Leakage risk assessment method and system in pipe jacking construction based on vibration signal

Abstract

The present application belongs to the technical field of leakage monitoring, and particularly relates to a pipe jacking construction leakage risk assessment method and system based on vibration signals, which comprises the following steps: synchronously collecting the instantaneous torque signal of the main shaft of a pipe jacking machine and the broadband vibration signal inside the pipe jacking machine shell, extracting a high-frequency feature sub-band set representing the weak jetting characteristics of the fluid by using a wavelet packet decomposition algorithm; constructing a high-frequency energy entropy containing a mechanical work decoupling mechanism according to the relative energy contribution rate and the fluctuation state of the instantaneous torque signal; introducing the ratio of the measured pressure of the mud tank to the static water pressure of the external reservoir as the core fluid mechanics constraint boundary, and combining the relative deviation degree of the high-frequency energy entropy to calculate the dynamic risk degree; comparing and verifying the dynamic risk degree with the alarm critical threshold value, and activating the multi-stage water stop curtain grouting action. The present application strips the strong mechanical noise and eliminates the high-frequency flutter artifact data, captures the initial small high-pressure jet flow, and realizes the early active defense closed loop.

Description

Technical Field

[0001] This invention relates to the field of leakage monitoring technology. More specifically, this invention relates to a method and system for assessing leakage risk during pipe jacking construction based on vibration signals. Background Technology

[0002] In underwater tunneling construction of rock pipe jacking projects in water conservancy engineering, facing the complex conditions of high-strength rock layers of hundreds of megapascals and water heads of tens of meters high, the sealing and water-stopping of the tunnel entrance are the core links that determine the success or failure of the project. Existing technologies usually use a physical sealing structure of high-strength rubber flanges and steel pressure plates at the tunnel entrance, supplemented by a mud balance system to maintain the pressure balance at the tunnel face. In order to monitor the effectiveness of such physical sealing structures and prevent high-pressure groundwater from flowing in, existing safety early warning methods mainly rely on contact-type water pressure sensors and electromagnetic flowmeters directly installed inside the pipe sections or on the mud circulation pipeline to monitor macroscopic fluid dynamics parameters. At the same time, some existing technologies attempt to introduce conventional single-point vibration monitoring methods for auxiliary early warning. The common practice is to use the fast Fourier transform algorithm to convert the time-domain vibration signal of the shield surface to the frequency domain, and to determine whether there is any fluid penetration anomaly by calculating the total amplitude or root mean square value of a specific frequency band.

[0003] However, the aforementioned existing technologies have significant drawbacks: First, the monitoring logic based on macroscopic fluid dynamics parameters suffers from severe physical lag. When contact water pressure sensors or electromagnetic flowmeters detect pressure drops or abnormal increases in water volume, it often means that high-pressure groundwater has completely breached the physical seal and formed a macroscopic seepage channel. Taking remedial measures at this point can easily trigger large-scale water and sand inrushes or even flooding of the working well. Second, composite rock jacking machines generate extremely intense mechanical impact noise with a very wide spectral distribution during the crushing of rock mass. This high-intensity rock-breaking background noise completely masks the weak high-frequency jet vibration signal generated when high-pressure water flows through tiny gaps in the frequency domain. The lack of conventional vibration analysis algorithms that effectively decouple the weak high-frequency characteristics of fluid-structure interaction makes it impossible to distinguish whether the sudden increase in amplitude is caused by natural and intense mechanical vibration from the cutterhead cutting locally hard rock or by abnormal fluid vibration caused by the friction of high-pressure water jets against gaps. This causes the early warning and assessment system to frequently generate false alarm signals and force the main control system to perform unplanned shutdown operations. Summary of the Invention

[0004] To address the technical problem that existing technologies cannot accurately and promptly identify early-stage micro-leakage pores under strong mechanical noise interference, the present invention provides solutions in the following aspects.

[0005] In a first aspect, the present invention provides a method for assessing leakage risk during pipe jacking construction based on vibration signals, comprising: simultaneously acquiring instantaneous torque signals of the main shaft of the pipe jacking machine and broadband vibration signals inside the casing of the pipe jacking machine; using a wavelet packet decomposition algorithm to subdivide the broadband vibration signals into frequency bands, and extracting a set of high-frequency feature sub-bands characterizing weak fluid jetting features; constructing a high-frequency energy entropy including a mechanical work decoupling mechanism based on the relative energy contribution rate of the high-frequency feature sub-bands and the fluctuation state of the instantaneous torque signals; wherein, the high-frequency energy entropy is obtained by measuring the high-frequency fluid energy in each high-frequency feature sub-band. The disorder of the distribution is used as the numerator, and an exponential penalty term based on the absolute deviation of the instantaneous torque signal from its average value is used as the denominator. The ratio of the measured pressure in the mud chamber to the static pressure in the external reservoir is introduced as the core fluid dynamics constraint boundary. Combined with the relative deviation of the high-frequency energy entropy from its historical benchmark average, the dynamic risk degree used to assess the current sealing failure state is calculated. The dynamic risk degree is compared and verified with the system's preset alarm threshold in real time. Based on the verification result, the underlying control system is triggered to activate the multi-level water-stop curtain grouting action to achieve a closed-loop active defense against leakage.

[0006] This invention extracts a set of high-frequency feature subbands based on wavelet packet decomposition and constructs a high-frequency energy entropy by combining the fluctuation state of the instantaneous torque signal. It further introduces the pressure difference ratio inside and outside the mud chamber to calculate the dynamic risk level to trigger grouting action. It removes low-frequency strong impact background noise from a physical mechanism perspective and uses the exponential fluctuation of the instantaneous torque signal to adaptively punish and eliminate high-frequency flutter artifact data caused by sudden changes in local rock hardness. Combined with macroscopic fluid dynamic constraints, the control system can capture the initial small high-pressure jet before the occurrence of macroscopic water inrush disaster, effectively avoiding misjudgment caused by single sensor drift and missed detection caused by severe mechanical vibration, and realizing early active defense closed loop.

[0007] Preferably, the extraction of the high-frequency feature subband set characterizing the weak jetting features of the fluid includes: applying a decomposition layer of [number] layers to the broadband vibration signal. Wavelet packet decomposition is used to obtain a sequence of node coefficients for multiple independent frequency bands; frequencies with centers lower than the rock-breaking cutoff frequency are filtered out. The low-frequency nodes are decomposed, and the remaining high-frequency nodes are retained to form a high-frequency feature sub-band set; the decomposition layer number The range of values ​​is to The rock-breaking cutoff frequency The range of values ​​is Hertz hertz.

[0008] Preferably, in the set of high-frequency feature subbands, the first... The relative energy contribution rate of each sub-band is equal to that of the first sub-band. The ratio of the sum of squared node coefficients of a sub-band to the sum of squared node coefficients of all sub-bands in the set of high-frequency feature sub-bands.

[0009] This invention uses wavelet packet decomposition to filter out low-frequency nodes while retaining high-frequency nodes. While preserving local time-domain resolution, it performs fine and uniform division of the frequency domain, stripping away the low-frequency noise base containing the main rock-breaking energy. The obtained relative energy contribution rate achieves a normalized description of the high-frequency energy distribution characteristics, providing a reliable data foundation for extracting pure weak fluid characteristics, thereby improving the accuracy of early identification of small seepage pores.

[0010] Preferably, the construction of the high-frequency energy entropy including the mechanical work decoupling mechanism includes: calculating the average value of the instantaneous torque signal within the current observation window using a moving average algorithm. ; In the formula, It is the high-frequency energy entropy; For high-frequency characteristic subband set The total number of subbands included; This is the index for the high-frequency subband; For the first The relative energy contribution rate of each sub-band; It is the natural logarithm function; It is an exponential function with the natural constant as its base; This is the torque ripple suppression coefficient; The instantaneous torque signal acquired at the current moment; This represents the absolute deviation of the current instantaneous torque signal from its average value. This indicates taking the absolute value.

[0011] This invention constructs the information entropy numerator by relative energy contribution rate and constructs the exponential penalty term denominator based on the absolute deviation of instantaneous torque signal, and calculates high-frequency energy entropy together; it amplifies weak leakage characteristics by using information entropy model, and at the same time achieves strong exponential suppression of high-frequency mechanical flutter artifact data caused by sudden changes in rock hardness at the pure mathematical level, separates mechanical dynamic noise source from fluid acoustic emission source, and eliminates interference from strong mechanical noise.

[0012] Preferably, the calculation for assessing the dynamic risk of the current sealing failure state includes: calculating the historical baseline average of the high-frequency energy entropy during a historical leak-free period; if the high-frequency energy entropy is greater than the historical baseline average, then calculating the dynamic risk. ,include: In the formula, For dynamic risk level; It is the high-frequency energy entropy; This represents the historical benchmark average of high-frequency energy entropy. The static water pressure of the external reservoir; This is the current measured pressure in the mud chamber; This is the pressure bias constant; It is a water head pressure sensitive factor.

[0013] This invention combines the deviation of high-frequency energy entropy from the historical benchmark average with a ratio term constructed based on the measured pressure of the mud chamber and the static water pressure of the external reservoir to calculate the dynamic risk level. It also nonlinearly combines microscopic acoustic characteristics with macroscopic fluid dynamic boundary conditions across physical domains, so that an extremely high risk assessment value is output only when the pressure barrier has a potential failure and is accompanied by abnormal high-frequency vibration entropy increase characteristics, thus eliminating the system assessment error caused by the long-term slow drift of a single sensor.

[0014] Preferably, the method further includes: if the high-frequency energy entropy is less than or equal to the historical baseline average, then the dynamic risk level is zero.

[0015] Preferably, the activation of the multi-stage water-stop curtain grouting action includes: when the dynamic risk level exceeds the alarm threshold, generating and sending an interruption command for the propulsion cylinder to the programmable controller; increasing the injection pressure holding parameters of the main mud pump; and activating the outer polyurethane grouting reinforcement circuit to implement grouting.

[0016] This invention sets up a system based on the comparison results between dynamic risk level and alarm threshold, triggering an interruption command and activating the grouting reinforcement circuit. This enables the underlying hardware to autonomously determine the response timing, implementing microsecond-level equipment blocking and proactive perimeter reinforcement interception before a catastrophic macroscopic water surge forms a true through-channel, ensuring proactive defense effectiveness from the execution end.

[0017] Preferably, the alarm threshold value range is [1.3, 1.7].

[0018] Preferably, the instantaneous torque signal is acquired by a high-frequency torque sensor deployed on the end of the main shaft drive hydraulic motor at a preset sampling frequency, and the wideband vibration signal is acquired synchronously by a piezoelectric accelerometer deployed close to the pressure-bearing housing behind the main bearing.

[0019] In a second aspect, the present invention provides a leakage risk assessment system for pipe jacking construction based on vibration signals, comprising a processor and a memory, wherein the memory stores computer program instructions, and when the computer program instructions are executed by the processor, the above-mentioned leakage risk assessment method for pipe jacking construction based on vibration signals is implemented.

[0020] By adopting the above technical solution, the leakage risk assessment method based on vibration signal in pipe jacking construction is generated into a computer program and stored in a memory so that it can be loaded and executed by a processor. A terminal device can then be made based on the memory and processor for convenient use.

[0021] The beneficial effects of this invention are as follows: This invention extracts a set of high-frequency feature subbands based on wavelet packet decomposition and constructs a high-frequency energy entropy by combining the fluctuation state of the instantaneous torque signal. It further introduces the pressure difference ratio inside and outside the mud chamber to calculate the dynamic risk level to trigger grouting action. It removes low-frequency strong impact background noise from a physical mechanism perspective and uses the exponential fluctuation of the instantaneous torque signal to adaptively punish and eliminate high-frequency flutter artifact data caused by sudden changes in local rock hardness. Combined with macroscopic fluid dynamic constraints, the control system can capture the initial small high-pressure jet before the occurrence of macroscopic water inrush disaster, effectively avoiding misjudgment caused by single sensor drift and missed detection caused by severe mechanical vibration, and realizing early active defense closed loop. Attached Figure Description

[0022] Figure 1 This is a flowchart illustrating the leakage risk assessment method for pipe jacking construction based on vibration signals in this invention. Figure 2 It is a schematic diagram illustrating the status of each operating stage and the monitoring of macroscopic physical parameters; Figure 3 This is a schematic diagram illustrating the feature detection effect based on high-frequency amplitude energy in the prior art; Figure 4 This is a schematic diagram illustrating the early and accurate warning effect of the present invention based on dynamic risk level. Detailed Implementation

[0023] 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, not all, of the embodiments of the present invention. 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.

[0024] The specific embodiments of the present invention will now be described in detail with reference to the accompanying drawings.

[0025] This invention discloses a method for assessing leakage risk during pipe jacking construction based on vibration signals, referring to... Figure 1 This includes steps S1-S4: S1: Synchronously acquire the instantaneous torque signal of the main shaft of the pipe jacking machine and the broadband vibration signal inside the casing of the pipe jacking machine. Use the wavelet packet decomposition algorithm to subdivide the broadband vibration signal into frequency bands and extract the high-frequency feature sub-band set that characterizes the weak jetting characteristics of the fluid.

[0026] It should be noted that when high-pressure water breaks through the initial tiny gaps in the sealed structure, it generates jet acoustic emission and weak vibrations with specific high-frequency characteristics due to the intense throttling effect. These physical characteristics are extremely weak in the time domain and are completely masked by the huge low-frequency impact noise generated by the high-intensity rock breaking, rendering conventional vibration extraction algorithms ineffective. Therefore, this invention utilizes wavelet packet decomposition technology to perform fine and uniform division of the frequency domain while preserving local time domain resolution. This forcibly strips away the low-frequency noise substrate containing the main rock-breaking energy and focuses on the high-frequency micro-perturbation frequency band that may contain leakage jet characteristics, providing a clean data foundation for subsequent enhancement of weak fluid characteristics.

[0027] Specifically, during the pipe jacking operation, a high-frequency torque sensor deployed at the end of the main shaft drive hydraulic motor acquires instantaneous torque signals at a preset sampling frequency. Simultaneously, a wideband vibration signal is acquired by a piezoelectric accelerometer deployed close to the pressure-bearing housing behind the main bearing. ; for the acquired broadband vibration signal Apply decomposition layer number Wavelet packet decomposition decomposes the full-band signal into a sequence of node coefficients in multiple independent frequency bands; it filters out frequencies whose center frequency is below the rock-breaking cutoff frequency. All low-frequency nodes are retained, and the remaining high-frequency nodes form a high-frequency feature subband set. ; Calculate the set of high-frequency characteristic subbands The Middle Relative energy contribution rate of individual sub-bands It is equal to the first The sum of squared nodal coefficients of each subband and the set of high-frequency feature subbands The ratio of the sum of squares of the coefficients of all sub-band nodes.

[0028] Specifically, by calculating the ratio of the energy of a single high-frequency sub-band to the total high-frequency energy, a normalized description of the high-frequency energy distribution characteristics was achieved. When a minor leak occurs in the sealed structure, generating a high-frequency jet, the energy of a specific high-frequency sub-band will change significantly, leading to an increase in the relative energy contribution rate. The distribution pattern changes; this calculation formula provides a core component for subsequent assessment of the disorder of energy distribution.

[0029] The number of decomposition layers This is an integer parameter set to control the fineness of frequency band division. If this value is set too small, the frequency band division will be too coarse, failing to effectively remove the low-frequency noise substrate and causing high-frequency feature aliasing to fail. If it is set too large, the computational complexity will increase exponentially, causing serious delays in real-time system monitoring and leading to hardware risks. Therefore, its value range is set to [range missing]. to In this embodiment, it is set as follows: This ensures an optimal balance between extracting pure high-frequency features and maintaining the real-time performance of the system's high-frequency operations; in other embodiments, implementers can set the number of decomposition layers based on the controller's computing power margin.

[0030] The rock-breaking cutoff frequency This is a frequency parameter set to define the boundary between low-frequency mechanical noise and high-frequency fluid jet. If this value is set too small, a large amount of mechanical cutting noise will mix into the feature subband set, resulting in an extremely low signal-to-noise ratio in subsequent evaluations and causing false alarms. If it is set too large, it will incorrectly filter out some true jet features in the low-frequency band, leading to feature loss and abnormal missed detections. Therefore, its value range is set to [range missing]. Hertz Hertz, in this embodiment it is set to Hertz, to ensure precise stripping of the rock-breaking main frequency energy of the cutting tool; in other embodiments, the implementer can set the rock-breaking cutoff frequency according to the cutting head rotation speed and the hardness of the formation rock.

[0031] For example, Figure 2 This diagram illustrates the status of monitoring macroscopic physical parameters at various operating stages. The horizontal axis represents time, the left vertical axis represents pressure, and the right vertical axis represents deviation rate. The diagram shows the dynamic changes of the curves corresponding to the measured pressure in the mud chamber, the static water pressure in the external reservoir, and the absolute deviation rate of instantaneous torque under four different operating stages: homogeneous safe formation, cutting of extremely hard dikes, initial minor leakage, and severe penetrating leakage. In the cutting of extremely hard dikes stage, the curve corresponding to the absolute deviation rate of instantaneous torque shows a significant numerical surge, indicating that strong mechanical vibration and a sharp increase in equipment workload occur at this time. In the initial minor leakage and severe penetrating leakage stages, as the high-pressure groundwater gradually breaks through the barrier, the curve corresponding to the measured pressure in the mud chamber shows a step-like downward trend and gradually approaches the curve corresponding to the constant static water pressure in the external reservoir, representing the physical process of the gradual failure of the macroscopic pressure defense line.

[0032] S2: Based on the relative energy contribution rate of the high-frequency characteristic sub-band set and the fluctuation state of the instantaneous torque signal, a high-frequency energy entropy containing a mechanical work decoupling mechanism is constructed to suppress the interference of geological abrupt changes on high-frequency vibration characteristics.

[0033] It should be noted that the intense high-frequency chatter generated when the cutterhead encounters a localized extremely hard rock vein is easily confused with the energy fluctuations caused by weak seepage jets, leading to false alarms in the early warning system. Hard rock cutting is inevitably accompanied by drastic changes in the instantaneous torque signal of the spindle, and the initial minor seepage will not cause transient impact load on the spindle torque. Therefore, this invention uses the instantaneous torque signal as the absolute physical reference benchmark for the mechanical load working state, and constructs a reverse penalty decoupling logic to achieve a non-intrusive and complete separation of the mechanical power noise source and the fluid acoustic emission source at a purely mathematical level: when the high-frequency energy shows an increasing trend and the torque fluctuates violently in sync, it is judged as normal local hard rock breaking fluctuation and is mathematically forced to suppress it; when the high-frequency energy distribution disorder increases significantly while the spindle torque remains in a relatively stable benchmark operating state, it highlights the abnormal crevice seepage jet characteristics.

[0034] Specifically, the instantaneous torque signal acquired at the current moment is defined as... The average value of the instantaneous torque signal within the current time window is calculated based on the time series using a moving average algorithm. According to the relative energy contribution rate and instantaneous torque signal Calculate the high-frequency energy entropy within the current observation window. This achieves the suppression of high-frequency mechanical flutter data caused by abrupt changes in rock hardness. The specific calculation formula is as follows:

[0035] In the formula, It is the high-frequency energy entropy; For high-frequency characteristic subband set The total number of subbands included; This is the index for the high-frequency subband; For the first The relative energy contribution rate of each sub-band; It is the natural logarithm function; It is an exponential function with the natural constant as its base; This is the torque ripple suppression coefficient; The instantaneous torque signal acquired at the current moment; This represents the average value of the instantaneous torque signal within the current observation window; This represents the absolute deviation of the current instantaneous torque signal; This indicates taking the absolute value.

[0036] Among them, if the first Relative energy contribution rate of individual sub-bands When the value is zero, it indicates that there is no acoustic energy generated by mechanical resonance or fluid jet within the high-frequency characteristic sub-band, meaning that the local frequency band is in a state of absolute acoustic silence. At this time, the sub-band corresponding to... Setting the value of the operation term to zero indicates that the absolutely silent subband will not increase the overall disorder of the high-frequency energy distribution.

[0037] Among them, if the average value of the instantaneous torque signal within the current observation window When the ratio equals zero, it indicates that the main shaft of the pipe jacking machine is in a completely idle, unstuck, or stopped state without load. At this time, since no basic mechanical work is being done, any minute mechanical disturbance will cause the ratio to become infinitely large, thus triggering a crash in the underlying hardware calculations. Therefore, when the average value is zero... When the value equals zero, the denominator term in the formula, i.e., the exponential form of the mechanical work decoupling penalty term, is directly decoupled. The overall value is set to 1, meaning that no mechanical work penalty is applied to the high-frequency energy distribution under this no-load condition.

[0038] The numerator of this calculation uses an information entropy model to measure the degree of disorder in the distribution of high-frequency fluid energy among various high-frequency characteristic subbands. With the initial occurrence of micro-leakage pores, the high-pressure jet exhibits typical broadband high-frequency white noise characteristics. This physical phenomenon causes the high-frequency energy distribution to tend towards homogenization, leading to a significant increase in the information entropy value of the numerator. The denominator of this calculation is an exponential mechanical work decoupling penalty term, which is applied when the front-end cutterhead encounters a sudden hard rock event causing an instantaneous torque signal. Deviation from the average When the absolute deviation is large, the ratio of the absolute deviation increases, causing the denominator value to grow exponentially. This strongly suppresses the false entropy increase in the numerator caused by high-frequency mechanical resonance, and completely eliminates the high-frequency mechanical flutter artifact data caused by sudden changes in rock hardness. Conversely, if the spindle torque runs smoothly, the denominator value approaches 1, and the entropy increase physical characteristics of the fluid jet extracted by the numerator are completely preserved. This achieves non-invasive separation of the mechanical power noise source and the fluid acoustic emission source, ensuring the physical uniqueness of the evaluation parameters for the fluid jet characteristics.

[0039] The torque fluctuation suppression coefficient This is a physical gain parameter set to adjust the penalty effect of mechanical torque fluctuations on the system entropy. If this value is set too small, the system cannot effectively suppress the high-frequency mechanical flutter artifacts caused by sudden hard rock changes, leading to failure of decoupling between mechanical power and fluid sound sources and causing frequent false alarms of shutdown. If it is set too large, even occasional slight mechanical disturbances will cause the system entropy to be excessively compressed, resulting in the true initial seepage jet characteristics being completely penalized and masked, leading to catastrophic missed alarms. Therefore, its value range is set to [range missing]. to In this embodiment, it is set as follows: This ensures that the physical characteristics of fluid jet entropy increase are fully preserved while eliminating artifacts of sudden changes in rock hardness; in other embodiments, the implementer can set the torque fluctuation suppression coefficient according to the rated torque fluctuation rate of the pipe jacking machine drive motor.

[0040] S3: The ratio of the measured pressure in the mud chamber to the static pressure in the external reservoir is introduced as the core fluid dynamics constraint boundary. Combined with the relative deviation of the high-frequency energy entropy, the dynamic risk level is calculated to assess the current sealing failure state.

[0041] It should be noted that while single-dimensional vibration acoustic characteristics decouple the mechanical interference at the front end, during long-term propulsion operations, the combined effects of shield shell wear, piezoelectric sensor aging, and background temperature drift may produce gradually changing data errors, causing the evaluation benchmark to shift. When the pipe jacking machine completely exits the tunnel and enters the deep-water high-head reservoir area, the internal support pressure provided by the mud balance system must be consistently higher than the external reservoir hydrostatic pressure to effectively prevent high-pressure water backflow. Therefore, this invention introduces the core macroscopic fluid dynamic boundary condition of the pressure difference between the inside and outside of the mud chamber and cascades it with the extracted microscopic acoustic characteristics across physical domains. This ensures that a high-risk value is only output when the pressure barrier is at risk of failure and accompanied by abnormal high-frequency vibration entropy increase, completely eliminating system misjudgment caused by single sensor drift.

[0042] Specifically, the current measured pressure in the mud chamber is obtained through pressure sensors deployed on the mud baffle. Simultaneously, the hydrostatic pressure of the external reservoir is calculated based on the data from the reservoir level gauge and the current elevation difference. Data from the historical leak-free phases recorded by the equipment during its traversal of initially homogeneous and safe rock formations were extracted to calculate the historical baseline mean of the high-frequency energy entropy. If the high-frequency energy entropy acquired at the current moment Less than or equal to the historical benchmark mean When the current sealing system is determined to be in a safe state, the dynamic risk level is directly adjusted. Set to zero; if the high-frequency energy entropy acquired at the current moment... Greater than the historical benchmark average Calculate dynamic risk level. The specific calculation formula is as follows:

[0043] In the formula, For dynamic risk level; It is the high-frequency energy entropy; This represents the historical benchmark average of high-frequency energy entropy. The static water pressure of the external reservoir; This is the current measured pressure in the mud chamber; This is the pressure bias constant; It is a water head pressure sensitive factor.

[0044] The first term of this formula is a relative variable term, which calculates the current high-frequency energy entropy. Exceeding the historical benchmark average The proportional amplitude characterizes the significance of the microscopic jet physical characteristics. The larger the value, the more significant the high-frequency weak jet physical characteristics are at the sealing layer gaps. The latter term of this calculation formula constitutes the cascade multiplication factor of the macroscopic state in fluid mechanics. When the mud balance system is operating normally, the current measured pressure in the mud chamber is... The supply is sufficient and its value is far greater than the static pressure of the external reservoir. At that time, the base of the fraction was relatively small. After exponential calculation, the impact of accidental environmental fluctuations caused by weak acoustic vibration characteristics in front was greatly reduced, improving the physical disturbance resistance of the evaluation system; once the current measured pressure of the mud chamber... An abnormal drop occurred, approaching the hydrostatic pressure of the external reservoir. The value of this exponential term increases sharply as the denominator decreases, nonlinearly and exponentially amplifying the deterioration of the acoustic vibration domain, thus increasing the dynamic risk. The explosive growth reflects the fact that the physical sealing barrier inside the tunnel entrance is on the verge of being completely penetrated and broken through by the external high-pressure water.

[0045] The pressure bias constant This is a safety parameter set to prevent computational crashes caused by the denominator becoming zero due to complete loss of pressure inside the chamber. If this value is set too small, it can easily trigger a hardware interrupt error of division by zero, causing the main control computer board to crash, under extreme conditions of severe depressurization and emptying of the mud chamber. If it is set too large, it will significantly weaken the sensitivity of the multiplier cascade amplification of dynamic risk due to changes in mud pressure, resulting in an extremely sluggish system warning response. Therefore, its value range is set to 0.01 MPa to 0.05 MPa. In this embodiment, it is set to 0.02 MPa to ensure the robust operation of the algorithm and the true numerical reflection of the fluid dynamics state. In other embodiments, the implementer can set the pressure bias constant according to the lower limit of the range accuracy of the mud pressure sensor.

[0046] The head pressure sensitive factor This is an exponential parameter set to control the weighting of the differential pressure ratio on the degree of risk. If this value is set too small, the multiplier cascade effect brought about by the differential pressure ratio will not be obvious, and it will not be able to form an explosive increase in the degree of risk when the mud pressure is about to fail, resulting in missing the optimal window period for liquid grouting and sealing. If it is set too large, even small normal pressure fluctuations in the mud pipeline network will be amplified exponentially and nonlinearly, causing the system to over-erroneously judge the risk defense line. Therefore, its value range is set to [range missing]. to In this embodiment, it is set as follows: This ensures that the risk level can reflect the physical nature of the external water body breaking through the sealing barrier with extreme sensitivity and scientific accuracy; in other embodiments, the implementer can set the head pressure sensitivity factor based on the absolute head height of the external reservoir's hydrostatic pressure.

[0047] S4: The dynamic risk level is compared and verified with the alarm threshold set by the system in real time. If the verification result exceeds the judgment boundary, the underlying control system is triggered to activate the multi-level water-stop curtain grouting action.

[0048] It should be noted that without a rigorous execution judgment logic, the underlying control system cannot autonomously determine the timing of hardware device response; blindly activating grouting or interrupting propulsion will severely delay the project progress or even cause mechanical seizure, while a slow response will cause high-pressure water to completely tear the seal, leading to surrounding rock collapse and equipment flooding accidents; therefore, this invention sets a trigger physical boundary for risk judgment, and compares the calculated dynamic risk level with this boundary in real time at high speed, enabling the system to implement early interception control before a catastrophic macroscopic water surge forms a real through channel, realizing early closed-loop active defense based on the precursors of micro-fluid vibration.

[0049] Specifically, the calculated dynamic risk level is obtained in real time. and dynamic risk level Compared with the preset alarm threshold in the control system Perform real-time numerical comparison: when dynamic risk level The value exceeds the alarm threshold. At that time, the programmable controller in the pipe jacking control room sends an interruption command to the propulsion cylinder and simultaneously increases the injection pressure holding parameters of the main mud pump, instantly activating the outer polyurethane grouting reinforcement circuit of the multi-stage sealing system at the tunnel entrance, completing the active defense closed loop for seepage prevention.

[0050] The alarm threshold This is a judgment boundary parameter set to control the emergency response sensitivity of the mud balance system. If the value is set too small, the evaluation system will overreact to the occasional small high-frequency sound waves brought by the high-speed friction seal of the normal hard rock strata, resulting in frequent triggering of false alarm signals and forced interruption of the mud pump and main thrust cylinder propulsion operations, seriously delaying the overall pipeline breakthrough project progress. If the value is set too large, the evaluation system will exhibit a sluggish physical response to the fluid dynamics formation process of the small high-pressure seepage channel at the tunnel entrance in the early stage, resulting in missing the optimal time window for initiating the dual-liquid grouting and sealing, and ultimately the external high-pressure water flow will completely tear the high-strength rubber water-stop ring, causing irreversible risk of surrounding rock collapse. Therefore, its value range is set to 1.3 to 1.7, and in this embodiment it is set to 1.5 to ensure that while completely filtering the residual impact noise of non-uniform cutting of granite, it can make instantaneous status confirmation of high-pressure micro-jet with a micron-level penetration diameter. In other embodiments, the implementers can set the alarm critical threshold according to the seasonal fluctuation of the actual water head height of the external reservoir.

[0051] For example, Figure 3 This diagram illustrates the effect of feature detection based on high-frequency amplitude energy in existing technologies. The horizontal axis represents time, and the vertical axis represents relative energy amplitude. The diagram shows the spatial distribution of the curves corresponding to existing technology features under varying operating conditions in the absence of a decoupling mechanism between mechanical work and fluid dynamics. During the cutting stage of extremely hard rock veins, the intense high-frequency mechanical vibration caused by the cutterhead cutting hard rock causes the curves corresponding to existing technology features to surge significantly and exceed the straight line corresponding to the existing technology's alarm threshold, falling into a severe false alarm zone, thus triggering unplanned shutdown false alarms in the control system. In the initial stage of minor leakage, the high-frequency amplitude energy generated by the small high-pressure water jet is extremely weak compared to the rock-breaking energy. The curve values ​​corresponding to existing technology features are severely suppressed and remain completely below the straight line corresponding to the existing technology's alarm threshold, falling into a fatal false alarm zone. This prevents existing technology from capturing true early signs of leakage, thus missing the optimal window for initiating dual-liquid grouting and sealing.

[0052] For example, Figure 4This diagram illustrates the early and accurate warning effect of the present invention based on dynamic risk level. The horizontal axis represents time, and the vertical axis represents dynamic risk level. The diagram shows the distribution of the curve corresponding to the dynamic risk level output by the present invention as the working conditions change after applying the mechanical work decoupling mechanism and macroscopic fluid dynamics constraint boundary proposed in this invention. During the cutting stage of extremely hard rock veins, due to the effective adaptive constraint penalty mechanism based on instantaneous torque signal fluctuations, the curve corresponding to the dynamic risk level is strictly suppressed and fits to the zero baseline, eliminating high-frequency chatter artifacts in the main equipment caused by sudden changes in local rock hardness, achieving zero false alarms in this stage. In the initial stage of minor leakage, due to the high-frequency energy distribution formed by the high-pressure microjet tends to... The homogenization process leads to a significant increase in information entropy, which is simultaneously accompanied by a cascaded amplification effect of hydrodynamic multipliers triggered by slight depressurization of the mud chamber. The curve corresponding to the dynamic risk level rapidly rises and exceeds the straight line corresponding to the alarm critical threshold, accurately falling into the precise early interception zone. This allows the system to capture the formation state of the initial small high-pressure jets in the orifice before the formation of catastrophic macroscopic water inrush. In the severe penetration and leakage stage, as the mud pressure defense line is on the verge of being breached, the curve corresponding to the dynamic risk level shows a nonlinear exponential explosive growth and falls into the disaster outbreak response zone. This reflects the dangerous nature of the physical sealing barrier inside the tunnel being completely breached by the external high-pressure water body, thus realizing early closed-loop active defense based on the precursors of micro-fluid vibrations.

[0053] This invention also discloses a leakage risk assessment system for pipe jacking construction based on vibration signals, including a processor and a memory. The memory stores computer program instructions, which, when executed by the processor, implement the leakage risk assessment method for pipe jacking construction based on vibration signals according to this invention.

[0054] The system also includes other components well known to those skilled in the art, such as communication buses and communication interfaces, the settings and functions of which are known in the art and will not be described in detail here.

Claims

1. A method for assessing leakage risk during pipe jacking construction based on vibration signals, characterized in that, include: The instantaneous torque signal of the main shaft of the pipe jacking machine and the broadband vibration signal inside the casing of the pipe jacking machine are acquired synchronously. The broadband vibration signal is subdivided into frequency bands using the wavelet packet decomposition algorithm to extract a set of high-frequency feature subbands that characterize the weak jetting features of the fluid. Based on the relative energy contribution rate of the high-frequency characteristic sub-band set and the fluctuation state of the instantaneous torque signal, a high-frequency energy entropy incorporating a mechanical work decoupling mechanism is constructed. The high-frequency energy entropy is calculated by using the disorder of the high-frequency fluid energy distribution among the various high-frequency characteristic sub-bands as the numerator and an exponential penalty term based on the absolute deviation of the instantaneous torque signal relative to its average value as the denominator. The formula for calculating the high-frequency energy entropy is: In the formula, It is the high-frequency energy entropy; High-frequency characteristic subband set The total number of subbands included; This is the index for the high-frequency subband; For the first The relative energy contribution rate of each sub-band; It is the natural logarithm function; It is an exponential function with the natural constant as its base; This is the torque ripple suppression coefficient; The instantaneous torque signal acquired at the current moment; This represents the absolute deviation of the current instantaneous torque signal from its average value. Indicates taking the absolute value; The ratio of the measured pressure in the mud chamber to the hydrostatic pressure in the external reservoir is introduced as the core fluid dynamics constraint boundary. Combined with the relative deviation of the high-frequency energy entropy from its historical benchmark average, the dynamic risk level for assessing the current seal failure state is calculated, including: if the high-frequency energy entropy is greater than the historical benchmark average, then the dynamic risk level is calculated. In the formula, For dynamic risk level; This represents the historical benchmark average of high-frequency energy entropy. The static water pressure of the external reservoir; This is the current measured pressure in the mud chamber; This is the pressure bias constant; The water head pressure sensitive factor; if the high-frequency energy entropy is less than or equal to the historical benchmark average, then the dynamic risk level is zero; The dynamic risk level is compared and verified with the alarm threshold preset by the system in real time. Based on the verification result, the underlying control system is triggered to activate the multi-level water-stop curtain grouting action to realize the active defense closed loop of seepage prevention.

2. The method for assessing leakage risk during pipe jacking construction based on vibration signals according to claim 1, characterized in that, The set of high-frequency feature subbands extracted to characterize the weak jetting features of the fluid includes: The broadband vibration signal is decomposed into layers with the following number of layers: Wavelet packet decomposition operation is used to obtain the node coefficient sequence of multiple independent frequency bands; The center of the screening frequency is lower than the rock breaking cutoff frequency. The low-frequency nodes are selected, and the remaining high-frequency nodes are retained to form a set of high-frequency feature subbands; The number of decomposition layers The range of values ​​is to The rock-breaking cutoff frequency The range of values ​​is Hertz hertz.

3. The method for assessing leakage risk during pipe jacking construction based on vibration signals according to claim 1, characterized in that, The high-frequency feature subband set of the first The relative energy contribution rate of each sub-band is equal to that of the first sub-band. The ratio of the sum of squared node coefficients of a sub-band to the sum of squared node coefficients of all sub-bands in the set of high-frequency feature sub-bands.

4. The method for assessing leakage risk during pipe jacking construction based on vibration signals according to claim 1, characterized in that, The method for calculating the average value of the instantaneous torque signal includes: The average value of the instantaneous torque signal within the current observation window is calculated using a moving average algorithm.

5. The method for assessing leakage risk during pipe jacking construction based on vibration signals according to claim 1, characterized in that, The method for calculating the historical benchmark mean of the high-frequency energy entropy includes: Calculate the historical baseline mean of the high-frequency energy entropy during the historical leak-free phase.

6. The method for assessing leakage risk during pipe jacking construction based on vibration signals according to claim 1, characterized in that, The activation of the multi-stage waterstop curtain grouting action includes: When the dynamic risk level exceeds the alarm threshold, an interruption command for the propulsion cylinder is generated and sent to the programmable controller. Increase the injection pressure holding parameters of the main mud pump and activate the outer polyurethane grouting reinforcement circuit to implement grouting.

7. The method for assessing leakage risk during pipe jacking construction based on vibration signals according to claim 1, characterized in that, The alarm threshold value range is [1.3, 1.7].

8. The method for assessing leakage risk during pipe jacking construction based on vibration signals according to claim 1, characterized in that, The instantaneous torque signal is acquired by a high-frequency torque sensor deployed at the end of the main shaft drive hydraulic motor at a preset sampling frequency, and the wideband vibration signal is acquired synchronously by a piezoelectric accelerometer deployed close to the pressure-bearing housing behind the main bearing.

9. A leakage risk assessment system for pipe jacking construction based on vibration signals, characterized in that, include: A processor and a memory, the memory storing computer program instructions that, when executed by the processor, implement the leakage risk assessment method for pipe jacking construction based on vibration signals according to any one of claims 1-8.

Images