Big data-based hydrogeological dynamic monitoring and analysis system and method

Abstract

The invention relates to the technical field of hydrogeological monitoring, in particular to a hydrogeological dynamic monitoring and analysis system and method based on big data. The method specifically comprises the steps that hydrogeological data are collected in real time, edge calculation is combined for data denoising, space-time calibration and error correction, variational mode decomposition, geologic model constraint and wavelet transform are adopted for eliminating noise and drift errors, storage and transmission efficiency are optimized through incremental coding and Huffman coding, and quadratic form calibration codes are generated to ensure data reliability; a hydrological evolution trend is predicted by fusing a space-time deep learning model and a self-attention mechanism, and adaptive anomaly detection is realized by combining DBSCAN clustering and prediction error analysis; and optimizing a water resource allocation strategy through reinforcement learning, hierarchically evaluating a risk level and generating a management scheme, and finally visually displaying monitoring data, an abnormal score and a decision result. According to the method, multi-source data fusion, space-time modeling and intelligent decision making are integrated, and hydrological monitoring precision and response efficiency are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of hydrogeological monitoring, and particularly to a hydrogeological dynamic monitoring and analysis system and method based on big data. Background Art

[0002] With the continuous impact of global climate change, urbanization process and human activities on natural resources, the dynamic changes of the hydrogeological environment have become increasingly complex, and traditional hydrogeological monitoring methods are difficult to meet the needs of modern society for efficient, accurate and real-time data analysis; the changes in hydrogeology not only involve important parameters such as groundwater level, flow velocity, and pressure, but also are closely related to water quality changes, meteorological conditions and geological structures; in order to effectively address this challenge, the hydrogeological monitoring and analysis method based on big data has emerged as the times require.

[0003] In a Chinese invention patent application with the publication number CN117609900A, a method for monitoring hydrogeological risks includes: constructing a risk monitoring model; using the risk monitoring model to monitor the mountain to be measured; the construction method of the risk monitoring model is: constructing a threshold evaluation sub-model for the mountain area to be measured to determine the threshold for disasters to occur in the mountain area to be measured; constructing a weather prediction sub-model to predict weather data within a preset time; obtaining the hydrogeological data of the current mountain area to be measured by means of automatic scanning; based on a neural network model, constructing a risk monitoring model according to the weather data, hydrogeological data and the threshold for disasters to occur in the mountain area to be measured, which solves the problem of low prediction accuracy of hydrogeological risks in mountainous areas.

[0004] However, the existing methods rely on a single sensor to collect hydrogeological parameters, are easily affected by noise interference and equipment drift, resulting in low data credibility; there is a lack of a dynamic weight adjustment mechanism during multi-source data fusion, and it is difficult to balance the spatio-temporal differences of sensor, remote sensing and meteorological data; anomaly detection mostly relies on fixed thresholds or single algorithms, and cannot adapt to the dynamic changes in complex geological environments, with high omission and misjudgment rates, which restricts the real-time performance, prediction accuracy and emergency response ability of hydrogeological monitoring. It is urgent to construct an intelligent monitoring system integrating edge computing, spatio-temporal deep learning and reinforcement learning to improve data reliability, anomaly recognition sensitivity and decision-making adaptability. Summary of the Invention

[0005] The object of the present invention is to propose a hydrogeological dynamic monitoring and analysis system and method based on big data for the problems existing in the background art.

[0006] The technical solution of the present invention: A hydrogeological dynamic monitoring and analysis method based on big data includes the following specific implementation steps:

[0007] S1. Collect hydrogeological data, perform denoising, error compensation, and spatio-temporal calibration using edge computing, remove low-frequency drift, introduce a geological model to correct spatial errors, remove high-frequency noise, compensate for sensor drift errors through dynamic deviation regression, and calculate the optimization processing efficiency. Finally, generate a quadratic calibration code.

[0008] S2. Use the quadratic calibration code to verify the availability and rationality of hydrogeological data, including data binary conversion, calculation of data reset codes, and construction and comparison of two calibration coefficients. Store historical data using HDFS and real-time data using a NoSQL database.

[0009] S3. Dynamically adjust data weights through an adaptive data fusion mechanism, construct a spatio-temporal deep learning model to fuse multi-scale convolution, LSTM, and self-attention mechanisms to predict the hydrogeological evolution trend. Based on an adaptive anomaly detection method, combine DBSCAN clustering and prediction error analysis for dynamic threshold anomaly detection, and generate a comprehensive score.

[0010] S4. Integrate the comprehensive score, historical anomaly frequency, and time factors through a hierarchical framework, calculate the risk score, divide the low, medium, and high risk levels, and automatically select monitoring or scheduling plans. Use the Q-learning reinforcement learning model to define the state and action spaces, optimize the decision-making by combining the reward mechanism and dynamic learning rate, output the hydrogeological management plan, and adjust the weights through intelligent feedback to improve the prediction accuracy.

[0011] S5. Visualize and display the processed hydrogeological data, comprehensive score, and hydrogeological management plan.

[0012] Preferably, the denoising process is as follows:

[0013] S21. Decompose the original hydrogeological data into n modal components, set the low-frequency threshold ω thresh , filter out the components with frequencies lower than this threshold, and merge the remaining modal components to form a preliminarily denoised signal:

[0014] Define the objective function as: For minimizing the bandwidth;

[0015] In the formula, u k (t) represents the kth modal component; w k represents the central frequency of the kth modal; δ(t) represents the Dirac impulse function; represents the first derivative with respect to time t; represents the complex exponential function; represents the square of the two-norm; j represents the imaginary unit;

[0016] S22. Establish a geological model to predict the theoretical hydrogeological parameter G model(t), calculate the error, determine the deviation between the measured data and the predicted data, and use the correction coefficient λ to adjust the data, and define the digital model: X corrected (t) = X raw (t) + λ(G model (t) - X raw (t)), and use the geological model to correct the error;

[0017] Among them, X raw (t) represents the original hydrogeological data collected by the sensor; G model (t) represents the theoretical value calculated based on the hydrogeological model; λ represents the correction coefficient, α represents the adjustment parameter;

[0018] S23. Decompose the signal into multi-level sub-signals using the optimal wavelet basis, remove the high-frequency noise using the improved soft-threshold denoising function, and perform inverse transformation on the denoised wavelet coefficients to obtain a smoothed signal: Adopt adaptive threshold control to denoise the smoothness;

[0019] In the formula, c j represents the wavelet coefficient; T r represents the classical threshold; β represents the adjustment parameter that controls the denoising smoothness; T′ represents the adaptive threshold;

[0020] S24. Define the long-term error compensation digital model: X final (t) = X filtered (t) + γ(P pred (t) - X filtered (t));

[0021]

[0022] Among them, X final (t) represents the finally corrected data; X filtered (t) represents the denoised data; P pred (t) represents the predicted value; γ represents the correction coefficient; β represents the parameter that controls the correction sensitivity.

[0023] Preferably, the generation process of the quadratic calibration code is as follows:

[0024] S31. Convert the processed hydrogeological data data into a binary string datab;

[0025] S32. Calculate the data encapsulation code CE = H(datab);

[0026] Among them, H() represents the predefined hash function, is the integer domain modulo p; p is a predefined large prime number of 1024 bits; n is a predefined integer dimension;

[0027] S33. Construct a system of linear equations A·s = CE - v mod p;

[0028] where A is a predefined randomly selected invertible matrix A ∈ GL(n,p); GL(n,p) is the general linear group over the finite field and v is a predefined randomly selected vector,

[0029] S34. Solve to obtain the quadratic form check code c = A -1 (CE - v) mod p.

[0030] Preferably, the process of using the quadratic form check code to verify the availability and rationality of hydrogeological data is as follows:

[0031] S41. Convert the processed hydrogeological data data into a binary string datab′;

[0032] S42. Calculate the data reset code CA = H(datab′);

[0033] S43. Construct the first check coefficient fⅠ = s T Bs + 2s T u;

[0034] where T represents the transpose matrix of the matrix; B is a predefined parsing code, B = A T A mod p, which is a symmetric matrix; u represents a predefined first parsing coefficient, u = Av mod p;

[0035] S44. Construct the second check coefficient fⅡ = (CA) T (CA) - w mod p;

[0036] where w represents a predefined second parsing coefficient, w = v T v mod p;

[0037] S45. If fⅠ = fⅡ, the check passes, indicating that the received processed hydrogeological data data is available and reasonable; otherwise, an alarm is immediately given.

[0038] Preferably, the spatio-temporal deep learning model includes:

[0039] An input layer, and the input data is a spatio-temporal feature matrix;

[0040] Multi-scale convolutional layer, which uses multi-scale convolution to extract spatial features through several convolutional kernels of sizes 3×3, 5×5, and 7×7, capturing multi-level feature information in space;

[0041] LSTM layer, which learns time series features and models time dependencies;

[0042] Spatio-temporal fusion layer, which fuses the spatial features extracted by the convolutional layer with the time features learned by the LSTM layer, and uses the self-attention mechanism to calculate the importance weights W f of the spatial and time features, and performs weighted fusion X;

[0043] Output layer, which sends the features X after spatio-temporal fusion into the fully connected layer to generate the final prediction result, and outputs the predicted value X of the future hydrogeological state pred .

[0044] Preferably, the implementation process of the adaptive anomaly detection method is as follows:

[0045] S61. Use the DBSCAN density clustering algorithm to separate abnormal data points from normal data points according to the density difference of the data, and measure the density of the data by the distance between data points. The area with lower density is the abnormal area: And calculate the anomaly score S DBSCAN ;

[0046] where C is the clustering center; X i represents the i-th data point; C(X i ) represents the clustering center of the i-th data point;

[0047] S62. Accurately judge whether it is abnormal by quantifying the error between the current observation value and the predicted value. The error calculation formula is:

[0048]

[0049] where X current represents the current observation value; X pred represents the model predicted value, which is predicted by the ST-ConvLSTM model; σ represents the standard deviation of the current data set;

[0050] When the error value S anomaly is greater than the preset threshold τ, the current data point is considered an abnormal point;

[0051] S63. Build a multi-dimensional comprehensive evaluation model, combine the confidence of the data source and the correlation between data sources, and identify abnormal points by calculating the comprehensive score of each abnormal point: S total =λ1S DBSCAN +λ2Sanomaly +λ3S confidence ;

[0052]

[0053] Among them, S total represents the comprehensive score; S DBSCAN represents the anomaly score based on DBSCAN clustering; S anomaly represents the anomaly score based on prediction error; S confidence represents the confidence score of the data source; λ1, λ2, and λ3 represent weight coefficients, that is, the importance of each evaluation dimension.

[0054] Preferably, the calculation process of the anomaly score S DBSCAN based on DBSCAN clustering is as follows:

[0055] S71. For each point X i , calculate the number of points in its neighborhood, that is, the neighborhood density ρ(X i ). For a point X i , if there are k points in its neighborhood, the density can be expressed as:

[0056] Among them, k represents the number of neighbors of point X i within the radius ∈; V ∈ (X i ) represents the neighborhood volume of point X i .

[0057] S72. For the core point X i , use its density ρ(X i ) as a reference value to calculate the density difference of other points. For other points X j , calculate their density difference Δρ ij : Δρ ij =ρ(X i ) - ρ(X j );

[0058] S73. According to the density difference Δρ ij of the points, combined with the clustering results of DBSCAN to calculate the anomaly score S DBSCAN (X i ):

[0059] Among them, N ∈ (X i ) represents all the points in the neighborhood of point X i .

[0060] Preferably, the hierarchical framework includes:

[0061] S81. Data Input Layer: Based on the comprehensive score S total , extract the historical abnormal occurrence frequency H and obtain the time factor T;

[0062] S82. Risk Assessment Layer: Based on the comprehensive score S total Calculate the abnormal risk score R of the hydrological event for determining whether intervention measures are needed: R = w1S total + w2H + w3T;

[0063] Among them, S total represents the comprehensive score; H represents the historical abnormal occurrence frequency; T represents the time factor; w1, w2, and w3 represent the weight coefficients;

[0064] Classify the hydrogeological conditions according to the risk score R;

[0065] S83. Optimization Inference Layer: Automatically select the optimal hydrogeological management plan D according to the risk level * , and automatically learn the relationship between the environment and the decision through the Q-learning reinforcement learning model to optimize the decision-making process;

[0066] S84. Intelligent Feedback Layer: Conduct retrospective analysis on historical decisions and adjust the decision weights:

[0067] In the formula, R real represents the true risk value; R pred represents the predicted risk value; η represents the learning rate;

[0068] The optimization logic is: If the past decision is accurate, maintain the current weight; if the prediction error is large, adjust the risk calculation weight to improve the accuracy of future decisions.

[0069] Preferably, the optimization process of automatically learning the relationship between the environment and the decision through the Q-learning reinforcement learning model is as follows:

[0070] S91. Define the state space S = {R, E, D history};

[0071] Among them, R represents the current risk score; E represents the current environmental state; D history represents the combined information of the historical decision-making plan and the effect;

[0072] S92. Define the action space A = {a1, a2, a3, a4};

[0073] Open the floodgate for flood discharge a1: When the water level is too high, take the measure of opening the floodgate for flood discharge;

[0074] Water storage scheduling a2: When the water level is too low, adjust the water level by storing water;

[0075] Enhanced monitoring a3: Increase the monitoring frequency;

[0076] Adjust the irrigation strategy a4: Adjust the irrigation measures during drought periods;

[0077] S93. Set the reward function R t+1 = f(S t , a t , S t+1 );

[0078] Among them, S t represents the current state; a t represents the current decision-making action taken; S t+1 represents the next state, that is, the change of the environment after executing a t ; f() evaluates the contribution of the currently taken action to the environmental change;

[0079] S94. In Q-learning, the Q value represents the long-term expected reward of taking a certain action in a certain state. The update formula is as follows: Q(s, a) = (1 - α')Q(s, a) + α'[r + γ'max a′ Q(s', a')];

[0080] Among them, Q(s, a) represents the Q value of executing action a in the current state s; α' represents the learning rate; r represents the reward after executing the action; γ' represents the discount factor; max a′ Q(s', a') represents the optimal Q value of the next state s';

[0081] S95. Adjust the dynamic learning rate and the intelligent exploration strategy.

[0082] The technical solution of the present invention: A hydrogeological dynamic monitoring and analysis system based on big data, which is used to execute the above-mentioned hydrogeological dynamic monitoring and analysis method based on big data, including:

[0083] A hydrogeological data acquisition module, which is used to obtain hydrogeological data by using an Internet of Things sensor network and perform local data preprocessing through edge computing technology;

[0084] A big data management and storage module, which is used to store historical hydrogeological data by using a distributed storage architecture;

[0085] A dynamic monitoring and analysis module, which is used to combine spatio-temporal big data analysis, use a deep learning model to predict the change trend of the groundwater level and evaluate the geological disaster risk, and provide multi-level data fusion analysis;

[0086] An intelligent decision-making support module, which is used to provide auxiliary decision-making support and optimize the water resource allocation strategy by using a reinforcement learning model;

[0087] User interaction and visualization module, used to display monitoring data, warning information, and historical trends.

[0088] Compared with the prior art, the above technical solutions of the present invention have the following beneficial technical effects:

[0089] The present invention designs a hydrogeological dynamic monitoring and analysis system and method based on big data. Through technological innovation and system integration, it solves problems such as poor data quality, low prediction accuracy, rigid anomaly detection, and insufficient decision-making efficiency in traditional hydrogeological monitoring, providing a high-precision, high-response, and intelligent solution for water resource management:

[0090] (1) Enhanced data quality and reliability: By adopting edge computing combined with variational mode decomposition (VMD), wavelet transform, and geological model constraint (GCF), multi-stage denoising and error correction are realized, effectively eliminating sensor noise, drift, and spatial offset problems; optimizing storage and transmission efficiency through incremental coding and Huffman coding, reducing the amount of redundant data; and introducing quadratic calibration codes to verify data integrity, and verifying data rationality through the comparison of symmetric matrices and predefined parsing codes, significantly improving the reliability of data transmission and storage, and avoiding data failure caused by interference or tampering;

[0091] (2) Improved monitoring accuracy and prediction ability: Construct a spatio-temporal deep learning model (ST-ConvLSTM) integrating self-attention mechanism, dynamically allocate spatio-temporal feature weights, and combine adaptive data fusion (ADF) to dynamically adjust the confidence of multi-source data, reducing the prediction error of groundwater level and improving the accuracy of geological disaster risk assessment;

[0092] (3) Optimized anomaly detection robustness: Construct a multi-dimensional comprehensive scoring model (S total ), through DBSCAN clustering, prediction error analysis, and data source confidence fusion, dynamically adjust the threshold to adapt to different hydrogeological conditions, reducing the false alarm rate and missed detection rate of anomaly detection;

[0093] (4) Efficient intelligent decision-making and resource management: Optimize the water resource allocation strategy based on the hierarchical reinforcement learning framework (Q-learning), combine risk scoring (R) to adaptively generate management plans, shortening the decision-making response time; continuously optimize the model through the intelligent feedback layer, enhancing the system's adaptability and significantly improving the long-term decision-making accuracy.

[0094] (5) Strengthened system scalability and practicality: Adopt a distributed storage architecture (HDFS + NoSQL) to support massive data management, and the modular design is compatible with multiple types of sensors and data sources, and can be extended to hydrogeological monitoring scenarios of different scales and geographical environments, reducing the operation and maintenance costs. Description of the Drawings

[0095] Figure 1 This is the system architecture diagram of a hydrogeological dynamic monitoring and analysis system based on big data proposed by the present invention;

[0096] Figure 2 This is the flowchart of a hydrogeological dynamic monitoring and analysis method based on big data proposed by the present invention. Specific embodiments

[0097] Embodiment 1, as Figure 1 shown, a hydrogeological dynamic monitoring and analysis system based on big data proposed by the present invention includes: a hydrogeological data acquisition module, a big data processing and storage module, a dynamic monitoring and analysis module, an intelligent decision-making support module, and a user interaction and visualization module;

[0098] The hydrogeological data acquisition module uses an Internet of Things sensor network, including but not limited to groundwater level monitoring sensors, water quality monitoring sensors, weather stations, and remote sensing data sources, to obtain hydrogeological data and perform local data preprocessing through edge computing technology;

[0099] The big data management and storage module uses a distributed storage architecture (including but not limited to HDFS, NoSQL databases) to store historical hydrogeological data;

[0100] The dynamic monitoring and analysis module combines spatio-temporal big data analysis and uses deep learning models to predict the change trend of groundwater level and geological disaster risk assessment, providing multi-level data fusion analysis, including but not limited to the linkage analysis of surface water and groundwater and the modeling of the dynamic relationship between precipitation and groundwater;

[0101] The intelligent decision-making support module provides auxiliary decision-making support and uses a reinforcement learning (RL) model to optimize the water resource allocation strategy and improve management efficiency;

[0102] The user interaction and visualization module displays monitoring data, early warning information, and historical trends.

[0103] Embodiment 2, as Figure 2 shown, a hydrogeological dynamic monitoring and analysis method based on big data proposed by the present invention is applied to a hydrogeological dynamic monitoring and analysis system based on big data proposed in Embodiment 1, and its specific implementation steps are as follows:

[0104] S1. The hydrogeological data acquisition module uses an Internet of Things sensor network, including but not limited to groundwater level monitoring sensors, water quality monitoring equipment, weather stations, and remote sensing images, to collect hydrogeological data in real time:

[0105] S11. At a certain moment t, the hydrogeological state vector in a certain area A is X(t) = [H(t), V(t), P(t), Q(t), T(t), C(t)];

[0106] Among them, H(t) represents the groundwater level (m), measured by a water level sensor; V(t) represents the groundwater flow velocity (m / s), measured by a flow velocity sensor; P(t) represents the aquifer pressure (Pa), directly measured by a pressure sensor; Q(t) represents the groundwater flow rate (m 3 / s), calculated by a flowmeter; T(t) represents the water temperature (°C), measured by a temperature sensor; C(t) represents water quality parameters, including but not limited to pH, conductivity, dissolved oxygen, measured by a multi-channel water quality sensor;

[0107] S12. Deploy an edge computing unit at the data acquisition end to perform denoising, data completion, and anomaly detection on the collected data, reduce invalid data, and improve data quality. Specifically:

[0108] S1201. Based on the variational mode decomposition method (VMD), decompose the original signal into different frequency modes, and adaptively optimize to select key frequency band signals to remove low-frequency drift, that is: decompose the original hydrogeological data into n modal components (IMFs), set a low-frequency threshold ω thresh , filter out the components with frequencies lower than this threshold, and merge the remaining modal components to form a denoised signal:

[0109] The VMD objective function is:

[0110] In the formula, u k (t) represents the kth modal component, that is, the signal in different frequency bands; w k represents the central frequency of the kth mode; δ(t) represents the Dirac impulse function; represents the first derivative with respect to time t, used to obtain the instantaneous frequency of the signal; represents the complex exponential function, that is, the frequency component of each mode; represents the square of the two-norm, that is, the sum of the squares of the vectors. This part calculates the energy error of each mode, and the goal is to minimize the overlap between modal components; j represents the imaginary unit;

[0111] Accordingly: The objective function is used to minimize the bandwidth and improve the separation effect of modal components;

[0112] S1202. Introduce geological structure constraints (GCF) and perform spatial error correction according to the geological model, that is: establish a geological model to predict the theoretical hydrogeological parameter G model(t), calculate the error, determine the deviation between the measured data and the predicted data, and use the correction coefficient λ to adjust the data to ensure that the measured data conforms to the geological characteristics;

[0113] The digital model is: X corrected (t) = X raw (t) + λ(G model (t) - X raw (t));

[0114] Among them, X raw (t) represents the original hydrogeological data collected by the sensor; G model (t) represents the theoretical value calculated based on the hydrogeological model, that is, the prediction of hydrogeological parameters at a specific time t based on known geological characteristics (including but not limited to the direction of groundwater flow, rock layer thickness, etc.); λ represents the correction coefficient, α represents the adjustment parameter, the greater the error, the stronger the correction intensity; X corrected (t) represents the corrected data, the hydrogeological data after spatial error correction;

[0115] Accordingly: Use the geological model to correct the error and improve the data reliability;

[0116] S1203. Sensor signals are usually affected by high-frequency noise (including but not limited to electromagnetic interference, equipment jitter), which affects the data quality. Use wavelet transform to decompose the signal into components of different scales, and use the adaptive threshold method to remove high-frequency noise, that is: use the optimal wavelet basis to decompose the signal into multi-level sub-signals, adopt the improved soft threshold denoising function to remove high-frequency noise, and perform inverse transformation on the denoised wavelet coefficients to obtain a smooth signal:

[0117] In the formula, c j represents the wavelet coefficient; T r represents the classical threshold; β represents the adjustment parameter, which controls the denoising smoothness; T′ represents the adaptive threshold;

[0118] Accordingly: Adopt the adaptive threshold to control the denoising smoothness and reduce the loss of high-frequency information;

[0119] S1204. Since there may still be sensor drift errors after denoising, long-term error compensation is required. Establish a digital model: X final (t) = X filtered (t) + γ(P pred (t) - X filtered (t));

[0120]

[0121] Among them, X final(t) represents the finally corrected data, which is the hydrogeological data after dynamic deviation regression compensation; X filtered (t) represents the denoised data, which is the hydrogeological data after removing low-frequency noise and high-frequency noise; P pred (t) represents the predicted value, which is the regression prediction value based on historical data and the model, that is, the sensor drift in the long-term trend; γ represents the correction coefficient, which is used to determine how to adjust the denoised data; β represents the parameter that controls the correction sensitivity, which is used to adjust the correction intensity of the sensor drift error;

[0122] S13. Since the data of different sensors have time delay and geographical offset, spatio-temporal calibration is required to ensure the spatial consistency of the data. Specifically:

[0123] Use NTP (Network Time Protocol) to unify the time of all sensors to UTC standard time to reduce the time drift error;

[0124] Adopt a coordinate transformation matrix to calibrate the positions of each sensor: P′ i =RP i +T v ;

[0125] Among them, P i represents the original coordinates (GPS / WGS84 format); P′ i represents the converted coordinates (UTM format); R represents the rotation matrix, which corrects the sensor tilt error; T v represents the translation vector, which corrects the position deviation;

[0126] S14. Adopt delta encoding to only store the change amount of hydrogeological parameters, reduce storage redundancy, use Huffman encoding to optimize the transmission efficiency, and use Flink / Spark Streaming to perform streaming calculation on the data to improve the real-time response ability;

[0127] S15. Output the processed hydrogeological data data, and generate a quadratic calibration code for the processed hydrogeological data data. The generation process is as follows:

[0128] S1501. Convert the processed hydrogeological data data into a binary string datab;

[0129] S1502. Calculate the data encapsulation code CE = H(datab);

[0130] Among them, H() represents a predefined hash function, (that is, the integer domain modulo p); p is a predefined large prime number of 1024 bits; n is a predefined integer dimension (n = 128);

[0131] S1503. Construct a system of linear equations \(A\cdot s = CE - v\bmod p\);

[0132] where \(A\) is a randomly selected invertible matrix \(A\in GL(n,p)\) (ensuring that the determinant is non - zero); \(GL(n,p)\) is the general linear group over the finite field and \(v\) is a randomly selected vector as predefined,

[0133] S1504. Solve to obtain the quadratic form calibration code \(c = A\) -1 (CE - v)\bmod p;

[0134] S15. Transmit {quadratic form calibration code \(s\), processed hydrogeological data \(data\)} to the big data management and storage module and the dynamic monitoring and analysis module;

[0135] S2. The big data management and storage module establishes an efficient and scalable data storage system to ensure the long - term storage, fast retrieval, and data consistency of a large amount of hydrogeological data. Specifically:

[0136] Obtain {quadratic form calibration code \(s\), processed hydrogeological data \(data\)}, extract the quadratic form calibration code \(s\) and the processed hydrogeological data \(data\) from it, and based on the quadratic form calibration code \(s\), check the availability and rationality of the processed hydrogeological data \(data\) transmitted by the hydrogeological data acquisition module. The checking process is as follows:

[0137] S21. Convert the processed hydrogeological data \(data\) into a binary string \(datab'\);

[0138] S22. Calculate the data reset code \(CA = H(datab')\);

[0139] S23. Construct the first check coefficient \(f_{Ⅰ}=s\) T Bs + 2s T u;

[0140] where \(T\) here represents the transpose matrix of the matrix; \(B\) is a predefined parsing code, \(B = A\) T A\bmod p, which is a symmetric matrix; \(u\) represents a predefined first parsing coefficient, \(u = Av\bmod p\);

[0141] S24. Construct the second check coefficient \(f_{Ⅱ}=(CA)\) T (CA)-w\bmod p;

[0142] where \(w\) represents a predefined second parsing coefficient, \(w = v\) T v\bmod p;

[0143] S25. If fⅠ = fⅡ, the calibration passes, indicating that the processed hydrogeological data data received is available and reasonable; otherwise, an alarm is immediately issued.

[0144] Adopt a distributed storage architecture of HDFS + NoSQL database, store historical data (including but not limited to multi-year hydrological records) in HDFS (Hadoop Distributed File System), and store real-time data (including but not limited to hydrological changes in the most recent 24 hours) in a NoSQL database (non-relational database) to improve query efficiency.

[0145] S3. The dynamic monitoring and analysis module constructs a set of efficient, accurate, and intelligent hydrogeological dynamic monitoring and analysis systems through big data analysis and intelligent algorithms to achieve real-time monitoring, trend prediction, anomaly detection, and decision support for key hydrogeological indicators. The specific implementation process is as follows:

[0146] S31. Obtain real-time hydrogeological data data = {X sensor (t), X remote (t), X weather (t)}, and obtain historical hydrogeological data X history (t);

[0147] To improve the monitoring accuracy, construct an Adaptive Data Fusion (ADF) mechanism to dynamically adjust the weights of different data sources, so that data with high confidence contributes more to the final monitoring result:

[0148] X fused (t) = w1X sensor (t) + w2X remote (t) + w3X weather (t) + w4X history (t);

[0149]

[0150] Among them, X sensor (t) represents the sensor data at time t; X remote (t) represents the remote sensing data at time t; X weather (t) represents the weather data at time t; X history (t) represents the historical data; w1, w2, w3, and w4 represent the adaptive weights; μ is used to control the adjustment amplitude; Q i represents the confidence level of the i-th data source; represents the average confidence level of all data sources; X fused (t) represents the fused data;

[0151] It should be noted that the performance of each data source is evaluated based on historical data, that is, the confidence level Q of the data source:

[0152] In the formula, Q i represents the confidence of the i-th data source; θ is used to control the sensitivity of the adjustment and determines the amplitude of the confidence change; error i represents the average error of the i-th data source within the past T time window; merror represents the average error of all data sources within the same time window;

[0153] S32. Construct a spatio-temporal deep learning model (ST-ConvLSTM) to predict hydrological evolution, specifically:

[0154] S3201. Construct a spatio-temporal feature matrix FX t ={X fused (t - 2), X fused (t - 1), X fused (t)};

[0155] S3202. Input the spatio-temporal feature matrix FX t into the ST-ConvLSTM model, and the model structure is:

[0156] (1) Input layer: The input data is the spatio-temporal feature matrix FX t ;

[0157] (2) Multi-scale convolutional layer: Adopt multi-scale convolution (Multi-Scale Convolution), and extract spatial features through convolutional kernels of different sizes (including but not limited to 3×3, 5×5, 7×7) to capture multi-level feature information in space. Each convolutional block processes data at different scales to help the model identify spatial patterns of different sizes;

[0158] (3) LSTM layer (learning time series features and modeling time-dependent relationships): The LSTM layer models the time dependence of the data and captures the long-term memory in the time series data;

[0159] (4) Spatio-temporal fusion layer (ST-fusion): Fuse the spatial features extracted by the convolutional layer with the time features learned by the LSTM layer. For this purpose, a self-attention mechanism (Self-Attention) is used to calculate the importance weights W f of the spatial and time features, and perform weighted fusion X;

[0160] Specifically: For the spatial feature X conv output by the convolutional layer and the time feature h t output by the LSTM layer, calculate their fusion weight W f through the self-attention mechanism: W f = softmax(Wa *[X conv ,h t +b a ), and generate spatio-temporal fusion feature X = W f ·X conv +(1 - W f )·h t ;

[0161] Among them, W a represents the weight matrix of the self-attention layer, which is used to calculate the fusion weight of spatial features and temporal features; b a represents the bias term of the self-attention layer; softmax() represents the softmax function;

[0162] Accordingly: Through the self-attention mechanism, the model can adaptively adjust the weights of spatial and temporal features according to the spatio-temporal dependence of the input data, so as to more accurately capture the evolution law of the hydrogeological system;

[0163] (5) Output layer: Send the spatio-temporally fused feature X into the fully connected layer to generate the final prediction result, and output the predicted value X of the future hydrogeological state (including but not limited to groundwater level, flow velocity, etc.) pred ;

[0164] S33. Construct an Adaptive Anomaly Detection (AAE) method to achieve dynamic and intelligent anomaly point identification by combining clustering analysis and prediction error analysis, specifically as follows:

[0165] S3301. Preliminary screening of anomaly points (based on DBSCAN density clustering): Use the DBSCAN (Density-Based Spatial Clustering of Applications with Noise) density clustering algorithm to separate abnormal data points from normal data points according to the density difference of the data, and measure the density of the data by the distance between data points. The area with lower density is the abnormal area:

[0166] Among them, C is the clustering center; X i represents the i-th data point; C(X i ) represents the clustering center of the i-th data point;

[0167] Accordingly: Using the DBSCAN algorithm for preliminary screening can distinguish abnormal points from normal data points according to the density change of data points, avoid the problem of artificially setting fixed thresholds, and calculate the anomaly score S based on DBSCAN clustering DBSCAN :

[0168] (1) Calculate the neighborhood density of each data point: For each point X i , first calculate the number of points within its neighborhood, that is, the neighborhood density ρ(X i ). For a point X i , if there are k points within its neighborhood, the density can be expressed as:

[0169] where k represents the number of neighbors of point X i within a radius ∈; V ∈ (X i ) represents the neighborhood volume of point X i , which depends on the dimension of the data space;

[0170] (2) Calculate the density difference of each point: Assume that point X i is a core point (i.e., it contains at least MinPts neighbors within its neighborhood), and use its density ρ(X i ) as a reference value to calculate the density difference of other points. For other points X j , calculate their density difference Δρ ij , and the formula is: Δρ ij = ρ(X i ) - ρ(X j );

[0171] (3) Calculate the anomaly score: According to the density difference Δρ ij of the point, combine the clustering results of DBSCAN to calculate the anomaly score S DBSCAN (X i ):

[0172] where N ∈ (X i ) represents all points within the neighborhood of point X i ;

[0173] S3302. Dynamic Anomaly Assessment Based on Prediction Error: After the preliminary screening of DBSCAN, some anomaly points may still not be accurately identified. Further combine the prediction error analysis, and quantify the error between the current observed value and the predicted value to accurately determine whether it is an anomaly. The error calculation formula is:

[0174]

[0175] where X current represents the current observed value; X pred represents the model predicted value, which is predicted by the ST-ConvLSTM model; σ represents the standard deviation of the current dataset;

[0176] When the error value S anomalyWhen it is greater than the preset threshold τ, the current data point is considered an outlier;

[0177] It should be noted that in order to cope with different hydrological conditions and data distribution changes, the setting of the threshold τ is not fixed but dynamically adjusted. By calculating based on the distribution of historical data, the threshold is adjusted so that the model can adapt to the data characteristics of different time periods and different regions; the threshold adjustment formula is: τ(t) = μ + η·σ(t);

[0178] Among them, μ represents the mean of historical data; σ(t) represents the standard deviation of data at the current moment; η represents the dynamic adjustment coefficient, which controls the sensitivity of the threshold;

[0179] S3303. Build a multi-dimensional comprehensive evaluation model, combine the confidence of data sources and the correlation between data sources, comprehensively evaluate the possibility of anomalies, and identify the most likely outliers first by calculating the comprehensive score of each outlier: S total = λ1S DBSCAN + λ2S anomaly + λ3S confidence ;

[0180]

[0181] Among them, S total represents the comprehensive score; S DBSCAN represents the outlier score based on DBSCAN clustering; S anomaly represents the outlier score based on prediction error; S confidence represents the confidence score of the data source; λ1, λ2, λ3 represent weight coefficients, that is, the importance of each evaluation dimension;

[0182] S34. Transmit the comprehensive score S total to the intelligent decision support module;

[0183] S4. The intelligent decision support module receives the comprehensive score S total of each monitoring point, which is used to measure the degree of data anomaly and credibility. Introduce the intelligent decision support system (IDSS, Intelligent Decision Support System), and use machine learning prediction, risk assessment and adaptive optimization strategies to generate the optimal hydrogeological management plan to assist decision-makers in precise intervention:

[0184] In order to adapt to different decision-making scenarios, a hierarchical decision-making framework is adopted, forming a closed-loop optimization from data analysis, risk assessment to final strategy generation, including:

[0185] Data input layer (input S total ): Use the hydrogeological analysis results of step S3 as the decision basis data;

[0186] Risk assessment layer (calculate the abnormal risk score R): Evaluate potential risks through big data analysis and historical data modeling;

[0187] Optimization inference layer (calculate the optimal decision D * ): Generate the optimal management plan based on the machine learning model;

[0188] Intelligent feedback layer (adaptive optimization): Continuously optimize the decision model through reinforcement learning to improve the decision accuracy;

[0189] Specifically:

[0190] S41. Data input layer: Receive the comprehensive score S total , extract the historical abnormal occurrence frequency H (including but not limited to hydrological observation data in the past similar environments for trend analysis), and obtain the time factor T (including but not limited to meteorological data affecting hydrogeology in a certain time period);

[0191] S42. Risk assessment layer: Based on the comprehensive score S total Calculate the abnormal risk score R of the hydrological event to determine whether intervention measures need to be taken: R = w1S total + w2H + w3T;

[0192] Among them, S total represents the comprehensive score; H represents the historical abnormal occurrence frequency (higher weight in high-frequency abnormal areas); T represents the time factor (higher weight in flood season anomalies); w1, w2, and w3 represent weight coefficients;

[0193] According to the risk score R, divide the hydrogeological situation into different levels:

[0194]

[0195] Among them, R low and R high respectively represent the predefined low threshold and high threshold;

[0196] S43. Optimization inference layer:

[0197] S4301. Automatically select the optimal hydrogeological management plan D according to the risk level * :

[0198] Plan 1: Low risk (R < R low ): D1: Continuously monitor, normal monitoring frequency, no additional intervention;

[0199] Plan 2: Medium risk (R low ≤ R < R high):D2: Release early warning information, strengthen monitoring, and notify relevant departments;

[0200] Scenario 3: High risk (R≥R high ):D3: Emergency dispatch, including but not limited to opening floodgates for flood discharge and storage regulation;

[0201] S4302. Automatically learn the relationship between the environment and decisions through the Q-learning reinforcement learning model, enabling the selection of the optimal solution based on risk assessment and optimizing the decision-making process through continuous learning. Specifically:

[0202] (1) Define the state space S: S = {R, E, D history};

[0203] Among them, R represents the current risk score; E represents the current environmental state; D history represents the combined information of historical decision-making schemes and effects;

[0204] (2) In hydrogeological decision-making, the system needs to select different operation schemes as decisions and define the action space A:

[0205] Open floodgates for flood discharge a1: When the water level is too high, take measures to open floodgates for flood discharge;

[0206] Storage regulation a2: When the water level is too low, adjust the water level by storage;

[0207] Strengthen monitoring a3: Reduce future unpredictable risks by increasing the monitoring frequency;

[0208] Adjust irrigation strategy a4: Adjust irrigation measures during droughts to improve water resource utilization efficiency;

[0209] Therefore, the action space A is: A = {a1, a2, a3, a4};

[0210] (3) Set the reward function R t+1 = f(S t , a t , S t+1 );

[0211] Among them, S t represents the current state; a t represents the current decision-making action taken; S t+1 represents the next state, that is, the change of the environment after executing a t ; f() evaluates the contribution of the current action taken to the environmental change (if an action leads to system stability or achieves the goal, the reward is positive; if it leads to adverse results, the reward is negative);

[0212] It should be noted that the reward mechanism is designed as follows: Reward for successfully mitigating risks: If the decision taken reduces the hydrological risks, the reward is positive; Penalty for failing to effectively mitigate risks: If the decision taken fails to reduce the risks, a negative reward is given; Penalty for over-scheduling or mismanagement: If the water resource management measures are excessive and result in waste of resources, a negative reward also needs to be imposed.

[0213] (4) In Q-learning, the Q-value represents the long-term expected reward for taking a certain action in a certain state, and the update formula is as follows: Q(s,a) = (1 - α′)Q(s,a) + α′[r + γ′max a′ Q(s′,a′)];

[0214] Among them, Q(s,a) represents the Q-value of executing action a in the current state s; α′ represents the learning rate; r represents the reward after executing the action; γ′ represents the discount factor; max a′ Q(s′,a′) represents the optimal Q-value of the next state s′;

[0215] (5) Dynamic learning rate and intelligent exploration strategy:

[0216] Dynamic learning rate: The learning rate is dynamically adjusted as the number of training times increases. It is relatively large in the initial stage to enable the model to learn quickly; when the training reaches a certain stage, it gradually decreases to avoid over-adjustment;

[0217] Intelligent exploration mechanism: In the initial stage, an exploration strategy (including but not limited to ε-greedy) is adopted to try new actions with a higher probability; as learning progresses, the utilization of the current optimal decision is gradually increased and exploration is reduced;

[0218] S44. Intelligent feedback layer: Used to optimize the decision-making system to continuously improve the decision-making accuracy as the data is updated. Specifically:

[0219] S4401. Conduct retrospective analysis on historical decisions and adjust decision weights:

[0220] In the formula, R real represents the true risk value (actual data feedback); R pred represents the predicted risk value; η represents the learning rate;

[0221] S4402. The optimization logic is: If the past decisions are accurate, maintain the current weights; if the prediction error is large, then adjust the risk calculation weights to improve the accuracy of future decisions;

[0222] S45. Transmit the hydrogeological management plan D * to the user interaction and visualization module;

[0223] S5. The user interaction and visualization module visually displays the processed hydrogeological data data and the comprehensive score S total , and the hydrogeological management plan D * .

[0224] The embodiments of the present invention have been described in detail with reference to the accompanying drawings. However, the present invention is not limited thereto. Various changes can be made without departing from the spirit of the present invention within the scope of knowledge possessed by those skilled in the art to which the present invention pertains.

Claims

1. A hydrogeological dynamic monitoring and analysis method based on big data, characterized in that: The specific implementation steps include the following: S1. Collect hydrogeological data, use edge computing for denoising, error compensation and spatiotemporal calibration, remove low-frequency drift, introduce geological models to correct spatial errors, remove high-frequency noise, and compensate for sensor drift errors through dynamic deviation regression, and calculate and optimize processing efficiency, and finally generate a quadratic calibration code; S2. Use quadratic calibration code to verify the availability and rationality of hydrogeological data, including data binary conversion, data reset code calculation, and the construction and comparison of two calibration coefficients. HDFS is used to store historical data, and NoSQL database is used to store real-time data. S3. Dynamically adjust data weights through adaptive data fusion mechanism, build spatiotemporal deep learning model integrating multi-scale convolution, LSTM and self-attention mechanism to predict hydrological evolution trend, perform dynamic threshold anomaly detection based on adaptive anomaly detection method, combine DBSCAN clustering and prediction error analysis, and generate comprehensive scores; S4. Integrate comprehensive scores, historical abnormal frequency and time factors through a hierarchical framework, calculate risk scores, divide low, medium and high risk levels and automatically select monitoring or scheduling plans, use Q-learning reinforcement learning model to define state and action space, combine reward mechanism and dynamic learning rate to optimize decision-making, output hydrogeological management plans, and adjust weights through intelligent feedback to improve prediction accuracy; S5. Visualize and display the processed hydrogeological data, comprehensive scores, and hydrogeological management plans.

2. A hydrogeological dynamic monitoring and analysis method based on big data according to claim 1, characterized in that: The denoising process is as follows: S21. Decompose the original hydrological data into n modal components and set the low-frequency threshold ω thresh , filter out the components with frequencies lower than the threshold, and merge the remaining modal components to form the primary denoised signal: The objective function is defined as: Used to minimize bandwidth; In the formula, u k (t) represents the kth modal component; w k represents the center frequency of the kth mode; δ(t) represents the Dirac pulse function; represents the first-order derivative with respect to time t; represents a complex exponential function; represents the square of the two-norm; j represents the imaginary unit; S22. Establish geological model and predict theoretical hydrological parameters G model (t), calculate the error, determine the deviation between the measured data and the predicted data, and use the correction factor λ to adjust the data and define the digital model: X corrected (t) = X raw (t)+λ(G model (t)-X raw (t)), error correction using geological models; Among them, X raw (t) represents the original hydrological data collected by the sensor; G model (t) represents the theoretical value calculated based on the hydrogeological model; λ represents the correction coefficient, α represents the adjustment parameter; S23, use the optimal wavelet basis to decompose the signal into multi-level sub-signals, use the improved soft threshold denoising function to remove high-frequency noise, and perform inverse transformation on the denoised wavelet coefficients to obtain a smooth signal: Adaptive threshold is used to control the denoising smoothness; In the formula, c j represents the wavelet coefficient; T r represents the classic threshold; β represents the adjustment parameter, which controls the denoising smoothness; T′ represents the adaptive threshold; S24. Define the long-term error compensation digital model: X final (t) = X filtered (t)+γ(P pred (t)-X filtered (t)); Among them, X final (t) indicates the final corrected data; X filtered (t) represents the denoised data; P pred (t) represents the predicted value; γ represents the correction coefficient; β represents the parameter for controlling the correction sensitivity.

3. The method for dynamic hydrogeological monitoring and analysis based on big data according to claim 1, characterized in that: The generation process of the quadratic check code is as follows: S31, converting the processed hydrogeological data data into a binary string datab; S32, calculate the data encapsulation code CE=H(datab); Where H() represents a predefined hash function. is the integer domain modulo p; p is a predefined large prime number of 1024 bits; n is a predefined integer dimension; S33, construct a linear equation system A·s=CE-v mod p; Where A is a predefined randomly selected reversible matrix A∈GL(n,p); GL(n,p) is a finite field The general linear group on ; v is a predefined randomly selected vector, S34, solve to obtain the quadratic check code c = A -1 (CE-v)mod p.

4. A hydrogeological dynamic monitoring and analysis method based on big data according to claim 3, characterized in that: The verification process of the availability and rationality of hydrogeological data using quadratic calibration codes is as follows: S41, converting the processed hydrogeological data data into a binary string datab′; S42, calculate the data reset code CA=H(datab′); S43, construct the first calibration coefficient fⅠ=s T Bs+2s T u; Where T represents the transposed matrix of the matrix; B is the predefined analytical code, B = A T A mod p is a symmetric matrix; u represents a predefined first analytical coefficient, u = Av mod p; S44, construct the second calibration coefficient fⅡ=(CA) T (CA)-w mod p; Wherein, w represents the predefined second analytical coefficient, w=v T v mod p; S45. If fⅠ=fⅡ, the calibration is passed, indicating that the received processed hydrogeological data data is usable and reasonable; otherwise, an alarm is immediately issued.

5. The method for dynamic hydrogeological monitoring and analysis based on big data according to claim 1, characterized in that: Spatiotemporal deep learning models include: Input layer, the input data is the spatiotemporal feature matrix; The multi-scale convolution layer uses multi-scale convolution to extract spatial features through several convolution kernels of sizes 3×3, 5×5, and 7×7 to capture multi-level feature information in space; LSTM layer, learning time series features and modeling time dependencies; The spatiotemporal fusion layer combines the spatial features extracted by the convolutional layer with the temporal features learned by the LSTM layer, and uses the self-attention mechanism to calculate the importance weights W of the spatial and temporal features. f , and perform weighted fusion X; The output layer sends the feature X after time-space fusion to the fully connected layer to generate the final prediction result and output the predicted value X of the future hydrogeological state. pred .

6. A hydrogeological dynamic monitoring and analysis method based on big data according to claim 5, characterized in that: The implementation process based on the adaptive anomaly detection method is as follows: S61. Use the DBSCAN density clustering algorithm to separate abnormal data points from normal data points based on the density difference of the data. The density of the data is measured by the distance between the data points. The area with lower density is the abnormal area: And calculate the anomaly score S based on DBSCAN clustering DBSCAN ; Among them, C is the cluster center; X i represents the i-th data point; C(X i ) represents the cluster center of the i-th data point; S62. By quantifying the error between the current observed value and the predicted value, it is accurately determined whether it is abnormal. The error calculation formula is: Among them, X current Indicates the current observation value; X pred Represents the model prediction value, which is predicted by the ST-ConvLSTM model; σ represents the standard deviation of the current data set; When the error value S anomaly When it is greater than the preset threshold τ, the current data point is considered an abnormal point; S63. Build a multi-dimensional comprehensive evaluation model, combine the confidence of data sources and the correlation between data sources, and identify outliers by calculating the comprehensive score of each outlier: S total =λ1S DBSCAN +λ2S anomaly +λ3S confidence ; Among them, S total Indicates the comprehensive score; S DBSCAN represents the anomaly score based on DBSCAN clustering; S anomaly represents the anomaly score based on prediction error; S confidence It represents the confidence score of the data source; λ1, λ2, and λ3 represent weight coefficients, that is, the importance of each evaluation dimension.

7. A hydrogeological dynamic monitoring and analysis method based on big data according to claim 6, characterized in that: Anomaly score S based on DBSCAN clustering DBSCAN The calculation process is as follows: S71. For each point X i , calculate the number of points in its neighborhood, that is, the neighborhood density ρ(X i ), for a point X i , if there are k points in its neighborhood, the density can be expressed as: Where k represents point X i The number of neighbors within radius ∈; V ∈ (X i ) represents point X i The neighborhood volume of S72, for the core point X i , its density ρ(X i ) is used as a reference value to calculate the density difference of other points. j , calculate their density difference Δρ ij :Δρ ij =ρ(X i )-ρ(X j ); S73, based on the density difference Δρ of the points ij , combined with the clustering results of DBSCAN to calculate the anomaly score S DBSCAN (X i ): Among them, N ∈ (X i ) represents point X i All points in the neighborhood of .

8. The method for dynamic hydrogeological monitoring and analysis based on big data according to claim 1, characterized in that: The layered framework includes: S81, data input layer: based on comprehensive score S total , extract the historical abnormal occurrence frequency H and obtain the time factor T; S82, Risk Assessment Layer: Based on the comprehensive score S total Calculate the abnormal risk score R of the hydrological event to determine whether intervention measures are needed: R = w1S total +w2H+w3T; Among them, S total represents the comprehensive score; H represents the frequency of historical anomalies; T represents the time factor; w1, w2, and w3 represent weight coefficients; According to the risk score R, the hydrogeological conditions are classified; S83, Optimization reasoning layer: Automatically select the optimal hydrogeological management plan based on risk level D * And automatically learn the relationship between the environment and decision-making through the Q-learning reinforcement learning model to optimize the decision-making process; S84, Intelligent Feedback Layer: Retrospective analysis of historical decisions and adjustment of decision weights: In the formula, R real Represents the true risk value; R pred represents the predicted risk value; η represents the learning rate; The optimization logic is: if past decisions were accurate, maintain the current weights; if the prediction error is large, adjust the risk calculation weights to improve the accuracy of future decisions.

9. A hydrogeological dynamic monitoring and analysis method based on big data according to claim 8, characterized in that: The optimization process of automatically learning the relationship between the environment and decision-making through the Q-learning reinforcement learning model is: S91. Define the state space S: S = {R, E, D history }; Among them, R represents the current risk score; E represents the current environment status; D history Represents the combined information of historical decision plans and effects; S92, define action space A = {a1, a2, a3, a4}; Open the gate to release flood a1: When the water level is too high, take measures to open the gate to release flood; Water storage scheduling a2: When the water level is too low, the water level is adjusted by storing water; Strengthen monitoring a3: Reduce future unpredictable risks by increasing the frequency of monitoring; Adjust irrigation strategy a4: Adjust irrigation measures during drought to improve water resource utilization efficiency; S93. Setting reward function R t+1 =f(S t ,a t ,S t+1 ); Among them, S t Indicates the current state; a t Indicates the current decision action; S t+1 Indicates the next state, that is, execute a t Changes in the environment; f() evaluates the contribution of the current action to the change in the environment; S94. In Q-learning, the Q value represents the long-term expected benefit of taking a certain action in a certain state. The update formula is as follows: Q(s,a)=(1-α′)Q(s,a)+α′[r+γ′max a′ Q(s′,a′)]; Among them, Q(s,a) represents the Q value of executing action a in the current state s; α′ represents the learning rate; r represents the reward after executing the action; γ′ represents the discount factor; max a′ Q(s′,a′) represents the optimal Q value of the next state s′; S95. Adjust the dynamic learning rate and intelligent exploration strategy.

10. A hydrogeological dynamic monitoring and analysis system based on big data, which is used to execute the hydrogeological dynamic monitoring and analysis method based on big data according to any one of claims 1 to 9, characterized in that: include: The hydrogeological data acquisition module is used to acquire hydrogeological data using an IoT sensor network and perform local data preprocessing using edge computing technology; Big data management and storage module, used to store historical hydrogeological data using a distributed storage architecture; Dynamic monitoring and analysis module, which combines spatiotemporal big data analysis, uses deep learning models to predict groundwater level trends and geological disaster risk assessment, and provides multi-level data fusion analysis; Intelligent decision support module, used to provide auxiliary decision support and optimize water resource allocation strategy using reinforcement learning model; User interaction and visualization module, used to display monitoring data, early warning information, and historical trends.

Images