Accurate water level measurement and monitoring management method based on geothermal well water level gauge

Abstract

The invention provides a geothermal well water level gauge-based water level accurate measurement and monitoring management method, which comprises the following steps of: performing integrated monitoring deployment by combining various sensors, probe fixing guide rails and double probes; deploying an edge calculation module to obtain underground measurement data; performing drift correction and anti-interference detection on the sensor, and deploying a sensor standby channel; data storage and visualization are carried out; a safety threshold value is preset, a water level prediction model is constructed, water level anomaly detection is carried out, and fault cause tracing and strategy suggestion are carried out on anomaly; equipment linkage is carried out, real-time monitoring and remote automatic control are carried out, and health scoring and service life prediction of the equipment are carried out. The precision and reliability of water level measurement can be improved, timely and effective continuous monitoring and intelligent management are achieved, meanwhile, through cooperation of anomaly detection and automatic control, faults and risks are reduced, geothermal resources are protected, and manpower and operation cost is saved.

Description

Technical Field

[0001] The present invention relates to the technical field of geothermal well water level measurement, and particularly to a method for precise measurement, monitoring and management of the water level based on a geothermal well water level gauge. Background Art

[0002] Geothermal well water level measurement refers to monitoring the liquid level height of hot water or steam in the well through technical means, reflecting the dynamic changes of the geothermal reservoir, including parameters such as water level rise and fall, pressure fluctuation, etc. Its water level change can directly reflect the exploitation volume and recharge balance of geothermal resources, guiding reasonable development. The existing geothermal well water level measurement methods include the pressure sensor method, the ultrasonic measurement method, the conductivity probe method, etc. A geothermal well water level gauge is a special instrument and equipment for measuring and monitoring the dynamic changes of the geothermal well water level. It is a special equipment integrating the above measurement technologies and usually includes a sensor, a data acquisition module and a transmission system. It is the core tool for realizing high-precision measurement, and its technical performance directly determines the accuracy and application value of the data.

[0003] Due to the special environment in the well (such as high temperature, high pressure, high salinity), at present, the existing measurement methods still have the problem of inaccurate measurement in deep well conditions. For example: 1) In the pressure sensor method, the temperature in the geothermal well can reach 110 - 180 °C, and the pressure is as high as 1 - 10 MPa. Ordinary pressure sensors are prone to aging and drift, and need to be calibrated frequently. Moreover, when converting the water level through pressure, it is necessary to assume that the fluid density is constant. In actual working conditions, fluid phase change, such as steam-water mixture, will introduce errors. 2) In the ultrasonic measurement method, bubbles, steam or suspended impurities in the well will scatter sound waves, resulting in attenuation or misjudgment of the measurement signal. 3) The float type water level gauge is only applicable to scenarios where the water level changes slowly and cannot capture sudden fluctuations, such as the sudden change of the water level during the instant of pumping and recharge. Moreover, enough space needs to be reserved in the wellbore, which is not applicable to small-diameter wells or multi-layer pipe wells. Generally speaking, the existing measurement methods are relatively single, it is difficult to obtain parameters such as water level, temperature, pressure, etc. simultaneously, limiting the comprehensive analysis ability, and during the measurement, there is a lack of timely and effective monitoring, resulting in a large amount of later maintenance work of the equipment and increasing the manpower input. Summary of the Invention

[0004] In order to overcome the deficiencies of the prior art, the purpose of the present invention is to provide a method for precise measurement, monitoring and management of the water level based on a geothermal well water level gauge, which can improve the accuracy and reliability of water level measurement, achieve timely and effective continuous monitoring and intelligent management, and at the same time, through the coordination of anomaly detection and automatic control, reduce faults and risks, protect geothermal resources, and save manpower and operation costs.

[0005] To achieve the above purpose, the present invention provides the following solution: A method for precise measurement, monitoring and management of the water level based on a geothermal well water level gauge, including:

[0006] Select multiple pressure sensors for combination to obtain the main monitoring component, select fiber optic sensors as the auxiliary monitoring component, and design probe fixing guides and dual probes in the geothermal well pipeline to complete the integrated monitoring deployment;

[0007] Set an embedded processor and a high-precision clock near the wellhead or well site to collect and preprocess the sensor data locally, and then perform multi-source data fusion on the preprocessed data to obtain downhole measurement data;

[0008] Perform drift correction on the pressure sensors, perform automated anti-interference detection on the pressure sensors at the same depth, and set two sensor nodes and two independent power supplies at the same depth for continuous monitoring during faults;

[0009] Transmit the downhole measurement data to the cloud database for hierarchical storage and management, and design a cloud API interface to build a data visualization platform to display real-time water level curves, temperature distributions, and alarm records;

[0010] Preset a water level safety threshold, perform real-time monitoring and alarms on the downhole measurement data, construct a water level prediction model, perform water level anomaly detection to obtain abnormal events, and then trace the root cause of the fault and provide strategy suggestions for the abnormal events;

[0011] Automatically adjust the operating frequency of the geothermal pump according to the downhole measurement data and strategy suggestions, design a mobile operation and maintenance APP for real-time monitoring and remote control of the geothermal well, and then construct a device health status assessment model based on device indicators to perform device health scoring and life prediction.

[0012] Optionally, select multiple pressure sensors for combination to obtain the main monitoring component, and select fiber optic sensors as the auxiliary monitoring component, including:

[0013] Based on the probe fixing guide, arrange a set of pressure sensors every 50 - 100 meters according to the depth gradient in the geothermal well pipeline to obtain the main monitoring component; among them, each set of the pressure sensors includes a main pressure sensor and a secondary pressure sensor;

[0014] Use the RS485 communication interface and differential amplifier to collect and transmit the signals of the pressure sensors, and configure a temperature sensor on the pressure sensors to dynamically correct the temperature error on the pressure readings;

[0015] Deploy fiber Bragg grating nodes around the geothermal well pipeline at a spacing of 1 meter or less than 1 meter, and embed metal waveguide grooves for fixing fiber Bragg gratings inside the geothermal well pipeline to obtain the auxiliary monitoring component;

[0016] The fiber Bragg grating is used to collect the pressure and temperature curves, and the wavelength data is output in real time through an optical wave demodulator. Based on the output wavelength data, the temperature and pressure fluctuations are monitored in real time by wavelength demodulation, and dynamic error compensation is provided by phase demodulation.

[0017] Optionally, a self-hydrophobic coating is provided on the surface of the dual probe. The dual probe includes a main probe for deep-water data collection and a secondary probe for detecting humidity and water droplets. The main probe is provided with a titanium alloy shell, and a temperature sensor is provided outside the main probe to directly contact the water body and sense the change of the water body temperature. The secondary probe is set as a short needle, and a humidity sensor or a capacitive sensor is provided outside the secondary probe for water contact alarm.

[0018] Optionally, an embedded processor and a high-precision clock are set near the wellhead or the well site to collect and locally preprocess the sensor data, and then multi-source data fusion is performed on the preprocessed data to obtain downhole measurement data, including:

[0019] An embedded sensor and a high-precision clock are deployed near the wellhead or the well site, and a communication interface is integrated to obtain an edge computing module. Then, the edge computing module is used to collect data to obtain initial measurement data;

[0020] The initial measurement data is filtered for noise using a low-pass filter and a median filter, the pressure data is compensated in real time using a compensation algorithm, and the continuous monitoring data of the sensor is segmented and normalized and reversibly compressed to complete the local preprocessing of the data; the expression of the compensation algorithm is:

[0021] P corrected =P measured ×(1+α(T-T 0 ))

[0022] where P corrected is the corrected pressure value, P measured is the measured pressure value, α is the material expansion coefficient, T 0 is the calibration temperature, and T is the actual temperature;

[0023] The high-precision clock and the PTP protocol are used to synchronize the clocks of the multi-source sensor data, and the time stamps of the multi-source sensor data per second are aligned to complete the data fusion and obtain the downhole measurement data.

[0024] Optionally, drift correction is performed on the pressure sensor, automatic anti-interference detection is performed on the pressure sensors at the same depth, and two sensor nodes and two independent power supplies are provided at the same depth for continuous monitoring during a fault, including:

[0025] Collect the reference water level and temperature values when the geothermal well is stable, establish a calibration baseline, and use the real-time data of the pressure sensor to dynamically update the calibration baseline;

[0026] Based on the calibration baseline, use Kalman filtering to predict, update, and calibrate the pressure sensor, and use the historical data of the pressure sensor to train a machine learning model to obtain a prediction calibration model. Use the prediction calibration model to predict the drift trend of the pressure sensor to complete machine learning-assisted calibration;

[0027] Based on the pressure sensors at the same depth, perform differential verification. When the difference exceeds the preset interference range, mark it as an interference or abnormal state to complete automatic anti-interference detection;

[0028] At the same depth of the geothermal well, two sensor nodes and two independent power supplies are set up to obtain a backup channel for each group of the pressure sensors. When it is detected that the pressure sensor has no signal or the drift exceeds the standard, it is automatically switched to the backup channel for continuous monitoring during a fault.

[0029] Optionally, the detection expression of the pressure sensor is:

[0030]

[0031] where P final is the final pressure value, n is the number of sensors participating in data fusion, ω i is the weighting coefficient of the i-th sensor, and P sensor,i is the pressure value measured by the i-th sensor.

[0032] Optionally, transmit the downhole measurement data to the cloud database for hierarchical storage and management, design a cloud API interface, and build a data visualization platform to display real-time water level curves, temperature distributions, and alarm records, including:

[0033] Use cellular networks, wireless transmission technologies, or fiber optic communications to transmit the downhole measurement data to a time series database in the cloud; the time series database includes a real-time storage layer and a historical storage layer, and the data structure in the time series database consists of a timestamp, water level, temperature, pressure, and sensor number;

[0034] Develop a standardized data interface, use the OAuth protocol to authenticate user access, and set permission levels for different roles;

[0035] Use a front-end framework to build a responsive web interface and integrate a streaming chart component to dynamically render the water level change trend and temperature field distribution to display real-time water level curves, temperature distributions, and alarm records.

[0036] Optionally, preset a water level safety threshold and a water level fluctuation threshold, monitor the downhole measurement data in real time and issue an alarm, and construct a water level prediction model to detect water level anomalies and obtain abnormal events, including:

[0037] Preset the upper and lower limits of the water level to obtain the water level safety threshold, and determine whether the downhole measurement data exceeds the water level safety threshold. If so, automatically trigger an alarm;

[0038] Use the sliding window statistical method to calculate the mean, maximum, minimum, and standard deviation of the water level change in the past 15 minutes to obtain the water level fluctuation range, set the water level fluctuation threshold, and then determine whether the current 15-minute water level change exceeds the water level fluctuation threshold. If so, automatically trigger an alarm;

[0039] Select and combine the LSTM model and the ARIMA model to obtain the initial model architecture, and based on historical downhole measurement data, extract the correlation characteristics between the water level, temperature, and pressure as features to train the initial model architecture to obtain the water level prediction model;

[0040] Use the water level prediction model to obtain the water level prediction value, calculate the water level error between the water level prediction value and the actual water level value. When the water level error exceeds the preset error threshold, mark it as an abnormal water level, and record the water level value, sensor ID, timestamp, temperature, and pressure information of the abnormal water level to obtain an abnormal event.

[0041] Optionally, trace the cause of the fault and provide policy suggestions for the abnormal event, including:

[0042] Combine the sensor drift state, historical fault information, pump power load, geothermal well equipment operation log, and wellhead environment data to obtain multi-modal associated data;

[0043] Use the causal analysis tree algorithm and the multi-modal associated data to analyze the cause of the fault of the abnormal event, and classify the abnormal event according to the cause of the fault;

[0044] Preset a maintenance suggestion library, and automatically generate different policy suggestions according to different types of the abnormal events.

[0045] Optionally, automatically adjust the operating frequency of the geothermal pump according to the downhole measurement data and policy suggestions, design a mobile operation and maintenance APP for real-time monitoring and remote control of the geothermal well, and then construct an equipment health status assessment model based on equipment indicators to perform equipment health scoring and life prediction, including:

[0046] Synchronize the downhole measurement data and strategic suggestions to the pumping system control unit to interface with the geothermal water pump, and preset an emergency control trigger mechanism and a dynamic adjustment control strategy. Use the PID algorithm to perform automatic control of the geothermal water pump. The emergency control trigger mechanism automatically shuts down the pump by determining whether the current water level is lower than the lower water level limit. The dynamic adjustment control strategy adjusts the pump speed by determining whether the water level change within the current 15 minutes exceeds the water level fluctuation threshold.

[0047] Design a mobile operation and maintenance APP for visual monitoring and anomaly push of geothermal wells, remote adjustment of the dynamic control strategy, and generation of operation and maintenance logs. The operation and maintenance logs include anomaly records, operation records, and equipment status.

[0048] Collect the working time, vibration value, and power consumption of the pump to obtain equipment indicators. Based on the equipment indicators, construct an equipment health assessment model. Use the equipment health status assessment model to regularly calculate the health score of the latest equipment, and then combine the health score to calculate the equipment life prediction curve in real time. Among them:

[0049] The expression of the health score is:

[0050]

[0051] Among them, H(t) is the health score, ω i is the weighting coefficient of the i-th sensor, X i (t) is the specific sensor value at time t, X i,std is the standard state expected value of the specific sensor, σ i is the standard deviation of the specific sensor, and n is the type of input sensor.

[0052] The expression of the life prediction curve is:

[0053] R(t) = L 0 ·H(t) β

[0054] Among them, R(t) is the predicted remaining life at the calculation moment, L 0 is the theoretical maximum life of the equipment, and β is the attenuation rate parameter.

[0055] By providing a method for accurate measurement and monitoring management of the water level based on a geothermal well water level gauge, the present invention discloses the following technical effects:

[0056] 1. Diversified measurement methods: 1) By designing a combined measurement method with multiple sensors, a multi-sensor array composed of a pressure sensor (for high-sensitivity water pressure measurement) and an optical fiber sensor (with high temperature resistance and corrosion prevention capabilities) is adopted to compensate and verify the data of the pressure sensor, improving the measurement accuracy; 2) By designing a dedicated dual probe, a main probe and a secondary probe are designed. The main probe is extended to detect deeper water levels, and a temperature sensor is equipped to quickly sense the temperature at the moment of contacting the water surface. The secondary probe only touches the water surface for alarm to avoid false alarms caused by humidity or surface water droplets. The functions of the main and secondary probes are separated, improving the reliability of the measurement.

[0057] 2. Intelligent data acquisition and processing: 1) Deploy an edge computing module near the wellhead or well site to achieve synchronous acquisition of multi-channel sensor data. Combined with an embedded microprocessor (such as an ARM or RISC-V architecture), signal filtering and temperature compensation algorithms are used to improve the data quality; 2) It can compare the sensor data with known reference values or optical fiber monitoring results, automatically correct the drift and error of each sensor, and can also combine the Kalman filter and prediction calibration model to multi-step predict the drift trend of the sensor and achieve the adaptive calibration of the sensor; 3) It can achieve anti-interference detection between the same group of sensors to enable the use of backup measurement channels to ensure that the system still has core monitoring capabilities in the case of failure of some sensors or power supplies.

[0058] 3. Real-time monitoring and fault tracing: 1) By setting multiple safety thresholds, rule-based automatic alarms for anomalies can be achieved. At the same time, combined with long-term and short-term predictions based on machine learning algorithms, the predicted water level values in the future can be predicted, enhancing the detection ability for various types of anomalies; 2) Through the correlation analysis of multi-modal data, the efficiency of anomaly location and fault troubleshooting is significantly improved.

[0059] 4. Efficient maintenance management: 1) By linking the data with the control unit of the pumping system, it can be directly connected to hardware such as the frequency converter of the pump and the valve drive, and then realize automatic control, quickly respond in case of emergencies, and at the same time have flexible and fine-grained operation capabilities (such as dynamically adjusting the pump speed); 2) It can build a health assessment model for major equipment (such as pumps, water level gauges, or filtration devices, etc.) based on sensors and historical data, perform equipment health scoring and life prediction, so as to better generate maintenance plans, spare part lists, etc., improve the later maintenance efficiency and management efficiency of the equipment, and reduce costs.

[0060] The technical solution of the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. Description of the Drawings

[0061] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings required for use in the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can be obtained based on these drawings.

[0062] Figure 1 Schematic diagram of the method flow provided by the embodiment of the present invention;

[0063] Figure 2 Schematic diagram of the sensor detection and correction process provided by the embodiment of the present invention;

[0064] Figure 3 Schematic diagram of the water level monitoring and abnormal alarm process provided by the embodiment of the present invention. Detailed implementation manners

[0065] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, rather than all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts belong to the scope of protection of the present invention.

[0066] To make the above objects, features, and advantages of the present invention more obvious and understandable, the present invention will be further described in detail below in conjunction with the drawings and specific implementation manners.

[0067] As Figure 1 shown, the present invention provides a method for accurate water level measurement and monitoring management based on a geothermal well water level gauge, including the following steps:

[0068] Step 1, as Figure 2 shown, select a variety of pressure sensors for combination to obtain a main monitoring component, select a fiber optic sensor as an auxiliary monitoring component, and design a probe fixing guide rail and a double probe in the geothermal well pipeline to complete the integrated monitoring deployment; specifically including:

[0069] 1.1 Based on the probe fixing guide rail, according to the depth gradient in the geothermal well pipeline, arrange a set of pressure sensors every 50 - 100 meters to obtain the main monitoring component; wherein, each set of the pressure sensors includes a main pressure sensor and a sub-pressure sensor, and the sub-pressure sensor is mainly used for redundancy support to prevent the main pressure sensor.

[0070] Pressure sensor: The operating pressure range should cover the hydrostatic pressure in the geothermal well (usually 20 - 25 MPa) and adapt to high temperature and high pressure conditions (the maximum operating temperature is 150 - 200 °C). A high-precision MEMS pressure sensor can be selected, with an accuracy of 0.1% FS and high vibration resistance. The housing material is selected as titanium alloy or 316L stainless steel to adapt to the high salinity corrosion environment. The internal processing of the chip adopts a vacuum potting process to protect the internal sensitive components from the adverse effects of high temperature and high pressure and extend the service life.

[0071] 1.2 Use the RS485 communication interface and differential amplifier to collect and transmit the signals of the pressure sensor, and configure a temperature sensor on the pressure sensor to dynamically correct the temperature error in the pressure reading.

[0072] 1.3 Deploy fiber Bragg grating nodes around the geothermal well pipeline at a spacing of 1 meter or less than 1 meter, and embed a metal waveguide groove for fixing the fiber Bragg grating inside the geothermal well pipeline (to avoid damage caused by vibration) to obtain an auxiliary monitoring component.

[0073] Fiber optic sensor (fiber Bragg grating, FBG): High temperature pressure (temperature resistance 200 °C, accuracy up to ±0.1 MPa), monitoring of the temperature field at different depths in deep wells (accuracy 0.1 °C), and the protective layer adopts a plasma-sprayed alumina ceramic protective layer (temperature resistance 800 °C) to avoid erosion by mineralized substances and impurities.

[0074] 1.4 Use the fiber Bragg grating to collect the pressure and temperature curves, and output the wavelength data in real time through an optical wave demodulator. Based on the output wavelength data, use wavelength demodulation to monitor the temperature and pressure fluctuations in real time, and use phase demodulation to provide dynamic error compensation.

[0075] 1.5 A self-hydrophobic coating is provided on the surface of the double probe. The double probe includes a main probe for deep water data collection and a sub-probe for detecting humidity and water droplets. The main probe is provided with a titanium alloy housing, and a temperature sensor (bare design) is provided outside the main probe to directly contact the water body and sense the change in the water body temperature. The sub-probe is set as a short needle, and a humidity sensor or a capacitive sensor is provided outside the sub-probe to perform water contact alarm to avoid mis-touch problems.

[0076] Step 2, as Figure 2 shown, set an embedded processor and a high-precision clock near the wellhead or the well site to collect and locally preprocess the sensor data, and then perform multi-source data fusion on the preprocessed data to obtain the downhole measurement data; specifically including:

[0077] 2.1 Deploy embedded sensors (such as ARM Cortex-A series chips or FPGA+ARM architecture, supporting multi-task parallel processing) and high-precision clocks (such as TCXO crystal oscillators, with clock drift <0.1ppm) near the wellhead or on the well site, and integrate communication interfaces (including RS485, Modbus, CAN bus, fiber optic interfaces, etc., facilitating the integration of multiple sensors) to obtain an edge computing module, and then use the edge computing module to collect data to obtain initial measurement data;

[0078] 2.2 Use a low-pass filter and a median filter to filter the noise of the initial measurement data, use a compensation algorithm to perform real-time compensation on the pressure data, and perform segmented standardization and reversible compression on the continuous monitoring data of the sensor (reducing the data transmission burden) to complete the local preprocessing of the data; the expression of the compensation algorithm is:

[0079] P corrected =P measured ×(1+α(T-T 0 ))

[0080] where P corrected is the corrected pressure value, P measured is the measured pressure value, α is the material expansion coefficient, T 0 is the calibration temperature, and T is the actual temperature.

[0081] 2.3 Use the high-precision clock and the PTP protocol (Precision Time Protocol) to synchronize the clocks of multi-source sensor data and align the timestamps of multi-source sensor data per second to complete data fusion, generate a unified data stream, and obtain downhole measurement data.

[0082] Step 3: Perform drift correction on the pressure sensor, perform automated anti-interference detection on the pressure sensors at the same depth, and set two sensor nodes and two independent power supplies at the same depth for continuous monitoring during a fault; specifically including:

[0083] 3.1 Collect the reference water level and temperature values when the geothermal well is stable, establish a calibration baseline, and use the real-time data of the pressure sensor to perform dynamic update of the calibration baseline;

[0084] 3.2 Based on the calibration baseline, use Kalman filtering to predict (predict the current state according to the sensor historical values), update (combine the actual measurement values and fuse and update the drift with the prediction results), and calibrate (re-enter the parameters after drift correction into the system) the pressure sensor, and use the historical data of the pressure sensor to train a machine learning model (such as LSTM time series analysis) to obtain a prediction calibration model, and use the prediction calibration model to predict the drift trend of the pressure sensor to complete machine learning-assisted calibration;

[0085] 3.3 Based on the pressure sensors at the same depth, perform differential verification. When the difference exceeds the preset interference range, mark it as an interference or abnormal state to complete the automated anti-interference detection.

[0086] 3.4 Two sensor nodes and two independent power supplies are set at the same depth of the geothermal well to obtain the backup channels of each group of the pressure sensors. When it is detected that the pressure sensor has no signal or the drift exceeds the standard, it will automatically switch to the backup channel for continuous monitoring during a fault. The detection expression of the pressure sensor is:

[0087]

[0088] where P final is the final pressure value, n is the number of sensors participating in data fusion, ω i is the weighting coefficient of the i-th sensor, and P sensor,i is the pressure value measured by the i-th sensor.

[0089] Step 4: Transmit the downhole measurement data to the cloud database for hierarchical storage and management, and design a cloud API interface to build a data visualization platform to display the real-time water level curve, temperature distribution, and alarm records; specifically including:

[0090] 4.1 Use cellular networks, wireless transmission technologies, or fiber optic communications to transmit the downhole measurement data to the time series database in the cloud; the time series database includes a real-time storage layer (storing high-frequency data for the most recent 7 days for real-time query) and a historical storage layer (saving historical data, using compressed storage for convenient long-term backtracking), and the data structure in the time series database consists of a timestamp, water level, temperature, pressure, and sensor number.

[0091] 4G / 5G cellular network: In areas with stable cellular signal coverage, give priority to using 4G / 5G modules to ensure fast data transmission and unrestricted broadband access, and use industrial-grade wireless communication modules (such as SIMCom, Quectel) to enhance anti-interference and high-temperature adaptability.

[0092] LoRa or LoRaWAN (long-distance wireless transmission): For areas with a large well site area or weak cellular signals, use LoRa modules to transmit data from multiple monitoring nodes back to a centralized gateway to achieve low-power, long-distance (>10km) transmission. Combining with the LoRaWAN protocol, it supports multi-node data management and reduces construction costs at the same time.

[0093] Fiber Optic Communication: When there are fixed network deployment conditions between the well site and the data center, fiber optic communication access is preferably promoted. An industrial fiber optic terminal is introduced at the wellhead to transmit data stably at high speed through the optical cable, which is especially suitable for areas with requirements for highly reliable transmission.

[0094] 4.2 Develop standardized data interfaces, authenticate user access using the OAuth protocol (such as OAuth2.0), and set permission levels for different roles (such as administrator, ordinary user, auditor).

[0095] 4.3 Use front-end frameworks (such as React, Vue) to build a responsive web interface and integrate streaming chart components (such as Chart.js, d3.js) to dynamically render the water level change trend and temperature field distribution to display the real-time water level curve, temperature distribution, and alarm records.

[0096] Real-time water level curve: Displays the water level change trend according to the depth, supports zooming and sliding to view the history;

[0097] Temperature distribution: Presents the temperature gradient change at different depths in the form of a heat map;

[0098] Alarm records: Displays the past alarm times and conditions and supports filtering.

[0099] Step 5, as Figure 3 shown, preset the water level safety threshold, conduct real-time monitoring and alarming of the downhole measurement data, construct a water level prediction model, conduct water level anomaly detection to obtain abnormal events, and then trace the root cause of the fault and give policy suggestions for the abnormal events; specifically including:

[0100] 5.1 Preset the upper water level and the lower water level to obtain the water level safety threshold, and judge whether the downhole measurement data exceeds the water level safety threshold. If so, an alarm is automatically triggered.

[0101] Upper water level (high water level risk): Prevent geothermal well leakage. For example, when the water level reaches 95% of the capacity → trigger a warning;

[0102] Lower water level (low water level risk): Avoid the pump running idly. For example, when the water level is below 10% of the capacity → stop pumping and give an alarm;

[0103] Alarm trigger logic: Combine the real-time collected data. If it exceeds the threshold, output multi-level alarms. For example, the first-level alarm (close to the threshold): Remind to pay attention and detect the trend; the second-level alarm (exceeding the limit): Send an SMS push or an App notification, and start the protection strategy (such as stopping the pump or reducing the speed).

[0104] 5.2 Use the sliding window statistical method to calculate the mean, maximum, minimum and standard deviation of the water level changes in the past 15 minutes, obtain the water level fluctuation range, and set the water level fluctuation threshold (such as triggering a short-term fluctuation alarm when the water level changes by >5% within 15 minutes), and then determine whether the water level change within the current 15 minutes exceeds the water level fluctuation threshold. If so, the alarm is automatically triggered.

[0105] 5.3 Select and combine the LSTM model and the ARIMA model to obtain the initial model architecture, and based on the historical downhole measurement data, extract the correlation characteristics between the water level, temperature and pressure as features, train the initial model architecture, obtain the water level prediction model, and output the water level prediction value for a period of time in the future.

[0106] LSTM (Long Short-Term Memory Network): A time series model based on deep learning, suitable for capturing the long-term and short-term dynamics of water level changes;

[0107] ARIMA (Autoregressive Integrated Moving Average): A traditional statistical model used for balanced forecasting of time series.

[0108] 5.4 Use the water level prediction model to obtain the water level prediction value, calculate the water level error between the water level prediction value and the actual water level value, and when the water level error exceeds the preset error threshold, it is marked as an abnormal water level, and the water level value, sensor ID, timestamp, temperature and pressure information of the abnormal water level are recorded to obtain an abnormal event.

[0109] Trace the cause of the abnormal event and provide strategic recommendations, including:

[0110] 5.5 Combine the sensor drift status and historical fault information, pump power load, geothermal well equipment operation log and wellhead environmental data (such as wellhead temperature, weather data, etc.) to obtain multimodal correlation data;

[0111] 5.6 Analyze the fault cause of the abnormal event by using the causal analysis tree algorithm and the multimodal association data, and classify the abnormal event according to the fault cause.

[0112] Causes of failure: abnormal sensor drift, rapid pumping of well water, excessive pump power, insufficient water supply, etc.; abnormal classification such as "low pressure abnormality", "pumping abnormality", "communication interruption" provide a clear basis for fault location.

[0113] 5.7 The preset maintenance suggestion library automatically responds to and generates different strategic suggestions according to different types of abnormal events, such as low water level abnormality: reduce the speed of the pump to maintain system stability. Severe water level fluctuations: it is recommended to inspect the surrounding water replenishment equipment.

[0114] Step 6: Automatically adjust the operating frequency of the geothermal water pump according to the downhole measurement data and strategic suggestions, design a mobile operation and maintenance APP for real-time monitoring and remote control of the geothermal well, and then construct an equipment health status evaluation model based on equipment indicators to perform equipment health scoring and life prediction; specifically including:

[0115] 6.1 Synchronize the downhole measurement data and strategic suggestions to the pumping system control unit to interface with the geothermal water pump, preset an emergency control trigger mechanism and a dynamic adjustment control strategy, and use the PID algorithm for automatic control of the geothermal water pump; the emergency control trigger mechanism determines whether the current water level is lower than the lower water level limit. If so, automatically shut down the pump. The dynamic adjustment control strategy determines whether the water level change within the current 15 minutes exceeds the water level fluctuation threshold. If so, adjust the pump speed (for example, if the current water level drops by 20%, adjust the speed to 80% for operation).

[0116] 6.2 Design a mobile operation and maintenance APP for visual monitoring of the geothermal well (displaying the current water level, temperature, pressure distribution, and historical trend charts), abnormal push (pushing abnormal alarms and popping up reminders in emergency situations), remote adjustment of the dynamic control strategy, and generation of operation and maintenance logs; the operation and maintenance logs include abnormal records (alarm time and processing results), operation records (operation control instructions and operator information, including timestamps), and equipment status (pump operation duration, equipment wear assessment, etc.).

[0117] 6.3 Collect the working time, vibration value, and power consumption of the pump to obtain equipment indicators. According to the equipment indicators, construct an equipment health assessment model. Use the equipment health status evaluation model to regularly calculate the health score of the latest equipment, and then combine the health score to calculate the equipment life prediction curve in real time; where:

[0118] The expression for the health score is:

[0119]

[0120] where H(t) is the health score, ω i is the weighting coefficient of the i-th sensor, X i (t) is the specific sensor value at time t, X i,std is the standard state expected value of the specific sensor, σ i is the standard deviation of the specific sensor, and n is the type of sensor input;

[0121] The expression for the life prediction curve is:

[0122] R(t) = L 0 ·H(t) β

[0123] Among them, R(t) is the predicted remaining life at the calculation moment, L 0 is the theoretical maximum life of the device, and β is the attenuation rate parameter.

[0124] Therefore, by providing a method for accurate water level measurement and monitoring management based on a geothermal well water level gauge, the present invention can improve the accuracy and reliability of water level measurement, achieve timely and effective continuous monitoring and intelligent management, and at the same time, through the coordination of anomaly detection and automatic control, reduce faults and risks, protect geothermal resources, and save manpower and operating costs.

[0125] The various embodiments in this specification are described in a progressive manner. Each embodiment focuses on the differences from other embodiments. For the same and similar parts between the various embodiments, reference can be made to each other.

[0126] Specific examples are used in this article to elaborate on the principle and implementation manner of the present invention. The description of the above embodiments is only used to help understand the method and its core idea of the present invention; at the same time, for those of ordinary skill in the art, based on the idea of the present invention, there will be changes in the specific implementation manner and application scope. In summary, the content of this specification should not be construed as a limitation to the present invention.

Claims

1. A method for accurate water level measurement and monitoring management based on a geothermal well water level gauge, characterized in that: include: Select a variety of pressure sensors to combine and obtain the main monitoring component, select the optical fiber sensor as the auxiliary monitoring component, and design the probe fixing rail and dual probes in the geothermal well pipeline to complete the integrated monitoring deployment; An embedded processor and a high-precision clock are set up near the wellhead or well site to collect and pre-process sensor data locally, and then multi-source data fusion is performed on the pre-processed data to obtain downhole measurement data; Drift correction is performed on the pressure sensor, automatic anti-interference detection is performed on the pressure sensor at the same depth, and two sensor nodes and two independent power supplies are set at the same depth to perform continuous monitoring in case of failure; The downhole measurement data is transmitted to the cloud database for hierarchical storage and management, and a cloud API interface is designed to build a data visualization platform to display real-time water level curves, temperature distribution and alarm records; Preset water level safety threshold, conduct real-time monitoring and alarm on the underground measurement data, build a water level prediction model, perform water level anomaly detection, obtain abnormal events, and then trace the cause of the abnormal event and provide strategy recommendations; According to the downhole measurement data and strategy recommendations, the operating frequency of the geothermal water pump is automatically adjusted, and a mobile operation and maintenance APP is designed to conduct real-time monitoring and remote control of the geothermal well. Based on the equipment indicators, an equipment health status assessment model is constructed to perform equipment health scoring and life prediction.

2. The method for accurate water level measurement and monitoring management based on a geothermal well water level gauge according to claim 1 is characterized in that: Select a variety of pressure sensors to combine to obtain the main monitoring component, and select optical fiber sensors as auxiliary monitoring components, including: Based on the probe fixing rail, according to the depth gradient in the geothermal well pipeline, a group of pressure sensors are arranged every 50-100 meters to obtain a main monitoring component; wherein each group of pressure sensors includes a main pressure sensor and a secondary pressure sensor; The RS485 communication interface and the differential amplifier are used to collect and transmit the signal of the pressure sensor, and a temperature sensor is configured on the pressure sensor to dynamically correct the temperature error in the pressure reading; Deploy fiber Bragg grating nodes around the geothermal well pipeline at a spacing of 1 meter or less, and embed a metal waveguide groove for fixing the fiber Bragg grating inside the geothermal well pipeline to obtain an auxiliary monitoring component; The fiber Bragg grating is used to collect pressure and temperature curves, and the wavelength data is output in real time through an optical wave demodulator. Based on the output wavelength data, wavelength demodulation is used to monitor temperature and pressure fluctuations in real time, and phase demodulation is used to provide dynamic error compensation.

3. The method for accurate water level measurement and monitoring management based on a geothermal well water level gauge according to claim 2 is characterized in that: The surface of the dual probe is provided with a self-hydrophobic coating. The dual probe includes a main probe for deep-water data collection and a sub-probe for detecting humidity and water droplets. The main probe is provided with a titanium alloy shell. A temperature sensor is externally provided on the main probe to directly contact the water body and sense changes in the water temperature. The sub-probe is set as a short needle. A humidity sensor or a capacitive sensor is externally provided on the sub-probe to perform a water contact alarm.

4. The method for accurate water level measurement and monitoring management based on a geothermal well water level gauge according to claim 3 is characterized in that: An embedded processor and a high-precision clock are set up near the wellhead or well site to collect and pre-process sensor data locally. Multi-source data fusion is then performed on the pre-processed data to obtain downhole measurement data, including: Deploy embedded sensors and high-precision clocks near the wellhead or well site, and integrate communication interfaces to obtain edge computing modules, and then use the edge computing modules to collect data to obtain initial measurement data; The initial measurement data is subjected to noise filtering using a low-pass filter and a median filter, the pressure data is compensated in real time using a compensation algorithm, and the sensor continuous monitoring data is segmented and normalized and reversibly compressed to complete the localization preprocessing of the data; the expression of the compensation algorithm is: P corrected =P measured ×(1+α(T-T0)) Among them, P corrected is the corrected pressure value, P measured is the measured pressure value, α is the material expansion coefficient, T0 is the calibration temperature, and T is the actual temperature; The high-precision clock and PTP protocol are used to synchronize the clocks of multi-source sensor data, and to align the timestamps of the multi-source sensor data every second, so as to complete data fusion and obtain downhole measurement data.

5. The method for accurate water level measurement and monitoring management based on a geothermal well water level gauge according to claim 4 is characterized in that: Drift correction is performed on the pressure sensor, automatic anti-interference detection is performed on the pressure sensor at the same depth, and two sensor nodes and two independent power supplies are arranged at the same depth to perform continuous monitoring in case of failure, including: Collecting the reference water level and temperature values ​​when the geothermal well is stable, establishing a calibration baseline, and dynamically updating the calibration baseline using the real-time data of the pressure sensor; Based on the correction baseline, the pressure sensor is predicted, updated and calibrated using Kalman filtering, and the historical data of the pressure sensor is used to train a machine learning model to obtain a prediction calibration model, and the prediction calibration model is used to predict the drift trend of the pressure sensor to complete machine learning assisted calibration; Based on the pressure sensors at the same depth, differential verification is performed. When the difference exceeds the preset interference range, it is marked as interference or abnormal state, completing automatic anti-interference detection; Two sensor nodes and two independent power supplies are set at the same depth of the geothermal well to obtain a backup channel for each group of pressure sensors. When it is detected that the pressure sensor has no signal or the drift exceeds the standard, it automatically switches to the backup channel for continuous monitoring in case of failure.

6. The method for accurate water level measurement and monitoring management based on a geothermal well water level gauge according to claim 5, characterized in that: The detection expression of the pressure sensor is: Among them, P final is the final pressure value, n is the number of sensors involved in data fusion, ω i is the weighting coefficient of the i-th sensor, P sensor,i is the pressure value measured by the i-th sensor.

7. The method for accurate water level measurement and monitoring management based on a geothermal well water level gauge according to claim 6 is characterized in that: The downhole measurement data is transmitted to the cloud database for hierarchical storage and management, and a cloud API interface is designed to build a data visualization platform to display real-time water level curves, temperature distribution and alarm records, including: The downhole measurement data is transmitted to a time series database in the cloud by using a cellular network, wireless transmission technology or optical fiber communication; the time series database includes a real-time storage layer and a historical storage layer, and the data structure in the time series database consists of a timestamp, water level, temperature, pressure and sensor number; Develop standardized data interfaces, use OAuth protocol to authenticate user access, and set permission levels for different roles; Utilize the front-end framework to build a responsive web interface and integrate streaming chart components to dynamically render water level change trends and temperature field distribution to display real-time water level curves, temperature distribution, and alarm records.

8. The method for accurate water level measurement and monitoring management based on a geothermal well water level gauge according to claim 7 is characterized in that: Preset water level safety thresholds and water level fluctuation thresholds, conduct real-time monitoring and alarms on the downhole measurement data, build a water level prediction model, perform water level anomaly detection, and obtain abnormal events, including: Preset the upper and lower limits of the water level to obtain a water level safety threshold, and determine whether the downhole measurement data exceeds the water level safety threshold. If so, automatically trigger an alarm; Use the sliding window statistical method to calculate the mean, maximum, minimum and standard deviation of the water level changes in the past 15 minutes, obtain the water level fluctuation range, and set the water level fluctuation threshold. Then determine whether the water level change in the current 15 minutes exceeds the water level fluctuation threshold. If so, automatically trigger an alarm. Select and combine the LSTM model and the ARIMA model to obtain an initial model architecture, and based on historical downhole measurement data, extract the correlation characteristics between water level, temperature and pressure as features, train the initial model architecture, and obtain a water level prediction model; The water level prediction model is used to obtain the water level prediction value, and the water level error between the water level prediction value and the actual water level value is calculated. When the water level error exceeds the preset error threshold, it is marked as an abnormal water level, and the water level value, sensor ID, timestamp, temperature and pressure information of the abnormal water level are recorded to obtain an abnormal event.

9. The method for accurate water level measurement and monitoring management based on a geothermal well water level gauge according to claim 8, characterized in that: Trace the cause of the abnormal event and provide strategic recommendations, including: Combining the sensor drift status and historical fault information, pump power load, geothermal well equipment operation log and wellhead environment data, multi-modal correlation data is obtained; Analyzing the fault cause of the abnormal event by using a causal analysis tree algorithm and the multimodal association data, and classifying the abnormal event according to the fault cause; A maintenance suggestion library is preset to automatically respond to and generate different strategy suggestions according to different types of abnormal events.

10. The method for accurate water level measurement and monitoring management based on a geothermal well water level gauge according to claim 9, characterized in that: According to the downhole measurement data and strategy recommendations, the operating frequency of the geothermal water pump is automatically adjusted, and a mobile operation and maintenance APP is designed to conduct real-time monitoring and remote control of the geothermal well. Based on the equipment indicators, an equipment health status assessment model is constructed to perform equipment health scoring and life prediction, including: The downhole measurement data and strategy recommendations are synchronized to the pumping system control unit to control the geothermal water pump, and an emergency control trigger mechanism and a dynamic adjustment control strategy are preset, and the PID algorithm is used to automatically control the geothermal water pump; the emergency control trigger mechanism determines whether the current water level is less than the water level lower limit, and if so, the pump is automatically shut down; the dynamic adjustment control strategy determines whether the current water level change within 15 minutes exceeds the water level fluctuation threshold, and if so, the pump speed is adjusted; Design a mobile operation and maintenance APP to perform visual monitoring of geothermal wells and push notifications of abnormalities, remotely adjust the dynamic control strategy, and generate operation and maintenance logs; the operation and maintenance logs include abnormal records, operation records, and equipment status; The working time, vibration value and power consumption of the water pump are collected to obtain equipment indicators. According to the equipment indicators, an equipment health assessment model is constructed. The equipment health status assessment model is used to regularly calculate the health score of the latest equipment. Combined with the health score, the equipment life prediction curve is calculated in real time; wherein: The expression of the health score is: Among them, H(t) is the health score, ω i is the weighting coefficient of the i-th sensor, X i (t) is the specific sensor value at time t, X i,std is the expected value of the standard state of a specific sensor, σ i is the standard deviation of a specific sensor, and n is the type of sensor input; The expression of the life prediction curve is: R(t)=L0·H(t) β Among them, R(t) is the estimated remaining life at the calculation time, L0 is the theoretical maximum life of the equipment, and β is the decay rate parameter.

Images