A wellhead data acquisition modeling system

Abstract

The present invention discloses a wellhead data acquisition modeling system, comprising a data acquisition module, a hydraulic feature processing module, a well-layer coupling modeling module and an interactive control module; the present invention belongs to the field of well water monitoring and management technology, and based on an adaptive time-frequency analysis method, introduces a dynamic window function with enhanced local spectrum entropy to perform feature extraction and abnormal warning on transient hydraulic fluctuation signals, thereby significantly improving the time-frequency resolution and change recognition capability of hydraulic signals; by combining a multi-scale parameter coordination mechanism, a dynamic coupling model between a pumping system and an aquifer is constructed, and a void compression hysteresis function and a joint parameter calibration method are introduced to achieve the accuracy and dynamic adaptability of well-layer response modeling; the interactive control module implements model prediction control according to a combination of hydraulic feature vectors, hydraulic abnormality warning signals and well-layer coupling prediction parameters, and outputs wellhead data acquisition decision support data.

Description

Technical Field

[0001] The invention belongs to the technical field of well water monitoring and management, and in particular relates to a wellhead data acquisition and modeling system. Background Art

[0002] A wellhead data acquisition and modeling system is used to monitor groundwater levels, temperature, water quality, and other environmental data in real time. By simultaneously collecting data from multiple sensors and performing preprocessing, denoising, and calibration, it ensures data accuracy and timeliness. Using advanced analytical methods and modeling techniques, the system integrates multi-source data to establish a dynamic groundwater monitoring model. It can also detect anomalies such as water quality changes and water level fluctuations in real time. This system can provide scientific decision-making support for water resource management and water environment protection, providing timely warnings of potential risks such as water pollution and groundwater resource changes, thereby ensuring the sustainable use and safe management of water resources.

[0003] However, in the existing wellhead data acquisition and modeling system, there are technical problems such as the inability to efficiently and synchronously collect multi-source data from the wellhead, the imperfect real-time feature extraction mechanism, and the lack of a response mechanism to hydraulic change trends; in the existing hydraulic feature processing process, there are technical problems such as the rapid change rate of transient hydraulic characteristics, fuzzy frequency domain information, difficulty in locating signal energy changes during the pumping process, and susceptibility to high-frequency noise interference; in the existing well-layer coupling data modeling process, there are technical problems such as the difficulty in quantifying the interaction between the pumping system and the aquifer, the delayed response of the coupling model, the lack of dynamic characterization capabilities for nonlinear seepage changes, and the rough model parameter calibration mechanism. Summary of the Invention

[0004] In view of the above situation, in order to overcome the defects of the prior art, the present invention provides a wellhead data acquisition and modeling system. In view of the technical problems in the existing wellhead data acquisition and modeling system, there are the wellhead multi-source data that cannot be efficiently and synchronously collected, the real-time feature extraction mechanism is not sound, and there is a lack of a response mechanism to the hydraulic change trend. This solution creatively adopts a comprehensive optimization method that combines hydraulic feature improvement extraction and identification with well-layer coupling modeling optimization. Through hydraulic feature improvement extraction and identification, the accuracy and real-time problems of hydraulic feature extraction are solved, so that hydraulic pattern changes can be captured in time and effective early warning can be carried out; at the same time, through well-layer coupling modeling optimization, the coupling between the pumping system and the aquifer is enhanced, ensuring that the model can dynamically adapt to changes in the downhole environment, and improving the accuracy and adaptability of modeling; in view of the technical problems in the existing hydraulic feature processing process, there are the transient hydraulic feature changes at a fast speed, the frequency domain information is fuzzy, the signal energy changes in the pumping process are difficult to locate, and are easily interfered by high-frequency noise. This solution creatively adopts a feature-enhanced transient hydraulic fluctuation identification method improved by combining adaptive time-frequency analysis to extract transient hydraulic features, and innovatively introduces local The spectral entropy-enhanced dynamic window function significantly improves the time-frequency resolution of transient hydraulic fluctuation signals by integrating local spectral entropy calculations and combining them with adaptive time-frequency analysis methods. In its implementation, the calculation of local spectral entropy parameters accurately represents the signal's complexity by analyzing the signal's local frequency distribution, and dynamically adjusts the window width to accommodate changes in the transient frequency. To address technical issues in existing well-layer coupled data modeling, such as the difficulty in quantifying the interaction between the pumping system and the aquifer, delayed coupled model response, a lack of dynamic characterization of nonlinear seepage changes, and a crude model parameter calibration mechanism, this proposal innovatively employs a joint pumping system and aquifer modeling approach incorporating a multi-scale parameter coordination mechanism for well-layer coupled data modeling. By constructing a dynamic transmission equation for the pumping system, a dual-scale seepage response model for the aquifer, and a coupling boundary setting mechanism, it achieves real-time modeling and feedback optimization of the coupling process between the pumping system and the aquifer. Furthermore, a joint parameter calibration method is introduced to dynamically adjust model parameters based on sensor data, improving the model's ability to predict aquifer compression response, water level disturbance transmission, and pumping efficiency changes.

[0005] The technical solution adopted by the present invention is as follows: the present invention provides a wellhead data acquisition and modeling system, including a data acquisition module, a hydraulic characteristic processing module, a well-layer coupling modeling module and an interactive control module;

[0006] The data acquisition module is used for synchronous data acquisition and processing, obtains wellhead multi-source optimization data through data synchronous acquisition and processing, and sends the wellhead multi-source optimization data to the hydraulic characteristic processing module and the well-layer coupling modeling module;

[0007] The hydraulic feature processing module is used for extracting transient hydraulic features, obtaining hydraulic feature vectors and hydraulic anomaly warning signal data through the transient hydraulic feature extraction, and sending the hydraulic feature vectors and hydraulic anomaly warning signal data to the interactive control module;

[0008] The well-layer coupling modeling module is used for well-layer coupling data modeling, obtains a well-layer coupling prediction parameter combination through well-layer coupling data modeling, and sends the well-layer coupling prediction parameter combination to the interactive control module;

[0009] The interactive control module is used for model predictive control, and obtains wellhead data acquisition decision support data through model predictive control.

[0010] Furthermore, the data synchronous acquisition and processing is used to provide synchronous and efficient wellhead data, specifically performing data synchronous acquisition to obtain wellhead raw data, and obtaining wellhead multi-source optimized data through preprocessing operations;

[0011] The wellhead raw data includes water level data, instantaneous flow data, water temperature data, conductivity data, wellhead vibration signal and well pipe displacement data;

[0012] The wellhead multi-source optimized data specifically refers to the wellhead original data after time synchronization, denoising filtering and basic error calibration.

[0013] Furthermore, the transient hydraulic feature extraction is used to extract hydraulic features and identify abnormal hydraulic fluctuation states. Specifically, based on the wellhead multi-source optimization data, a feature-enhanced transient hydraulic fluctuation identification method improved by combining adaptive time-frequency analysis is used to perform transient hydraulic feature extraction to obtain hydraulic feature vectors and hydraulic anomaly warning signal data. The method includes the following steps: adaptive time-frequency analysis improvement, transient feature extraction, dynamic feature enhancement, adaptive hydraulic pattern classification, hydraulic feature vector construction, and hydraulic pattern anomaly detection and warning.

[0014] The improved adaptive time-frequency analysis is used to improve the time-frequency resolution of transient hydraulic fluctuation signals. Specifically, it is achieved by constructing a dynamic window function that integrates local spectral entropy enhancement and adopting a time-varying sampling strategy. The wellhead sampling step size is adjusted according to the change of instantaneous frequency. The time-frequency distribution data of the hydrodynamic signal is obtained through adaptive time-frequency distribution calculation.

[0015] The dynamic window function with integrated local spectrum entropy enhancement performs transient signal adjustment according to the instantaneous frequency in the wellhead multi-source optimization data, and improves the window width adjustment of the dynamic window function by constructing local spectrum entropy parameters to obtain a dynamic window adjustment calculation function, and the calculation formula is:

[0016] ;

[0017] Where, is the dynamic window adjustment calculation function, t is the time index, is the local time difference variable, is the dynamic window width, which is used to adjust the time-frequency resolution over time. The calculation formula is , where a is the basic window width, b is the adjustment coefficient, H f (t) is the local spectral entropy parameter;

[0018] The local spectrum entropy parameter is constructed by calculating the local frequency distribution of the signal at each sampling moment and performing a complexity representation of the signal spectrum based on the local spectrum distribution;

[0019] The calculation formula of the local spectrum distribution is:

[0020] ;

[0021] Wherein, P(f,t) is the local spectrum distribution function, X(f,t) is the complex spectrum coefficient after time-frequency transformation, and the complex spectrum coefficient X(f,t) after time-frequency transformation represents the complex frequency domain response of the original signal at time t and instantaneous frequency f, which is specifically calculated by short-time Fourier transform. |X(f,t)| is the amplitude of the original signal at time t and instantaneous frequency f, f is the instantaneous frequency, and F is the set of hydraulic characteristic instantaneous frequency bands;

[0022] The specific calculation formula of the local spectrum entropy parameter is:

[0023] ;

[0024] Where H f (t) is the local spectrum entropy parameter, P(f,t) is the local spectrum distribution function, and the local spectrum distribution function P(f,t) is used to represent the normalized frequency distribution as a whole. It is the calculation item of the frequency component information amount, specifically using the natural logarithm base;

[0025] The time-varying sampling strategy is used to adaptively adjust the data sampling period according to the instantaneous frequency changes caused by the pumping behavior. Specifically, the sampling step length is calculated based on the maximum instantaneous frequency in the wellhead multi-source optimization data to obtain the time-varying sampling step length;

[0026] The adaptive time-frequency distribution calculation is specifically performed based on the dynamic window adjustment calculation function and the time-varying sampling step, by maximizing the time-frequency aggregation index to obtain the time-frequency distribution data of the hydrodynamic signal;

[0027] The transient feature extraction is used to extract transient change features from the hydraulic mode signal, specifically by performing multi-scale gradient feature calculation and transient energy ratio extraction based on the time-frequency distribution data of the hydrodynamic signal to obtain transient gradient feature data and transient energy ratio data, and by constructing a transient feature matrix based on the transient gradient feature data and transient energy ratio data to obtain transient feature data;

[0028] The multi-scale gradient feature calculation is used to extract the local transient change rate from the time-frequency distribution of the hydraulic signal, and reflect the energy mutation characteristics caused by pumping disturbance, water level mutation, aquifer response or equipment abnormality in the hydraulic system, and obtain transient gradient feature data;

[0029] The transient energy ratio extraction is used to detect sudden changes, surges, pumping fluctuations and pump interference in hydraulic signals to obtain transient energy ratio data;

[0030] The transient characteristic matrix is constructed by combining transient gradient characteristic data and transient energy ratio data to obtain transient characteristic data;

[0031] The dynamic feature enhancement is specifically to adopt a dynamic feature weighting method based on statistics, perform weighted calculation on the transient feature data according to the variance and change rate of the feature, and perform feature selection through a variance screening method to obtain the screening feature set data;

[0032] The adaptive hydraulic pattern classification is used to perform real-time hydraulic pattern classification and identify hydraulic pattern changes based on screening features. Specifically, it constructs a basic classification model and performs dynamic hydraulic pattern data category center design based on real-time data to optimize the adaptability of hydraulic pattern classification and obtain hydraulic pattern data category center feature data. Based on the hydraulic pattern data category center feature data, the similarity between the features in the screening feature set data and the hydraulic pattern data category center feature data is calculated, and hydraulic pattern classification is performed through a Gaussian model to obtain hydraulic pattern classification label data.

[0033] The hydraulic feature vector is constructed by constructing the hydraulic feature vector based on the hydraulic mode classification label data and the feature mean, feature standard deviation and feature change rate values in the filtered feature set data to obtain the hydraulic feature vector;

[0034] The hydraulic pattern anomaly detection and warning is specifically based on the characteristic change rate value in the hydraulic feature vector and the hydraulic pattern classification label data, and adopts a dynamic threshold mechanism to achieve real-time identification and warning of pumping mutations and unstable behavior of the hydraulic system, thereby obtaining hydraulic anomaly warning signal data.

[0035] Furthermore, the well-layer coupling data modeling is used to establish a dynamic response model between the pumping system and the aquifer and realize real-time quantitative modeling and optimization of the coupling relationship. Specifically, based on the wellhead multi-source optimization data, a pumping system-aquifer joint modeling method combined with a multi-scale parameter coordination mechanism is adopted to perform well-layer coupling data modeling to obtain a well-layer coupling prediction parameter combination, including the following steps: pumping system modeling, aquifer modeling, coupling boundary setting and joint parameter calibration;

[0036] The pumping system modeling is used to describe the water flow behavior inside the water pump and well pipe. Specifically, a dynamic transmission equation between pumping flow, well pipe water level, and system load is established, and a state residual feedback mechanism is introduced to improve the model's adaptability to disturbance changes.

[0037] The aquifer modeling is used to characterize the seepage response behavior of the aquifer during the pumping process. Specifically, a dual-scale seepage response model is used to jointly represent the macroscopic and microscopic aquifer characteristics. The dual-scale seepage response model includes a macroscopic layer model and a microscopic layer model. The macroscopic layer model is constructed based on the unsteady groundwater flow governing equation. The microscopic layer model introduces a void compression hysteresis function to calculate the corresponding hysteresis of the aquifer.

[0038] The coupling boundary setting specifically involves establishing coupling boundary conditions between the pumping system model and the aquifer model, and calculating the flow continuity at the well wall boundary by establishing the coupling boundary conditions. By introducing a multi-source data weight coefficient mechanism, parameters are adjusted in real time to obtain a well-layer joint initial model.

[0039] The joint parameter calibration is used to optimize and adjust the parameters of the pumping system model and the aquifer model. Specifically, based on the actually collected water level, flow rate and conductivity sensor data, a parameter calibration method combined with a minimum error criterion is used to perform parameter updates and iterative model training to obtain a well-layer coupling comprehensive model. By using the well-layer coupling comprehensive model, a well-layer coupling prediction parameter combination is obtained.

[0040] The well-layer coupling prediction parameter combination specifically includes a real-time hydraulic characteristic vector, a pumping system water level parameter, an aquifer response flow parameter, and an aquifer dynamic water storage parameter.

[0041] Furthermore, the model predictive control is used to dynamically optimize the well-layer coupling prediction model based on real-time monitoring data during the operation of the groundwater well, and to guide the adjustment and optimization of the wellhead data acquisition plan. Specifically, the model predictive control is performed based on the hydraulic characteristic vector, the hydraulic abnormality warning signal data and the well-layer coupling prediction parameter combination to obtain wellhead data acquisition decision support data;

[0042] The wellhead data acquisition decision support data specifically includes wellhead data acquisition priority decision data, data acquisition time attribute data, and wellhead data quality adjustment and optimization suggestions;

[0043] The wellhead data acquisition priority decision data is used to determine the data type to be acquired first and its corresponding data acquisition frequency, including the priority acquisition parameter type and data acquisition frequency;

[0044] The data collection time attribute data is used to optimize the collection scheduling strategy, including the collection window size and collection duration data;

[0045] The wellhead data quality adjustment optimization suggestion is used to provide a quality assessment result of the currently acquired data, specifically a quality assessment result of real-time data.

[0046] The beneficial effects achieved by the present invention using the above scheme are as follows:

[0047] (1) In view of the technical problems in the existing wellhead data acquisition and modeling system, such as the inability to efficiently and synchronously collect multi-source data at the wellhead, the imperfect real-time feature extraction mechanism, and the lack of a response mechanism to hydraulic change trends, this solution creatively adopts a comprehensive optimization method that combines improved extraction and identification of hydraulic features with well-layer coupling modeling optimization. Through improved extraction and identification of hydraulic features, the accuracy and real-time performance issues of hydraulic feature extraction are solved, so that changes in hydraulic patterns can be captured in a timely manner and effective early warnings can be issued. At the same time, through well-layer coupling modeling optimization, the coupling between the pumping system and the aquifer is enhanced, ensuring that the model can dynamically adapt to changes in the downhole environment, thereby improving the accuracy and adaptability of the modeling.

[0048] The accuracy and adaptability of modeling are improved by adapting to changes in the environment.

[0049] (2) In view of the technical problems in the existing hydraulic feature processing process, such as the rapid change of transient hydraulic features, fuzzy frequency domain information, difficult to locate the energy change of the pumping process signal, and easy interference by high-frequency noise, this scheme creatively adopts the feature-enhanced transient hydraulic fluctuation identification method improved by adaptive time-frequency analysis to extract transient hydraulic features. It innovatively introduces the dynamic window function enhanced by local spectrum entropy. By integrating the local spectrum entropy calculation and combining it with the adaptive time-frequency analysis method, the time-frequency resolution of the transient hydraulic fluctuation signal is significantly improved. In the specific implementation, the calculation of the local spectrum entropy parameter can accurately represent the complexity of the signal by analyzing the local frequency distribution of the signal, and on this basis, the window width is dynamically adjusted to adapt to the change of the instantaneous frequency.

[0050] (3) In view of the technical problems in the existing well-layer coupling data modeling process, such as the difficulty in quantifying the interaction between the pumping system and the aquifer, the delayed response of the coupling model, the lack of dynamic characterization capability for nonlinear seepage changes, and the rough model parameter calibration mechanism, this scheme creatively adopts a joint modeling method of the pumping system and the aquifer combined with a multi-scale parameter coordination mechanism to carry out well-layer coupling data modeling. By constructing the dynamic transmission equation of the pumping system, the two-scale seepage response model of the aquifer and the coupling boundary setting mechanism, the real-time modeling and feedback optimization of the coupling process between the pumping system and the aquifer are realized. At the same time, the joint parameter calibration method is introduced to dynamically adjust the model parameters in combination with the sensor data, thereby improving the model's prediction capability for the aquifer compression response, water level disturbance transmission and pumping efficiency changes. BRIEF DESCRIPTION OF THE DRAWINGS

[0051] Figure 1 A structural diagram of a wellhead data acquisition and modeling system provided by the present invention;

[0052] Figure 2 A schematic flow chart of the steps performed by the system is provided for the present invention;

[0053] Figure 3 A flow chart of the steps performed by the hydraulic feature processing module;

[0054] Figure 4 Flowchart illustrating the steps performed for the well-layer coupling modeling module.

[0055] The accompanying drawings are used to provide further understanding of the present invention and constitute a part of the specification. They are used to explain the present invention together with the embodiments of the present invention and do not constitute a limitation of the present invention. DETAILED DESCRIPTION

[0056] The technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all the embodiments; based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention.

[0057] Example 1, see Figure 1 ,The technical solution adopted by the present invention is as follows: ,The present invention provides a wellhead data acquisition and modeling system, which includes a data acquisition module, a hydraulic characteristic processing module, a well-layer coupling modeling module and an interactive control module;

[0058] The data acquisition module is used for synchronous data acquisition and processing, obtains wellhead multi-source optimization data through data synchronous acquisition and processing, and sends the wellhead multi-source optimization data to the hydraulic characteristic processing module and the well-layer coupling modeling module;

[0059] The hydraulic feature processing module is used for extracting transient hydraulic features, obtaining hydraulic feature vectors and hydraulic anomaly warning signal data through the transient hydraulic feature extraction, and sending the hydraulic feature vectors and hydraulic anomaly warning signal data to the interactive control module;

[0060] The well-layer coupling modeling module is used for well-layer coupling data modeling, obtains a well-layer coupling prediction parameter combination through well-layer coupling data modeling, and sends the well-layer coupling prediction parameter combination to the interactive control module;

[0061] The interactive control module is used for model predictive control, and obtains wellhead data acquisition decision support data through model predictive control.

[0062] By performing the above operations, in order to address the technical problems in the existing wellhead data acquisition and modeling system, such as the inability to efficiently and synchronously collect multi-source data at the wellhead, the imperfect real-time feature extraction mechanism, and the lack of a response mechanism to hydraulic change trends, this solution creatively adopts a comprehensive optimization method that combines improved extraction and identification of hydraulic features with well-layer coupling modeling optimization. Through improved extraction and identification of hydraulic features, the problems of accuracy and real-time performance of hydraulic feature extraction are solved, so that changes in hydraulic patterns can be captured in a timely manner and effective early warnings can be issued. At the same time, through well-layer coupling modeling optimization, the coupling between the pumping system and the aquifer is enhanced, ensuring that the model can dynamically adapt to changes in the downhole environment and improving the accuracy and adaptability of modeling.

[0063] Example 2: This example is based on the above example. Figure 1 、 Figure 2 The data synchronous acquisition and processing is used to provide synchronous and efficient wellhead data, specifically to perform data synchronous acquisition to obtain wellhead raw data, and obtain wellhead multi-source optimized data through preprocessing operations;

[0064] The wellhead raw data includes water level data, instantaneous flow data, water temperature data, conductivity data, wellhead vibration signal and well pipe displacement data;

[0065] The wellhead multi-source optimized data specifically refers to the wellhead original data after time synchronization, denoising filtering and basic error calibration.

[0066] Example 3: This example is based on the above example. Figure 1 、 Figure 2 and Figure 3The transient hydraulic feature extraction is used to extract hydraulic features and identify abnormal hydraulic fluctuation states. Specifically, based on the wellhead multi-source optimization data, a feature-enhanced transient hydraulic fluctuation identification method improved by combining adaptive time-frequency analysis is used to perform transient hydraulic feature extraction to obtain hydraulic feature vectors and hydraulic anomaly warning signal data. The method includes the following steps: adaptive time-frequency analysis improvement, transient feature extraction, dynamic feature enhancement, adaptive hydraulic pattern classification, hydraulic feature vector construction, and hydraulic pattern anomaly detection and warning.

[0067] The improved adaptive time-frequency analysis is used to improve the time-frequency resolution of transient hydraulic fluctuation signals. Specifically, it is achieved by constructing a dynamic window function that integrates local spectral entropy enhancement and adopting a time-varying sampling strategy. The wellhead sampling step size is adjusted according to the change of instantaneous frequency. The time-frequency distribution data of the hydrodynamic signal is obtained through adaptive time-frequency distribution calculation.

[0068] The dynamic window function with integrated local spectrum entropy enhancement performs transient signal adjustment according to the instantaneous frequency in the wellhead multi-source optimization data, and improves the window width adjustment of the dynamic window function by constructing local spectrum entropy parameters to obtain a dynamic window adjustment calculation function, and the calculation formula is:

[0069] ;

[0070] Where, is the dynamic window adjustment calculation function, t is the time index, is the local time difference variable, is the dynamic window width, which is used to adjust the time-frequency resolution over time. The calculation formula is , where a is the basic window width, the default value is 0.1, b is the adjustment coefficient, the default value is 0.3, H f (t) is the local spectral entropy parameter;

[0071] The local spectrum entropy parameter is constructed by calculating the local frequency distribution of the signal at each sampling moment and performing a complexity representation of the signal spectrum based on the local spectrum distribution;

[0072] The calculation formula of the local spectrum distribution is:

[0073] ;

[0074] Wherein, P(f,t) is the local spectrum distribution function, X(f,t) is the complex spectrum coefficient after time-frequency transformation, and the complex spectrum coefficient X(f,t) after time-frequency transformation represents the complex frequency domain response of the original signal at time t and instantaneous frequency f, which is specifically calculated by short-time Fourier transform. |X(f,t)| is the amplitude of the original signal at time t and instantaneous frequency f, f is the instantaneous frequency, and F is the set of hydraulic characteristic instantaneous frequency bands;

[0075] The specific calculation formula of the local spectrum entropy parameter is:

[0076] ;

[0077] Where H f (t) is the local spectrum entropy parameter, P(f,t) is the local spectrum distribution function, and the local spectrum distribution function P(f,t) is used to represent the normalized frequency distribution as a whole. It is the calculation item of the frequency component information amount, specifically using the natural logarithm base;

[0078] The time-varying sampling strategy is used to adaptively adjust the data sampling period according to the instantaneous frequency changes caused by pumping behavior. Specifically, the sampling step length is calculated based on the maximum instantaneous frequency in the wellhead multi-source optimization data to obtain the time-varying sampling step length. The calculation formula is:

[0079] ;

[0080] Where, is the time-varying sampling step, where k is the sampling time index, f max (t k ) is the maximum instantaneous frequency corresponding to the kth sampling moment;

[0081] The adaptive time-frequency distribution calculation is specifically performed based on the dynamic window adjustment calculation function and the time-varying sampling step, by maximizing the time-frequency aggregation index to obtain the time-frequency distribution data of the hydrodynamic signal. The calculation formula is:

[0082] ;

[0083] Where, TFR[t k ,f] is the output of the function that maximizes the time-frequency aggregation, which is implemented in the form of improved short-time Fourier transform and used to represent the time-frequency distribution data of hydrodynamic signals, where f is the instantaneous frequency, is the original transient signal corresponding to the local time difference variable, which is used to represent the original sampling signal, where n is the discretized time index, The time-varying sampling step Adjust the calculation function for the dynamic window at the time center, The overall is the Fourier basis function, which is used to perform frequency transformation on the local signal;

[0084] The transient feature extraction is used to extract transient change features from the hydraulic mode signal, specifically by performing multi-scale gradient feature calculation and transient energy ratio extraction based on the time-frequency distribution data of the hydrodynamic signal to obtain transient gradient feature data and transient energy ratio data, and by constructing a transient feature matrix based on the transient gradient feature data and transient energy ratio data to obtain transient feature data;

[0085] The multi-scale gradient feature calculation is used to extract the local transient change rate from the time-frequency distribution of the hydraulic signal and reflect the energy mutation characteristics caused by pumping disturbance, water level mutation, aquifer response or equipment abnormality in the hydraulic system to obtain transient gradient feature data. The calculation formula is:

[0086] ;

[0087] Where G(t,s) is the transient gradient characteristic data, where t is the time index, s is the multi-scale parameter, which is specifically adjusted by adjusting the window width of the dynamic window adjustment calculation function, and TFR(t,f) is the time-frequency distribution data of the hydrodynamic signal corresponding to time t and instantaneous frequency f;

[0088] The transient energy ratio extraction is used to detect sudden changes, surges, pumping fluctuations and pump interference in hydraulic signals to obtain transient energy ratio data. The calculation formula is:

[0089] ;

[0090] Where TER(t) is the transient energy ratio data, f is the instantaneous frequency, F is the set of instantaneous frequency bands of hydraulic characteristics, T is the total integration time, and TFR(t,f) is the time-frequency distribution data of the hydrodynamic signal corresponding to time t and instantaneous frequency f;

[0091] The transient characteristic matrix is constructed by combining transient gradient characteristic data and transient energy ratio data to obtain transient characteristic data. The calculation formula is:

[0092] ;

[0093] Where, F trans is the transient characteristic data, G(t,s max ) is the transient gradient characteristic data at the optimal scale, s max is the optimal scale index, TER(t) is the transient energy ratio data, The overall is the transient energy ratio change rate parameter;

[0094] The dynamic feature enhancement is specifically to adopt a dynamic feature weighting method based on statistics, perform weighted calculation on the transient feature data according to the variance and change rate of the feature, and perform feature selection through a variance screening method to obtain the screening feature set data;

[0095] The adaptive hydraulic pattern classification is used to classify hydraulic patterns in real time and identify hydraulic pattern changes based on screening features. Specifically, the adaptive hydraulic pattern classification is performed by constructing a basic classification model and performing a dynamic hydraulic pattern data category center design based on real-time data to optimize the adaptability of the hydraulic pattern classification and obtain hydraulic pattern data category center feature data. Based on the hydraulic pattern data category center feature data, the similarity between the features in the screening feature set data and the hydraulic pattern data category center feature data is calculated, and hydraulic pattern classification is performed using a Gaussian model to obtain hydraulic pattern classification label data.

[0096] The basic classification model specifically adopts a Gaussian mixture model, receives a transient feature vector as input, calculates the response probability of the input under each Gaussian component, and uses a soft assignment strategy to assign the input samples to each category, and finally outputs the category corresponding to the maximum probability as the classification result;

[0097] The dynamic hydraulic model data category center design introduces an attenuation factor and adjusts the category center in real time. The calculation formula is:

[0098] ;

[0099] Where C c (t+1) is the cluster center of the cth dynamic hydraulic mode data category at time t+1, is the attenuation factor, the default value is 0.9, C c (t) is the cluster center of the cth dynamic hydraulic mode data category at time t, X c is the set of data samples classified as class c in the current time step, is the transient feature data of the data sample set belonging to the cth category;

[0100] The hydraulic mode classification adopts a standard Gaussian model to output the category probability, and selects the hydraulic mode type with the highest classification probability as the hydraulic mode classification output to obtain hydraulic mode classification label data;

[0101] The specific hydraulic mode types of the hydraulic mode classification label data include continuous pumping state, pulsed pumping state, sudden interference state and intermittent recharge state;

[0102] The hydraulic feature vector is constructed by constructing the hydraulic feature vector based on the hydraulic mode classification label data and the feature mean, feature standard deviation and feature change rate in the filtered feature set data to obtain the hydraulic feature vector. The calculation formula is:

[0103] ;

[0104] Where V(t) is the hydraulic characteristic vector, t is the time index, is the hydraulic mode classification label data, is the feature mean in the filtered feature set data, is the characteristic standard deviation, is the characteristic rate of change value;

[0105] The hydraulic pattern anomaly detection and warning is specifically based on the characteristic change rate value in the hydraulic feature vector and the hydraulic pattern classification label data, and adopts a dynamic threshold mechanism to achieve real-time identification and warning of pumping mutations and unstable behavior of the hydraulic system, thereby obtaining hydraulic anomaly warning signal data.

[0106] By performing the above operations, in order to address the technical problems in the existing hydraulic feature processing process, such as the rapid change rate of transient hydraulic features, fuzzy frequency domain information, difficult to locate the energy changes of pumping process signals, and susceptibility to high-frequency noise interference, this scheme creatively adopts a feature-enhanced transient hydraulic fluctuation identification method improved by adaptive time-frequency analysis to extract transient hydraulic features. It innovatively introduces a dynamic window function enhanced by local spectral entropy. By integrating local spectral entropy calculation and combining it with the adaptive time-frequency analysis method, it significantly improves the time-frequency resolution of transient hydraulic fluctuation signals. In the specific implementation, the calculation of the local spectral entropy parameters can accurately represent the complexity of the signal by analyzing the local frequency distribution of the signal, and on this basis, dynamically adjust the window width to adapt to changes in the instantaneous frequency.

[0107] Example 4: This example is based on the above example. Figure 1 、 Figure 2 and Figure 4 The well-layer coupling data modeling is used to establish a dynamic response model between the pumping system and the aquifer and realize real-time quantitative modeling and optimization of the coupling relationship. Specifically, based on the wellhead multi-source optimization data, a pumping system-aquifer joint modeling method combined with a multi-scale parameter coordination mechanism is adopted to perform well-layer coupling data modeling to obtain a well-layer coupling prediction parameter combination, including the following steps: pumping system modeling, aquifer modeling, coupling boundary setting and joint parameter calibration;

[0108] The pumping system modeling is used to describe the water flow behavior inside the water pump and well pipe. Specifically, a dynamic transmission equation between pumping flow, well pipe water level, and system load is established, and a state residual feedback mechanism is introduced to improve the model's adaptability to disturbance changes.

[0109] The calculation formula of the dynamic transmission equation is:

[0110] ;

[0111] Where h w (t) is the well water level parameter, Q pump (t) is the pump flow rate parameter, Q loss (t) is the hydraulic system loss, specifically used to represent the pipeline leakage parameter, A w is the cross-sectional area parameter of the well pipe, is the residual correction coefficient, and the specific calculation formula is ,in, It is the residual adjustment parameter, the default value is 0.4, h meas (t) is the actual water level parameter, h pred (t) is the predicted water level parameter;

[0112] The aquifer modeling is used to characterize the seepage response behavior of the aquifer during the pumping process. Specifically, a dual-scale seepage response model is used to jointly represent the macroscopic and microscopic aquifer characteristics. The dual-scale seepage response model includes a macroscopic layer model and a microscopic layer model. The macroscopic layer model is constructed based on the unsteady groundwater flow governing equation. The microscopic layer model introduces a void compression hysteresis function to calculate the corresponding hysteresis of the aquifer.

[0113] The calculation formula of the dual-scale seepage response model is:

[0114] ;

[0115] Where, is the dynamic water storage parameter of the aquifer, and the specific calculation formula is: , where S s is the static water yield parameter of the aquifer, is the gap compression weight, The overall is the derivative of the void compression hysteresis function, which is used to model the micro-layer model. is the pore compression hysteresis term, which is calculated by fitting the empirical function based on the water level change. h is the equivalent height parameter of the water level in the aquifer, K is the permeability coefficient, is the water level equivalent height permeability gradient term, which is used to represent the groundwater flow modeling, Q well (t) is the pumping source parameter per unit volume;

[0116] The coupling boundary setting specifically involves establishing coupling boundary conditions between the pumping system model and the aquifer model, and calculating the flow continuity at the well wall boundary by establishing the coupling boundary conditions. By introducing a multi-source data weight coefficient mechanism, parameters are adjusted in real time to obtain a well-layer joint initial model.

[0117] The calculation formula of the coupling boundary condition is:

[0118] ;

[0119] Where Q pump (t) is the pump flow rate parameter. In the calculation formula of the coupled boundary condition, the pump flow rate parameter is used to express the flow conservation relationship at the coupled boundary between the pumping system and the aquifer model. K is the permeability coefficient. A b is the total area of the well wall, The whole is the normal aquifer level gradient, r w is the well radius, r is the radial coordinate index, is the boundary normal direction parameter;

[0120] The calculation formula of the multi-source data weight coefficient mechanism is:

[0121] ;

[0122] Where, is a weighted model parameter, which is used to represent the model parameters of the well-layer joint initial model after real-time adjustment. is the initial parameter weight of the model, the default value is 0.4, are the initial parameter values of the pumping system model and the aquifer model, is the real-time parameter weight, the default value is 0.6, It is the optimal parameter modified based on real-time data observation;

[0123] The joint parameter calibration is used to optimize and adjust the parameters of the pumping system model and the aquifer model. Specifically, based on the actually collected water level, flow rate and conductivity sensor data, a parameter calibration method combined with a minimum error criterion is used to perform parameter updates and iterative model training to obtain a well-layer coupling comprehensive model. By using the well-layer coupling comprehensive model, a well-layer coupling prediction parameter combination is obtained.

[0124] The calculation formula of the parameter calibration method combined with the minimum error criterion is:

[0125] ;

[0126] Where, is the well-layer coupling model parameter, which is used to represent the internal parameter set of the well-layer coupling comprehensive model. is the well-layer coupling model parameter of the previous iteration in the iterative model training process, is the update step size, the default value is 0.2, It is the optimal model parameter value estimated by inversion based on the actual water level, flow and conductivity parameters. It is the model parameter of the well-layer joint initial model after real-time adjustment;

[0127] The well-layer coupling prediction parameter combination specifically includes a real-time hydraulic characteristic vector, a pumping system water level parameter, an aquifer response flow parameter, and an aquifer dynamic water storage parameter.

[0128] By performing the above operations, in order to address the technical problems in the existing well-layer coupling data modeling process, such as the difficulty in quantifying the interaction between the pumping system and the aquifer, the delayed response of the coupling model, the lack of dynamic characterization capability for nonlinear seepage changes, and the rough model parameter calibration mechanism, this scheme creatively adopts a pumping system-aquifer joint modeling method combined with a multi-scale parameter coordination mechanism to carry out well-layer coupling data modeling. By constructing the dynamic transmission equation of the pumping system, the two-scale seepage response model of the aquifer, and the coupling boundary setting mechanism, real-time modeling and feedback optimization of the coupling process between the pumping system and the aquifer are achieved. At the same time, a joint parameter calibration method is introduced to dynamically adjust the model parameters in combination with sensor data, thereby improving the model's prediction ability for aquifer compression response, water level disturbance transmission, and pumping efficiency changes.

[0129] Example 5: This example is based on the above example. Figure 1 and Figure 2 The model predictive control is used to dynamically optimize the well-layer coupling prediction model based on real-time monitoring data during the operation of the groundwater well, and to guide the adjustment and optimization of the wellhead data acquisition plan. Specifically, the model predictive control is performed based on the hydraulic characteristic vector, the hydraulic abnormality warning signal data and the well-layer coupling prediction parameter combination to obtain wellhead data acquisition decision support data;

[0130] The wellhead data acquisition decision support data specifically includes wellhead data acquisition priority decision data, data acquisition time attribute data, and wellhead data quality adjustment and optimization suggestions;

[0131] The wellhead data acquisition priority decision data is used to determine the data type to be acquired first and its corresponding data acquisition frequency, including the priority acquisition parameter type and data acquisition frequency;

[0132] The data collection time attribute data is used to optimize the collection scheduling strategy, including the collection window size and collection duration data;

[0133] The wellhead data quality adjustment optimization suggestion is used to provide a quality assessment result of the currently acquired data, specifically a quality assessment result of real-time data.

[0134] It should be noted that, in this document, relational terms such as first and second, etc., are used only to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations. Moreover, the terms "include," "comprise," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process or method comprising a set of elements includes not only those elements, but also other elements not explicitly listed, or elements inherent to such process or method.

[0135] While the embodiments of the present invention have been shown and described, it will be apparent to those skilled in the art that various changes, modifications, substitutions, and alterations can be made to the embodiments without departing from the principles and spirit of the invention.

[0136] The present invention and its embodiments are described above. This description is not restrictive. The drawings show only one embodiment of the present invention, and the actual structure is not limited thereto. In short, if a person skilled in the art is inspired by this and, without departing from the purpose of the present invention, designs structures and embodiments similar to this technical solution without inventiveness, they shall fall within the scope of protection of the present invention.

Claims

1. A wellhead data acquisition and modeling system, characterized by: It includes data acquisition module, hydraulic characteristic processing module, well-layer coupling modeling module and interactive control module; The data acquisition module is used for synchronous data acquisition and processing, obtains wellhead multi-source optimization data through data synchronous acquisition and processing, and sends the wellhead multi-source optimization data to the hydraulic characteristic processing module and the well-layer coupling modeling module; The hydraulic feature processing module is used for transient hydraulic feature extraction, and adopts a feature-enhanced transient hydraulic fluctuation identification method improved by combining adaptive time-frequency analysis to obtain hydraulic feature vectors and hydraulic anomaly warning signal data, including the following steps: adaptive time-frequency analysis improvement, transient feature extraction, dynamic feature enhancement, adaptive hydraulic pattern classification, hydraulic feature vector construction and hydraulic pattern anomaly detection and warning; the adaptive time-frequency analysis improvement is used to improve the time-frequency resolution of transient hydraulic fluctuation signals by constructing a dynamic window function that integrates local spectrum entropy enhancement and adopts a time-varying sampling strategy to adjust the wellhead sampling step according to the change of instantaneous frequency, and calculate through adaptive time-frequency distribution; the local spectrum entropy parameter is constructed by calculating the local frequency distribution of the signal at each sampling moment and performing a complexity representation of the signal spectrum based on the local spectrum distribution; The well-layer coupling modeling module is used for well-layer coupling data modeling. It adopts a pumping system aquifer joint modeling method combined with a multi-scale parameter coordination mechanism to obtain a well-layer coupling prediction parameter combination, including the following steps: pumping system modeling, aquifer modeling, coupling boundary setting and joint parameter calibration; The interactive control module is used for model predictive control, and obtains wellhead data acquisition decision support data through model predictive control.

2. A wellhead data acquisition and modeling system according to claim 1, characterized in that: The data synchronous acquisition and processing is used to provide synchronous and efficient wellhead data, specifically performing data synchronous acquisition to obtain wellhead raw data, and obtaining wellhead multi-source optimized data through preprocessing operations; The wellhead raw data includes water level data, instantaneous flow data, water temperature data, conductivity data, wellhead vibration signal and well pipe displacement data; The wellhead multi-source optimized data specifically refers to the wellhead original data after time synchronization, denoising filtering and basic error calibration.

3. A wellhead data acquisition and modeling system according to claim 2, characterized in that: The transient hydraulic feature extraction is used to extract hydraulic features and identify abnormal hydraulic fluctuation states. Specifically, based on the wellhead multi-source optimization data, a feature-enhanced transient hydraulic fluctuation identification method improved by combining adaptive time-frequency analysis is used to perform transient hydraulic feature extraction to obtain hydraulic feature vectors and hydraulic anomaly warning signal data. The method includes the following steps: adaptive time-frequency analysis improvement, transient feature extraction, dynamic feature enhancement, adaptive hydraulic pattern classification, hydraulic feature vector construction, and hydraulic pattern anomaly detection and warning. The improved adaptive time-frequency analysis is used to improve the time-frequency resolution of transient hydraulic fluctuation signals. Specifically, it is achieved by constructing a dynamic window function that integrates local spectral entropy enhancement and adopting a time-varying sampling strategy. The wellhead sampling step size is adjusted according to the change of instantaneous frequency. The time-frequency distribution data of the hydrodynamic signal is obtained through adaptive time-frequency distribution calculation. The dynamic window function with integrated local spectrum entropy enhancement performs transient signal adjustment according to the instantaneous frequency in the wellhead multi-source optimization data, and improves the window width adjustment of the dynamic window function by constructing local spectrum entropy parameters to obtain a dynamic window adjustment calculation function, and the calculation formula is: ； Where, is the dynamic window adjustment calculation function, t is the time index, is the local time difference variable, is the dynamic window width, which is used to adjust the time-frequency resolution over time. The calculation formula is , where a is the basic window width, b is the adjustment coefficient, H f (t) is the local spectral entropy parameter; The time-varying sampling strategy is used to adaptively adjust the data sampling period according to the instantaneous frequency changes caused by the pumping behavior. Specifically, the sampling step length is calculated based on the maximum instantaneous frequency in the wellhead multi-source optimization data to obtain the time-varying sampling step length; The adaptive time-frequency distribution calculation is specifically performed based on the dynamic window adjustment calculation function and the time-varying sampling step, by maximizing the time-frequency aggregation index to obtain the time-frequency distribution data of the hydrodynamic signal.

4. A wellhead data acquisition and modeling system according to claim 3, characterized in that: The local spectrum entropy parameter is constructed by calculating the local frequency distribution of the signal at each sampling moment and performing a complexity representation of the signal spectrum based on the local spectrum distribution; The calculation formula of the local spectrum distribution is: ； Wherein, P(f,t) is the local spectrum distribution function, X(f,t) is the complex spectrum coefficient after time-frequency transformation, and the complex spectrum coefficient X(f,t) after time-frequency transformation represents the complex frequency domain response of the original signal at time t and instantaneous frequency f, which is specifically calculated by short-time Fourier transform. |X(f,t)| is the amplitude of the original signal at time t and instantaneous frequency f, f is the instantaneous frequency, and F is the set of hydraulic characteristic instantaneous frequency bands; The specific calculation formula of the local spectrum entropy parameter is: ； Where H f (t) is the local spectrum entropy parameter, P(f,t) is the local spectrum distribution function, and the local spectrum distribution function P(f,t) is used to represent the normalized frequency distribution as a whole. It is a calculation item for the amount of frequency component information, specifically using the natural logarithm base.

5. A wellhead data acquisition and modeling system according to claim 4, characterized in that: The transient feature extraction is used to extract transient change features from the hydraulic mode signal, specifically by performing multi-scale gradient feature calculation and transient energy ratio extraction based on the time-frequency distribution data of the hydrodynamic signal to obtain transient gradient feature data and transient energy ratio data, and by constructing a transient feature matrix based on the transient gradient feature data and transient energy ratio data to obtain transient feature data; The multi-scale gradient feature calculation is used to extract the local transient change rate from the time-frequency distribution of the hydraulic signal, and reflect the energy mutation characteristics caused by pumping disturbance, water level mutation, aquifer response or equipment abnormality in the hydraulic system, and obtain transient gradient feature data; The transient energy ratio extraction is used to detect sudden changes, surges, pumping fluctuations and pump interference in hydraulic signals to obtain transient energy ratio data; The transient characteristic matrix is constructed by combining transient gradient characteristic data and transient energy ratio data to obtain transient characteristic data; The dynamic feature enhancement is specifically to adopt a dynamic feature weighting method based on statistics, perform weighted calculation on the transient feature data according to the variance and change rate of the feature, and perform feature selection through a variance screening method to obtain the screening feature set data; The adaptive hydraulic pattern classification is used to perform real-time hydraulic pattern classification and identify hydraulic pattern changes based on screening features. Specifically, it constructs a basic classification model and performs dynamic hydraulic pattern data category center design based on real-time data to optimize the adaptability of hydraulic pattern classification and obtain hydraulic pattern data category center feature data. Based on the hydraulic pattern data category center feature data, the similarity between the features in the screening feature set data and the hydraulic pattern data category center feature data is calculated, and hydraulic pattern classification is performed through a Gaussian model to obtain hydraulic pattern classification label data.

6. A wellhead data acquisition and modeling system according to claim 5, characterized in that: The hydraulic feature vector is constructed by constructing the hydraulic feature vector based on the hydraulic mode classification label data and the feature mean, feature standard deviation and feature change rate values in the filtered feature set data to obtain the hydraulic feature vector; The hydraulic pattern anomaly detection and warning is specifically based on the characteristic change rate value in the hydraulic feature vector and the hydraulic pattern classification label data, and adopts a dynamic threshold mechanism to achieve real-time identification and warning of pumping mutations and unstable behavior of the hydraulic system, thereby obtaining hydraulic anomaly warning signal data.

7. The wellhead data acquisition and modeling system according to claim 6, characterized in that: The well-layer coupling data modeling is used to establish a dynamic response model between the pumping system and the aquifer and realize real-time quantitative modeling and optimization of the coupling relationship. Specifically, based on the wellhead multi-source optimization data, a pumping system-aquifer joint modeling method combined with a multi-scale parameter coordination mechanism is adopted to perform well-layer coupling data modeling to obtain a well-layer coupling prediction parameter combination, including the following steps: pumping system modeling, aquifer modeling, coupling boundary setting, and joint parameter calibration; The pumping system modeling is used to describe the water flow behavior inside the water pump and well pipe. Specifically, a dynamic transmission equation between pumping flow, well pipe water level, and system load is established, and a state residual feedback mechanism is introduced to improve the model's adaptability to disturbance changes. The aquifer modeling is used to characterize the seepage response behavior of the aquifer during the pumping process. Specifically, a dual-scale seepage response model is used to jointly represent the macroscopic and microscopic aquifer characteristics. The dual-scale seepage response model includes a macroscopic layer model and a microscopic layer model. The macroscopic layer model is constructed based on the unsteady groundwater flow governing equation. The microscopic layer model introduces a void compression hysteresis function to calculate the corresponding hysteresis of the aquifer. The coupling boundary setting specifically involves establishing coupling boundary conditions between the pumping system model and the aquifer model, and calculating the flow continuity at the well wall boundary by establishing the coupling boundary conditions. By introducing a multi-source data weight coefficient mechanism, parameters are adjusted in real time to obtain a well-layer joint initial model. The joint parameter calibration is used to optimize and adjust the parameters of the pumping system model and the aquifer model. Specifically, based on the actually collected water level, flow rate and conductivity sensor data, a parameter calibration method combined with a minimum error criterion is used to perform parameter updates and iterative model training to obtain a well-layer coupling comprehensive model. By using the well-layer coupling comprehensive model, a well-layer coupling prediction parameter combination is obtained. The well-layer coupling prediction parameter combination specifically includes a real-time hydraulic characteristic vector, a pumping system water level parameter, an aquifer response flow parameter, and an aquifer dynamic water storage parameter.

8. The wellhead data acquisition and modeling system according to claim 7, characterized in that: The model predictive control is used to dynamically optimize the well-layer coupling prediction model based on real-time monitoring data during the operation of the groundwater well, and to guide the adjustment and optimization of the wellhead data acquisition plan. Specifically, the model predictive control is performed based on the combination of the hydraulic characteristic vector, the hydraulic anomaly warning signal data and the well-layer coupling prediction parameter to obtain wellhead data acquisition decision support data; The wellhead data acquisition decision support data specifically includes wellhead data acquisition priority decision data, data acquisition time attribute data, and wellhead data quality adjustment and optimization suggestions; The wellhead data acquisition priority decision data is used to determine the data type to be acquired first and its corresponding data acquisition frequency, including the priority acquisition parameter type and data acquisition frequency; The data collection time attribute data is used to optimize the collection scheduling strategy, including the collection window size and collection duration data; The wellhead data quality adjustment optimization suggestion is used to provide a quality assessment result of the currently acquired data, specifically a quality assessment result of real-time data.

Images