A method and system for monitoring micro-displacement and vibration decoupling of a downhole reference point

Abstract

The present application belongs to the technical field of structural health monitoring, and relates to a kind of underground reference point micro-displacement and vibration decoupling monitoring method and system, comprising the following steps: based on trigger signal synchronous acquisition three-dimensional micro-displacement response data stream and vibration excitation data stream under the same clock source, based on short-time energy envelope definition coherent excitation window;Energy transfer efficiency parameter is calculated to generate the motion conversion efficiency;Based on free oscillation signal calculation subsequent residual oscillation energy parameter;Structural dynamic response feature vector is stored to historical diagnosis database;According to the evaluation result, reference point structural health deterioration early warning result is output.The present application solves the problem of how to synchronously collect dynamic excitation and micro-displacement response data and effectively decouple the physical correlation between impact vibration and micro-displacement to monitor the structural health state.

Description

Technical Field

[0001] This invention belongs to the technical field of structural health monitoring, and relates to a method and system for decoupling monitoring of micro-displacement and vibration at downhole reference points. Background Technology

[0002] In underground engineering activities such as mining and tunnel construction, the stability of the surrounding rock or support structure is a core element for ensuring safe production. These underground structures are subjected to complex stress environments over long periods and are frequently impacted by dynamic loads such as blasting, which can easily induce the generation and propagation of micro-cracks, gradually accumulating damage and leading to deformation and instability. Long-term monitoring of the micro-displacements and vibration states of key reference points is of great significance for understanding the structural degradation process and early warning of potential risks. The current core challenge lies in distinguishing the transient structural response caused by a single dynamic excitation event from complex monitoring signals and quantifying the impact of the event on the structural health.

[0003] Currently, the solutions commonly used in the industry mainly fall into two categories. The first category, such as the Chinese patent document with application number CN115655130A, discloses a method and device for monitoring tunnel deformation based on three-dimensional scanning and positioning markers. This type of method primarily uses geodesy principles to obtain the cumulative deformation trend of a structure over a longer time scale. The second category, such as the Chinese patent document with application number CN101865664B, discloses a portable dynamic deflection and displacement measurement device and method. This type of method continuously or trigger-basedly records dynamic response data by installing sensors such as displacement gauges or accelerometers on the structure, thereby assessing vibration intensity and other related indicators.

[0004] However, the aforementioned existing technologies have limitations in dealing with complex downhole dynamic monitoring scenarios. Periodic methods based on geodetic surveys such as 3D scanning have a low measurement frequency, making it difficult to capture the transient response process of structures during short-term dynamic events. The results are often long-term superposition effects of multiple events, making it difficult to accurately attribute damage from a single event. Conventional sensor-based dynamic monitoring methods often evaluate vibration excitation signals and displacement response signals as independent indicators, rarely analyzing the intrinsic physical relationship between them. At the data acquisition level, if the excitation and response signal acquisition systems lack a strict hardware synchronization mechanism, even small deviations in the time reference can introduce uncertainty into subsequent causal relationship analysis. Furthermore, micrometer-level displacement measurements are easily affected by gradual baseline drift caused by environmental factors such as temperature changes, thus interfering with the accuracy of the analysis results.

[0005] In summary, how to simultaneously collect dynamic excitation and micro-displacement response data and effectively decouple the physical relationship between impact vibration and micro-displacement in order to accurately monitor the structural health status is a technical problem that urgently needs to be solved by those skilled in the art. Summary of the Invention

[0006] In a first aspect, the present invention provides a method for decoupling monitoring of micro-displacement and vibration at a downhole reference point, comprising the following steps: S1. Monitor the piezoelectric analog output signal of the downhole reference point. When the piezoelectric analog output signal exceeds the preset trigger level, a trigger signal is generated. Based on the trigger signal, the three-dimensional micro-displacement response data stream and vibration excitation data stream under the same clock source are synchronously acquired. S2. Process the vibration excitation data stream to calculate the short-time energy envelope, and define the coherent excitation window based on the short-time energy envelope; S3. Integrate the three-dimensional micro-displacement response data stream and vibration excitation data stream using the coherent excitation window as the interval, and calculate and generate the energy transfer efficiency parameter characterizing the motion conversion efficiency. S4. Extract the free oscillation signal of the three-dimensional micro-displacement response data stream based on the boundary information of the coherent excitation window, and calculate the subsequent residual oscillation energy parameters based on the free oscillation signal. S5. The energy transfer efficiency parameter is fused with the subsequent residual oscillation energy parameter to construct the structural dynamic response feature vector corresponding to a single excitation event, and the structural dynamic response feature vector is stored in the historical diagnostic database. S6. Extract multi-phase structural dynamic response feature vectors from the historical diagnostic database to assess the evolution trend, and output the baseline structural health deterioration warning result based on the assessment results.

[0007] A further aspect of this invention involves acquiring a three-dimensional micro-displacement response data stream and a vibration excitation data stream under the same clock source, comprising the following steps: When the piezoelectric analog output signal is detected to momentarily exceed the preset hardware trigger level, a hardware interrupt signal is generated to synchronously start the first analog-to-digital converter and the second analog-to-digital converter that share the same high-precision clock source. A three-dimensional micro-displacement response data stream is generated by capturing three-dimensional signals at a preset sampling rate using a first analog-to-digital converter, and a synchronous vibration excitation data stream is generated at the same sampling rate using a second analog-to-digital converter.

[0008] A further aspect of the present invention defines a coherent excitation window based on a short-time energy envelope, comprising the following steps: A frame-by-frame sliding window calculation is performed on the vibration excitation data stream to extract energy features within the bandpass filter frequency band and generate a continuous short-time energy envelope. The time stamp at which the energy in the short-time energy envelope first exceeds a preset multiple of the background noise energy benchmark is locked as the excitation start point.

[0009] A further aspect of the present invention, which defines the coherent excitation window based on a short-time energy envelope, includes the following steps: After the short-term energy envelope reaches its peak, the time stamp when the energy value first decays to below a preset percentage of that peak value is marked as the excitation main lobe end point; The time interval between the excitation start point and the excitation main lobe end point is defined as the coherent excitation window.

[0010] A further aspect of the present invention involves calculating and generating energy transfer efficiency parameters characterizing motion conversion efficiency, including the following steps: Using the excitation start point and excitation main lobe end point of the coherent excitation window as the upper and lower limits of integration, the total energy of the response stage is calculated by full-window amplitude integration of the three-dimensional micro-displacement response data stream; The same coherent excitation window is used to perform amplitude integration, thereby extracting the total energy of the excitation stage from the vibration excitation data stream; The energy transfer efficiency parameter is generated by calculating the ratio of the total energy in the response phase to the total energy in the excitation phase.

[0011] A further aspect of the present invention involves extracting the free oscillation signal from the three-dimensional micro-displacement response data stream, comprising the following steps: Starting from the end point of the main lobe of the coherent excitation window and extending it by a preset time along the time axis, a residual oscillation time window representing the free oscillation phase is defined. Extract the time-series sampling signal of the three-dimensional micro-displacement response data stream within the residual oscillation time window; A high-pass filter is used to remove the low-frequency baseline drift component from the extracted time-series sampled signal to output the target free oscillation signal.

[0012] A further aspect of this invention involves constructing a structural dynamic response feature vector corresponding to a single excitation event and storing the structural dynamic response feature vector in a historical diagnostic database, comprising the following steps: Extreme values ​​were extracted from the historical diagnostic database, and the energy transfer efficiency parameter and the subsequent residual oscillation energy parameter were normalized to eliminate dimensional differences. The normalized two features are concatenated to generate a two-dimensional dynamic response feature vector corresponding to the current event. The structural dynamic response feature vector is bound to the allocation event number containing timestamps and stored in the historical diagnostic database of the memory.

[0013] A further aspect of this invention involves extracting multi-period structural dynamic response feature vectors from a historical diagnostic database to assess evolutionary trends, including the following steps: Read the structural dynamic response feature vectors from multiple consecutive event windows in the historical diagnostic database according to the time series. The vectors of each period are separated and the rising slope of the normalized energy transfer efficiency parameter of each period is calculated using the least squares regression method. The distribution variance of the normalized subsequent residual oscillation energy parameters within the same window is calculated.

[0014] A further aspect of this invention involves outputting a baseline structural health degradation early warning result based on the evaluation results, including the following steps: Input the calculated ascent slope and distribution variance indices into the preset structural stability determination model; When the upward slope is greater than the first safety threshold and the distribution variance is greater than the second safety threshold, it is determined to be in an unstable state. Send baseline structural health deterioration warning results with microcrack development classification labels to the monitoring level, indicating that the structure is becoming unstable.

[0015] Secondly, the present invention provides a downhole reference point micro-displacement and vibration decoupling monitoring system, comprising the following modules: The synchronous data acquisition module monitors the piezoelectric analog output signal of the downhole reference point. When the piezoelectric analog output signal exceeds the preset trigger level, a trigger signal is generated. Based on the trigger signal, the three-dimensional micro-displacement response data stream and vibration excitation data stream under the same clock source are synchronously acquired. The excitation window definition module processes the vibration excitation data stream to calculate the short-time energy envelope and defines the coherent excitation window based on the short-time energy envelope. The energy transfer efficiency calculation module integrates the three-dimensional micro-displacement response data stream and vibration excitation data stream with the coherent excitation window as the interval, and calculates and generates energy transfer efficiency parameters that characterize motion conversion efficiency. The residual oscillation feature extraction module extracts the free oscillation signal of the three-dimensional micro-displacement response data stream based on the boundary information of the coherent excitation window, and calculates the subsequent residual oscillation energy parameters based on the free oscillation signal. The dynamic response feature vector construction module integrates the energy transfer efficiency parameter and the subsequent residual oscillation energy parameter to construct the structural dynamic response feature vector corresponding to a single excitation event, and stores the structural dynamic response feature vector in the historical diagnostic database. The time-series trend assessment and early warning module extracts multi-period structural dynamic response feature vectors from the historical diagnostic database to assess the evolution trend, and outputs early warning results of structural health deterioration at the baseline point based on the assessment results.

[0016] In summary, the present invention has the following beneficial technical effects: 1. A nanosecond-level synchronous triggering mechanism based on a hardware comparator is adopted. When the output signal of the vibration sensing unit exceeds the preset hardware trigger level, the first and second analog-to-digital converters, sharing the same high-precision clock source, are synchronously activated. This reduces the time delay and uncertainty introduced by software polling or independent triggering systems, ensuring the alignment of the acquired three-dimensional micro-displacement response data stream and the vibration excitation data stream on the time base. This provides underlying data support with consistent timestamps for subsequent accurate analysis of the physical causal relationship between dynamic excitation and structural response.

[0017] 2. By performing short-time energy envelope calculations on the synchronously acquired vibration excitation data stream, a coherent excitation window characterizing the physical process of the actual excitation event is dynamically defined. Using this time window as a unified integration interval, the total energy of the response stage of the three-dimensional micro-displacement and the total energy of the excitation stage of vibration are calculated separately, and their ratio is obtained. By placing the external energy input and the structural response energy output within the same time domain determined by the physical characteristics of the event itself for correlation analysis, this overcomes the limitations of traditional methods that rely solely on isolated peak values ​​of displacement or vibration, and provides an energy transfer efficiency parameter that can objectively quantify the structural motion conversion efficiency.

[0018] 3. After analyzing the main lobe excitation process, an independent residual oscillation time window was constructed, starting from the end point of the main lobe excitation. By extracting the micro-displacement response data within this window and performing high-pass filtering, the free oscillation stage of the structure after the forced vibration ended was separated, and the low-frequency baseline drift component caused by non-mechanically varying factors was removed. The subsequent residual oscillation energy parameters calculated in this way can objectively reflect the internal damping characteristics of the benchmark anchoring system, providing an evaluation index that is less affected by environmentally varying factors for assessing the structure's ability to dissipate residual energy.

[0019] 4. The extracted single-event energy transfer efficiency parameters and subsequent residual oscillation energy parameters are fused into a structural dynamic response feature vector, and a historical diagnostic database is introduced to perform time-series trend analysis on the feature vectors across multiple periods. Evolutionary indices such as the rise slope and distribution variance of the parameters are extracted, extending the state assessment system from a static snapshot based on a single event to a dynamic evolutionary analysis based on continuous events. When the evolutionary trend indicators trigger a preset safety threshold, the system can output a baseline structural health deterioration warning and support providing auxiliary classification labels for the causes of deterioration based on the combined change patterns of the feature vectors, which is beneficial for early monitoring of the structural instability evolution process. Attached Figure Description

[0020] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the accompanying drawings used in the description of the embodiments or the prior art will be briefly introduced below. The drawings are used to provide a further understanding of the present invention.

[0021] Figure 1 A flowchart illustrating an embodiment of this application is disclosed.

[0022] Figure 2 A schematic diagram of the module architecture in the embodiments of this application is disclosed.

[0023] Figure 3 The present application discloses characteristic curves of the short-time energy envelope of vibration excitation and the definition of the coherent excitation window in the embodiments thereof.

[0024] Figure 4 The structural health evolution trend and instability determination curves in the embodiments of this application are disclosed. Detailed Implementation

[0025] The following is in conjunction with the appendix Figure 1 - Figure 4 A preferred description of the present invention is provided below.

[0026] See attached document Figure 1 This invention proposes a method for decoupling monitoring of micro-displacement and vibration at downhole reference points, comprising the following steps: S1. Monitor the piezoelectric analog output signal of the downhole reference point. When the piezoelectric analog output signal exceeds the preset trigger level, a trigger signal is generated. Based on the trigger signal, the three-dimensional micro-displacement response data stream and vibration excitation data stream under the same clock source are synchronously acquired. S2. Process the vibration excitation data stream to calculate the short-time energy envelope, and define the coherent excitation window based on the short-time energy envelope; S3. Integrate the three-dimensional micro-displacement response data stream and vibration excitation data stream using the coherent excitation window as the interval, and calculate and generate the energy transfer efficiency parameter characterizing the motion conversion efficiency. S4. Extract the free oscillation signal of the three-dimensional micro-displacement response data stream based on the boundary information of the coherent excitation window, and calculate the subsequent residual oscillation energy parameters based on the free oscillation signal. S5. The energy transfer efficiency parameter is fused with the subsequent residual oscillation energy parameter to construct the structural dynamic response feature vector corresponding to a single excitation event, and the structural dynamic response feature vector is stored in the historical diagnostic database. S6. Extract multi-phase structural dynamic response feature vectors from the historical diagnostic database to assess the evolution trend, and output the baseline structural health deterioration warning result based on the assessment results.

[0027] In one embodiment of the present invention, step S1 includes the following steps: When the piezoelectric analog output signal is detected to momentarily exceed the preset hardware trigger level, a hardware interrupt signal is generated to synchronously start the first analog-to-digital converter and the second analog-to-digital converter that share the same high-precision clock source. A three-dimensional micro-displacement response data stream is generated by capturing three-dimensional signals at a preset sampling rate using a first analog-to-digital converter, and a synchronous vibration excitation data stream is generated at the same sampling rate using a second analog-to-digital converter.

[0028] Specifically, the first step of the method of this invention is to capture the original dynamic response of the downhole reference point when it is subjected to external excitation through an event-driven hardware synchronization mechanism. In a specific implementation scenario, this monitoring method is executed by a monitoring terminal deployed near the downhole reference point. The monitoring terminal integrates a control unit, a three-dimensional micro-displacement sensing unit, a vibration sensing unit, and corresponding signal processing circuits.

[0029] In system standby mode, the control unit puts most circuits into sleep mode, keeping only a hardware trigger circuit based on a low-power comparator active. The positive input of the low-power comparator continuously receives the piezoelectric analog output signal from the vibration sensing unit, while its inverting input is connected to a stable reference voltage generated by a voltage divider network, which is the preset hardware trigger level.

[0030] The preset hardware trigger level is a specific voltage value, for example, set between 50 mV and 200 mV. This value is set based on statistical analysis of the long-term background noise of the target monitoring environment, and is usually taken as 1.5 to 2 times the peak-to-peak value of the background noise signal, in order to strike a balance between effectively filtering out environmental noise interference and ensuring sensitivity to real excitation events.

[0031] When a vibration event such as blasting occurs underground, if the instantaneous voltage amplitude of the piezoelectric analog output signal from the vibration sensing unit exceeds the preset hardware trigger level, the comparator's output logic level will undergo a rapid nanosecond-level flip, thereby generating a hardware interrupt signal with a steep edge.

[0032] The aforementioned nanosecond-level hardware synchronization triggering mechanism refers to using a hardware comparator circuit to achieve signal threshold detection. Its response delay is typically in the range of 10 ns to 100 ns. This response delay range is set to balance system power consumption and synchronization accuracy. If the delay is less than 10 ns, a high-speed comparator with a static current in the milliampere range is usually required, which violates the extremely low power consumption requirements of downhole battery-powered equipment during long-term dormancy. If the delay is greater than 100 ns, it will cause a non-negligible phase error between the displacement and vibration signals when capturing high-frequency transient oscillation signals such as blast shock waves, thus affecting the accuracy of subsequent energy integration. 10 ns to 100 ns is an engineering selection range that balances microwatt-level standby power consumption and microsecond-level high-frequency synchronous sampling, thereby achieving the capture of the starting point of the excitation event.

[0033] A hardware interrupt signal is fed to the synchronization trigger pins of the first and second analog-to-digital converters (ADCs). Here, both the first and second ADCs can be 24-bit delta-sigma type ADCs, whose high-resolution characteristics can capture subtle signal variations. The first and second ADCs share a sampling clock provided by the same high-precision clock source, ensuring consistency in their time base. The high-precision clock source can be a temperature-compensated crystal oscillator (TCXO), with a frequency stability of ±2 ppm, providing a unified and stable sampling clock reference for the ADCs.

[0034] Upon receiving a hardware interrupt signal, both analog-to-digital converters (ADCs) are simultaneously awakened and transition from sleep to active status, beginning to perform analog-to-digital conversion on their respective input signals. The first ADC continuously samples the analog signals in the X, Y, and Z orthogonal directions output by the three-dimensional micro-displacement sensing unit at a preset sampling rate, generating a data sequence that constitutes the three-dimensional micro-displacement response data stream.

[0035] The preset high sampling rate is a preset sampling frequency value, for example, set in the range of 5 kHz to 50 kHz. Its selection must follow the Nyquist theorem, ensuring that this frequency is at least twice the highest frequency component of the effective signal in the vibration excitation data stream to avoid signal aliasing. Furthermore, the three-dimensional micro-displacement response data stream is a time-series dataset composed of multiple data points. Each data point contains a timestamp and its corresponding three-dimensional spatial displacement coordinates (x, y, z), with the physical unit being mm.

[0036] Simultaneously, a second analog-to-digital converter samples the output signal of the vibration sensing unit at the same preset high sampling rate, generating a time-synchronized vibration excitation data stream. This vibration excitation data stream is also a time-series dataset, with each data point containing a timestamp and its corresponding vibration signal amplitude, typically in mV or associated with acceleration g. This step ensures that the displacement and vibration data used in subsequent analysis originate from the same triggering event and have an unbiased causal relationship in the time dimension.

[0037] For example, in one embodiment, the monitoring system is deployed at a reference point on the roof of the roadway where stability needs to be monitored. The preset hardware trigger level threshold is set to 100 mV. This setting is based on previous field measurements, which showed that the peak value of the background micro-vibration noise of the roadway roof under normal operating conditions (without blasting or other abnormal operations) is typically between 30 mV and 40 mV. 100 mV is approximately 2.5 times the peak value of the background noise. This effectively avoids random false triggering caused by routine production activities such as the operation of coal mining machines and the passage of heavy vehicles, while also ensuring a high sensitivity response to the first arrival wave of far-field blasting shock waves reaching the reference point.

[0038] The preset high sampling rate of 10 kHz is based on the fact that when the shock wave from an underground blast in a coal mine is transmitted through the rock mass to the anchoring reference point, the characteristic frequency band of the high-frequency effective structural response induced by it is mainly concentrated between 1000 Hz and 2000 Hz. According to the Nyquist-Shannon sampling theorem, a sampling rate of 10 kHz not only far exceeds twice the highest frequency of the effective signal, leaving sufficient anti-aliasing margin, but also can reconstruct the main lobe shape of the subsequent short-time energy envelope with high precision at a time resolution of 0.1 ms, meeting the transient high-frequency energy integration requirements after nanosecond-level triggering.

[0039] Shock waves from blasting operations at a distant location underground are transmitted to a reference point and, at a certain triggering moment... The piezoelectric analog output signal voltage from the vibration sensing unit rapidly jumps from 20 mV (background noise level) to 105 mV. Since 105 mV exceeds the preset hardware trigger level threshold of 100 mV, the comparator in the hardware trigger circuit flips, generating a hardware interrupt signal.

[0040] An interrupt signal is simultaneously sent to the trigger pins of both the first and second analog-to-digital converters (ADCs) within nanoseconds. Both are then activated and begin synchronous sampling at a frequency of 10 kHz. The first ADC begins acquiring signals from the three-dimensional micro-displacement sensor. The three-dimensional displacement data generated by the first sampling point after a certain time may be (0.001, -0.002, 0.001) mm, and the displacement data generated by the second sampling point after 100 μs becomes (0.025, -0.031, 0.018) mm. These continuous data points constitute the three-dimensional micro-displacement response data stream.

[0041] Simultaneously, a second analog-to-digital converter acquires the vibration signal. The amplitude obtained at the first sampling point corresponds to 105 mV, while the amplitude obtained at the second sampling point may have jumped to 350 mV. These continuous amplitude data constitute the vibration excitation data stream. In this way, two data streams synchronized in terms of time start and sampling interval are obtained, laying the foundation for subsequent decoupling analysis.

[0042] In one embodiment of the present invention, step S2 includes the following steps: A frame-by-frame sliding window calculation is performed on the vibration excitation data stream to extract energy features within the bandpass filter frequency band and generate a continuous short-time energy envelope. The time stamp at which the energy in the short-time energy envelope first exceeds a preset multiple of the background noise energy benchmark is locked as the excitation start point.

[0043] Specifically, following the vibration excitation data stream acquired in the previous step, this step processes it to dynamically define the core time interval of an excitation event. This process is executed by the control unit of the monitoring terminal, which performs frame-by-frame sliding window energy calculation on the vibration excitation data stream.

[0044] The control unit divides the vibration excitation data stream with a preset frame length N and frame shift M. For each data frame, a digital bandpass filter with a preset frequency band is applied to process it in order to filter out low-frequency geological noise and high-frequency electromagnetic interference that are not related to the structural response, while retaining the frequency components of the core excitation signal.

[0045] The preset frequency band is set according to the underground environment and the characteristics of the excitation source. For example, for blasting excitation, it can be set from 50Hz to 1500Hz. The upper and lower limits of this preset frequency band have clear physical directionality. The lower limit is set at 50Hz to effectively avoid low-frequency slow displacement caused by natural geological creep, which is common in the underground environment, as well as 50Hz electromagnetic interference from power frequency AC equipment. The upper limit is set at 1500Hz because the main frequency energy of typical blasting shock waves or rock fracture micro-vibrations in coal mines is concentrated below 1000Hz. The upper limit of 1500Hz can completely envelop the target excitation signal and filter out local high-frequency resonance noise caused by the sensor housing or mounting bracket. This frequency band covers the main frequency components generated by the explosion. The sum of the squares of the amplitudes of all sampling points in the filtered data frame is calculated and used as the energy value of the frame. The entire calculation process slides along the entire data stream to generate a continuous short-time energy envelope that varies with time. The energy of the k-th frame after filtering. The calculation formula is as follows:

[0046] In the above formula, The summation index represents the local sampling point number within the sliding window; The vibration excitation data stream, after being bandpass filtered, is in the k-th frame. The amplitude of each sampling point, and its corresponding global time point number is N is the frame length of the sliding window used to calculate energy, i.e., the number of sampling points; M is the frame shift, i.e., the sampling point interval between the starting points of two adjacent windows; and k is the frame number. and Values ​​start from 0. This formula calculates the total energy of the filtered signal within the k-th time window. Since energy is the square of the amplitude, its unit is the square of the signal amplitude in units, for example... or Furthermore, the generated short-time energy envelope is a one-dimensional time series, where each element represents the vibrational energy within the corresponding time window.

[0047] After acquiring continuous short-time energy envelopes, the control unit begins dynamically searching for the physical boundaries of the event. The control unit compares the energy value of the short-time energy envelopes with a preset background noise energy benchmark. The preset background noise energy benchmark is the statistical average value of the environmental vibration energy collected by the system when no significant excitation event occurs. This benchmark value can be determined based on the average energy of several frames in the initial stage of the acquired data stream, or obtained through statistical analysis of historical monitoring data.

[0048] When the energy value of a data frame in the energy envelope sequence first exceeds a multiple of the background noise energy benchmark, the control unit locks the timestamp corresponding to that frame and marks it as the excitation start point. It should be understood that the multiple is a dimensionless coefficient. The value typically ranges from 5 to 20. The specific value of this multiple needs to be dynamically adjusted based on the background noise of the monitoring area. In deep return airway tunnels with low background noise and a stable environment, a lower limit of 5 can be used to ensure sensitive detection of weak far-field rupture events. However, near coal mining faces with continuous mechanical operations such as large coal mining machines and scraper conveyors, where background random vibrations are strong, an upper limit of 20 is required to prevent the random noise peaks generated by mechanical operations from causing false triggering of the system. This setting ensures that the detected event is higher than the background noise, avoiding misjudging random noise fluctuations as excitation events. The locked excitation starting point... It is a specific timestamp.

[0049] The control unit searches for the energy peak and its corresponding time point within the entire short-time energy envelope, and then continues traversing the energy envelope backward along the time axis from that peak point. When the energy value is detected to first decay below a preset percentage of the peak value, the timestamp at that moment is marked as the end point of the excitation main lobe. Similarly, the preset percentage of the peak value is another dimensionless coefficient. The value range is usually between 0.1 and 0.2. The setting of this ratio is based on the energy dissipation characteristics in structural dynamics. When the vibration energy of the input system decays to 10% to 20% of the peak value, which is about 7 dB to 10 dB, it means that the work process of external forced excitation, such as blast shock wave, has basically ended, and the reference point structure begins to transition from the forced vibration stage to the free decay oscillation stage.

[0050] This coefficient As the end point of the excitation main lobe, it can physically separate the external input energy from the energy dissipated by the structure's own damping in the time domain. This parameter is used to define the main release stage of the excitation energy, excluding the wake portion containing complex reflections and whose energy has almost completely decayed from the main analysis window; excitation main lobe end point It is also a specific timestamp.

[0051] The time interval between the excitation start point and the excitation main lobe end point is defined as the coherent excitation window, which can characterize the main physical processes of the real excitation event; that is, the closed interval defined by these two timestamps. The start and end timestamps of this time window are stored for later use. In subsequent comparison operations, whether comparing with the background noise energy benchmark or with the peak energy, the comparisons are made between physical quantities of the same dimensions.

[0052] For example, the control unit in the foregoing embodiments processes... The vibration excitation data stream is collected starting from a specific time. Assuming a sliding window frame length N of 256 sampling points, a frame shift M of 128 sampling points, and a sampling rate of 10 kHz, each window duration is 25.6 ms, and the overlap between windows is 12.8 ms. The control unit first calculates the preset background noise energy benchmark of 450 from the first 1000 sampling points. The set multiplier is 10, and the peak preset ratio is 0.1.

[0053] The control unit begins calculating the short-time energy envelope frame by frame. Assuming the energy value is 4600 in frame 5... Because 4600 >10×450 The control unit will set the start timestamp of the 5th frame, i.e. +(5-1)×12.8ms= +51.2 ms, marked as the excitation start point. .

[0054] The control unit continued calculating and found that the energy peaked at 85,000 in the 12th frame. Searching backwards from frame 12, it was found that the energy decayed to 8450 at frame 27. Since 8450 < 0.1 × 85000 = 8500, the control unit will set the end timestamp of the 27th frame, i.e. +(27-1)×12.8 ms+25.6 ms= +358.4 ms, marked as the end point of the excitation main lobe. Therefore, the coherent excitation window generated in this embodiment is a time interval []. +51.2 ms, +358.4ms).

[0055] See Figure 3 The dynamic definition process of the excitation window is explained. The solid black line in the figure represents the short-time energy envelope curve extracted through frame-by-frame sliding window calculation; the horizontal dashed line below represents the preset background noise level threshold of 4500. The horizontal dotted line above represents the peak ratio threshold of 8500. As can be seen from the figure, when the short-time energy envelope first breaks through the background noise level, the system accurately locks the excitation start point. The value is set to 0, corresponding to 51.2 ms. Subsequently, the envelope curve reaches its peak, and when it first falls below the peak proportion threshold during the descent phase, the system truncates and locks the end point of the excitation main lobe, corresponding to 358.4 ms. The gray shaded area between these two key time stamps is the coherent excitation window dynamically defined by the system. This window strips away the main lobe of the effective physical impact, eliminates interference from early environmental random noise and later invalid tail waves, and provides a time-domain anchor point for subsequent energy integration.

[0056] In one embodiment of the present invention, step S3 includes the following steps: Using the excitation start point and excitation main lobe end point of the coherent excitation window as the upper and lower limits of integration, the total energy of the response stage is calculated by full-window amplitude integration of the three-dimensional micro-displacement response data stream; The same coherent excitation window is used to perform amplitude integration, thereby extracting the total energy of the excitation stage from the vibration excitation data stream; The energy transfer efficiency parameter is generated by calculating the ratio of the total energy in the response phase to the total energy in the excitation phase.

[0057] Specifically, after defining the coherent excitation window, the core step of the method of this invention lies in using this time window as the sole analysis anchor point to synchronously analyze the synchronously captured response and excitation data streams in order to calculate the key parameters characterizing the dynamic transmission properties of the structure. This step is still automatically executed by the control unit of the monitoring terminal. The control unit reads the start and end timestamps of the coherent excitation window stored in the previous step, i.e., the excitation start point and the excitation main lobe end point.

[0058] The control unit processes the three-dimensional micro-displacement response data stream using these two timestamps as strict upper and lower limits for integration. The control unit iterates through all sampling points within the time window of the three-dimensional micro-displacement response data stream, calculating the square of the transient amplitude of the three-dimensional displacement vector at each sampling point, i.e., summing the squares of the X, Y, and Z components. Then, this sum of squares is accumulated across all sampling points within the window to obtain a scalar value, which is defined as the total energy of the response stage. The total energy of the response stage is then calculated. The specific formula is:

[0059] Among them, the total energy of the response phase It is a cumulative scalar value, with the dimension of the square of the length, for example... It quantifies the total deformation of the benchmark structure during the core incentive period; This is a summation index representing the global discrete sampling point index; , and These are the three-dimensional micro-displacement response data streams at the 1st... The three orthogonal component values ​​of each sampling point; and These are the starting points of the incentive. and the end point of the main lobe The corresponding discrete sampling point number. Considering the data addressing of the underlying hardware, the calculation formula here is: and ,in, This is the initial timestamp triggered by the hardware interrupt signal, i.e., the time of the 0th sampling point; Sampling rate; This is a function that takes the integer part downwards.

[0060] The control unit uses the same coherent excitation window as the integration interval and performs similar operations on the vibration excitation data stream. It accumulates the squares of the amplitudes of all vibration signal sampling points within this time window to calculate and extract the total energy of the excitation phase. The specific formula is:

[0061] Among them, the total energy of the incentive phase It is also a cumulative scalar value, the dimension of which is the square of the signal amplitude, for example... This quantifies the total vibrational energy applied to the reference point; The preprocessed vibration excitation data stream is in the first... The amplitude of each sampling point.

[0062] The control unit generates a specific composite dimension by calculating the ratio of the total energy in the response phase to the total energy in the excitation phase, such as... The value of this value is the energy transfer efficiency parameter. The calculation formula is:

[0063] Among them, the energy transfer efficiency parameter It is a key structural health indicator, which, in a physical sense, characterizes the efficiency with which a reference-point anchoring system converts the energy of external vibration excitation into its own macroscopic displacement motion. Parameters Although not a purely dimensionless efficiency value, as a composite dimensional parameter with stable physical meaning, its relative variation trend among different excitation events provides a reliable quantitative basis for assessing the evolution of the structural health status at the benchmark point. Its baseline value is related to the initial state of the benchmark point installation. If this parameter shows a continuous upward trend in consecutive monitoring events, it usually indicates that the anchoring system may be loosening or microcracks may be propagating, leading to a larger displacement response in the structure under the same excitation.

[0064] For example, continuing from the previous embodiment, the control unit calls the determined coherent excitation window, the time interval of which is [ +51.2 ms, [+358.4 ms], sampling rate 10 kHz. The starting sampling point number for integration is calculated based on this. ,Right now ), End sampling point number ,Right now .

[0065] The control unit processes the three-dimensional micro-displacement response data stream by summing the squares of the three-dimensional displacement components from the 512th to the 3584th sampling point (3073 data points in total), and assumes that the total energy of the response stage is calculated. The control unit performs cumulative calculations of the squared amplitudes of the vibration excitation data stream within the same sampling point interval [512, 3584], assuming the total energy of the excitation phase is obtained. The energy transfer efficiency parameter is calculated by dividing the two values ​​to obtain the result. This value will be recorded and used for subsequent trend analysis and status assessment.

[0066] In one embodiment of the present invention, step S4 includes the following steps: Starting from the end point of the main lobe of the coherent excitation window and extending it by a preset time along the time axis, a residual oscillation time window representing the free oscillation phase is defined. Extract the time-series sampling signal of the three-dimensional micro-displacement response data stream within the residual oscillation time window; A high-pass filter is used to remove the low-frequency baseline drift component from the extracted time-series sampled signal to output the target free oscillation signal.

[0067] Specifically, after calculating the energy transfer efficiency of the core excitation and response stages in the preceding steps, this method further extracts subsequent features of the three-dimensional micro-displacement response data stream based on the boundary information of the coherent excitation window, which is used to quantify the damping and stability of the system. This step is executed by the control unit of the monitoring terminal.

[0068] The control unit reads and locks the end point of the main lobe of the coherent excitation window determined in the preceding steps. Starting from this point in time, the control unit extends backward along the time axis by a fixed preset duration, thereby constructing a residual oscillation time window at the software level to characterize the free oscillation phase of the reference point system after the forced excitation ends.

[0069] The preset duration is set based on the typical decay time of the target reference point structure, usually between 200 ms and 1 s. The specific value of this preset duration is strongly related to the lithology of the surrounding rock and the stiffness of the anchoring medium. For reference points with hard tops and tight anchoring, the structural damping is large and the free oscillation decays quickly, so the preset duration should be between 200 ms and 500 ms. For reference points with fractured soft rock or initial loosening of the anchoring system, the ringing effect is obvious and the energy dissipation is slow, so the preset duration should be extended to 500 ms to 1 s to ensure that all residual energy dissipated by structural damping can be fully integrated and recorded. Therefore, the residual oscillation time window is a time interval of fixed length starting from the end point of the excitation main lobe.

[0070] The control unit extracts all time-series sampled signals whose timestamps fall within the newly established residual oscillation time window from the three-dimensional micro-displacement response data stream. Since this signal may contain slow quasi-static drift caused by temperature changes or sensor aging, the control unit processes the extracted time-series sampled signals using a digital high-pass filter to ensure that only the oscillation components determined by the structural dynamics are analyzed.

[0071] The high-pass filter can be implemented using a second-order or higher-order Butterworth or Chebyshev infinite impulse response (IIR) digital filter with a low cutoff frequency, for example, in the range of 0.5 Hz to 2 Hz. This setting is based on the fact that the frequency of free oscillation of the structure is typically much higher than 2 Hz, while the baseline drift frequency caused by factors such as temperature is much lower than 0.5 Hz. This filter can effectively separate the two. The filter aims to remove the extremely low-frequency baseline drift component from the signal, i.e., the slow changes in displacement signal caused by non-vibration factors, while retaining the higher-frequency residual vibration information related to the structure's natural frequency, thus obtaining a segment of free oscillation signal.

[0072] The control unit calculates the energy integral of the time-series sampled signal after high-pass filtering. Specifically, it accumulates the squares of the transient amplitudes of the three-dimensional displacement vectors at all sampling points within the signal segment; the result is defined as the subsequent residual oscillation energy parameter. The specific calculation formula is as follows:

[0073] In the formula, It inherits and stimulates the end point of the main lobe from the preceding steps. The corresponding discrete sampling point number; It is the sampling point number corresponding to the end point of the residual oscillation time window, and its value is determined by... The number of sampling points within the preset time period is determined in addition to this. , and These represent the three-dimensional micro-displacement response data streams after high-pass filtering at the [number]th [timeframe]. The formula quantifies the total vibrational energy of a given sampling point by calculating the sum of squares of the displacement amplitudes during the free oscillation phase. Its dimension is the square of length, for example... This is consistent with the total energy dimension of the response stage calculated in the aforementioned steps, ensuring that subsequent numerical comparisons and fusion analyses can be performed.

[0074] Furthermore, it should be noted that the subsequent residual oscillation energy parameter It is a scalar value. The subsequent residual oscillation energy parameter can effectively quantify the ability of a reference point to recover to a stable state after being subjected to an impact. It is a direct indicator for evaluating its internal damping characteristics and the stability of the anchoring system. The magnitude of the value is inversely proportional to the damping ratio of the reference point anchoring system. A system with good damping and structural stability... The value is low.

[0075] For example, continuing with the previous embodiment, the control unit obtains the excitation main lobe end point as... +358.4ms, the corresponding sampling point number is The preset duration is set to 300 ms. Therefore, the residual oscillation time window is [ +358.4ms, [+658.4 ms], at a sampling rate of 10 kHz, the corresponding sampling point interval is [3584, 6584]. The control unit extracts these 3001 data points from the three-dimensional micro-displacement response data stream.

[0076] The control unit uses a second-order Butterworth high-pass filter with a cutoff frequency of 1 Hz to process this signal in order to remove the slow drift of about 0.1 Hz that may be caused by the slight change in ambient temperature after the explosion.

[0077] After filtering, the control unit performs a cumulative calculation of the squared transient amplitudes of the 3001 processed three-dimensional displacement data points. Assuming the final obtained residual oscillation energy parameter is... 。 This value reflects the intensity of the aftershocks at the reference point after this excitation event, and will be stored together with the energy transfer efficiency parameter to construct the feature vector of this event.

[0078] In one embodiment of the present invention, step S5 includes the following steps: Extreme values ​​were extracted from the historical diagnostic database, and the energy transfer efficiency parameter and the subsequent residual oscillation energy parameter were normalized to eliminate dimensional differences. The normalized two features are concatenated to generate a two-dimensional dynamic response feature vector corresponding to the current event. The structural dynamic response feature vector is bound to the allocation event number containing timestamps and stored in the historical diagnostic database of the memory.

[0079] Specifically, after calculating the energy transfer efficiency parameter and the subsequent residual oscillation energy parameter, which characterize different physical properties, in the previous steps, this step fuses these two heterogeneous parameters to construct a unified feature vector that can comprehensively characterize the dynamic response of the reference point structure under a single explosion excitation event. This process is executed by the control unit of the monitoring terminal.

[0080] Because the energy transfer efficiency parameter and the subsequent residual oscillation energy parameter have different physical dimensions and numerical ranges, the control unit first normalizes these two parameters separately to eliminate this difference and ensure equal weighting in subsequent model analysis. The normalization process uses a min-max normalization method, linearly mapping the original value of each parameter to the interval [0,1]. If the calculated parameter value for the current event exceeds the historical extreme value range, the corresponding extreme value in the historical diagnostic database is dynamically updated using the current value, or a truncation function is introduced to clamp the normalization result to the upper limit of 1, and a higher anomaly weight is assigned to this feature vector to accommodate data mutations caused by severe structural deterioration. Specifically, the normalization calculation formula for the energy transfer efficiency parameter is:

[0081] The normalized formula for calculating the energy parameter of the subsequent residual oscillation is:

[0082] In the above formula, and These are the normalized parameter values. and These are the historical minimum and maximum values ​​of the energy transfer efficiency parameter. and These are the historical minimum and maximum values ​​of the subsequent residual oscillation energy parameter. These extreme values ​​are dynamically queried and updated from the established historical diagnostic database. Through this method, parameters with different physical dimensions are unified into the same [0,1] numerical space, the dimensional differences are eliminated, and the effectiveness of subsequent analysis is guaranteed.

[0083] After normalization, the control unit performs feature concatenation of the normalized energy transfer efficiency parameter and the normalized subsequent residual oscillation energy parameter in a predetermined order. This operation is represented in the data structure as combining two scalar values ​​into a two-dimensional vector, which is defined as the structural dynamic response feature vector.

[0084] The structural dynamic response feature vector is a vector containing information in multiple dimensions; in this embodiment, it is a two-dimensional vector. superscript in the formula This represents the transpose operation of a matrix or vector. To ensure the traceability of data for each event, the control unit assigns a unique event number to the currently triggered event. This event number is an integer used to uniquely identify each stimulus event, such as an incrementing counter, and obtains a high-precision timestamp of the event from the hardware clock module.

[0085] The control unit uses the generated event number to bind the structural dynamic response feature vector with the corresponding high-precision timestamp, forming a complete event record. This event record is then written as a new entry into the historical diagnostic database of the non-volatile memory configured within the monitoring terminal for persistent storage, to be used for subsequent time-series analysis and trend assessment.

[0086] Non-volatile memory can be onboard Flash memory chips or electrically erasable programmable read-only memory (EEPROM), which can ensure that data is not lost after the device is powered off; while the historical diagnostic database is a structured data storage area, which can be organized into a table, with each row representing an event, and containing at least the following fields: event number, timestamp, normalized energy transfer efficiency, and normalized subsequent residual oscillation energy.

[0087] For example, following the previous steps, the control unit obtains the energy transfer efficiency parameters calculated for this event. and subsequent residual oscillation energy parameters Suppose we retrieve historical extreme values ​​from the historical diagnostic database. , , as well as .

[0088] Control unit Perform normalization calculations: .

[0089] right Perform normalization calculations: The control unit concatenates these two normalized values ​​to generate the structural dynamic response feature vector for this event. 。The system assigns a unique event number to this event, such as 1034, and obtains the high-precision timestamp of the trigger, i.e. A line containing (1034, The complete event record of [0.5, 0.791] is written to the database in non-volatile memory, awaiting subsequent calls.

[0090] In one embodiment of the present invention, step S6 includes the following steps: Input the calculated ascent slope and distribution variance indices into the preset structural stability determination model; When the upward slope is greater than the first safety threshold and the distribution variance is greater than the second safety threshold, it is determined to be in an unstable state. Send baseline structural health deterioration warning results with microcrack development classification labels to the monitoring level, indicating that the structure is becoming unstable.

[0091] Specifically, after accumulating structural dynamic response feature vectors from multiple excitation events in the historical diagnostic database, the final step of the method in this invention is to assess the structural health evolution trend of the baseline point based on these time-series data and output early warning results with clear classification labels. This analysis process can be executed periodically by the control unit of the monitoring terminal, or by the host computer software after centralized data reading.

[0092] The analysis module reads consecutive data from the historical diagnostic database in chronological order. The structural dynamic response feature vector of the periodic event. The data from multiple periods form a sliding window for time series analysis. Here, the multi-period time series parameters, i.e., continuous... The data for periodic events refers to a series of time-ordered structural dynamic response feature vectors extracted from historical diagnostic databases.

[0093] For this window The analysis module extracts the energy transfer efficiency parameters for each period from the eigenvectors and uses the least squares method to analyze these parameters. Linear regression analysis was performed on the parameter values ​​as a function of time or event sequence.

[0094] This analysis allows us to calculate the slope of the fitted line, which represents the rising slope of the energy transfer efficiency parameter. The rate of change of this parameter was quantified, quantitatively describing the trend of energy transfer efficiency changes during recent consecutive excitation events. Simultaneously, the analysis module also extracted data within the same window. The subsequent residual oscillation energy parameters of the periodic event are calculated. The variance of the distribution of each parameter value. It describes the degree of dispersion of the parameter within the statistical window, and the variance value quantifies the instability or volatility of the structural damping characteristics in the near term.

[0095] The analysis module inputs the calculated upward slope and distribution variance—two higher-order characteristic indicators—into the preset structural stability determination model. The structural stability determination model is a rule-based decision system, its core being a set of logical judgment conditions.

[0096] The structural stability determination model defines two key safety thresholds: the first safety threshold... Second security threshold First safety threshold Second security threshold These are critical values ​​set based on extensive experimental data, simulation analysis, or expert experience, used to distinguish between normal and abnormal structural states. For example, assuming the application scenario is the reference point for roof anchor bolts in a coal mine roadway, its... and The setting needs to take into account factors such as rock mass properties and anchoring technology.

[0097] The structural stability determination model compares the input ramp rate with a first safety threshold. The comparison is performed, and the variance of the input distribution is compared with the second safety threshold. A comparison is performed. The decision logic detects that the upward slope consistently exceeds the first safety threshold. Furthermore, the distribution variance also exceeded the second safety threshold. If the time is right, then the reference point structure is determined to be in a deterioration stage, transitioning from stability to instability.

[0098] At this point, the system will generate a clear baseline structural health degradation warning result. This result not only includes warning level information, but also includes classification labels indicating the root cause of degradation, such as microcrack development labels.

[0099] The classification label provides a preliminary inference about the cause of degradation. For example, when both energy transfer efficiency and subsequent residual oscillatory energy exhibit unstable growth, the physical mechanism most likely corresponds to the initiation and propagation of microcracks within the anchor body, thus assigning a microcrack development label. The warning results are ultimately output via a wireless communication module or local display interface to alert maintenance personnel and prompt them to take appropriate measures.

[0100] For example, suppose the analysis module sets the analysis window size. For 20 events, combined Figure 4The assessment process for the aforementioned structural health evolution trend is explained. The analysis module retrieved 20 consecutive event records from event numbers 1015 to 1034 from the historical diagnostic database. For the normalized sequence of energy transfer efficiency parameters, its upward slope was calculated using linear regression. The unit is per event. For the normalized sequence of subsequent residual oscillation energy parameters, calculate its distribution variance. .

[0101] The preset first security threshold is The second safety threshold is During the decision-making process, the model found... ,at the same time Since both conditions are met, the reference point structure is determined to be in an unstable state.

[0102] Therefore, a baseline structural health deterioration early warning result, including a Level II warning level and a microcrack development classification label, is generated and reported to the mine safety monitoring center platform via a 4G communication module. Upon receiving the warning, the platform automatically marks the baseline point in red on the electronic map and sends an alert SMS to the relevant responsible personnel, stating: "Structural instability trend detected at baseline point G087, inferred to be caused by microcrack development. Please dispatch personnel for immediate verification."

[0103] See appendix Figure 2 The present invention also proposes a downhole reference point micro-displacement and vibration decoupling monitoring system, comprising the following modules: The synchronous data acquisition module monitors the piezoelectric analog output signal of the downhole reference point. When the piezoelectric analog output signal exceeds the preset trigger level, a trigger signal is generated. Based on the trigger signal, the three-dimensional micro-displacement response data stream and vibration excitation data stream under the same clock source are synchronously acquired. The excitation window definition module processes the vibration excitation data stream to calculate the short-time energy envelope and defines the coherent excitation window based on the short-time energy envelope. The energy transfer efficiency calculation module integrates the three-dimensional micro-displacement response data stream and vibration excitation data stream with the coherent excitation window as the interval, and calculates and generates energy transfer efficiency parameters that characterize motion conversion efficiency. The residual oscillation feature extraction module extracts the free oscillation signal of the three-dimensional micro-displacement response data stream based on the boundary information of the coherent excitation window, and calculates the subsequent residual oscillation energy parameters based on the free oscillation signal. The dynamic response feature vector construction module integrates the energy transfer efficiency parameter and the subsequent residual oscillation energy parameter to construct the structural dynamic response feature vector corresponding to a single excitation event, and stores the structural dynamic response feature vector in the historical diagnostic database. The time-series trend assessment and early warning module extracts multi-period structural dynamic response feature vectors from the historical diagnostic database to assess the evolution trend, and outputs early warning results of structural health deterioration at the baseline point based on the assessment results.

[0104] Each of the modules can be implemented in whole or in part through software, hardware, or a combination thereof. It supports hardware embedded in or independent of the processor in the computer device, and also supports software stored in the memory of the computer device, so that the processor can call and execute the operations corresponding to each of the above modules.

[0105] The above embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention, and should all be included within the protection scope of the present invention.

Claims

1. A method for decoupling monitoring of micro-displacement and vibration at a downhole reference point, characterized in that, Includes the following steps: S1. Monitor the piezoelectric analog output signal of the downhole reference point. When the piezoelectric analog output signal exceeds the preset trigger level, a trigger signal is generated. Based on the trigger signal, the three-dimensional micro-displacement response data stream and vibration excitation data stream under the same clock source are synchronously acquired. S2. Process the vibration excitation data stream to calculate the short-time energy envelope, and define the coherent excitation window based on the short-time energy envelope; S3. Integrate the three-dimensional micro-displacement response data stream and vibration excitation data stream using the coherent excitation window as the interval, and calculate and generate the energy transfer efficiency parameter characterizing the motion conversion efficiency. S4. Extract the free oscillation signal of the three-dimensional micro-displacement response data stream based on the boundary information of the coherent excitation window, and calculate the subsequent residual oscillation energy parameters based on the free oscillation signal. S5. The energy transfer efficiency parameter is fused with the subsequent residual oscillation energy parameter to construct the structural dynamic response feature vector corresponding to a single excitation event, and the structural dynamic response feature vector is stored in the historical diagnostic database. S6. Extract multi-phase structural dynamic response feature vectors from the historical diagnostic database to assess the evolution trend, and output the baseline structural health deterioration warning result based on the assessment results.

2. The method for decoupling monitoring of micro-displacement and vibration at a downhole reference point according to claim 1, characterized in that, Acquiring the three-dimensional micro-displacement response data stream and vibration excitation data stream under the same clock source includes the following steps: When the piezoelectric analog output signal is detected to momentarily exceed the preset hardware trigger level, a hardware interrupt signal is generated to synchronously start the first analog-to-digital converter and the second analog-to-digital converter that share the same high-precision clock source. A three-dimensional micro-displacement response data stream is generated by capturing three-dimensional signals at a preset sampling rate using a first analog-to-digital converter, and a synchronous vibration excitation data stream is generated at the same sampling rate using a second analog-to-digital converter.

3. The method for decoupling monitoring of micro-displacement and vibration at a downhole reference point according to claim 1, characterized in that, Defining the coherent excitation window based on the short-time energy envelope includes the following steps: A frame-by-frame sliding window calculation is performed on the vibration excitation data stream to extract energy features within the bandpass filter frequency band and generate a continuous short-time energy envelope. The time stamp at which the energy in the short-time energy envelope first exceeds a preset multiple of the background noise energy benchmark is locked as the excitation start point.

4. The method for decoupling monitoring of micro-displacement and vibration at a downhole reference point according to claim 3, characterized in that, The coherent excitation window is defined based on the short-time energy envelope, and the following steps are also included: After the short-term energy envelope reaches its peak, the time stamp when the energy value first decays to below a preset percentage of that peak value is marked as the excitation main lobe end point; The time interval between the excitation start point and the excitation main lobe end point is defined as the coherent excitation window.

5. The method for decoupling monitoring of micro-displacement and vibration at a downhole reference point according to claim 4, characterized in that, Calculating and generating energy transfer efficiency parameters characterizing motion conversion efficiency includes the following steps: Using the excitation start point and excitation main lobe end point of the coherent excitation window as the upper and lower limits of integration, the total energy of the response stage is calculated by full-window amplitude integration of the three-dimensional micro-displacement response data stream; The same coherent excitation window is used to perform amplitude integration, thereby extracting the total energy of the excitation stage from the vibration excitation data stream; The energy transfer efficiency parameter is generated by calculating the ratio of the total energy in the response phase to the total energy in the excitation phase.

6. The method for decoupling monitoring of micro-displacement and vibration at a downhole reference point according to claim 4, characterized in that, Extracting the free oscillation signal from the three-dimensional micro-displacement response data stream includes the following steps: Starting from the end point of the main lobe of the coherent excitation window and extending it by a preset time along the time axis, a residual oscillation time window representing the free oscillation phase is defined. Extract the time-series sampling signal of the three-dimensional micro-displacement response data stream within the residual oscillation time window; A high-pass filter is used to remove the low-frequency baseline drift component from the extracted time-series sampled signal to output the target free oscillation signal.

7. The method for decoupling monitoring of micro-displacement and vibration at a downhole reference point according to claim 1, characterized in that, Constructing the structural dynamic response feature vector corresponding to a single excitation event and storing the structural dynamic response feature vector in a historical diagnostic database includes the following steps: Extreme values ​​were extracted from the historical diagnostic database, and the energy transfer efficiency parameter and the subsequent residual oscillation energy parameter were normalized to eliminate dimensional differences. The normalized two features are concatenated to generate a two-dimensional dynamic response feature vector corresponding to the current event. The structural dynamic response feature vector is bound to the allocation event number containing timestamps and stored in the historical diagnostic database of the memory.

8. The method for decoupling monitoring of micro-displacement and vibration at a downhole reference point according to claim 1, characterized in that, Extracting multi-period structural dynamic response feature vectors from historical diagnostic databases to assess evolutionary trends includes the following steps: Read the structural dynamic response feature vectors from multiple consecutive event windows in the historical diagnostic database according to the time series. The vectors of each period are separated and the rising slope of the normalized energy transfer efficiency parameter of each period is calculated using the least squares regression method. The distribution variance of the normalized subsequent residual oscillation energy parameters within the same window is calculated.

9. The method for decoupling monitoring of micro-displacement and vibration at a downhole reference point according to claim 8, characterized in that, Based on the assessment results, output the baseline structural health deterioration early warning results, including the following steps: Input the calculated ascent slope and distribution variance indices into the preset structural stability determination model; When the upward slope is greater than the first safety threshold and the distribution variance is greater than the second safety threshold, it is determined to be in an unstable state. Send baseline structural health deterioration warning results with microcrack development classification labels to the monitoring level, indicating that the structure is becoming unstable.

10. A decoupled monitoring system for micro-displacement and vibration of a downhole reference point, characterized in that, Includes the following modules: The synchronous data acquisition module monitors the piezoelectric analog output signal of the downhole reference point. When the piezoelectric analog output signal exceeds the preset trigger level, a trigger signal is generated. Based on the trigger signal, the three-dimensional micro-displacement response data stream and vibration excitation data stream under the same clock source are synchronously acquired. The excitation window definition module processes the vibration excitation data stream to calculate the short-time energy envelope and defines the coherent excitation window based on the short-time energy envelope. The energy transfer efficiency calculation module integrates the three-dimensional micro-displacement response data stream and vibration excitation data stream with the coherent excitation window as the interval, and calculates and generates energy transfer efficiency parameters that characterize motion conversion efficiency. The residual oscillation feature extraction module extracts the free oscillation signal of the three-dimensional micro-displacement response data stream based on the boundary information of the coherent excitation window, and calculates the subsequent residual oscillation energy parameters based on the free oscillation signal. The dynamic response feature vector construction module integrates the energy transfer efficiency parameter and the subsequent residual oscillation energy parameter to construct the structural dynamic response feature vector corresponding to a single excitation event, and stores the structural dynamic response feature vector in the historical diagnostic database. The time-series trend assessment and early warning module extracts multi-period structural dynamic response feature vectors from the historical diagnostic database to assess the evolution trend, and outputs early warning results of structural health deterioration at the baseline point based on the assessment results.

Images