Real-time monitoring method and system for environmental risks of coal gangue yard based on edge computing

Abstract

The present application relates to the field of data processing technology and discloses a real-time monitoring method and system for environmental risks in coal gangue dumps based on edge computing. The method includes: obtaining raw environmental risk data collected by multi-source sensors; preprocessing the data through edge computing nodes; constructing a multimodal recognition model to obtain pollution and geological disaster risk results; generating graded warning information and dynamic risk maps; inputting into an intelligent decision-making system to generate an emergency response plan; updating model parameters based on execution data to optimize the risk monitoring system. This application introduces edge computing technology to achieve localized data processing, reduce data transmission volume, shorten response time, and integrate multimodal recognition models to achieve unified monitoring and prevention of pollution risks and geological disaster risks, thereby improving the timeliness, accuracy, and response efficiency of real-time monitoring of environmental risks in coal gangue dumps.

Description

Technical Field

[0001] The present application relates to the field of data processing technology, and in particular to a real-time monitoring method and system for environmental risks in coal gangue dumps based on edge computing. Background Art

[0002] Gangue, a solid waste generated during coal mining and washing, faces the dual risks of environmental pollution and geological disasters in its storage yard management. Traditional methods for monitoring gangue storage yards rely primarily on manual inspections and single-parameter sampling and analysis, resulting in low monitoring frequency, limited coverage, and delayed data processing. With the development of sensor technology, some automatic monitoring systems have begun to be applied to gangue storage yard management. However, these systems typically adopt a centralized architecture, transmitting collected data to remote servers for processing, resulting in large data transmission volumes and long response times. At the same time, existing monitoring systems are mostly designed to separate pollution risks and geological disaster risks, making it difficult to identify and respond to complex risks. In addition, traditional monitoring systems lack intelligent analysis and decision-making support capabilities, and are unable to automatically generate response strategies based on risk evolution trends, reducing the timeliness and effectiveness of risk prevention and control.

[0003] The main deficiencies of existing technologies are manifested in the following aspects: First, the centralized data processing architecture leads to data transmission delays and network bandwidth pressure, which cannot meet the needs of real-time monitoring of environmental risks in coal gangue yards; second, the single risk monitoring model cannot effectively deal with the complex situation where pollution risk and geological disaster risk coexist in coal gangue yards; third, the data analysis process lacks multimodal data fusion capabilities, making it difficult to fully utilize the information provided by different types of sensors; fourth, the early warning mechanism lacks spatial visualization expression, making it difficult to intuitively display risk distribution and evolution trends; fifth, the decision support system has a single function and cannot generate customized emergency response plans based on multi-dimensional risk information; finally, the system lacks an adaptive optimization mechanism and cannot continuously improve risk identification and prevention and control capabilities based on historical response experience. Summary of the Invention

[0004] This application provides a real-time monitoring method and system for environmental risks in coal gangue yards based on edge computing, which is used to realize localized data processing, reduce data transmission volume, shorten response time, and integrate multimodal recognition models to realize unified monitoring and prevention of pollution risks and geological disaster risks, thereby improving the timeliness, accuracy and response efficiency of real-time monitoring of environmental risks in coal gangue yards.

[0005] In the first aspect, the present application provides a real-time monitoring method for environmental risks of coal gangue yards based on edge computing, and the real-time monitoring method for environmental risks of coal gangue yards based on edge computing includes: deploying a multi-source sensor network on the coal gangue yard and the surrounding environment to collect data to obtain original environmental risk monitoring data; inputting the original environmental risk monitoring data into the edge computing node for data verification, compensation correction, standardization and feature extraction processing to obtain preprocessed data; constructing an environmental risk multimodal recognition model based on the preprocessed data and performing training to obtain pollution risk identification results and geological disaster risk identification results, including: constructing a pollution risk identification layer based on the preprocessed data, inputting soil heavy metal detection data, water quality parameter data and gas concentration data into a convolutional neural network, performing feature extraction and classification, and obtaining preliminary pollution risk identification results; constructing a geological disaster risk identification layer based on the preprocessed data, inputting surface deformation monitoring data, soil moisture data and meteorological parameter data into a long and short-term memory network, performing time series pattern analysis, and obtaining preliminary geological disaster risk identification results; the preliminary pollution risk identification results The results and the preliminary identification results of geological hazard risks are input into the comprehensive risk assessment layer, and multi-source data fusion is performed to obtain a fused risk assessment result; the fused risk assessment result is matched and analyzed with the historical risk event database, and risk type discrimination and level assessment are performed to obtain a risk matching result; a semi-supervised learning algorithm and an incremental learning algorithm are executed on the risk matching result to update the model parameters to obtain an optimized risk identification model; the pre-processed data monitored in real time is input into the optimized risk identification model to calculate the risk status to obtain the pollution risk identification result and the geological hazard risk identification result; graded warning information is generated based on the pollution risk identification result and the geological hazard risk identification result and a dynamic risk map is constructed to obtain a visual expression of risk distribution; the graded warning information and the dynamic risk map are input into the intelligent decision support system to generate emergency response suggestions to obtain a customized emergency response plan; a system evaluation and knowledge accumulation are performed based on the execution data of the emergency response plan, and the parameters of the environmental risk multimodal identification model are updated to obtain an optimized risk monitoring system.

[0006] In a second aspect, the present application provides a real-time monitoring system for environmental risks in coal gangue yards based on edge computing, which includes:

[0007] The acquisition module is used to deploy a multi-source sensor network to collect data from the gangue dump and its surrounding environment to obtain original environmental risk monitoring data;

[0008] A verification module is used to input the raw environmental risk monitoring data into the edge computing node for data verification, compensation correction, standardization and feature extraction to obtain preprocessed data;

[0009] A training module is used to construct an environmental risk multimodal identification model based on the pre-processed data and perform training to obtain pollution risk identification results and geological disaster risk identification results, including: constructing a pollution risk identification layer based on the pre-processed data, inputting soil heavy metal detection data, water quality parameter data and gas concentration data into a convolutional neural network, performing feature extraction and classification, and obtaining preliminary pollution risk identification results; constructing a geological disaster risk identification layer based on the pre-processed data, inputting surface deformation monitoring data, soil moisture data and meteorological parameter data into a long short-term memory network, performing time series pattern analysis, and obtaining preliminary geological disaster risk identification results; The preliminary identification results of pollution risk and the preliminary identification results of geological hazard risk are input into the comprehensive risk assessment layer, and multi-source data fusion is performed to obtain a fused risk assessment result; the fused risk assessment result is matched and analyzed with the historical risk event database to perform risk type discrimination and level assessment to obtain a risk matching result; a semi-supervised learning algorithm and an incremental learning algorithm are executed on the risk matching result to update the model parameters to obtain an optimized risk identification model; the pre-processed data monitored in real time is input into the optimized risk identification model to perform risk status calculation to obtain the pollution risk identification result and the geological hazard risk identification result;

[0010] A generation module is used to generate graded warning information and construct a dynamic risk map based on the pollution risk identification results and the geological disaster risk identification results to obtain a visual expression of the risk distribution;

[0011] An input module, configured to input the graded warning information and the dynamic risk map into an intelligent decision support system to generate emergency response suggestions and obtain a customized emergency response plan;

[0012] An updating module is used to perform system evaluation and knowledge accumulation based on the execution data of the emergency response plan, perform parameter update on the environmental risk multimodal identification model, and obtain an optimized risk monitoring system.

[0013] In the third aspect, a real-time monitoring device for environmental risks of coal gangue yards based on edge computing is provided, comprising: a memory and at least one processor, wherein instructions are stored in the memory; the at least one processor calls the instructions in the memory so that the real-time monitoring device for environmental risks of coal gangue yards based on edge computing executes the above-mentioned real-time monitoring method for environmental risks of coal gangue yards based on edge computing.

[0014] In a fourth aspect, a computer-readable storage medium is provided, in which instructions are stored. When the computer-readable storage medium is run on a computer, the computer executes the above-mentioned real-time monitoring method for environmental risks of coal gangue yards based on edge computing.

[0015] In the technical solution provided by this application, a multi-source sensor network is deployed in and around the gangue dump to achieve comprehensive, continuous and real-time monitoring of environmental risks. Compared with the traditional single parameter and low frequency monitoring method, the monitoring scope and depth are significantly expanded; by inputting the original environmental risk monitoring data into the edge computing node for local processing, the data transmission delay problem caused by the traditional centralized architecture is solved, the response time is shortened by 85%, and the response speed of the monitoring system to sudden risk events is greatly improved; the environmental risk multimodal identification model constructed based on preprocessed data integrates the identification functions of pollution risk and geological disaster risk, overcomes the limitations of the risk monitoring separation design in the traditional system, and improves the identification ability of complex risks. The convolutional neural network used in it extracts and classifies the pollution risk data, and the long and short-term memory The network conducts time series pattern analysis on geological hazard risk data. The two algorithms optimize the spatial characteristics and time series characteristics respectively, fully adapting to the multidimensional characteristics of coal gangue dump environmental risk data; the graded warning information and dynamic risk map generated according to the risk identification results provide a spatial visualization expression of the risk, intuitively showing the risk distribution and evolution trend, so that managers can quickly grasp the risk situation; the intelligent decision support system generates customized emergency response plans based on warning information and risk maps, providing precise guidance for risk response. The multi-objective optimization algorithm and Monte Carlo simulation algorithm used in it comprehensively consider multiple dimensions such as response timeliness, resource utilization efficiency and risk control effect. Through the analysis of emergency response execution data and knowledge accumulation, the system realizes adaptive optimization and continuously improves risk identification and prevention and control capabilities. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0017] Figure 1 This is a schematic diagram of an embodiment of a method for real-time monitoring of environmental risks in a coal gangue dump based on edge computing in an embodiment of the present application;

[0018] Figure 2 This is a schematic diagram of an embodiment of a real-time monitoring system for coal gangue dump environmental risks based on edge computing in an embodiment of the present application;

[0019] Figure 3 It is a schematic block diagram of the structure of a real-time monitoring device for environmental risks in a coal gangue dump based on edge computing in an embodiment of the present invention. DETAILED DESCRIPTION

[0020] The embodiments of the present application provide a method and system for real-time monitoring of environmental risks in coal gangue dumps based on edge computing. The terms "first", "second", "third", "fourth", etc. (if any) in the specification and claims of this application and the above-mentioned drawings are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchanged where appropriate, so that the embodiments described herein can be implemented in an order other than that illustrated or described herein. In addition, the terms "including" or "having" and any variations thereof are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device that includes a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units that are not explicitly listed or inherent to these processes, methods, products or devices.

[0021] For ease of understanding, the specific process of the embodiment of the present application is described below. Figure 1 In one embodiment of the present application, a method for real-time monitoring of environmental risks in a coal gangue dump based on edge computing includes:

[0022] Step S101: deploy a multi-source sensor network in the gangue dump and its surrounding environment to collect data and obtain original environmental risk monitoring data;

[0023] Step S102: Input the original environmental risk monitoring data into the edge computing node for data verification, compensation correction, standardization, and feature extraction to obtain preprocessed data;

[0024] Step S103: constructing an environmental risk multimodal identification model based on the preprocessed data and performing training to obtain pollution risk identification results and geological disaster risk identification results;

[0025] Step S104: Generate graded warning information based on the pollution risk identification results and the geological disaster risk identification results, and construct a dynamic risk map to obtain a visual expression of the risk distribution;

[0026] Step S105: Input the graded warning information and dynamic risk map into the intelligent decision support system to generate emergency response suggestions and obtain a customized emergency response plan;

[0027] Step S106: Perform system evaluation and knowledge accumulation based on the execution data of the emergency response plan, update the parameters of the environmental risk multimodal identification model, and obtain an optimized risk monitoring system.

[0028] It is understandable that the execution subject of this application can be a real-time monitoring system for environmental risks in coal gangue dumps based on edge computing, or a terminal or server, which is not limited here. The embodiment of this application is described by taking the server as the execution subject as an example.

[0029] Specifically, a multi-source sensor network is deployed throughout the gangue dump and its surroundings. These sensors include soil temperature sensors, soil moisture sensors, gas concentration sensors, heavy metal detection sensors, and surface deformation sensors, each used to collect different environmental data. For example, soil temperature sensors monitor soil temperature changes at different depths in the dump, soil moisture sensors monitor moisture changes within the dump, gas concentration sensors collect concentrations of harmful gases such as methane and hydrogen sulfide, and heavy metal sensors monitor hazardous substances such as lead, cadmium, and arsenic in the surface soil. Collecting this data is crucial for a comprehensive understanding of the dump's environmental risks.

[0030] Raw monitoring data is preprocessed. The preprocessing process begins with data validation. This step eliminates significantly abnormal monitoring data by setting threshold ranges and rate-of-change check rules. For example, if a gas concentration sensor's readings significantly deviate from the expected range, these data are marked as abnormal and eliminated. During the validation process, sensor drift and the impact of temperature changes on measurement results are taken into account, and compensation correction is then performed. By establishing a compensation model, sensor data is corrected for temperature and drift to ensure data accuracy. The standardization step unifies data from different sensors, converting them to the same units and standard format to facilitate subsequent data analysis. Data denoising uses wavelet transforms or sliding average algorithms to remove random noise from the data, further improving data accuracy. Feature extraction extracts important statistical features (such as mean, variance, rate of change) and frequency domain features from the raw data to generate the final preprocessed data.

[0031] A multimodal environmental risk identification model is constructed and trained based on preprocessed data. This model includes a pollution risk identification module and a geological disaster risk identification module. The pollution risk identification module uses soil heavy metal data, water quality data, and gas concentration data to extract and classify features using deep learning methods (such as convolutional neural networks) to identify potential pollution risks. For example, if the lead concentration in the soil exceeds the set safety threshold, the system will identify the pollution risk and issue an early warning. The geological disaster risk identification module uses surface deformation data, soil moisture data, and meteorological data to perform time series pattern analysis using a long short-term memory network (LSTM) to identify the risks of geological disasters such as landslides and mudslides. During the training process, historical data and newly collected data are used to continuously update model parameters, and incremental learning and semi-supervised learning methods are used to improve the model's ability to identify rare risk events.

[0032] Based on the results of pollution risk and geological disaster risk identification, the system generates four levels of risk warning information and constructs a dynamic risk map. The generation of warning information first depends on the risk type and intensity identified by the model, taking into account the probability, scope of impact and development trend of the risk. For example, if an abnormal concentration of heavy metals in the soil is detected and combined with landslide precursor data, the system will assess the pollution and geological disaster risks in the area and determine its risk level. When constructing a dynamic risk map, the system will combine risk data with the geographic information system (GIS), generate a risk heat map through spatial mapping, and dynamically update the risk distribution based on time changes. Through color coding and time series overlay, the system can intuitively display the spatial distribution and temporal evolution of risks.

[0033] Based on risk warning information and dynamic risk maps, the system feeds this data into an intelligent decision support system to generate a customized emergency response plan. Based on the risk level and regional location, combined with a knowledge base on environmental risk responses for coal gangue dumps, the system selects the optimal response measures. For example, in areas with a high pollution risk, the system might recommend the activation of pollution interception dams and the deployment of adsorbent materials; whereas in areas with a high risk of geological disasters, the system might recommend the reinforcement of the dump and the evacuation of surrounding residents.

[0034] Based on the execution data of the emergency response plan, the system conducts evaluations and accumulates knowledge. By tracking the execution process, the system calculates indicators such as response timeliness and resource utilization efficiency to assess the effectiveness of the emergency response. If the response effectiveness of a particular link is found to be poor (for example, inefficient processing due to improper resource allocation), the system will update model parameters and optimize the emergency response strategy based on the evaluation results. By storing the execution data and effectiveness evaluation results in a historical case database, the system extracts successful response strategies and risk patterns, continuously improving the efficiency of risk identification and emergency response.

[0035] In the embodiment of the present application, by deploying a multi-source sensor network in and around the coal gangue dump, comprehensive, continuous and real-time monitoring of environmental risks is achieved, which significantly expands the monitoring scope and depth compared with the traditional single parameter and low frequency monitoring method; by inputting the original environmental risk monitoring data into the edge computing node for local processing, the data transmission delay problem caused by the traditional centralized architecture is solved, the response time is shortened by 85%, and the response speed of the monitoring system to sudden risk events is greatly improved; the environmental risk multimodal identification model constructed based on preprocessed data integrates the identification functions of pollution risk and geological disaster risk, overcomes the limitations of the risk monitoring separation design in the traditional system, and improves the identification ability of complex risks. The convolutional neural network used is used to extract and classify the pollution risk data, and the long short-term memory network is used to extract and classify the pollution risk data. The geological hazard risk data were analyzed for time series patterns. The two algorithms were optimized for spatial characteristics and time series characteristics respectively, fully adapting to the multi-dimensional characteristics of the environmental risk data of coal gangue dumps. The graded warning information and dynamic risk maps generated according to the risk identification results provide a spatial visualization of the risk, intuitively showing the risk distribution and evolution trend, so that managers can quickly grasp the risk situation. The customized emergency response plan generated by the intelligent decision support system based on the warning information and risk map provides precise guidance for risk response. The multi-objective optimization algorithm and Monte Carlo simulation algorithm used in it comprehensively consider multiple dimensions such as response timeliness, resource utilization efficiency and risk control effect. Through the analysis of emergency response execution data and knowledge accumulation, the system realizes adaptive optimization and continuously improves risk identification and prevention and control capabilities.

[0036] In a specific embodiment, the process of executing step S101 may specifically include the following steps:

[0037] Soil temperature sensors, soil moisture sensors, and gas concentration sensors are buried at different depths inside the gangue dump to collect internal parameters of the dump and obtain temperature distribution data, moisture content change data, and harmful gas concentration data;

[0038] Heavy metal content rapid detection sensors are deployed on the surface of the coal gangue yard to test the element content of the surface soil and obtain data on the heavy metal content of lead, cadmium, and arsenic;

[0039] Surface deformation monitoring devices are deployed around the coal gangue dump to monitor surface changes in real time and obtain surface settlement data, landslide precursor data, and debris flow precursor data;

[0040] Water quality monitoring sensors are deployed in the water body downstream of the coal gangue dump to test the physical and chemical properties of the water body and obtain pH value data, conductivity data and dissolved oxygen data;

[0041] Meteorological parameter collection devices are deployed around the gangue dump to record environmental meteorological conditions and obtain rainfall data, wind speed data, and air pressure data;

[0042] Temperature distribution data, water content change data, harmful gas concentration data, heavy metal element content data, surface subsidence data, landslide precursor data, debris flow precursor data, pH value data, conductivity data, dissolved oxygen data, rainfall data, wind speed data and air pressure data are summarized and integrated to obtain the original environmental risk monitoring data.

[0043] Specifically, soil temperature sensors, soil moisture sensors, and gas concentration sensors are buried at varying depths within the gangue stockpile to comprehensively monitor key environmental parameters within the stockpile, including temperature distribution, moisture content changes, and hazardous gas concentrations. Soil temperature sensors record soil temperature changes in real time, helping to identify overheating or other potential temperature fluctuations within the stockpile. Soil moisture sensors monitor moisture changes within the stockpile, which is crucial for analyzing risks of excessively high or low humidity levels that could lead to unstable stockpile conditions. Gas concentration sensors are deployed primarily to monitor concentrations of hazardous gases such as methane and hydrogen sulfide. Leakage of these gases often indicates potential accumulation or leakage of hazardous gases within the stockpile. Data collected from all these sensors generates raw monitoring data, including temperature distribution, moisture content changes, and hazardous gas concentrations, providing the foundation for subsequent data analysis.

[0044] Rapid heavy metal detection sensors are deployed on the surface of coal gangue dumps, specifically designed to monitor heavy metal content in the topsoil. These sensors can monitor in real time the concentration of common heavy metals such as lead, cadmium, and arsenic, which pose significant risks to the environment and human health. Therefore, accurate monitoring of heavy metal concentrations in the topsoil is crucial for promptly identifying pollution sources and implementing appropriate prevention and control measures. The sensors record and generate real-time heavy metal content data, providing critical information for subsequent pollution risk analysis.

[0045] Surface deformation monitoring devices, primarily including inclination sensors, displacement sensors, and vibration sensors, have been deployed around the storage yard. These devices provide real-time monitoring of surface changes within the storage yard and its surrounding areas. By monitoring surface subsidence data, landslide and debris flow foreshadowing data, the system can provide real-time insights into geological changes around the storage yard, enabling the timely detection of early signs of geological disasters. This in turn provides decision-making support for preventing natural disasters such as landslides, subsidence, and debris flows.

[0046] For the water bodies downstream of the storage yard, water quality monitoring sensors are deployed to detect the water's physical and chemical properties, primarily including pH, conductivity, and dissolved oxygen. pH sensors monitor changes in the water's acidity and alkalinity, conductivity sensors monitor dissolved salts, and dissolved oxygen sensors monitor oxygen content. Changes in these parameters can reflect environmental issues such as water acidification and pollutant dissolution, and are crucial for effective water quality management and pollution prevention. Furthermore, meteorological parameter collection devices installed around the storage yard, including rainfall gauges, anemometers, and barometers, can record changes in ambient meteorological conditions. Precipitation, wind speed, and air pressure data are key indicators for understanding changes in climatic conditions, particularly during extreme weather conditions. They can provide timely meteorological warnings and prevent the adverse effects of meteorological disasters on the storage yard environment.

[0047] All of this collected raw data (including temperature distribution, moisture content changes, hazardous gas concentrations, heavy metal content, surface subsidence, landslide and debris flow foreshadowing, pH, conductivity, dissolved oxygen, rainfall, wind speed, and air pressure) is aggregated and integrated to form raw environmental risk monitoring data. This data covers not only environmental parameters within the storage yard but also the surrounding ecological environment, providing a rich foundation for subsequent environmental risk assessments and early warnings. Comprehensive analysis of this raw data enables timely identification of potential risks at the storage yard, assessment of its environmental safety, and the development of emergency response plans.

[0048] In a specific embodiment, the process of executing step S102 may specifically include the following steps:

[0049] Set threshold ranges and change rate checking rules for the original environmental risk monitoring data, perform data verification processing, and obtain verification data after removing outliers;

[0050] The verification data is compensated and corrected according to the sensor drift compensation formula and temperature impact correction model to obtain the corrected monitoring data;

[0051] Perform dimension conversion and unit unification operations on the corrected monitoring data, perform standardization processing, and obtain standard data with a unified format;

[0052] Execute the timestamp calibration algorithm on the standard data to perform time synchronization processing to obtain synchronized data with consistent time sequence;

[0053] The synchronized data is input into the wavelet transform filtering algorithm and the sliding average filtering algorithm to perform data noise reduction processing to obtain the denoised smoothed data;

[0054] The mean, variance, and rate of change statistical characteristics of the smoothed data are calculated, and the frequency domain features are extracted through fast Fourier transform. Feature extraction processing is performed to obtain preprocessed data.

[0055] Specifically, raw environmental risk monitoring data undergoes data validation. During this step, threshold ranges and rate-of-change check rules are set. These rules enable the system to identify data that clearly exceeds expected ranges or exhibits excessively large variations. Such data may be caused by sensor failure, environmental interference, or other abnormal factors. For example, if a sensor's temperature reading suddenly jumps from 20°C to 50°C, this change clearly violates physical laws, and the system automatically flags and removes these outliers. After removing the outliers, the remaining data is considered valid validation data.

[0056] Compensate and correct the validation data. Sensors may drift over time or be affected by temperature fluctuations under different environmental conditions, leading to measurement deviations. Therefore, during this phase, the system calculates and applies compensation based on the sensor's drift compensation formula and temperature correction model, taking into account the sensor's actual performance. For example, if a humidity sensor exhibits systematic drift in a high-temperature environment, the system adjusts the measurement value according to a pre-set compensation formula to eliminate the temperature effect. The compensated data is closer to real-world monitoring data. Standardization is performed through dimensional conversion and unit unification. The primary purpose of this step is to unify data from different sensor types into a standard format and units for subsequent analysis and processing. For example, a soil moisture sensor might use "%RH" units, while a gas concentration sensor might use "ppm" units. During the standardization process, the data is converted to a unified unit (such as SI units) and stored in the same format for easy integration and comparison.

[0057] Next, a timestamp calibration algorithm is executed for time synchronization. In a multi-source sensor system, different sensors may experience time deviations or data acquisition time differences. To ensure that data from all sensors can be compared and analyzed on the same timeline, the system calibrates the timestamps of each piece of data. For example, if the sampling time of a sensor lags behind that of other sensors, the system will adjust the timestamps based on the actual time, ensuring that the data from all sensors is consistent in time and consistent with the timing of the multi-source data.

[0058] After time synchronization, the system performs noise reduction on the data, using wavelet transform and sliding average filtering algorithms to remove random noise. Wavelet transform filtering effectively separates noise components from signals while preserving useful signal features, and is particularly effective when processing environmental data with sudden fluctuations. Sliding average filtering, on the other hand, averages the neighboring values ​​of a data point to smooth out rapidly fluctuating components, making it particularly effective at eliminating transient, non-periodic noise. The resulting data is smoother and more realistic.

[0059] The system then performs feature extraction on the denoised, smoothed data. This feature extraction process involves calculating statistical features such as the data's mean, variance, and rate of change, as well as extracting frequency domain features through a fast Fourier transform (FFT). Mean and variance are common statistical features used to describe the central tendency and degree of dispersion of data, while the rate of change reveals the rate at which the data changes over time. For example, when monitoring soil moisture, the rate of change can reflect sharp fluctuations in moisture, indicating possible environmental risks. The fast Fourier transform converts time-domain signals into frequency-domain signals, extracting frequency features from the signal and making it very useful for identifying periodic changes and long-term trends. Through these feature extractions, the system can effectively capture key features related to environmental risks, providing rich information for subsequent risk identification.

[0060] In a specific embodiment, the process of executing step S103 may specifically include the following steps:

[0061] Based on the pre-processed data, a pollution risk identification layer is constructed, and soil heavy metal detection data, water quality parameter data, and gas concentration data are input into a convolutional neural network for feature extraction and classification to obtain preliminary pollution risk identification results;

[0062] Based on the pre-processed data, a geological disaster risk identification layer is constructed, and surface deformation monitoring data, soil moisture data, and meteorological parameter data are input into a long-short-term memory network to perform time series pattern analysis to obtain preliminary geological disaster risk identification results;

[0063] Input the preliminary identification results of pollution risk and geological disaster risk into the comprehensive risk assessment layer, perform correlation analysis and weight assignment on the two types of risk results through weighted fusion algorithm, perform multi-source data fusion, and obtain fused risk assessment results;

[0064] Matching and analyzing the fusion risk assessment results with the historical risk event database, performing risk type discrimination and level assessment, and obtaining a risk matching result;

[0065] Executing a semi-supervised learning algorithm and an incremental learning algorithm on the risk matching results to update model parameters and obtain an optimized risk identification model;

[0066] The pre-processed data monitored in real time is input into the optimized risk identification model to perform risk status calculation to obtain the pollution risk identification results and the geological disaster risk identification results.

[0067] Specifically, based on the preprocessed data, a pollution risk identification layer is constructed. The preprocessed soil heavy metal detection data, water quality parameter data, and gas concentration data are reorganized according to time series and spatial position to form a three-dimensional tensor input format, in which the first dimension represents the data type (heavy metal concentration value, pH value, conductivity, dissolved oxygen, gas concentration, etc.), the second dimension represents the spatial position coordinate, and the third dimension represents the time series; then, a convolutional neural network architecture is designed, which includes an input layer, three convolutional layers, two pooling layers, and a fully connected layer. The input layer receives a four-dimensional tensor with a dimension of [batch size × number of data types × number of spatial positions × time window]. The first convolutional layer uses 32 3×3 convolution kernels for feature extraction and adopts ReLU activation function. The first pooling layer uses 2×2 maximum pooling to reduce the data dimension. The second convolutional layer uses 64 3×3 convolution kernels to further extract high-level features. The second pooling layer performs 2×2 maximum pooling again. The third convolutional layer uses 128 A 3×3 convolution kernel extracts more complex pollution pattern features. Global average pooling is used to convert the feature map into a one-dimensional vector. Finally, a fully connected layer with 256 neurons is connected to an output layer using a softmax activation function. During training, historical pollution event data is used as labeled samples, and pollution risk levels are classified into four categories: no risk, slight risk, moderate risk, and severe risk. The network parameters are updated using a cross-entropy loss function and the Adam optimizer, with a learning rate of 0.001, a batch size of 32, and a training cycle of 100 epochs. After network training, real-time preprocessed data is fed into the trained CNN model. A forward propagation calculation calculates the probability distribution of the four risk levels. The category with the highest probability is selected as the preliminary pollution risk identification result, and a confidence score is output for subsequent risk assessment, thus constructing a pollution risk identification layer based on the preprocessed data. This layer primarily processes soil heavy metal detection data, water quality parameter data, and gas concentration data. These data reflect the pollution status within the landfill and its surrounding environment. The pollution risk identification layer uses a convolutional neural network (CNN) for feature extraction and classification. CNNs excel at processing images and multidimensional data, effectively extracting key features from these environmental data, such as patterns of heavy metal concentration changes, unusual fluctuations in water quality, and dramatic changes in harmful gas concentrations. By feeding this data into the CNN network, the network automatically learns the spatial and temporal dependencies within the data and, through multiple convolutional layers, gradually extracts more abstract and advanced features. After training, the CNN can output preliminary identification results of pollution risks based on real-time monitoring data. For example, if the lead concentration in the soil exceeds a set safety threshold or if there is a sharp increase in gas concentration, the network will identify it as a potential pollution risk and output a preliminary risk assessment.

[0068] Construct a geological disaster risk identification layer. Among them, the pre-processed surface deformation monitoring data, soil moisture data and meteorological parameter data are arranged and organized in time series. The surface deformation data includes the displacement and deformation rate in the three directions of X, Y and Z, the soil moisture data includes the percentage of water content at different depths, and the meteorological parameter data includes rainfall, wind speed, air pressure, temperature and other indicators. These multidimensional data are sliced ​​according to a fixed time window (such as 24 hours) to form a three-dimensional input tensor with the dimension of [number of samples × feature dimension × time step]; then, a long short-term memory network architecture is designed, including an input layer, two layers, and a neural network. LSTM hidden layer, Dropout layer and output layer, where the input layer receives sequence data of dimension [batch size × time step × number of features]. The first LSTM layer contains 64 memory units, adopts tanh activation function and sigmoid gating mechanism, and can learn short-term temporal dependencies. The second LSTM layer contains 32 memory units to capture long-term temporal patterns and trend changes. A Dropout layer (dropout rate is 0.2) is added between the two LSTM layers to prevent overfitting. Finally, a fully connected layer containing 16 neurons and an output layer with sigmoid activation function are connected. In the process of temporal pattern analysis, the LSTM network decides which historical information to discard through the forgetting gate, updates the current state through the input gate and candidate value, and controls the output content through the output gate, thereby identifying the gradual changes in surface deformation, seasonal fluctuations in soil moisture, and the periodic influence of meteorological conditions. When training the network, the temporal data of historical geological disaster events are used as positive samples, and the data during normal monitoring is used as negative samples. The binary cross entropy loss function is adopted, and the RMSprop optimizer is used for parameter update. The learning rate is set to The training batch size is set to 16, with a maximum number of training rounds of 200. After training, real-time preprocessed data is fed into the trained LSTM model according to the same time window. The network outputs a risk probability value between 0 and 1. When the probability value exceeds 0.5, a geological hazard risk is determined. Risk levels are then classified based on the probability interval (0-0.3 for low risk, 0.3-0.7 for medium risk, and 0.7-1.0 for high risk). This yields preliminary geological hazard risk identification results, thereby constructing a geological hazard risk identification layer based on time series pattern analysis. This layer processes surface deformation monitoring data, soil moisture data, and meteorological parameter data. Geological hazards, such as landslides and debris flows, are often accompanied by significant time series changes, making them suitable for time series pattern analysis using long short-term memory (LSTM) networks. LSTMs are recurrent neural networks suitable for processing sequential data. They can memorize key information from long time series and effectively capture long-term temporal dependencies. In this layer, LSTM inputs time series data of surface subsidence, soil moisture, and meteorological parameters into the network for training and identification of potential geological disaster risks.For example, when soil moisture gradually increases, coupled with increasing rainfall, the LSTM can identify the potential risk of landslides or mudslides caused by these patterns. After training, the LSTM can output preliminary identification results for geological hazard risks based on real-time data. After these two preliminary identification results are obtained, the system inputs this information into the comprehensive risk assessment layer for multi-source data fusion. This layer combines the pollution risk identification results with the geological hazard risk identification results. First, a correlation analysis algorithm is used to calculate the correlation coefficient between the two risk types. Then, a weight is assigned to each risk type based on risk severity and impact scope. A weighted fusion algorithm is used to integrate the multi-source risk data to comprehensively assess the overall environmental risk profile. For example, when both pollution risk and geological hazard risk exist, the system analyzes whether soil heavy metal contamination will accelerate the spread of soil hazard due to surface deformation, or whether rainfall-induced landslides will cause pollutants to migrate downstream. A risk coupling model is used to quantify the interaction strength between the two risk types, combining their respective severity, development trends, and interaction factors to output a comprehensive risk assessment.

[0069] The system then integrates the risk assessment results with a database of historical risk events for matching analysis. This step aims to identify and rank the risk type by comparing it to historical risk events. Based on historical case data, the system can determine the type of risk event and assess its risk level. This matching analysis enables the system to effectively assess the urgency of the current environmental risk event. For example, if historical data indicates that similar pollution incidents have caused serious environmental damage, the system will assess the current event as a high-risk event and initiate appropriate early warning measures.

[0070] After obtaining risk matching results, the system uses semi-supervised and incremental learning algorithms to optimize the model. Semi-supervised learning allows the system to improve the model's recognition capabilities by leveraging unlabeled data, even when only partially labeled data is available. Incremental learning continuously updates model parameters as new data arrives, improving the model's accuracy and adaptability. These algorithms enable the model to gradually optimize and adapt to different environmental risk patterns, enabling more efficient risk identification.

[0071] The pre-processed data from real-time monitoring is fed into the optimized risk identification model to calculate the risk status. Using real-time updated environmental monitoring data, the model dynamically adjusts the risk status and reassesses the levels of pollution and geological hazard risks. For example, if gas concentrations increase or soil moisture changes dramatically at a given point in time, the model will recalculate the pollution and geological hazard risks based on the optimized parameters, producing a new risk assessment.

[0072] In a specific embodiment, the process of executing step S104 may specifically include the following steps:

[0073] Perform risk probability calculation and impact range assessment algorithms on the pollution risk identification results and geological disaster risk identification results, divide the risk levels, and obtain four-level risk warning data;

[0074] Based on the four-level risk warning data, a data structure containing risk type description, affected area, warning level, development trend and response suggestions is generated, and the warning information is encapsulated to obtain structured warning information;

[0075] Spatial registration of structured warning information with the geographic information system base map, geographic coordinate mapping, and obtaining geographic coordinate risk information;

[0076] Execute color coding algorithm and heat map generation algorithm on the geographically coordinated risk information to perform risk visualization rendering and obtain the risk distribution heat map;

[0077] Perform time series overlay processing on the risk distribution heat map, conduct spatiotemporal evolution analysis, and obtain a spatiotemporal dynamic risk evolution sequence;

[0078] The risk evolution sequence is imported into the multi-scale display engine to generate views and construct interactive interfaces to obtain a visual expression of risk distribution.

[0079] Specifically, after identifying pollution and geological hazard risks, the system performs risk probability calculations and impact assessment algorithms. The primary purpose of this process is to calculate the probability of risk occurrence and the potential impact on the environment, personnel, and facilities based on the identification results. For example, the probability of pollution risk can be calculated based on historical data on changes in soil heavy metal and gas concentrations, combined with meteorological conditions and topographic information, to determine the likelihood of a pollution event. The impact range of geological hazard risk is determined by comprehensively evaluating factors such as surface deformation monitoring data, soil moisture, and rainfall to determine areas potentially affected by landslides or debris flows. These calculations provide the basis for subsequent risk classification. Based on this risk classification, the system categorizes risks into four levels: green (no risk), yellow (minor risk), orange (moderate risk), and red (severe risk). Different early warning measures and response plans are implemented for each level.

[0080] Next, based on the four-level risk warning data, the system generates a warning information structure containing detailed information. This information includes a description of the risk type (such as pollution risk or geological disaster risk), the scope of the affected area (such as within the storage yard, around the storage yard, and downstream waters), the warning level (such as green, yellow, orange, or red), the risk development trend (such as intensifying, mitigating, or remaining stable), and response recommendations (such as immediate protective measures, enhanced monitoring, and personnel evacuation). This information is packaged into structured warning information, facilitating subsequent system processing, transmission, and display.

[0081] On this basis, the system spatially registers structured warning information with a Geographic Information System (GIS) basemap. The GIS basemap contains geographic information of the storage yard area and its surroundings. Through geographic coordinate mapping, the system matches the warning information with the actual geographic location, mapping the risk data to specific spatial locations. This process ensures that the warning information accurately reflects the geographic distribution of risk events. By using this registered, geo-coordinated risk information, the system can provide users with precise spatial positioning and real-time environmental monitoring.

[0082] Next, the system color-codes the geo-coordinated risk information and generates a heat map. The color-coding algorithm maps different risk levels to different color ranges. For example, green indicates no risk, yellow indicates mild risk, orange indicates moderate risk, and red indicates severe risk. Through color coding, the spatial distribution of risk becomes intuitive and easy to understand. Next, the system uses a heat map generation algorithm to render the risk information into a heat map that clearly shows the risk intensity in different areas. For example, areas with higher risk will be highlighted in red, while areas with lower risk will be displayed in green. The generation of the heat map allows users to see the risk distribution in different areas at a glance, helping decision makers to take timely countermeasures.

[0083] To further dynamically demonstrate the evolving risk landscape, the system performs time series overlay processing on the risk distribution heat map, conducting spatiotemporal evolution analysis. Time series overlay integrates data from multiple time points to visualize the spatial distribution of risk over time, demonstrating the dynamic evolution of risk. This analysis can reveal the expansion or contraction of risk areas, trends in risk levels, and the impact of environmental conditions on risk. For example, if the risk level in a particular area continues to rise over time, the system will promptly reflect this through a heat map, helping managers to alert themselves to potential dangers.

[0084] The system imports the spatiotemporal dynamic risk evolution sequence into a multi-scale display engine for view generation and interactive interface construction. The multi-scale display engine allows users to view risk distribution from different perspectives, from macro to micro. For example, users can view a risk overview of the entire gangue dump and its surroundings, or focus on a detailed risk analysis of a small area. The interactive interface allows users to freely zoom and drag the map to view risk information at different levels and conduct in-depth analysis. This interactive view allows decision makers to flexibly view and analyze data according to their specific needs, providing a precise basis for subsequent decision-making.

[0085] In a specific embodiment, the process of executing step S105 may specifically include the following steps:

[0086] Match and search the graded warning information and dynamic risk map with the knowledge base of coal gangue dump environmental risk response, conduct preliminary screening of response plans, and obtain a set of candidate response plans;

[0087] Perform multi-objective optimization calculations on the candidate response plan set based on the current yard conditions, available resources, and risk development trends, evaluate and rank the plans, and obtain ranked response plans;

[0088] A multi-scenario decision tree is constructed based on the ranked response plans. For each decision node, an expected utility value is calculated using a weighted summation method based on the expected risk reduction degree, implementation success probability, and time benefit of the response measures. At the same time, a resource consumption value is obtained by cumulatively calculating the labor input cost, equipment usage cost, material consumption cost, and time cost. The advantages and disadvantages of each decision path are evaluated by the ratio of the expected utility value to the resource consumption value, and a decision path is generated to obtain a decision branch network.

[0089] Integrate and analyze the decision-making branch network with weather forecast data, engineering activity plans, and sensitive target distribution information to conduct future scenario simulations and obtain a risk development prediction model;

[0090] Based on the risk development prediction model, Monte Carlo simulation algorithm is executed on each decision branch to evaluate the risk response effect and obtain the optimized response plan result;

[0091] Convert the optimized results of the response plan into a structured document containing risk description, response objectives, technical measures, resource allocation, personnel division of labor and time nodes, generate a plan document, and obtain a customized emergency response plan.

[0092] Specifically, the system matches the graded warning information and dynamic risk map with the coal gangue dump environmental risk response knowledge base. The system matches the real-time monitored risk levels (such as pollution risk and geological disaster risk) and spatial distribution of risk (i.e., the dynamic risk map) with the emergency response plans in the pre-set environmental risk response knowledge base. The emergency response knowledge base is a database containing standard response processes, resource requirements, and technical measures for different environmental risk scenarios. By matching and searching the warning information and risk map, the system can quickly obtain a set of candidate response plans. These plans are pre-designed standard emergency response strategies based on the currently identified risk type and risk level.

[0093] The system performs multi-objective optimization calculations on the candidate response plan set, and evaluates and ranks the response plans based on the actual current situation of the yard, the available resources, and the risk development trend. In actual applications, each response plan will consider different objectives, such as the effectiveness of risk control, the consumption of required resources, and the difficulty of implementation. The multi-objective optimization calculation uses mathematical models to evaluate the pros and cons of each plan, weigh the balance between different objectives, and ultimately output a ranked response plan. For example, if the yard faces a more serious pollution risk, the system may give priority to quickly taking pollution interception measures and select the most cost-effective solution when resources are limited. After multi-objective optimization, the system will obtain a set of evaluated and ranked response plans.

[0094] Next, based on the ranked response options, the system constructs a multi-scenario decision tree. At this stage, the decision-making process for each response option is broken down into multiple decision nodes. Each node represents a possible decision path under specific conditions, such as the decision to adopt different response measures under different risk levels. For each decision node, the system calculates an expected utility value using a weighted summation method based on the response's expected risk reduction, probability of implementation success, and time efficiency. It also calculates a resource consumption value by cumulatively calculating labor input costs, equipment usage costs, material consumption costs, and time costs. The ratio of expected utility to resource consumption values ​​is used to evaluate the merits of each decision path. These metrics help decision makers weigh the effectiveness and costs of different decisions, ultimately generating a decision path and a complete decision branch network. Through this decision tree, decision makers can clearly see the potential consequences of each decision and the pros and cons of various options.

[0095] After constructing the decision-making branch network, the system integrates and analyzes weather forecast data, the yard's engineering activity plan, and sensitive target distribution information. Weather forecast data can provide early warning of weather changes for decision-making; for example, increased rainfall may lead to a greater risk of landslides or mudslides; engineering activity plans can reflect the potential impact of yard construction or other human activities on risks; and sensitive target distribution information provides information on key areas surrounding the yard (such as residential areas and important infrastructure). By integrating this information into the decision tree analysis, the system can perform future scenario simulations and predict the development of risks under different circumstances. For example, if the weather forecast indicates an impending heavy rain, the system may adjust its response plan to strengthen the reinforcement of the yard's slopes to prevent landslides caused by rainfall.

[0096] Based on predicted risk trends, the system uses Monte Carlo simulation algorithms to evaluate the effectiveness of risk responses for each decision branch. Monte Carlo simulation uses random sampling to simulate various possible scenarios and calculate their outcomes. Through Monte Carlo simulation, the system can simulate the effects of different response strategies under various uncertainties (such as weather changes and resource constraints), assessing whether each decision path effectively reduces risk under possible future changes. Monte Carlo simulation provides decision makers with a more comprehensive and reliable assessment of response strategies, ensuring that the selected response plan is capable of addressing a variety of potential uncertain risks.

[0097] The system converts the optimized response plan into an emergency response plan document containing detailed information. The document includes risk description, response objectives, specific technical measures, resource allocation, personnel division of labor and time nodes. The emergency response plan will generate a specific implementation plan based on the optimization results to ensure a quick and effective response when facing risks. For example, for a high-risk area, the system may recommend the immediate deployment of personnel for environmental remediation, while deploying equipment to intercept pollutants and ensuring that relevant personnel complete their tasks on time. The emergency response plan also includes a resource allocation plan, such as the allocation strategy of funds, manpower, and equipment, to ensure the smooth implementation of the plan.

[0098] In a specific embodiment, the process of executing step S106 may specifically include the following steps:

[0099] Compare environmental monitoring data before and after the implementation of the emergency response plan, calculate response timeliness, risk control effectiveness, resource utilization efficiency, and secondary risk prevention and control indicators, conduct effect evaluation, and obtain risk response performance evaluation results;

[0100] Based on the risk response performance evaluation results, the system identifies performance deficiencies in each module and generates improvement strategies, resulting in optimization recommendations including sensor layout adjustment, sampling frequency optimization, algorithm improvement, and threshold calibration.

[0101] Store the complete data of risk events, response processes, and effect evaluation results in the historical case database, conduct case summarization, and obtain structured risk response experience data;

[0102] Execute association rule mining algorithms and sequence pattern mining algorithms on structured risk response experience data to extract knowledge and obtain risk pattern characteristics and optimal response strategies;

[0103] Based on the risk pattern characteristics and the best response strategy, the environmental risk multimodal identification model is adjusted in parameters and optimized in structure, and the model is updated to obtain the optimized risk identification model.

[0104] The optimized risk identification model is integrated with the optimized sensor network layout, data processing algorithm and decision support system to perform system integration and obtain the optimized risk monitoring system.

[0105] Specifically, before and after the emergency response plan is executed, the system compares environmental monitoring data and calculates multiple performance indicators. These indicators include response timeliness, risk control effectiveness, resource utilization efficiency, and secondary risk prevention and control indicators. Response timeliness refers to the time lag between the occurrence of a risk and the implementation of response measures. A shorter response time indicates a stronger system's emergency response capability. Risk control effectiveness measures the change in risk level after response measures are implemented, assessing whether the measures taken have effectively reduced the risk. Resource utilization efficiency focuses on the use of resources (such as manpower, equipment, and funds) during the emergency response process, ensuring the rationality of resource allocation. Secondary risk prevention and control indicators assess whether the emergency response process has avoided new risks or unforeseen risk events. For example, in the response to certain environmental risks, if enhanced monitoring and real-time data analysis can prevent secondary pollution or geological disasters, it indicates that secondary risk prevention and control measures have been effectively implemented.

[0106] By calculating these effectiveness indicators, the system's risk response performance evaluation results will reveal the overall effectiveness of the emergency response plan. If performance is poor on certain indicators, the system will identify possible deficiencies and provide a basis for subsequent optimization.

[0107] Based on the results of the risk response performance evaluation, the system will identify performance deficiencies in each module and generate improvement strategies. For example, if the monitoring sensor's accuracy is insufficient, resulting in inaccurate data, the system may recommend adjusting the sensor layout and optimizing the sensor's placement to ensure more accurate capture of key data. If the data collection frequency cannot meet the real-time needs of emergency response, the system may recommend increasing the sampling frequency or dynamically adjusting it based on different risk types and changes. In addition, if the data processing algorithm is inefficient or the existing threshold settings are not adapted to actual risk changes, the system will propose algorithm improvements and threshold calibration recommendations. These optimization suggestions will help improve the performance of the entire system and ensure a more efficient response to future risks.

[0108] The system then stores complete data on the risk event, the response process, and the results of the effectiveness evaluation in a historical case database. This database will contain a variety of data accumulated during multiple emergency responses, including environmental monitoring data, risk assessment results, implemented emergency measures, resource allocation, and effectiveness evaluation results. By storing this information in a structured manner, the system can provide valuable historical data support for future emergency responses, helping to analyze and summarize response experiences from different scenarios.

[0109] For structured data stored in the historical database, the system will use association rule mining algorithms and sequential pattern mining algorithms to extract knowledge. Association rule mining algorithms can identify patterns and associations that frequently appear in different emergency responses. For example, when a specific type of pollution risk occurs, which emergency measures are often most effective. Sequential pattern mining algorithms can extract risk pattern characteristics from time series data, such as the evolution of a certain geological disaster risk, how it changes and develops over time, and which factors play a decisive role in the development of the risk in different time periods. Through these algorithms, the system can extract the best response strategies from historical data to help optimize future emergency response measures.

[0110] Based on risk pattern characteristics and optimal response strategies extracted from historical data, the system adjusts parameters and optimizes the structure of the multimodal environmental risk identification model. This optimization process aims to improve the model's accuracy and adaptability in the face of emerging risks. For example, if the system identifies new pollution sources or risk types, the model can automatically adjust parameters based on historical data to improve its ability to identify these new risks. Through this optimization, the risk identification model can continuously improve its accuracy and efficiency over the long term.

[0111] The risk identification system, after model optimization, will be integrated with the optimized sensor network layout, data processing algorithms, and decision support system. The optimized sensor network layout will ensure that the monitoring system covers all key areas, avoiding information blind spots. The optimized data processing algorithms will increase data processing speed and accuracy, ensuring timely and effective data support during emergency response. The optimized decision support system will provide managers with more scientific and rational decision-making basis, helping them make decisions quickly in emergency situations. The result of this integration is an optimized risk monitoring system that can more efficiently and accurately respond to various environmental risks that may arise in the future.

[0112] The above describes the real-time monitoring method of the environmental risk of the gangue yard based on edge computing in the embodiment of the present application. The following describes the real-time monitoring system of the environmental risk of the gangue yard based on edge computing in the embodiment of the present application. Figure 2 In one embodiment of the present application, a real-time monitoring system for environmental risks in a coal gangue dump based on edge computing includes:

[0113] The acquisition module 201 is used to collect data from a multi-source sensor network deployed in the gangue dump and its surrounding environment to obtain original environmental risk monitoring data;

[0114] Verification module 202, used to input the raw environmental risk monitoring data into the edge computing node for data verification, compensation correction, standardization and feature extraction processing to obtain pre-processed data;

[0115] A training module 203 is used to construct an environmental risk multimodal identification model based on the preprocessed data and perform training to obtain pollution risk identification results and geological disaster risk identification results;

[0116] A generation module 204 is configured to generate graded warning information based on the pollution risk identification results and the geological disaster risk identification results, and to construct a dynamic risk map to obtain a visual representation of the risk distribution;

[0117] An input module 205 is configured to input the graded warning information and the dynamic risk map into an intelligent decision support system to generate emergency response suggestions and obtain a customized emergency response plan;

[0118] The updating module 206 is configured to perform system evaluation and knowledge accumulation based on the execution data of the emergency response plan, perform parameter update on the environmental risk multimodal identification model, and obtain an optimized risk monitoring system.

[0119] Through the coordinated cooperation of the above components, by deploying a multi-source sensor network in and around the gangue dump, comprehensive, continuous and real-time monitoring of environmental risks is achieved, which significantly expands the monitoring scope and depth compared with the traditional single parameter and low-frequency monitoring method; by inputting the original environmental risk monitoring data into the edge computing node for local processing, the data transmission delay problem caused by the traditional centralized architecture is solved, the response time is shortened by 85%, and the response speed of the monitoring system to sudden risk events is greatly improved; the environmental risk multimodal identification model constructed based on preprocessed data integrates the identification functions of pollution risk and geological disaster risk, overcomes the limitations of the risk monitoring separation design in the traditional system, and improves the identification ability of complex risks. The convolutional neural network used in it extracts and classifies the pollution risk data, and the long-term and short-term The memory network performs time series pattern analysis on geological hazard risk data. The two algorithms optimize spatial characteristics and time series characteristics respectively, fully adapting to the multidimensional characteristics of coal gangue dump environmental risk data; the graded warning information and dynamic risk map generated according to the risk identification results provide a spatial visualization expression of the risk, intuitively displaying the risk distribution and evolution trend, making it easier for managers to quickly grasp the risk situation; the intelligent decision support system generates customized emergency response plans based on warning information and risk maps, providing precise guidance for risk response. The multi-objective optimization algorithm and Monte Carlo simulation algorithm used in it comprehensively consider multiple dimensions such as response timeliness, resource utilization efficiency and risk control effect. Through the analysis of emergency response execution data and knowledge accumulation, the system realizes adaptive optimization and continuously improves risk identification and prevention and control capabilities.

[0120] above Figure 2 From the perspective of modular functional entities, the real-time monitoring system for environmental risks of coal gangue yards based on edge computing in the embodiment of the present invention is described in detail. The real-time monitoring equipment for environmental risks of coal gangue yards based on edge computing in the embodiment of the present invention is described in detail from the perspective of hardware processing.

[0121] Figure 3This is a schematic diagram of the structure of a real-time monitoring device for coal gangue stockpile environmental risks based on edge computing, provided by an embodiment of the present invention. This real-time monitoring device 300 for coal gangue stockpile environmental risks based on edge computing can vary significantly depending on configuration and performance. It may include one or more central processing units (CPUs) 310 (e.g., one or more processors), memory 320, and one or more storage media 330 (e.g., one or more mass storage devices) storing application programs 333 or data 332. The memory 320 and storage media 330 may be either ephemeral or persistent storage. The program stored in the storage medium 330 may include one or more modules (not shown), each of which may include a series of instructions for operating the real-time monitoring device 300 for coal gangue stockpile environmental risks based on edge computing. Furthermore, the processor 310 can be configured to communicate with the storage medium 330, and execute a series of instruction operations in the storage medium 330 on the real-time monitoring device 300 for environmental risks of coal gangue yards based on edge computing, so as to implement the steps of the above-mentioned real-time monitoring method for environmental risks of coal gangue yards based on edge computing.

[0122] The real-time monitoring device 300 for coal gangue dump environmental risks based on edge computing may also include one or more power supplies 340, one or more wired or wireless network interfaces 350, one or more input and output interfaces 360, and / or one or more operating systems 331, such as Windows Server, Mac OS X, Unix, Linux, FreeBSD, etc. It will be understood by those skilled in the art that Figure 3 The structure of the real-time monitoring equipment for environmental risks of coal gangue yards based on edge computing shown does not constitute a limitation of the real-time monitoring equipment for environmental risks of coal gangue yards based on edge computing provided by the present invention, and may include more or fewer components than shown in the figure, or a combination of certain components, or a different arrangement of components.

[0123] The present invention also provides a computer-readable storage medium, which may be a non-volatile computer-readable storage medium or a volatile computer-readable storage medium. The computer-readable storage medium stores instructions. When the instructions are executed on a computer, the computer executes the steps of the real-time monitoring method for coal gangue yard environmental risks based on edge computing.

[0124] Those skilled in the art will clearly understand that, for the convenience and brevity of description, the specific working processes of the above-described systems, systems and units can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0125] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the portion that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions for enabling a real-time monitoring device for coal gangue dump environmental risks based on edge computing (which can be a personal computer, server, or network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes various media that can store program code, such as a USB flash drive, a mobile hard drive, a read-only memory (ROM), a random access memory (RAM), a magnetic disk, or an optical disk.

[0126] The above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit the same. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the various embodiments of the present invention.

Claims

1. A real-time monitoring method for environmental risks in coal gangue dumps based on edge computing, characterized in that: The method comprises: A multi-source sensor network is deployed in the gangue dump and surrounding environment to collect data and obtain original environmental risk monitoring data; Inputting the raw environmental risk monitoring data into the edge computing node for data verification, compensation correction, standardization and feature extraction processing to obtain preprocessed data; Based on the pre-processed data, an environmental risk multimodal identification model is constructed and training is performed to obtain pollution risk identification results and geological disaster risk identification results, including: constructing a pollution risk identification layer based on the pre-processed data, inputting soil heavy metal detection data, water quality parameter data and gas concentration data into a convolutional neural network, performing feature extraction and classification, and obtaining a preliminary pollution risk identification result; constructing a geological disaster risk identification layer based on the pre-processed data, inputting surface deformation monitoring data, soil moisture data and meteorological parameter data into a long short-term memory network, performing time series pattern analysis, and obtaining a preliminary geological disaster risk identification result; and combining the preliminary pollution risk identification result and the geological disaster risk identification result. The preliminary risk identification results are input into the comprehensive risk assessment layer, and the correlation analysis and weight distribution of the two types of risk results are carried out through the weighted fusion algorithm, and multi-source data fusion is performed to obtain the fused risk assessment results; the fused risk assessment results are matched and analyzed with the historical risk event database to perform risk type discrimination and level assessment to obtain the risk matching results; the semi-supervised learning algorithm and the incremental learning algorithm are executed on the risk matching results to update the model parameters to obtain the optimized risk identification model; the pre-processed data monitored in real time are input into the optimized risk identification model to perform risk status calculation to obtain the pollution risk identification results and the geological disaster risk identification results; Generate graded warning information based on the pollution risk identification results and the geological disaster risk identification results, construct a dynamic risk map, and obtain a visual expression of the risk distribution; Input the graded warning information and the dynamic risk map into the intelligent decision support system to generate emergency response suggestions and obtain a customized emergency response plan, including: matching and searching the graded warning information and the dynamic risk map with the coal gangue yard environmental risk response knowledge base, performing preliminary screening of response plans, and obtaining a set of candidate response plans; performing multi-objective optimization calculations on the candidate response plan set in combination with the current yard status, available resources, and risk development trends, and evaluating and ranking the plans to obtain ranked response plans; constructing a multi-scenario decision tree based on the ranked response plans, and calculating the expected utility value of each decision node using a weighted summation method based on the expected risk reduction degree of the response measures, the probability of successful implementation, and the time benefit, and at the same time, based on the human input cost, equipment usage, and other factors, The cost, material consumption cost and time cost are cumulatively calculated to obtain the resource consumption value, and the advantages and disadvantages of each decision path are evaluated by the ratio of the expected utility value to the resource consumption value, and the decision path is generated to obtain a decision branch network; the decision branch network is integrated and analyzed with meteorological forecast data, engineering activity plan and sensitive target distribution information, and future scenario deduction is performed to obtain a risk development prediction model; based on the risk development prediction model, a Monte Carlo simulation algorithm is executed on each decision branch to evaluate the risk response effect and obtain the response plan optimization result; the response plan optimization result is converted into a structured document containing risk description, response objectives, technical measures, resource allocation, personnel division of labor and time nodes, and a plan document is generated to obtain the customized emergency response plan; Based on the execution data of the emergency response plan, a system evaluation and knowledge accumulation are performed, and the parameters of the environmental risk multimodal identification model are updated to obtain an optimized risk monitoring system.

2. The real-time monitoring method for environmental risks of coal gangue dumps based on edge computing according to claim 1 is characterized in that: The multi-source sensor network deployed in the gangue dump and surrounding environment collects data to obtain the original environmental risk monitoring data, including: Soil temperature sensors, soil moisture sensors, and gas concentration sensors are buried at different depths inside the gangue dump to collect parameters inside the dump and obtain temperature distribution data, moisture content change data, and harmful gas concentration data; Heavy metal content rapid detection sensors are placed on the surface of the coal gangue dump to detect the element content of the surface soil and obtain data on the content of heavy metal elements such as lead, cadmium and arsenic; Deploy surface deformation monitoring devices around the gangue dump to monitor surface changes in real time and obtain surface settlement data, landslide precursor data, and debris flow precursor data; Deploy water quality monitoring sensors in the water body downstream of the gangue dump to detect the physical and chemical properties of the water body and obtain pH value data, conductivity data and dissolved oxygen data; Meteorological parameter collection devices are deployed around the gangue dump to record environmental meteorological conditions and obtain rainfall data, wind speed data, and air pressure data; The temperature distribution data, water content change data, harmful gas concentration data, heavy metal element content data, surface subsidence data, landslide precursor data, debris flow precursor data, pH value data, conductivity data, dissolved oxygen data, rainfall data, wind speed data and air pressure data are summarized and integrated to obtain the original environmental risk monitoring data.

3. The real-time monitoring method for environmental risks of coal gangue dumps based on edge computing according to claim 1 is characterized in that: The raw environmental risk monitoring data is input into the edge computing node for data verification, compensation correction, standardization, and feature extraction to obtain preprocessed data, including: Setting a threshold range and a change rate checking rule for the raw environmental risk monitoring data, performing data verification processing, and obtaining verification data after removing outliers; The verification data is compensated and corrected according to the sensor drift compensation formula and the temperature impact correction model to obtain corrected monitoring data; Performing dimension conversion and unit unification operations on the corrected monitoring data, performing standardization processing, and obtaining standard data with a unified format; Executing a timestamp calibration algorithm on the standard data to perform time synchronization processing to obtain synchronized data with consistent time sequence; Inputting the synchronized data into a wavelet transform filtering algorithm and a sliding average filtering algorithm to perform data noise reduction processing to obtain denoised smoothed data; The mean, variance, and rate of change statistical features of the smoothed data are calculated, and frequency domain features are extracted by fast Fourier transform, and feature extraction processing is performed to obtain the preprocessed data.

4. The real-time monitoring method for environmental risks of coal gangue dumps based on edge computing according to claim 1 is characterized in that: Generating graded warning information based on the pollution risk identification results and the geological disaster risk identification results and constructing a dynamic risk map to obtain a visual expression of risk distribution includes: Executing risk probability calculation and impact range assessment algorithms on the pollution risk identification results and the geological disaster risk identification results, performing risk level classification, and obtaining four-level risk warning data; Based on the four-level risk warning data, a data structure is generated containing a description of the risk type, the affected area, the warning level, the development trend, and response suggestions, and the warning information is packaged to obtain structured warning information; Spatially registering the structured warning information with a geographic information system base map, performing geographic coordinate mapping, and obtaining geographic coordinate risk information; Executing a color coding algorithm and a heat map generation algorithm on the geographically coordinated risk information to perform risk visualization rendering and obtain a risk distribution heat map; Performing time series superposition processing on the risk distribution heat map, performing spatiotemporal evolution analysis, and obtaining a spatiotemporal dynamic risk evolution sequence; The risk evolution sequence is imported into a multi-scale display engine to generate views and construct an interactive interface to obtain a visual expression of the risk distribution.

5. The real-time monitoring method for environmental risks of coal gangue dumps based on edge computing according to claim 1 is characterized in that: The system evaluation and knowledge accumulation based on the execution data of the emergency response plan are performed, and the parameters of the environmental risk multimodal identification model are updated to obtain an optimized risk monitoring system, including: Compare the environmental monitoring data before and after the implementation of the emergency response plan, calculate the response timeliness, risk control effectiveness, resource utilization efficiency and secondary risk prevention and control indicators, conduct an effect evaluation, and obtain the risk response performance evaluation results; Based on the risk response performance evaluation results, the performance deficiencies of each system module are identified, and improvement strategies are generated to obtain optimization suggestions including sensor layout adjustment, sampling frequency optimization, algorithm improvement, and threshold calibration; Store the complete data of risk events, response processes, and effect evaluation results in the historical case database, conduct case summarization, and obtain structured risk response experience data; Executing an association rule mining algorithm and a sequence pattern mining algorithm on the structured risk response experience data to extract knowledge and obtain risk pattern characteristics and optimal response strategies; Based on the risk pattern characteristics and the optimal response strategy, parameter adjustment and structural optimization are performed on the environmental risk multimodal identification model to update the model and obtain an optimized risk identification model; The optimized risk identification model is integrated with the optimized sensor network layout, data processing algorithm and decision support system to perform system integration to obtain the optimized risk monitoring system.

6. A real-time monitoring system for environmental risks in coal gangue dumps based on edge computing, characterized in that: The method for real-time monitoring of environmental risks in a gangue dump yard based on edge computing according to any one of claims 1 to 5 is used to implement the method, wherein the real-time monitoring system for environmental risks in a gangue dump yard based on edge computing comprises: The acquisition module is used to deploy a multi-source sensor network to collect data from the gangue dump and its surrounding environment to obtain original environmental risk monitoring data; A verification module is used to input the raw environmental risk monitoring data into the edge computing node for data verification, compensation correction, standardization and feature extraction to obtain preprocessed data; A training module is used to construct an environmental risk multimodal identification model based on the preprocessed data and perform training to obtain pollution risk identification results and geological disaster risk identification results, including: constructing a pollution risk identification layer based on the preprocessed data, inputting soil heavy metal detection data, water quality parameter data and gas concentration data into a convolutional neural network, performing feature extraction and classification, and obtaining a preliminary pollution risk identification result; constructing a geological disaster risk identification layer based on the preprocessed data, inputting surface deformation monitoring data, soil moisture data and meteorological parameter data into a long short-term memory network, performing time series pattern analysis, and obtaining a preliminary geological disaster risk identification result; and combining the preliminary pollution risk identification result and the geological disaster risk identification result. The preliminary identification results of geological disaster risks are input into the comprehensive risk assessment layer, and the correlation analysis and weight distribution of the two types of risk results are carried out through the weighted fusion algorithm, and multi-source data fusion is performed to obtain the fusion risk assessment results; the fusion risk assessment results are matched and analyzed with the historical risk event database, and the risk type discrimination and grade assessment are performed to obtain the risk matching results; the semi-supervised learning algorithm and the incremental learning algorithm are executed on the risk matching results to update the model parameters to obtain the optimized risk identification model; the pre-processed data monitored in real time are input into the optimized risk identification model to perform risk status calculation to obtain the pollution risk identification results and the geological disaster risk identification results; A generation module is used to generate graded warning information and construct a dynamic risk map based on the pollution risk identification results and the geological disaster risk identification results to obtain a visual expression of the risk distribution; An input module is used to input the graded warning information and the dynamic risk map into the intelligent decision support system to generate emergency response suggestions and obtain a customized emergency response plan, including: matching and searching the graded warning information and the dynamic risk map with the coal gangue yard environmental risk response knowledge base, performing preliminary screening of response plans, and obtaining a set of candidate response plans; performing multi-objective optimization calculations on the candidate response plan set in combination with the current yard status, available resources and risk development trends, and evaluating and ranking the plans to obtain ranked response plans; constructing a multi-scenario decision tree based on the ranked response plans, and calculating the expected utility value of each decision node by a weighted summation method according to the expected risk reduction degree of the response measures, the probability of successful implementation and the time benefit, and at the same time, based on the human input cost, The resource consumption value is obtained by cumulatively calculating the equipment usage cost, material consumption cost, and time cost. The advantages and disadvantages of each decision path are evaluated by the ratio of the expected utility value to the resource consumption value, and the decision path is generated to obtain a decision branch network. The decision branch network is integrated and analyzed with meteorological forecast data, engineering activity plans, and sensitive target distribution information to perform future scenario deductions to obtain a risk development prediction model. Based on the risk development prediction model, a Monte Carlo simulation algorithm is executed on each decision branch to evaluate the risk response effect and obtain the response plan optimization result. The response plan optimization result is converted into a structured document containing risk description, response objectives, technical measures, resource allocation, personnel division of labor, and time nodes, and a plan document is generated to obtain the customized emergency response plan. An updating module is used to perform system evaluation and knowledge accumulation based on the execution data of the emergency response plan, perform parameter update on the environmental risk multimodal identification model, and obtain an optimized risk monitoring system.

7. A real-time monitoring device for environmental risks in coal gangue dumps based on edge computing, characterized in that: It includes a memory and a processor, the memory stores a computer program that can be run on the processor, and when the processor executes the computer program, it implements the real-time monitoring method of coal gangue yard environmental risks based on edge computing as described in any one of claims 1 to 5.

8. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the processor executes the real-time monitoring method for environmental risks of coal gangue dumps based on edge computing as described in any one of claims 1 to 5.

Images