter, determined based on historical samples during training. When the model output value exceeds a set critical value (e.g., the red tide occurrence critical index of 0.8 determined by actual measurements), the system marks the corresponding spatial area as a suspected red tide area.

[0034] The spectral characteristics of the water body in the suspected red tide area are compared and verified with remote sensing image data. The remote sensing data preferably uses reflectance information in the visible and near-infrared bands (such as the B2, B3, and B4 bands of satellite sensors). The spectral reflectance curves extracted from the suspected area are normalized and cosine similarity is calculated with a typical red tide spectral template library. When the spectral matching degree is higher than a preset threshold (preferably above 0.85), the system confirms that there is an abnormal red tide event in the area.

[0035] By combining the above-mentioned red tide anomaly event identification process with a three-level identification mechanism of time series mutation detection, multi-parameter threshold model, and remote sensing spectral verification, the accuracy and real-time performance of red tide monitoring can be improved, and misjudgments caused by a single monitoring parameter can be avoided.

[0036] A red tide impact transmission model is constructed to predict the extent of damage caused by red tides to the ecosystem. Preferably, the steps for constructing a red tide impact transmission model include: Based on the spatiotemporal distribution information of red tide anomalies in the fused dataset, red tide intensity, water quality degradation indicators, and ecosystem response parameters are extracted. A correlation model between red tide intensity and ecosystem damage was established. The correlation model used the rate of change of chlorophyll a concentration, the rate of decrease of dissolved oxygen concentration, and the magnitude of change of nutrient concentration as input features, and the vegetation mortality rate and the rate of decrease of benthic organism density as output indicators. The mapping relationship between red tide impact factors and the degree of ecosystem damage was obtained by training on historical red tide data based on multiple regression analysis. A red tide impact transmission model is constructed using the aforementioned mapping relationship.

[0037] Specifically, spatiotemporal regions identified as red tide anomalies are selected from the fused dataset, and key water quality and ecological characteristic parameters of these regions are extracted during consecutive time periods before and after the event. The extracted parameters include: Red tide intensity indicators: peak chlorophyll a concentration, duration, and spatial diffusion area; Water quality degradation indicators: rate of decrease in dissolved oxygen concentration, cumulative change in nutrients (such as nitrates and phosphates), and pH trend. Ecosystem response parameters: vegetation cover decline rate, vegetation mortality rate, benthic organism density decline rate, and soil oxygen content change.

[0038] Based on the extracted feature parameters, a quantitative correlation model between red tide intensity and ecosystem damage degree is established. Model input features ,in, The change rate of chlorophyll a concentration. The rate of decrease in dissolved oxygen concentration. This represents the range of change in nutrient concentration. For the duration of the red tide, This represents the red tide coverage area. (Model output index) ,in, For vegetation mortality, The rate of decline in benthic organism density was used. By statistically analyzing historical red tide data, the correlation coefficient matrix between input variables and output indicators was calculated; after filtering out multicollinear variables, a stable set of input features was formed.

[0039] A training method combining multiple linear regression and ridge regression is used to establish the mapping relationship between input and output: ; in, The intercept term of the regression equation, These are the model weight coefficients. This is the error term.

[0040] A red tide impact transmission model is constructed using the mapping relationship obtained through training. The model takes the red tide intensity parameter from real-time monitoring data as input, and outputs the ecosystem damage prediction results in real time by substituting the parameter into the regression equation. Furthermore, the prediction output is dynamically compared and corrected with the data from the ecological monitoring module to form a closed-loop self-learning mechanism.

[0041] By constructing the red tide impact transmission model as described above, a quantitative mapping between the intensity of red tide events and the degree of ecosystem degradation was achieved, as well as spatial propagation simulation and trend prediction of red tide ecological impacts.

[0042] A coupled evaluation model of ecological and structural indicators is established, and the weights of ecological and structural indicators in the model are dynamically adjusted based on the prediction results of the red tide anomaly event and the red tide impact transmission model. Preferably, the steps for establishing the coupled evaluation model include: The initial weight relationship between ecological and structural indicators was determined based on the analytic hierarchy process, forming a multi-factor comprehensive evaluation framework. The degree of ecological damage output by the red tide impact transmission model is used as a dynamic adjustment factor to adjust the weights of ecological indicators in real time. The coupling coefficient between ecological and structural indicators is calculated using the fuzzy comprehensive evaluation method to reflect the coupling effect of ecosystem degradation on structural stability. The parameters of the coupling evaluation model are updated based on the dynamic changes in the coupling coefficient.

[0043] Specifically, key indicators reflecting the health status of the ecosystem and seawall structure were extracted from the fused dataset. Ecological indicators included vegetation cover, soil moisture content, soil salinity, vegetation biomass change rate, and biodiversity index; structural indicators included tilt angle, strain value, settlement displacement, and crack propagation rate. To ensure comparability among different indicators, each indicator was normalized to unify its numerical range.

[0044] The initial weighting relationships between ecological and structural indicators were determined using the analytic hierarchy process (AHP). Specifically, based on expert experience and historical monitoring data, pairwise comparison matrices were constructed to determine the relative importance of each indicator. Then, through eigenvector calculation and consistency checks, initial weight vectors were obtained, thereby establishing a multi-factor comprehensive evaluation framework for ecological and structural indicators.

[0045] Based on this, the degree of ecological damage output from the red tide impact transmission model is used as a dynamic adjustment factor to adjust the weights of ecological indicators in real time. That is, when the system detects that a red tide event has led to an increase in the degree of ecological damage, the weights of ecological indicators in the overall model are automatically increased, making the model pay more attention to the impact of ecological degradation on structural safety.

[0046] Finally, the fuzzy comprehensive evaluation method is used to achieve the coupled calculation of ecological and structural indicators.

[0047] Preferably, the implementation process of the fuzzy comprehensive evaluation method includes: Construct a fuzzy evaluation matrix of ecological and structural indicators, where the elements in the fuzzy evaluation matrix represent the degree of membership of each indicator to the health status of the seawall. The membership values ​​of ecological and structural indicators are calculated using triangular membership functions. Based on the dynamic adjustment factor and the initial weight relationship, calculate the fuzzy weight vector of the ecological index and the structural index; By performing fuzzy synthesis operations, the fuzzy evaluation matrix and the fuzzy weight vector are synthesized to obtain the fuzzy comprehensive evaluation result of the coupling coefficient.

[0048] Specifically, a fuzzy evaluation matrix of ecological and structural indicators is first constructed, where each element represents the membership degree of each indicator at different health levels. The membership function is preferably a triangular membership function, and the corresponding health level membership degree is determined based on the monitoring value of each indicator. For example, when the tilt angle is within a small range, the corresponding health membership degree is higher; when the tilt angle exceeds the safety threshold, the membership degree decreases.

[0049] Based on the weighted relationships adjusted by the dynamic adjustment factor, the fuzzy weight vectors of each indicator are calculated. Through fuzzy synthesis operations, the weight vectors and the fuzzy evaluation matrix are comprehensively calculated to obtain the comprehensive fuzzy evaluation result of the seawall's ecological and structural status. This result reflects the comprehensive membership distribution of different health levels and is used to represent the overall health level of the seawall at the current moment.

[0050] To obtain quantitative evaluation results, the fuzzy evaluation results were defuzzified. A comprehensive health score was obtained by calculating the weighted average of each health level. This score is the coupling coefficient between the ecological and structural systems, which can quantitatively reflect the degree of impact of ecological degradation on the stability of the seawall structure.

[0051] Finally, the system adaptively updates the model parameters based on the dynamic changes in the coupling coefficient. When the coupling coefficient drops below a preset threshold, the system automatically triggers the model recalibration process and readjusts the weight matrix; when the changes in the weights of ecological indicators exceed a set proportion, the system automatically performs rolling updates to ensure that the evaluation model can continuously adapt to environmental changes and the impact of ecological disturbances.

[0052] Through the above methods, this invention achieves dynamic coupling modeling of ecological and structural indicators, which can reflect the comprehensive impact of red tides and other ecological anomalies on the structural safety of seawalls in real time.

[0053] The comprehensive health index of the seawall is calculated based on the coupled evaluation model, and graded early warning information is generated according to the preset threshold.

[0054] Preferably, the tiered early warning information includes: When the comprehensive health index of the seawall is greater than 0.8, a low-risk warning is issued; when the comprehensive health index of the seawall is less than 0.5, a high-risk warning is issued; otherwise, a medium-risk warning is issued.

[0055] Example 2 During the summer high-temperature season, a certain ecological seawall in a coastal area experienced red tides in some sea areas due to eutrophication and reduced water exchange. This resulted in increased turbidity, decreased dissolved oxygen, and withered vegetation in the waters along the seawall's edge. Traditional monitoring systems can only detect water quality anomalies and cannot quantitatively assess the impact of red tides on the ecosystem and the stability of the seawall structure, thus leading to technical problems such as delayed early warning and untimely response.

[0056] To address the aforementioned problems, this invention provides a real-time monitoring system for the ecological protection of seawalls based on multi-source data fusion, the structure of which is as follows: Figure 2 As shown. The specific implementation process of this system is as follows: Multiple types of sensor nodes are deployed along typical sections of the seawall, including structural monitoring equipment such as tilt sensors, strain gauges, and settlement monitoring points, as well as ecological monitoring equipment such as vegetation cover sensors, soil moisture sensors, and soil salinity sensors. Simultaneously, multi-parameter water quality monitors are deployed in the waters in front of the seawall to collect real-time data on indicators such as dissolved oxygen, chlorophyll a, water temperature, pH, turbidity, and nutrient concentration.

[0057] The system utilizes remote sensing satellites to periodically acquire NDVI vegetation index and chlorophyll distribution maps of water bodies in the monitored area, and combines this with UAV inspection imagery to visually detect vegetation degradation and surface structural damage on embankments. The system performs spatiotemporal registration of the aforementioned sensor data and remote sensing data through timestamp standardization and spatial coordinate matching, and generates a unified multi-source fusion dataset using Kriging interpolation and time series alignment methods.

[0058] The system detected a significant increase in chlorophyll a concentration and a sharp decrease in dissolved oxygen concentration within a short period of time in the fused dataset, which was identified as a suspected red tide event using a multi-parameter threshold model. Combining the trends in water temperature and nutrient concentration, the system determined that the area had a high risk of red tide and found high reflectance characteristics in the same area in the remote sensing image, thus confirming it as an abnormal red tide event.

[0059] The system invokes a red tide impact transmission model to calculate red tide intensity, water quality degradation, and ecological response parameters, predicting vegetation mortality and benthic organism density decline rates caused by red tides. Based on the model output, the degree of ecological damage is quantified as a reduction in ecological health score, and the system adjusts the weights of ecological indicators in the coupled evaluation model accordingly.

[0060] The system uses the fuzzy comprehensive evaluation method to calculate the coupling coefficient between ecological and structural indicators. Through dynamic weight adjustment and membership analysis, it obtains the comprehensive health index of the seawall at the current moment. The results show that the index has decreased from 0.83 under normal conditions to 0.57, indicating that ecosystem degradation has had a moderate impact on the structural stability of the seawall.

[0061] The system generates tiered early warning information based on changes in the comprehensive health index. When the health index is between 0.5 and 0.8, the system classifies it as a medium-risk level, automatically increases the monitoring frequency, dispatches drones for high-frequency inspections, and sends early warning alerts to the operations and maintenance center. If the health index subsequently drops further below 0.5, the system will automatically enter a high-risk mode, triggering ecological restoration and structural reinforcement plans.

[0062] Through the application of this embodiment, it is possible to jointly identify and predict the ecological degradation and structural safety risks of seawalls in the early stages of red tide occurrence, providing real-time, quantitative, and intelligent technical support for the ecological operation and maintenance of seawalls.

[0063] Finally, it should be noted that the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions described in the foregoing embodiments or make equivalent substitutions for some of the technical features. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A real-time monitoring method for the ecological protection of seawalls based on multi-source data fusion, characterized in that, include: The structure monitoring data, ecological monitoring data and water quality monitoring data of the seawall are collected by IoT sensors, and remote sensing data of the seawall area is obtained by remote sensing equipment. The structural monitoring data, ecological monitoring data, water quality monitoring data, and remote sensing data are spatiotemporally registered to unify the spatiotemporal resolution of different data sources and generate a fused dataset. Based on the water quality characteristic parameters in the fused dataset, red tide anomalies are identified. A red tide impact transmission model is constructed to predict the extent of damage caused by red tides to the ecosystem. A coupled evaluation model of ecological and structural indicators is established, and the weights of ecological and structural indicators in the model are dynamically adjusted based on the prediction results of the red tide anomaly event and the red tide impact transmission model. The comprehensive health index of the seawall is calculated based on the coupled evaluation model, and graded early warning information is generated according to the preset threshold.

2. The method for real-time monitoring of seawall ecology based on multi-source data fusion according to claim 1, characterized in that, The structural monitoring data includes tilt angle, strain value, and settlement displacement; the ecological monitoring data includes vegetation coverage, soil moisture content, and soil salinity; the water quality monitoring data includes chlorophyll a concentration, dissolved oxygen concentration, water temperature, and nutrient content; and the remote sensing data includes the vegetation index NDVI, water chlorophyll concentration distribution data, and high-resolution remote sensing images.

3. The method for real-time monitoring of seawall ecology based on multi-source data fusion according to claim 1, characterized in that, The steps to generate the fused dataset include: Spatial coordinate unification and timestamp standardization were performed on structural monitoring data, ecological monitoring data, water quality monitoring data, and remote sensing data, respectively. The Kriging interpolation algorithm is used to perform interpolation calculations on data with inconsistent spatial resolutions to generate unified spatial grid data. Synchronize data with different sampling frequencies based on time series alignment algorithm to eliminate time delay error; After spatial and temporal registration, the multi-source data is normalized and features are extracted to form a fusion dataset containing unified spatiotemporal features.

4. The method for real-time monitoring of seawall ecology based on multi-source data fusion according to claim 1, characterized in that, The steps to identify red tide anomalies include: Based on the time series mutation detection algorithm, abnormal fluctuation ranges of chlorophyll a concentration and dissolved oxygen concentration are identified; By combining a multi-parameter threshold model of water temperature and nutrient content, a secondary determination is made on the abnormal fluctuation range; When the determination result meets the characteristics of red tide formation, the corresponding spatial area is marked as a suspected red tide area; The spectral characteristics of the water body in the suspected area are compared and verified with remote sensing image data. If the matching degree is higher than the preset threshold, it is confirmed as a red tide anomaly.

5. The method for real-time monitoring of seawall ecology based on multi-source data fusion according to claim 1, characterized in that, The steps to construct a red tide impact transmission model include: Based on the spatiotemporal distribution information of red tide anomalies in the fused dataset, red tide intensity, water quality degradation indicators, and ecosystem response parameters are extracted. A correlation model between red tide intensity and ecosystem damage was established. The correlation model used the rate of change of chlorophyll a concentration, the rate of decrease of dissolved oxygen concentration, and the magnitude of change of nutrient concentration as input features, and the vegetation mortality rate and the rate of decrease of benthic organism density as output indicators. The mapping relationship between red tide impact factors and the degree of ecosystem damage was obtained by training on historical red tide data based on multiple regression analysis. A red tide impact transmission model is constructed using the aforementioned mapping relationship.

6. The method for real-time monitoring of seawall ecology based on multi-source data fusion according to claim 1, characterized in that, The steps to establish a coupled evaluation model include: The initial weight relationship between ecological and structural indicators was determined based on the analytic hierarchy process, forming a multi-factor comprehensive evaluation framework. The degree of ecological damage output by the red tide impact transmission model is used as a dynamic adjustment factor to adjust the weights of ecological indicators in real time. The coupling coefficient between ecological and structural indicators is calculated using the fuzzy comprehensive evaluation method to reflect the coupling effect of ecosystem degradation on structural stability. The parameters of the coupling evaluation model are updated based on the dynamic changes in the coupling coefficient.

7. The method for real-time monitoring of seawall ecology based on multi-source data fusion according to claim 6, characterized in that, The implementation process of the fuzzy comprehensive evaluation method includes: Construct a fuzzy evaluation matrix of ecological and structural indicators, where the elements in the fuzzy evaluation matrix represent the degree of membership of each indicator to the health status of the seawall. The membership values ​​of ecological and structural indicators are calculated using triangular membership functions. Based on the dynamic adjustment factor and the initial weight relationship, calculate the fuzzy weight vector of the ecological index and the structural index; By performing fuzzy synthesis operations, the fuzzy evaluation matrix and the fuzzy weight vector are synthesized to obtain the fuzzy comprehensive evaluation result of the coupling coefficient.

8. The method for real-time monitoring of seawall ecology based on multi-source data fusion according to claim 1, characterized in that, The tiered early warning information includes: When the comprehensive health index of the seawall is greater than 0.8, a low-risk warning is issued; when the comprehensive health index of the seawall is less than 0.5, a high-risk warning is issued; otherwise, a medium-risk warning is issued.

9. A real-time monitoring system for the ecological protection of seawalls based on multi-source data fusion, characterized in that, The system for implementing the real-time monitoring method for the ecological protection of seawalls according to any one of claims 1-8 includes: The data acquisition module is used to collect structural monitoring data, ecological monitoring data and water quality monitoring data of the seawall through IoT sensors, and to acquire remote sensing data of the seawall area through remote sensing equipment. The data fusion module is used to perform spatiotemporal registration on the structural monitoring data, ecological monitoring data, water quality monitoring data and remote sensing data, unify the spatiotemporal resolution of different data sources, and generate a fused dataset. The red tide event identification module is used to identify abnormal red tide events based on the water quality characteristic parameters in the fused dataset; An ecological damage assessment module is used to construct a red tide impact transmission model, which is used to predict the degree of damage that red tides cause to the ecosystem. The coupled evaluation module is used to establish a coupled evaluation model of ecological indicators and structural indicators, and dynamically adjust the weights of ecological indicators and structural indicators in the model based on the prediction results of the red tide anomaly event and the red tide impact transmission model. The graded early warning module is used to calculate the comprehensive health index of the seawall based on the coupled evaluation model and generate graded early warning information according to preset thresholds.

Images