A Data Acquisition Method for Subsea Tunnel Based on Internet of Things Technology

By deploying IoT sensors and logistic regression models in the subsea tunnel, multidimensional data is collected in real time, and the sampling frequency and pump speed are dynamically adjusted. This solves the problem of difficult monitoring of cavitation risk in subsea tunnel pump station units, and achieves efficient fault early warning and improved safety.

CN123110084APending Publication Date: 2026-11-17QINGDAO GUOXIN JIAOZHOU BAY SECOND SUBSEA TUNNEL CO LTD +2
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202611089066.1
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2026-07-22
Publication Date
2026-11-17

AI Technical Summary

Technical Problem

Under the influence of tidal changes and structural leakage, the cavitation risk of subsea tunnel pump station units is difficult to monitor in real time and accurately, which leads to difficulties in fault early warning and maintenance decision-making. Traditional monitoring methods are inadequate to cope with the complex seabed environment.

Method used

Based on IoT technology, by dividing the submarine tunnel into monitoring unit areas, deploying wireless pump group status monitoring nodes and environmental sensing nodes, and combining logistic regression models and edge devices, multidimensional data is collected in real time to construct pump group cavitation judgment factors and structure-pump group coupled leakage index, and dynamically adjusting sampling frequency and pump speed to identify anomalies.

Benefits of technology

It enables accurate identification and timely early warning of pump unit cavitation and structural leakage risks, reduces the risk of equipment damage, and improves the operational safety and maintenance efficiency of the submarine tunnel drainage system.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN123110084A_ABST
    Figure CN123110084A_ABST
Patent Text Reader

Abstract

This invention discloses a data acquisition method for large-scale equipment in subsea tunnels based on Internet of Things (IoT) technology, relating to the field of pump station equipment operation monitoring in subsea tunnel engineering. Addressing the risk of pump cavitation, this method constructs a pump cavitation judgment factor for the i-th monitoring unit by real-time acquisition of pump vibration spectrum and cavitation characteristic micro-acoustic signals, enabling timely identification of cavitation anomalies and improving the accuracy and response speed of fault warnings. Subsequently, the pump cavitation judgment factor for the i-th monitoring unit is corrected by combining the structure-pump coupling leakage index of the i-th monitoring unit area, effectively eliminating interference signals caused by structural leakage and improving the accuracy and robustness of cavitation identification. This joint correction method not only reduces the false alarm rate but also enhances the monitoring system's adaptability to the multi-factor coupling effects in complex seabed environments.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] This invention relates to the field of operation monitoring technology for pump station equipment in submarine tunnel engineering, specifically a data acquisition method for large-scale equipment in submarine tunnels based on Internet of Things (IoT) technology. Background Technology

[0002] In undersea tunnel engineering, because the tunnel is located below sea level, it is constantly affected by multiple hydraulic factors such as high water pressure, seepage, and tidal flow. Therefore, it is essential to install an efficient and reliable drainage system to ensure structural safety and traffic safety. Among them, the pumping station unit, as the core equipment of the entire drainage system, is mainly responsible for tasks such as draining accumulated water, dealing with water surges, and emergency flood drainage. It is one of the key electromechanical equipment to ensure the stable operation of the tunnel.

[0003] Pumping station units in undersea tunnels typically include: large submersible pumps, vertical or horizontal centrifugal pumps, pump drive motors, valve control devices, water level control modules, and corresponding electrical control cabinets, remote communication and protection units. Pumping stations are generally located at the lowest point of the tunnel or at drainage manholes, capable of automatically starting and stopping according to water level changes, and responding quickly to extreme weather conditions and sudden water inrushes, playing a role in flood prevention and hazard mitigation.

[0004] Cavitation is one of the most serious operational risks for subsea tunnel pump units. Tidal changes cause water pressure fluctuations, leading to sudden drops in localized pressure within the pump body. This creates air bubbles that burst in high-pressure areas, generating strong impacts that cause impeller wear and equipment damage. Traditional monitoring methods, due to the complex environment and mixed signals, struggle to accurately detect cavitation anomalies in real time, affecting fault warnings and maintenance decisions. Furthermore, humidity and water pressure changes caused by leakage in the subsea tunnel structure can also interfere with the pump station load through coupling effects, further exacerbating the cavitation risk of the pump units. Summary of the Invention

[0005] To address the shortcomings of existing technologies, this invention provides a method for large-scale data acquisition in submarine tunnels based on Internet of Things (IoT) technology, in order to solve the problems mentioned in the background section.

[0006] To achieve the above objectives, the present invention provides the following technical solution: a large-scale data acquisition method for submarine tunnels based on Internet of Things (IoT) technology, comprising the following steps: S1. Divide the pump station units distributed in the subsea tunnel into several monitoring unit areas, and deploy wireless pump unit status monitoring nodes and environmental sensing nodes in each monitoring unit area, using wireless communication modules to set the initial sampling frequency. Collect the tidal pressure sequence and pump start / stop status sequence of the i-th monitoring unit area; S2. Based on the tidal pressure sequence and the pump start-stop state sequence, a tidal pressure response model is constructed using the Logistic Regression machine learning model to predict the start-stop trend of the pump. When the risk of frequent start-stop induced by tidal pressure fluctuation is identified, the first regulation command is generated. S3. Set up edge devices in each monitoring unit area to collect the pump group vibration spectrum in real time. and cavitation characteristic microacoustic signal sequence And construct the pump group cavitation judgment factor for the i-th monitoring unit. After identifying abnormal cavitation signal events during pump unit operation, a second adjustment command is triggered; S4. Synchronously collect leakage parameters of structures adjacent to the pumping station, including the temperature change sequence of the tunnel wall surface. Water pressure change rate sequence The humidity change sequence was compared with the structure-pump coupling leakage index of the i-th monitoring unit area. Preset coupling penetration determination threshold ,like If the pump station load is affected by structural seepage, it is marked as a "high-sensitivity area for coupled leakage," triggering the structural collaborative analysis sub-strategy and feeding back to step S3 to correct the pump group cavitation judgment factor of the i-th monitoring unit. .

[0007] Preferably, the pump station units distributed in the submarine tunnel are divided into several monitoring unit areas. Wireless pump unit status monitoring nodes and environmental sensing nodes are deployed in each monitoring unit area. Operating status sensors and environmental parameter acquisition devices are set at key unit locations and structural interfaces in each monitoring unit area, including: pump unit current / frequency acquisition devices, air pressure sensors, water pressure sensors, high-frequency vibration and micro-sound monitoring devices, and seepage conductivity monitoring devices. Set a sampling time window containing N sampling points at different times to collect multi-source monitoring sequences, including: Tidal pressure sequence , ; This represents the tidal pressure value of the i-th monitoring unit area at the j-th sampling time, in kPa. Specifically, it is the tunnel confining pressure caused by seawater tidal changes, expressed as: ; in, The density of seawater, The acceleration due to gravity is set to 9.81 m / s². 2 , The tidal height of the corresponding pump group reference surface for the i-th monitoring unit area at the j-th sampling time; Pump start-stop state sequence , ; Where 1 indicates the pump is running, and 0 indicates the pump is stopped. This represents the start / stop binary status value of the pump group at the j-th sampling time for the i-th monitoring unit.

[0008] Preferably, based on the tidal pressure sequence and the pump start-up and shutdown state sequence, a logistic regression machine learning model is used to construct a tidal pressure response model to predict the pump start-up and shutdown trend. Specifically, this includes: Tidal pressure sequences and pump start-stop state sequences were collected to construct a training sample set. The tidal pressure sequences were then processed and standardized to the [0,1] interval. The pump start-stop state sequences were encoded using binary encoding, with running state recorded as 1 and stopped state recorded as 0. Extracting tidal pressure sequences Pump start-stop state sequence Construct the average pressure sequence of the i-th monitoring unit. and the average value of the start-stop state sequence : ; ; The above statistical characteristics are the average pressure sequence values ​​of the i-th monitoring unit. and the average value of the start-stop state sequence The tidal pressure sequence serves as the input feature vector for training a logistic regression model. Pump start-stop state sequence Calculate the Pearson correlation coefficient between the two variables. This is used to measure the driving force of tidal pressure on pump behavior: ; in, This represents the average value of the tidal pressure series. This represents the average value of the pump set start-up and shutdown sequence; and the Pearson correlation coefficient is used. The absolute value is used as the tidal pressure response index of the i-th monitoring unit area. , ; Preset moisture pressure sensitivity threshold ,when This indicates that there is a risk of frequent start-stop cycles induced by tidal pressure fluctuations in the i-th monitoring unit area, and generates the first adjustment command, including: at the initial sampling frequency Based on this, shorten the period by 10% to form the first sampling frequency. And execute; the adjustment command can improve the pump unit's operating status response sensitivity to tidal pressure disturbances, which helps to capture abnormal trends in a timely manner and optimize the control response; when This indicates that there is no risk of frequent start-stop cycles induced by tidal pressure fluctuations in the i-th monitoring unit area, and no adjustment is needed; the initial sampling frequency should be maintained. .

[0009] Preferably, edge devices are installed in each monitoring unit area to collect the pump unit vibration spectrum in real time through accelerometers and micro-acoustic sensors deployed at key parts of the pump body. and cavitation characteristic microacoustic signal sequence ; Vibration spectrum of pump unit Perform a Fast Fourier Transform (FFT) and select the energy peak value in the 2-15kHz frequency band as the feature input value to analyze the vibration spectrum of the pump unit. The process is updated to obtain the processed pump unit vibration spectrum. : ; This represents the peak vibration value of the pump housing after FFT processing for the i-th monitoring unit at the j-th sampling time; Microacoustic signal sequence of cavitation characteristics collected by microacoustic sensor Bandpass filtering and time-frequency joint analysis were performed to extract the power spectral density in the cavitation frequency band of 20-50kHz, forming a processed microacoustic signal sequence characteristic of cavitation. ; ; This represents the micro-acoustic vector of the pump group cavitation characteristics at the j-th sampling time for the i-th monitoring unit.

[0010] Preferably, the specific steps for identifying abnormal cavitation signal events during pump unit operation are as follows: identifying abnormal cavitation signal events and generating a second adjustment command, including: based on the processed pump unit vibration spectrum... and cavitation characteristic microacoustic signal sequence Set the sliding window length R, where n is the total number of sampling points within the nearest window length R. Start abrupt change identification from the k-th time step: for the k-th sampling time step, calculate the average of the total number of the previous R nearest sampling points, and construct the average vibration peak value of the i-th monitoring unit. and the average value of micro-sound characteristics : ; ; Taking the k-th sampling point as the endpoint, the mean of R points counting forward, the interval is from Starting from the kth sampling point and ending at the kth sampling point; The average value of the vibration peak value for the i-th monitoring unit and the average value of micro-sound characteristics After Min-Max normalization, the cavitation judgment factor of the pump group for the i-th monitoring unit is obtained using a weighted formula. : ; In the formula, and The average value of the vibration peak value of the i-th monitoring unit after normalization and the average value of micro-sound characteristics The weight values, and the sum of the weights equals 1; Preset perturbation threshold If within the sliding window length R, If so, it is determined that there is an abnormal cavitation disturbance event in the pump unit at the current moment, and a first mark is made, and it is identified whether a second adjustment command is triggered. This indicates that there is no risk of cavitation disturbance at the current moment, and monitoring will continue.

[0011] Preferably, the triggering condition steps for the second adjustment command are as follows: When the cavitation disturbance anomaly event is identified as 1 and the system returns to normal, then at the first sampling frequency... Based on this, the period is shortened by another 10% to form a second sampling frequency. ; If the cavitation disturbance anomaly is detected and continues for 3 cycles, a second adjustment command is triggered and executed, including: reducing the current pump speed of the monitoring unit by 5%-10%; If, after the second adjustment command is executed, if it is recognized After three consecutive cycles, starting from the second sampling frequency Restore to the first sampling frequency If, after the second adjustment command is executed, identification is performed... Still more If this occurs, a manual inspection command is triggered, uploaded to the remote operation and maintenance platform, and maintenance personnel are prompted to intervene.

[0012] Preferably, a structural environmental monitoring module is installed in each monitoring unit area. This module uses tunnel wall temperature sensors, structural gap water pressure monitors, and wall humidity change sensing devices deployed in the area adjacent to the pumping station to collect real-time leakage response characteristic parameters of the structures near the pumping station, including: Tunnel wall surface temperature change sequence , ; to This represents the change in tunnel wall surface temperature from the first adjacent time sampling point to the Mth adjacent time sampling point; Tunnel wall fissure water pressure change rate sequence , ; to This represents the rate of change of water pressure in the tunnel wall gaps from the first adjacent time sampling point to the Mth adjacent time sampling point; Tunnel wall humidity change sequence , ; to This represents the change in tunnel wall humidity from the first adjacent time sampling point to the Mth adjacent time sampling point; Temperature variation sequence of tunnel wall surface Tunnel wall fissure water pressure change rate sequence and the sequence of humidity changes in the tunnel walls Normalization is performed.

[0013] Preferably, a leakage weighting coefficient is set. And W=1; where, The weight value represents the temperature change of the tunnel wall surface, specifically representing the fluctuation of heat exchange in the wall caused by structural leakage. The temperature difference reflects the impact of water vapor infiltration or leakage. Based on the temperature difference sensitivity of the pump group pipeline, the initial weight is set to 0.2-0.3. The weight value of the rate of change of water pressure in the tunnel wall gaps represents the risk of structural stress and hydraulic coupling caused by the pressure change at the leakage point. It is given priority as the dominant term, with an initial weight of 0.5-0.6. The weight value representing the change in humidity of the tunnel wall indicates the medium-term reaction of the structure after it becomes damp and water accumulates, which has a long-term erosive effect on the pump unit. It is set as the secondary dominant factor, and the initial weight is set to 0.2-0.3. Calculate the structure-pump coupling leakage index for the i-th monitoring unit area. : ; in, This represents the normalized change in tunnel wall humidity at the m-th normalized sampling point adjacent to the previous time. This represents the rate of change of water pressure in the tunnel wall gaps at the m-th adjacent time sampling point after normalization. The normalized value of the change in humidity of the tunnel wall at the m-th adjacent time sampling point.

[0014] Preferably, a preset coupling penetration determination threshold is used. ,like If the i-th monitoring unit area is determined to have a structural leakage that could interfere with the operation of the pump unit, then the monitoring unit area is marked as a "highly sensitive area for coupled leakage". like If the i-th monitoring unit area does not pose a risk of structural leakage interfering with the operation of the pump unit, monitoring will continue. After being marked as a "highly sensitive area for coupled leakage", the structural collaborative analysis sub-strategy is triggered, which feeds back to step S3 to correct the pump group cavitation judgment factor of the i-th monitoring unit. Obtain the corrected cavitation judgment factor for the second pump group of the i-th monitoring unit. The corrected expression is: .

[0015] Preferably, the corrected cavitation judgment factor of the second pump group of the i-th monitoring unit is simultaneously applied. With disturbance threshold For comparison, if it is within the sliding window length R, If a cavitation disturbance anomaly is detected, it is determined that the pump unit currently experiences a cavitation disturbance anomaly, and a second flag is set. If the cavitation disturbance anomaly persists for three consecutive cycles, a second adjustment command is triggered and executed. If, within any sliding window period R after the execution of the second adjustment command, the cavitation disturbance anomaly detection rate continuously exceeds 60%, i.e., at least two anomalies occur in more than three detection cycles, a third adjustment command is triggered, including: further reducing the pump speed in the current monitoring unit area by 15% based on the second adjustment command; if the current pump speed is... Then adjust to: , This indicates the adjusted pump speed; And at the second sampling frequency Based on this, the period is shortened by another 25% to form a third sampling frequency. ; If the situation fails to recover after two consecutive cycles following the execution of the third adjustment command, a manual inspection command will be triggered and uploaded to the remote operation and maintenance platform to prompt maintenance personnel to intervene.

[0016] This invention provides a method for large-scale data acquisition in submarine tunnels based on Internet of Things (IoT) technology. It offers the following advantages: (1) This data acquisition method for large subsea tunnel pumps based on Internet of Things (IoT) technology constructs a multi-dimensional, multi-parameter real-time data acquisition network by deploying multiple types of sensor nodes within the tunnel, including pump current and start / stop status sensors, environmental pressure sensors (capturing confining pressure changes caused by tides), water pressure, humidity, and wall temperature sensors related to structural leakage, as well as pump vibration and micro-sound sensors. Through a wireless communication module, continuous data transmission of the pump unit and its surrounding environment is achieved, and the sampling frequency can be dynamically adjusted to meet data demands under different operating conditions, improving the timeliness and accuracy of data acquisition. Simultaneously, edge computing nodes preprocess and detect anomalies in the acquired raw data, reducing network bandwidth pressure and improving the speed and accuracy of anomaly signal identification. This multi-sensor, multi-time-series synchronous acquisition technology comprehensively captures abnormal changes caused by the interaction of multiple factors such as tidal fluctuations, structural leakage, and pump cavitation, enabling early warning of pump cavitation and structural safety monitoring, significantly improving the operational safety and maintenance efficiency of subsea tunnel pump station units.

[0017] (2) The data acquisition method for large subsea tunnel machinery based on Internet of Things technology addresses the cavitation risk of pump units. This technology constructs the pump unit cavitation judgment factor for the i-th monitoring unit by real-time acquisition of pump unit vibration spectrum and cavitation characteristic micro-acoustic signals. This enables timely identification of cavitation anomalies, improving the accuracy and response speed of fault warnings. Subsequently, the structure-pump coupling leakage index of the i-th monitoring unit area is combined with... Cavitation judgment factor for the pump group of the i-th monitoring unit The correction effectively eliminates interference signals caused by structural leakage, improving the accuracy and robustness of cavitation identification. This joint correction method not only reduces the false alarm rate but also enhances the monitoring system's adaptability to the multi-factor coupling effects in complex seabed environments, thereby ensuring the stability of pump unit operation and extending equipment lifespan.

[0018] (3) This data acquisition method for large subsea tunnel equipment based on Internet of Things (IoT) technology enables dynamic response and effective control of risks induced by pump cavitation and tidal fluctuations. Specifically, timely shortening of the data acquisition cycle and improving monitoring sensitivity helps to capture abnormal signals and sudden trends more quickly and accurately, avoiding the omission of potential risks. At the same time, reducing the pump speed can alleviate internal pressure fluctuations in the pump body, reduce the impact of bubble rupture on the impeller and pump body, and reduce equipment wear and the probability of failure. The synergistic effect of both promotes the improvement of the accuracy and timeliness of fault early warning. Attached Figure Description

[0019] Figure 1 This is a schematic diagram illustrating the steps of the large-scale data acquisition method for submarine tunnels based on Internet of Things (IoT) technology according to the present invention. Detailed Implementation

[0020] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0021] Example 1 Subsea tunnels are located underwater, and the external pressure they experience mainly comes from seawater. During the ebb and flow of the tides, the fluctuations in hydrostatic pressure caused by changes in seawater level will generate periodically changing pressure loads on the tunnel structure. When the tide rises, the external water pressure increases, making the tunnel structure more prone to micro-cracks and water seepage, or exacerbating existing seepage. When the tide falls, the external water pressure decreases, which may reduce the seepage rate. Rising tides → increased seepage load → rising water level → triggering pump unit startup; Tidal drop → reduced seepage load → water level drop → pump unit stops after reaching threshold.

[0022] As the tide rises, the sea level gradually increases, leading to an increase in seepage load within the tunnel, causing the water level to rise accordingly. When the water level reaches the control condition for starting the pump unit, the pump unit automatically starts to drain water; as the tide falls, the water level decreases, the seepage load decreases, and the water level drops back to the control condition for stopping the pump unit, at which point the pump unit automatically stops, thus achieving dynamic adjustment and protection of the tunnel drainage system.

[0023] This invention provides a method for large-scale data acquisition in submarine tunnels based on Internet of Things (IoT) technology. Please refer to [link / reference]. Figure 1 This includes the following steps: S1. Divide the pump station units distributed in the subsea tunnel into several monitoring unit areas, and deploy wireless pump unit status monitoring nodes and environmental sensing nodes in each monitoring unit area, using wireless communication modules to set the initial sampling frequency. Collect the tidal pressure sequence and pump start / stop status sequence of the i-th monitoring unit area; S2. Based on the tidal pressure sequence and the pump start-stop state sequence, a tidal pressure response model is constructed using the Logistic Regression machine learning model to predict the start-stop trend of the pump. When the risk of frequent start-stop induced by tidal pressure fluctuation is identified, the first regulation command is generated. S3. Set up edge devices in each monitoring unit area to collect the pump group vibration spectrum in real time. and cavitation characteristic microacoustic signal sequence And construct the pump group cavitation judgment factor for the i-th monitoring unit. After identifying abnormal cavitation signal events during pump unit operation, a second adjustment command is triggered; S4. Synchronously collect leakage parameters of structures adjacent to the pumping station, including the temperature change sequence of the tunnel wall surface. Water pressure change rate sequence The humidity change sequence was compared with the structure-pump coupling leakage index of the i-th monitoring unit area. Preset coupling penetration determination threshold ,like If the pump station load is affected by structural seepage, it is marked as a "high-sensitivity area for coupled leakage," triggering the structural collaborative analysis sub-strategy and feeding back to step S3 to correct the pump group cavitation judgment factor of the i-th monitoring unit. .

[0024] In this embodiment, the present invention, based on Internet of Things (IoT) technology, achieves synchronous real-time acquisition and intelligent analysis of multi-dimensional parameters of the subsea tunnel pumping station units and the environment. It can accurately capture key influencing factors such as tidal pressure fluctuations, pump start-up and shutdown status, and structural leakage, improving the accuracy and response speed in identifying pump cavitation and structural coupling leakage risks. By predicting pump start-up and shutdown trends through a logistic regression model and combining edge computing for in-depth analysis of vibration spectra and microacoustic signals, it effectively provides early warning of cavitation anomalies and promptly adjusts sampling frequency and pump speed parameters, reducing the risk of equipment damage. Simultaneously, the introduction of a structure-pump coupling leakage index enables dynamic feedback and correction of the impact of structural leakage on pump operation, enhancing the monitoring system's ability to perceive pump status under complex hydraulic environments. The overall solution significantly improves the operational safety and maintenance efficiency of the subsea tunnel drainage system, reduces the risk of sudden failures, and promotes the development of intelligent subsea tunnel operation and maintenance management.

[0025] Example 2 This embodiment is an explanation of Embodiment 1. Specifically, the pump station units distributed in the undersea tunnel are divided into several monitoring unit areas. Wireless pump unit status monitoring nodes and environmental sensing nodes are deployed in each monitoring unit area. Operating status sensors and environmental parameter acquisition devices are set at key unit locations and structural interfaces in each monitoring unit area, including: pump unit current / frequency acquisition devices, air pressure sensors, water pressure sensors, high-frequency vibration and micro-sound monitoring devices, and seepage conductivity monitoring devices. Set a sampling time window containing N sampling points at different times to collect multi-source monitoring sequences, including: Tidal pressure sequence , ; This represents the tidal pressure value of the i-th monitoring unit area at the j-th sampling time, in kPa. Specifically, it is the tunnel confining pressure caused by seawater tidal changes, expressed as: ; in, The density of seawater, The acceleration due to gravity is set to 9.81 m / s². 2 , The tidal height of the corresponding pump group reference surface for the i-th monitoring unit area at the j-th sampling time; Pump start-stop state sequence , ; Where 1 indicates the pump is running, and 0 indicates the pump is stopped. This represents the start / stop binary status value of the pump group at the j-th sampling time for the i-th monitoring unit.

[0026] Based on tidal pressure sequences and pump start-up / shutdown state sequences, a tidal pressure response model is constructed using a logistic regression machine learning model to predict pump start-up / shutdown trends, specifically including: Tidal pressure sequences and pump start-stop state sequences were collected to construct a training sample set. The tidal pressure sequences were then processed and standardized to the [0,1] interval. The pump start-stop state sequences were encoded using binary encoding, with running state recorded as 1 and stopped state recorded as 0. Extracting tidal pressure sequences Pump start-stop state sequence Construct the average pressure sequence of the i-th monitoring unit. and the average value of the start-stop state sequence : ; ; The above statistical characteristics are the average pressure sequence values ​​of the i-th monitoring unit. and the average value of the start-stop state sequence The tidal pressure sequence serves as the input feature vector for training a logistic regression model. Pump start-stop state sequence Calculate the Pearson correlation coefficient between the two variables. This is used to measure the driving force of tidal pressure on pump behavior: ; in, This represents the average value of the tidal pressure series. This represents the average value of the pump set start-up and shutdown sequence; and the Pearson correlation coefficient is used. The absolute value is used as the tidal pressure response index of the i-th monitoring unit area. , ; Preset moisture pressure sensitivity threshold ,when This indicates that there is a risk of frequent start-stop cycles induced by tidal pressure fluctuations in the i-th monitoring unit area, and generates the first adjustment command, including: at the initial sampling frequency Based on this, shorten the period by 10% to form the first sampling frequency. And execute; the adjustment command can improve the pump unit's operating status response sensitivity to tidal pressure disturbances, which helps to capture abnormal trends in a timely manner and optimize the control response; when This indicates that there is no risk of frequent start-stop cycles induced by tidal pressure fluctuations in the i-th monitoring unit area, and no adjustment is needed; the initial sampling frequency should be maintained. .

[0027] Moisture pressure sensitivity threshold The tidal pressure sensitivity threshold is determined through statistical analysis of historical operational and environmental monitoring data, combined with expert experience and engineering specifications. The threshold is set at a tidal pressure response index (the absolute value of the Pearson correlation coefficient) greater than 0.6 to 0.8. For example, when the tidal pressure response index > 0.7, it is considered a risk of frequent start-ups and shutdowns induced by tidal pressure fluctuations.

[0028] In this embodiment, this step involves deploying multiple sensors at key locations of the subsea tunnel pumping station unit to collect multi-source data in real time, including tidal pressure sequences and pump start-up / shutdown status sequences. A logistic regression model is then used to establish a response model of tidal pressure to pump start-up / shutdown behavior. This model effectively quantifies the driving force of tidal pressure changes on pump start-up / shutdown frequency, enabling accurate prediction of pump start-up / shutdown trends. Based on the model output and a preset tidal pressure sensitivity threshold, the system can intelligently determine whether the pump unit is at risk of frequent start-up / shutdown due to tidal fluctuations, thereby automatically adjusting the data sampling frequency and shortening the initial sampling period by 10%. This significantly improves the temporal resolution of data acquisition and the system's ability to perceive rapid pressure changes, ensuring more accurate and timely monitoring data and avoiding the shortcomings of traditional monitoring systems that miss key abnormal signals due to excessively long sampling periods. Shortening the sampling period also improves the real-time processing efficiency of edge devices, enabling pump control commands to respond more quickly to tidal disturbances, reducing pump mechanical wear and energy consumption, and extending equipment lifespan. This strategy effectively improves the safety and stability of pump station operation, provides a reliable data foundation for subsequent intelligent operation and maintenance and fault early warning, and promotes efficient management and risk control of the submarine tunnel pump station system.

[0029] Example 3 This embodiment is an explanation of Embodiment 1. Specifically, edge devices are set up in each monitoring unit area, and the vibration spectrum of the pump group is collected in real time through accelerometers and micro-acoustic sensors deployed at key parts of the pump body. and cavitation characteristic microacoustic signal sequence ; Vibration spectrum of pump unit Perform a Fast Fourier Transform (FFT) and select the energy peak value in the 2-15kHz frequency band as the feature input value to analyze the vibration spectrum of the pump unit. The process is updated to obtain the processed pump unit vibration spectrum. : ; This represents the peak vibration value of the pump housing after FFT processing for the i-th monitoring unit at the j-th sampling time; Microacoustic signal sequence of cavitation characteristics collected by microacoustic sensor Bandpass filtering and time-frequency joint analysis were performed to extract the power spectral density in the cavitation frequency band of 20-50kHz, forming a processed microacoustic signal sequence characteristic of cavitation. ; ; This represents the micro-acoustic vector of the pump group cavitation characteristics at the j-th sampling time for the i-th monitoring unit.

[0030] The specific steps for identifying abnormal cavitation signal events during pump unit operation are as follows: Identifying abnormal cavitation signal events and generating a second adjustment command, including: based on the processed pump unit vibration spectrum... and cavitation characteristic microacoustic signal sequence Set the sliding window length R, where n is the total number of sampling points within the nearest window length R. Start abrupt change identification from the k-th time step: for the k-th sampling time step, calculate the average of the total number of the previous R nearest sampling points, and construct the average vibration peak value of the i-th monitoring unit. and the average value of micro-sound characteristics : ; ; Taking the k-th sampling point as the endpoint, the mean of R points counting forward, the interval is from Starting from the kth sampling point and ending at the kth sampling point; The average value of the vibration peak value for the i-th monitoring unit and the average value of micro-sound characteristics After Min-Max normalization, the cavitation judgment factor of the pump group for the i-th monitoring unit is obtained using a weighted formula. : ; In the formula, and The average value of the vibration peak value of the i-th monitoring unit after normalization and the average value of micro-sound characteristics The weights are assigned to values ​​such that the sum of the weights equals 1. Based on the cavitation failure mechanism of the pump unit and field engineering experience, it is determined which signal is more sensitive to cavitation identification and assigned a larger weight. Through data analysis and expert experience, it was found that the micro-acoustic signal is more sensitive to cavitation anomaly identification, followed by the vibration signal; therefore, the weight is set to... , .

[0031] Preset perturbation threshold If within the sliding window length R, If so, it is determined that there is an abnormal cavitation disturbance event in the pump unit at the current moment, and a first mark is made, and it is identified whether a second adjustment command is triggered. This indicates that there is currently no risk of cavitation disturbance events affecting the pump unit, and monitoring will continue. Disturbance threshold. Referring to industry standards (such as API and ISO vibration limits for pump equipment) and long-term on-site monitoring data, the critical values ​​of vibration peak value and microsound power spectrum under normal operation and cavitation conditions are extracted. When the vibration peak normalization index exceeds 0.5 to 0.7, it is considered an abnormal disturbance. The comprehensive cavitation disturbance index threshold is initially set at around 0.6. If it exceeds, an early warning will be triggered.

[0032] In this embodiment, edge devices are deployed in each monitoring unit area, combined with accelerometers and micro-acoustic sensors at key parts of the pump body, to achieve real-time and accurate acquisition of the pump group's vibration spectrum and cavitation characteristic micro-acoustic signals. The vibration energy peak value in the 2-15kHz frequency band is extracted using Fast Fourier Transform (FFT), and combined with the power spectral density characteristics of the 20-50kHz cavitation frequency band, a comprehensive pump group cavitation judgment factor is formed. Based on a set sliding window, the mean values ​​of vibration peak values ​​and micro-acoustic characteristics are calculated in real time and normalized. Combined with a weighted algorithm, abnormal cavitation signals are efficiently identified, enabling timely early warning and marking of cavitation disturbance events. This method significantly improves the sensitivity and accuracy of pump group cavitation status, reduces the risk of misjudgment from traditional single-indicator monitoring, and effectively avoids equipment damage and performance degradation caused by cavitation. Simultaneously, a second adjustment command is triggered in real time to dynamically adjust the pump group's operating parameters, achieving intelligent operation and maintenance and fault prevention. This technology not only ensures the stability and safety of pump group operation but also extends equipment lifespan and improves the overall reliability and maintenance efficiency of the subsea tunnel drainage system.

[0033] Example 4 This embodiment is an explanation based on embodiment 3. Specifically, the triggering condition steps for the second adjustment command are as follows: When the cavitation disturbance anomaly event is identified as 1 and the system returns to normal, then at the first sampling frequency... Based on this, the period is shortened by another 10% to form a second sampling frequency. ; If the cavitation disturbance anomaly event is identified and continues for 3 cycles, the second adjustment command is triggered and executed, including: reducing the current monitoring unit pump speed by 5%-10% to mitigate the impact of bubble rupture caused by pressure fluctuations; If, after the second adjustment command is executed, if it is recognized After three consecutive cycles, starting from the second sampling frequency Restore to the first sampling frequency If, after the second adjustment command is executed, identification is performed... Still more If this occurs, a manual inspection command is triggered, uploaded to the remote operation and maintenance platform, and maintenance personnel are prompted to intervene.

[0034] In this embodiment, by gradually adjusting the sampling frequency and pump speed, the system can more flexibly adapt to the pump set's operating status, avoiding misjudgments and equipment damage caused by insufficient monitoring or excessive adjustment. Shortening the sampling cycle allows the system to capture abnormal changes more promptly, reducing the possibility of missed reports; moderately reducing the pump speed can effectively alleviate the impact of cavitation and reduce mechanical wear. Simultaneously, by restoring the sampling frequency and pump speed in stages, the system ensures that the equipment gradually returns to normal, avoiding secondary problems caused by drastic fluctuations. If the abnormality persists, manual inspection is promptly notified to ensure that the problem is addressed in a timely manner and to prevent the fault from escalating. Overall, this hierarchical, multi-step adjustment strategy improves the accuracy of monitoring and the scientific nature of the response, enhancing the stability of pump set operation and maintenance efficiency.

[0035] Example 5 This embodiment is an explanation of Embodiment 1. Specifically, a structural environmental monitoring module is set up in each monitoring unit area. Through tunnel wall temperature sensors, structural gap water pressure monitors, and wall humidity change sensing devices deployed in the area adjacent to the pumping station, leakage response characteristic parameters of the structures adjacent to the pumping station are collected in real time, including: Tunnel wall surface temperature change sequence , ; to This represents the change in tunnel wall surface temperature from the first adjacent time sampling point to the Mth adjacent time sampling point; Tunnel wall fissure water pressure change rate sequence , ; to This represents the rate of change of water pressure in the tunnel wall gaps from the first adjacent time sampling point to the Mth adjacent time sampling point; Tunnel wall humidity change sequence , ; to This represents the change in tunnel wall humidity from the first adjacent time sampling point to the Mth adjacent time sampling point; Temperature variation sequence of tunnel wall surface Tunnel wall fissure water pressure change rate sequence and the sequence of humidity changes in the tunnel walls Normalization is performed.

[0036] Set the leakage weight coefficient. And W=1; where, The weight value represents the temperature change of the tunnel wall surface, specifically representing the fluctuation of heat exchange in the wall caused by structural leakage. The temperature difference reflects the impact of water vapor infiltration or leakage. Based on the temperature difference sensitivity of the pump group pipeline, the initial weight is set to 0.2-0.3. The weight value of the rate of change of water pressure in the tunnel wall gaps represents the risk of structural stress and hydraulic coupling caused by the pressure change at the leakage point. It is given priority as the dominant term, with an initial weight of 0.5-0.6. The weight value representing the change in humidity of the tunnel wall indicates the medium-term reaction of the structure after it becomes damp and water accumulates, which has a long-term erosive effect on the pump unit. It is set as the secondary dominant factor, and the initial weight is set to 0.2-0.3. Calculate the structure-pump coupling leakage index for the i-th monitoring unit area. : ; in, This represents the normalized change in tunnel wall humidity at the m-th normalized sampling point adjacent to the previous time. This represents the rate of change of water pressure in the tunnel wall gaps at the m-th adjacent time sampling point after normalization. The normalized value of the change in humidity of the tunnel wall at the m-th adjacent time sampling point.

[0037] Preset coupling penetration determination threshold ,like If the i-th monitoring unit area is determined to have a structural leakage risk interfering with pump unit operation, this monitoring unit area is marked as a "high-sensitivity area for coupled leakage"; the coupled leakage determination threshold. This involves simulating leakage conditions in a laboratory, measuring changes in relevant environmental parameters, determining the typical leakage response range, and then basing this on practical engineering experience and data analysis. ; like If the i-th monitoring unit area does not pose a risk of structural leakage interfering with the operation of the pump unit, monitoring will continue. After being marked as a "highly sensitive area for coupled leakage", the structural collaborative analysis sub-strategy is triggered, which feeds back to step S3 to correct the pump group cavitation judgment factor of the i-th monitoring unit. Obtain the corrected cavitation judgment factor for the second pump group of the i-th monitoring unit. The corrected expression is: .

[0038] In this embodiment, a multi-dimensional environmental monitoring module is deployed near the pump station structure to collect key leakage response parameters such as temperature, gap water pressure, and humidity in real time. Combined with a rationally designed weighting coefficient, a structure-pump coupling leakage index is established to achieve a quantitative assessment of structural leakage behavior. This index reflects the comprehensive impact of leakage on the pump station load, particularly capturing the localized heat exchange anomalies, hydraulic pressure disturbances, and long-term wetting corrosion effects caused by leakage.

[0039] Based on the structure-pump coupling leakage index of the i-th monitoring unit area Cavitation judgment factor for the pump group of the i-th monitoring unit Dynamic correction enhances the targeting and accuracy of cavitation anomaly detection, effectively filters environmental interference factors, and improves the sensitivity of anomaly signal identification. Specifically, changes in wall water pressure caused by structural leakage will affect the pump inlet pressure fluctuation through hydraulic coupling, while temperature and humidity changes reflect the thermal and corrosion states during the leakage process. The combined effect of these factors directly influences the mechanism of cavitation.

[0040] This method achieves closed-loop coupled monitoring of structural status and pump unit operating status, enabling early warning of cavitation risks induced by leakage, guiding targeted adjustment of pump speed and maintenance strategies, significantly reducing equipment damage rate and probability of sudden failure, extending pump unit service life, and improving the safety and operational stability of the submarine tunnel drainage system.

[0041] Example 6 This embodiment is an explanation of Embodiment 5. Specifically, the cavitation judgment factor of the second pump group of the i-th monitoring unit is simultaneously modified. With disturbance threshold For comparison, if it is within the sliding window length R, If a cavitation disturbance anomaly is detected, it is determined that the pump unit currently experiences a cavitation disturbance anomaly, and a second flag is set. If the cavitation disturbance anomaly persists for three consecutive cycles, a second adjustment command is triggered and executed. If, within any sliding window period R after the execution of the second adjustment command, the cavitation disturbance anomaly detection rate continuously exceeds 60%, i.e., at least two anomalies occur in more than three detection cycles, a third adjustment command is triggered, including: further reducing the pump speed in the current monitoring unit area by 15% based on the second adjustment command; if the current pump speed is... Then adjust to: This indicates the adjusted pump speed; And at the second sampling frequency Based on this, the period is shortened by another 25% to form a third sampling frequency. ; If the situation fails to recover after two consecutive cycles following the execution of the third adjustment command, a manual inspection command will be triggered and uploaded to the remote operation and maintenance platform to prompt maintenance personnel to intervene.

[0042] In this embodiment, the modified pump group cavitation judgment factor is based on the structure-pump group coupled leakage index, dynamically adjusting the original cavitation index to eliminate the impact of environmental interference caused by structural leakage on cavitation signal identification. By introducing the actual operating state of structural leakage, the modified judgment factor more accurately reflects the true cavitation risk of the pump group, avoiding misjudgments and omissions, and improving the reliability of abnormal event identification. The modified judgment factor can effectively distinguish between environmental fluctuations caused by structural leakage and actual pump group cavitation disturbances, making abnormal event detection more accurate and stable, thereby reducing false alarms caused by environmental noise and ensuring the scientific and necessary nature of adjustment command triggering. Continuous multi-cycle abnormal identification based on the modified judgment factor can better reflect the persistence and severity of cavitation disturbances, providing a more solid basis for the implementation of the second and third adjustment commands. The modified judgment not only improves the accuracy of pump group operating status monitoring but also optimizes the response effect of adjustment strategies, helping to achieve more detailed risk classification and more accurate operation and maintenance decisions, minimizing equipment damage risks, extending equipment life, and ensuring the safe and stable operation of the subsea tunnel pump station. The third adjustment command, based on continuous and high-frequency cavitation anomaly identification, specifically increases the pump speed adjustment range (e.g., a 15% reduction), effectively mitigating internal pressure fluctuations and bubble bursting impacts within the pump unit. This significantly reduces the risk of mechanical damage to the impeller and other critical components from cavitation, fundamentally improving the operational safety of the pump unit. Secondly, by further shortening the sampling cycle (e.g., by 25% compared to the second sampling frequency), the timeliness and resolution of the monitoring data are improved, enabling the system to more quickly and accurately capture the dynamic changes of cavitation disturbances, respond promptly to abnormal pump unit conditions, and prevent the rapid deterioration of faults.

[0043] The threshold is set to facilitate comparison. The size of the threshold depends on the amount of sample data and the number of bases set by those skilled in the art for each set of sample data; as long as it does not affect the ratio between the parameter and the quantized value, it is acceptable.

[0044] The above formulas are all derived from software simulation using a large amount of data and are selected to be close to the actual values. The coefficients in the formulas are set by those skilled in the art according to the actual situation. The above description is only a preferred embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Any equivalent substitutions or changes made by those skilled in the art within the technical scope disclosed in the present invention, based on the technical solution and inventive concept of the present invention, should be covered within the protection scope of the present invention.

Claims

1. A method for large-scale data acquisition in submarine tunnels based on Internet of Things (IoT) technology, characterized in that: Includes the following steps: S1. Divide the pump station units distributed in the subsea tunnel into several monitoring unit areas, and deploy wireless pump unit status monitoring nodes and environmental sensing nodes in each monitoring unit area, using wireless communication modules to set the initial sampling frequency. Collect the tidal pressure sequence and pump start / stop status sequence of the i-th monitoring unit area; S2. Based on the tidal pressure sequence and the pump start-stop state sequence, a tidal pressure response model is constructed using the Logistic Regression machine learning model to predict the start-stop trend of the pump. When the risk of frequent start-stop induced by tidal pressure fluctuation is identified, the first regulation command is generated. S3. Set up edge devices in each monitoring unit area to collect the pump group vibration spectrum in real time. and cavitation characteristic microacoustic signal sequence And construct the pump group cavitation judgment factor for the i-th monitoring unit. After identifying abnormal cavitation signal events during pump unit operation, a second adjustment command is triggered; S4. Synchronously collect leakage parameters of structures adjacent to the pumping station, including the temperature change sequence of the tunnel wall surface. Water pressure change rate sequence The humidity change sequence was compared with the structure-pump coupling leakage index of the i-th monitoring unit area. Preset coupling penetration determination threshold ,like If the pump station load is affected by structural seepage, it is marked as a "high-sensitivity area for coupled leakage," triggering the structural collaborative analysis sub-strategy and feeding back to step S3 to correct the pump group cavitation judgment factor of the i-th monitoring unit. .

2. The method for large-scale data acquisition in submarine tunnels based on Internet of Things technology according to claim 1, characterized in that, The pump station units distributed in the undersea tunnel are divided into several monitoring unit areas. Wireless pump unit status monitoring nodes and environmental sensing nodes are deployed in each monitoring unit area. Operating status sensors and environmental parameter acquisition devices are set at key unit locations and structural interfaces in each monitoring unit area, including: pump unit current / frequency acquisition devices, air pressure sensors, water pressure sensors, high-frequency vibration and micro-sound monitoring devices, and seepage conductivity monitoring devices. Set a sampling time window containing N sampling points at different times to collect multi-source monitoring sequences, including: Tidal pressure sequence , ; This represents the tidal pressure value of the i-th monitoring unit area at the j-th sampling time, in kPa. Specifically, it is the tunnel confining pressure caused by seawater tidal changes, expressed as: ; in, The density of seawater, The acceleration due to gravity is set to 9.81 m / s². 2 , The tidal height of the corresponding pump group reference surface for the i-th monitoring unit area at the j-th sampling time; Pump start-stop state sequence , ; Where 1 indicates the pump is running, and 0 indicates the pump is stopped. This represents the start / stop binary status value of the pump group at the j-th sampling time for the i-th monitoring unit.

3. The method for large-scale data acquisition in submarine tunnels based on Internet of Things technology according to claim 1, characterized in that, Based on tidal pressure sequences and pump start-up / shutdown state sequences, a tidal pressure response model is constructed using a logistic regression machine learning model to predict pump start-up / shutdown trends, specifically including: Tidal pressure sequences and pump start-stop state sequences were collected to construct a training sample set. The tidal pressure sequences were then processed and standardized to the [0,1] interval. The pump start-stop state sequences were encoded using binary encoding, with running state recorded as 1 and stopped state recorded as 0. Extracting tidal pressure sequences Pump start-stop state sequence Construct the average pressure sequence of the i-th monitoring unit. and the average value of the start-stop state sequence : ; ; The average pressure sequence of the i-th monitoring unit and the average value of the start-stop state sequence The tidal pressure sequence serves as the input feature vector for training a logistic regression model. Pump start-stop state sequence Calculate the Pearson correlation coefficient between the two variables. This is used to measure the driving force of tidal pressure on pump behavior: ; in, This represents the average value of the tidal pressure series. This represents the average value of the pump set start-up and shutdown sequence; and the Pearson correlation coefficient is used. The absolute value is used as the tidal pressure response index of the i-th monitoring unit area. , ; Preset moisture pressure sensitivity threshold ,when This indicates that there is a risk of frequent start-stop cycles induced by tidal pressure fluctuations in the i-th monitoring unit area, and generates the first adjustment command, including: at the initial sampling frequency Based on this, shorten the period by 10% to form the first sampling frequency. And execute; when This indicates that there is no risk of frequent start-stop cycles induced by tidal pressure fluctuations in the i-th monitoring unit area, and no adjustment is needed; the initial sampling frequency should be maintained. .

4. The method for large-scale data acquisition in submarine tunnels based on Internet of Things technology according to claim 1, characterized in that, Edge devices are installed in each monitoring unit area to collect the pump unit's vibration spectrum in real time using accelerometers and micro-acoustic sensors deployed at key parts of the pump body. and cavitation characteristic microacoustic signal sequence ; Vibration spectrum of pump unit Perform a Fast Fourier Transform (FFT) and select the energy peak value in the 2-15kHz frequency band as the feature input value to analyze the vibration spectrum of the pump unit. The process is updated to obtain the processed pump unit vibration spectrum. : ; This represents the peak vibration value of the pump housing after FFT processing for the i-th monitoring unit at the j-th sampling time; Microacoustic signal sequence of cavitation characteristics collected by microacoustic sensor Bandpass filtering and time-frequency joint analysis were performed to extract the power spectral density in the cavitation frequency band of 20-50kHz, forming a processed microacoustic signal sequence characteristic of cavitation. ; ; This represents the micro-acoustic vector of the pump group cavitation characteristics at the j-th sampling time for the i-th monitoring unit.

5. The method for large-scale data acquisition in submarine tunnels based on Internet of Things technology according to claim 1, characterized in that, The specific steps for identifying abnormal cavitation signal events during pump unit operation are as follows: Identifying abnormal cavitation signal events and generating a second adjustment command, including: based on the processed pump unit vibration spectrum... and cavitation characteristic microacoustic signal sequence Set the sliding window length R, where n is the total number of sampling points within the nearest window length R. Start abrupt change identification from the k-th time step: for the k-th sampling time step, calculate the average of the total number of the previous R nearest sampling points, and construct the average vibration peak value of the i-th monitoring unit. and the average value of micro-sound characteristics : ; ; Taking the k-th sampling point as the endpoint, the mean of R points counting forward, the interval is from Starting from the kth sampling point and ending at the kth sampling point; The average value of the vibration peak value for the i-th monitoring unit and the average value of micro-sound characteristics After Min-Max normalization, the cavitation judgment factor of the pump group for the i-th monitoring unit is obtained using a weighted formula. : ; In the formula, and The average value of the vibration peak value of the i-th monitoring unit after normalization and the average value of micro-sound characteristics The weight values, and the sum of the weights equals 1; Preset perturbation threshold If within the sliding window length R, If so, it is determined that there is an abnormal cavitation disturbance event in the pump unit at the current moment, and a first mark is made, and it is identified whether a second adjustment command is triggered. This indicates that there is no risk of cavitation disturbance at the current moment, and monitoring will continue.

6. The method for large-scale data acquisition in submarine tunnels based on Internet of Things technology according to claim 5, characterized in that, The triggering conditions for the second adjustment command are as follows: When the cavitation disturbance anomaly event is identified as 1 and the system returns to normal, then at the first sampling frequency... Based on this, the period is shortened by another 10% to form a second sampling frequency. ; If the cavitation disturbance anomaly is detected and continues for 3 cycles, a second adjustment command is triggered and executed, including: reducing the current pump speed of the monitoring unit by 5%-10%; If, after the second adjustment command is executed, if it is recognized After three consecutive cycles, starting from the second sampling frequency Restore to the first sampling frequency If, after the second adjustment command is executed, identification is performed... Still more If this occurs, a manual inspection command is triggered, uploaded to the remote operation and maintenance platform, and maintenance personnel are prompted to intervene.

7. The method for large-scale data acquisition in submarine tunnels based on Internet of Things technology according to claim 1, characterized in that, A structural environmental monitoring module is installed in each monitoring unit area. Through tunnel wall temperature sensors, structural gap water pressure monitors, and wall humidity change sensing devices deployed in the vicinity of the pumping station, leakage response characteristic parameters of the structures adjacent to the pumping station are collected in real time, including: Tunnel wall surface temperature change sequence , ; to This represents the change in tunnel wall surface temperature from the first adjacent time sampling point to the Mth adjacent time sampling point; Tunnel wall fissure water pressure change rate sequence , ; to This represents the rate of change of water pressure in the tunnel wall gaps from the first adjacent time sampling point to the Mth adjacent time sampling point; Tunnel wall humidity change sequence , ; to This represents the change in tunnel wall humidity from the first adjacent time sampling point to the Mth adjacent time sampling point; Temperature variation sequence of tunnel wall surface Tunnel wall fissure water pressure change rate sequence and the sequence of humidity changes in the tunnel walls Normalization is performed.

8. The method for large-scale data acquisition in submarine tunnels based on Internet of Things technology according to claim 1, characterized in that, Set the leakage weight coefficient. And W=1; where, The weighted values ​​represent the temperature changes on the tunnel wall surface, with the initial weights set to 0.2-0.3; The weighted value represents the rate of change of water pressure in the tunnel wall gaps, with the initial weight set to 0.5-0.6; The weighting values ​​represent the changes in humidity in the tunnel walls, with the initial weighting set to 0.2-0.3; Calculate the structure-pump coupling leakage index for the i-th monitoring unit area. : ; in, This represents the normalized change in tunnel wall humidity at the m-th normalized sampling point adjacent to the previous time. This represents the rate of change of water pressure in the tunnel wall gaps at the m-th adjacent time sampling point after normalization. The normalized value of the change in humidity of the tunnel wall at the m-th adjacent time sampling point.

9. The method for large-scale data acquisition in submarine tunnels based on Internet of Things technology according to claim 8, characterized in that, Preset coupling penetration determination threshold ,like If the i-th monitoring unit area is determined to have a structural leakage that could interfere with the operation of the pump unit, then the monitoring unit area is marked as a "highly sensitive area for coupled leakage". like If the i-th monitoring unit area does not pose a risk of structural leakage interfering with the operation of the pump unit, monitoring will continue. After being marked as a "highly sensitive area for coupled leakage", the structural collaborative analysis sub-strategy is triggered, which feeds back to step S3 to correct the pump group cavitation judgment factor of the i-th monitoring unit. Obtain the corrected cavitation judgment factor for the second pump group of the i-th monitoring unit. The corrected expression is: 。 10. The method for large-scale data acquisition in submarine tunnels based on Internet of Things technology according to claim 9, characterized in that, The corrected cavitation judgment factor for the second pump group of the i-th monitoring unit will be simultaneously applied. With disturbance threshold For comparison, if it is within the sliding window length R, If a cavitation disturbance anomaly is detected, it is determined that the pump unit currently experiences a cavitation disturbance anomaly, and a second flag is set. If the cavitation disturbance anomaly persists for three consecutive cycles, a second adjustment command is triggered and executed. If, within any sliding window period R after the execution of the second adjustment command, the cavitation disturbance anomaly detection rate continuously exceeds 60%, i.e., at least two anomalies occur in more than three detection cycles, a third adjustment command is triggered, including: further reducing the pump speed in the current monitoring unit area by 15% based on the second adjustment command; if the current pump speed is... Then adjust to: , This indicates the adjusted pump speed; And at the second sampling frequency Based on this, the period is shortened by another 25% to form a third sampling frequency. ; If the situation fails to recover after two consecutive cycles following the execution of the third adjustment command, a manual inspection command will be triggered and uploaded to the remote operation and maintenance platform to prompt maintenance personnel to intervene.