Multi-source sensing based structural monitoring data adaptive acquisition control method and system and inclinometer

Abstract

The application discloses a structure monitoring data adaptive acquisition control method and system based on multi-source sensing and a inclinometer, and relates to the technical field of monitoring control. The structure monitoring data adaptive acquisition control method based on multi-source sensing comprises the following steps: S1, performing multi-source sensing synchronous layer position data time binding, state checking and standardization processing; S2, performing sampling correction credible analysis based on the multi-source sensing synchronous layer position data, performing layer position compensation control, sampling gate determination and credible sampling layer position generation; S3, performing structure state credible analysis on the credible sampling layer position; and S4, performing data uploading and remote archiving, executing dormancy switching and state writing storage. The method solves the problem that the existing acquisition control method is dependent on encoder counting and fixed layer distance parking, and it is difficult to correct the layer position deviation and steady-state misjudgment caused by friction change, inclination disturbance and wireless communication delay, thereby improving the reliability of the layer position corresponding to the layered acquisition data and the accuracy of abnormal identification.

Description

Technical Field

[0001] This invention belongs to the field of monitoring and control, specifically an adaptive acquisition and control method, system, and inclinometer for structural monitoring data based on multi-source sensing. Background Technology

[0002] With the continuous expansion of infrastructure, structural safety monitoring plays an increasingly important role in engineering construction, operation and maintenance, and disaster early warning. Existing structural monitoring technologies typically involve deploying various types of sensors to conduct long-term observations of structural deformation, attitude changes, displacement changes, vibration response, and environmental changes, using the collected data for structural safety assessment, trend analysis, and early warning. In recent years, sensor technology, automated control technology, wireless communication technology, and data management platforms have been increasingly applied to the field of structural monitoring, transforming traditional manual inspections and periodic readings into an intelligent monitoring model that combines automatic data acquisition, remote transmission, centralized storage, and visual analysis. By uniformly collecting, organizing, storing, and displaying multi-source monitoring data, the monitoring system can provide engineering managers with continuous structural status information, providing data support for structural deformation trend analysis, abnormal condition identification, and maintenance decisions.

[0003] For example, invention patent CN110080261B discloses a control system for a barrel-shaped structure, including a host computer, a barrel-shaped structure, a PLC control module, a data acquisition module, and a drainage pump. The PLC control module is connected to the barrel-shaped structure and is used to control the attitude and sinking of the barrel-shaped structure. The data acquisition module is connected to the barrel-shaped structure and is used to collect the attitude and position data of the barrel-shaped structure, and transmit the collected data to the PLC control module and the host computer. One end of the drainage pump is connected to the barrel-shaped structure, and the other end is connected to the PLC control module, used to generate negative pressure to control the negative pressure sinking of the barrel-shaped structure. The host computer is used to receive the data collected by the data acquisition module, and analyze the collected data to adjust the attitude of the barrel-shaped structure according to the analysis results, thereby obtaining the attitude information of the barrel-shaped structure and automatically correcting the deviation, ensuring the floating balance and smooth sinking of the barrel-shaped structure.

[0004] For example, invention patent CN108604360B discloses a method and system for monitoring facility anomalies. The system includes a data acquisition unit, a learning model selection unit, and an anomaly alarm unit. The data acquisition unit collects facility data; the learning model selection unit selects multiple models to predict facility data values; and the anomaly alarm unit includes a prediction algorithm unit equipped with multiple prediction algorithms, an ensemble learning unit that performs ensemble learning based on the prediction data output by the prediction algorithm unit and outputs final prediction data, and alarm logic that determines whether the facility is abnormal by comparing the final prediction data with the data collected by the data acquisition unit. This method utilizes multiple prediction models with different characteristics to learn from the real-time collected facility data, generating prediction values ​​with higher prediction accuracy, and diagnosing facility anomalies accordingly, thereby achieving accurate monitoring and early warning of facility anomalies.

[0005] While existing structural monitoring and automatic control systems can collect data on structural attitude, position, or operational status, and perform anomaly identification, early warning output, or remote display, they still have the following shortcomings in deep sequence monitoring scenarios such as inclinometer layered data acquisition:

[0006] First, the acquisition process lacks unified time binding and status verification for depth position, tilt angle, displacement, drive status, communication status, and homing status, resulting in unclear correspondence between multi-source data in the same sampling round.

[0007] Second, layer positioning relies heavily on encoder counting, fixed layer spacing, or single position feedback, making it difficult to comprehensively judge friction changes, stepper motor operation fluctuations, sensor tilt angle disturbances, Bluetooth energy-saving communication reading delays, and zero-return trigger consistency. This can easily lead to layer offset, unstable stopping, and steady-state misjudgment.

[0008] Third, the structural state analysis after data collection lacks further verification of reliable sampling layers, making it difficult to distinguish layer data with inconsistent multi-source responses in a timely manner. This can lead to abnormal layer data directly affecting the structural deformation trend analysis and anomaly identification results.

[0009] Fourth, the remote upload and archiving process lacks classification processing of trusted structural state data, abnormal layer markers, and fused abnormal records. The state writing and low-power sleep after device upload also lack continuous closed loop, affecting the data traceability and operational stability of long-term multi-round monitoring. Summary of the Invention

[0010] To address the above problems, this invention provides an adaptive acquisition and control method, system, and inclinometer for structural monitoring data based on multi-source sensing. This addresses the issues that existing acquisition and control methods rely heavily on encoder counting and fixed-layer-distance docking, making it difficult to correct for layer shifts and steady-state misjudgments caused by friction changes, tilt disturbances, and wireless communication delays. Consequently, these methods suffer from insufficient reliability of layer-level correspondence and accuracy of anomaly identification in the acquired data.

[0011] To achieve the above objectives, the technical solution adopted by this invention is: an adaptive acquisition and control method for structural monitoring data based on multi-source sensing, comprising: S1, acquiring multi-source sensing synchronous layer data, performing time binding, state verification, and standardization processing of the multi-source sensing synchronous layer data; S2, performing sampling correction reliability analysis based on the multi-source sensing synchronous layer data, and performing layer compensation control, sampling gating judgment, and reliable sampling layer generation based on the sampling correction reliability analysis results; S3, performing structural state reliability analysis on the reliable sampling layers, and generating reliable structural state data, marking abnormal layers, and archiving fusion anomaly records based on the structural state reliability analysis results; S4, uploading and remotely archiving the reliable structural state data, abnormal layer markings, and fusion anomaly records, and performing multi-level power domain sleep switching and state writing storage.

[0012] Further, the specific process of collecting multi-source sensor synchronous layer data and performing time binding, status verification, and standardization of the multi-source sensor synchronous layer data is as follows: During the inclinometer tube layer acquisition stage, the main control microcontroller controls the sensor to descend along the inclinometer tube through the stepper motor drive interface circuit, stepper motor power supply, and brake control circuit. The main control microcontroller confirms the adjacent sampling layer based on the encoder count change and triggers a stop sampling. The multi-source sensor synchronous layer data includes: sensor depth position data collected through the encoder interface and stepper motor pulse count; sensor tilt angle data, sensor displacement data, stepper motor current data, and stepper motor voltage data; Bluetooth power-saving communication status data read through the Bluetooth power-saving wireless communication circuit, including the number of lost packets, the number of successful reads, sensor buffer timestamps, main control reading completion timestamps, and reading completion status; and laser switch trigger status data collected through the laser switch detection interface circuit. The system collects state data, including the laser switch trigger state data (including the zero-return trigger level and zero-return trigger timestamp); generates local acquisition records based on sampling rounds and sampling layers; removes outliers and smooths noise in sensor depth position data, sensor tilt angle data, and sensor displacement data; performs missing value completion and peak removal on stepper motor current data and stepper motor voltage data using a linear interpolation algorithm; corrects and completes missing markers in Bluetooth power-saving communication state data; performs logical verification on the laser switch trigger state data to confirm sensor zero-return completion and retains the zero-return trigger record; standardizes multi-source sensor synchronization layer data using a zero-mean unit variance standardization algorithm; and normalizes sensor depth position data, sensor tilt angle data, sensor displacement data, stepper motor current data, stepper motor voltage data, and Bluetooth power-saving communication state data using a maximum-minimum linear normalization algorithm, while retaining the binary logical state of the laser switch trigger state data.

[0013] Furthermore, the specific process of sampling correction and reliability analysis based on multi-source sensor synchronous layer data is as follows: The sensor depth position data is converted to the encoder depth recorded by the main control system through encoder pulse counting; the sensor layer position confirmation depth is obtained by subtracting the zero-return deviation, aligning the reading time, and confirming the layer position dwell time for the sensor depth position data, laser switch trigger status data, and Bluetooth power-saving communication status data; the instantaneous current change of the stepper motor is calculated by the absolute value of the difference between adjacent sampling times for the stepper motor current data; the instantaneous voltage change of the stepper motor is calculated by the absolute value of the difference between adjacent sampling times for the stepper motor voltage data; the standard deviation of the sensor tilt angle is calculated by the standard deviation within the dwell time window corresponding to the sensor layer position confirmation depth for the sensor tilt angle data; the sensor layer position confirmation displacement is obtained by extracting values ​​within the dwell time window and verifying the layer position correspondence for the sensor displacement data; and the sensor layer position confirmation depth is obtained by extracting values ​​within the dwell time window and verifying the layer position correspondence for the sensor tilt angle data. The process involves: recognizing the tilt angle; calculating the Bluetooth power-saving communication status index by the ratio of lost packets to successful reads and linearly normalizing it using the maximum and minimum values; obtaining the laser zero-return consistency value by valid trigger status verification and zero-return time consistency verification for the zero-return trigger level and zero-return trigger timestamp; calculating the absolute value of the difference between the master control record encoder depth and the sensor layer confirmation depth, multiplying it by the sum of the instantaneous current change of the stepper motor, the instantaneous voltage change of the stepper motor, and one, to obtain the encoding depth deviation weighting term; performing an exponential operation on the negative of the encoding depth deviation weighting term to obtain the depth deviation index suppression term; calculating the sum of the Bluetooth power-saving communication status index, the laser zero-return consistency value, and one, to obtain the communication zero-return support term; calculating the inverse hyperbolic sine of the sensor tilt angle fluctuation standard deviation plus one, to obtain the tilt angle fluctuation suppression term; calculating the ratio of the hyperbolic tangent of the communication zero-return support term to the tilt angle fluctuation suppression term, to obtain the communication zero-return nonlinear response term; and calculating the product of the depth deviation index suppression term and the communication zero-return nonlinear response term to obtain the sampling correction reliability value.

[0014] Furthermore, the specific process of performing layer compensation control, sampling gating judgment, and reliable sampling layer generation based on the sampling correction reliability analysis results is as follows: Real-time comparison of the sampling correction reliability value and the sampling correction reliability threshold: When the sampling correction reliability value is less than the sampling correction reliability threshold, the main control microcontroller issues a layer compensation control command through the stepper motor power supply and drive interface circuit, and controls the brake release and relocking through the stepper motor power supply and brake control circuit. At the current sampling layer, it performs extended dwell time, rereads multi-source sensor synchronous layer data, and reconfirms the laser switch's return to zero state. If the sampling correction reliability value is still less than the sampling correction reliability threshold after reprocessing, an abnormal layer is marked and abnormal layer information is generated. The main control record encoder is then deeply... The sensor layer confirmation depth, stepper motor instantaneous current change, stepper motor instantaneous voltage change, sensor tilt angle fluctuation standard deviation, Bluetooth power-saving communication status index, and laser zero-return consistency are created and archived in the calibration anomaly database. These data are not used as input for trusted structural state data; they only carry anomaly layer information into the trusted structural state analysis. When the sampled calibration confidence value is greater than or equal to the sampled calibration confidence threshold, it is marked as a trusted sampling layer. The sensor layer confirmation depth, sensor layer confirmation tilt angle, sensor layer confirmation displacement, stepper motor current data, stepper motor voltage data, Bluetooth power-saving communication status index, and laser zero-return consistency of the sampling layer are created and archived in the structural state database and then entered into the trusted structural state analysis.

[0015] Furthermore, the specific process of performing structural state reliability analysis on the reliable sampling layer is as follows: The sliding median is used to calculate the sensor layer confirmation depth, sensor layer confirmation tilt angle, and sensor layer confirmation displacement reference values ​​for adjacent layers of the reliable sampling layer; the standard deviation of the stepper motor current data is calculated using the adjacent layer window standard deviation to obtain the stepper motor current fluctuation standard deviation; the standard deviation of the stepper motor voltage data is calculated using the adjacent layer window standard deviation to obtain the stepper motor voltage fluctuation standard deviation; the Bluetooth power-saving reading latency index is obtained by calculating and normalizing the sensor buffer timestamp and the main control read completion timestamp through time difference; the difference between the sensor layer confirmation depth and the sensor layer confirmation depth reference value, and the difference between the sensor layer confirmation tilt angle and the sensor layer confirmation tilt angle reference value are calculated. The difference between the test values ​​and the difference between the sensor layer confirmed displacement and the sensor layer confirmed displacement reference value are squared, summed, and then the square root is taken to obtain the Euclidean quantity of the multi-source sensor deviation. The laser zero-return consistency quantity is calculated and one is added to obtain the zero-return normalization denominator. The multi-source sensor deviation Euclidean quantity is calculated and divided by the zero-return normalization denominator to obtain the zero-return adjustment deviation. The zero-return adjustment deviation is negative and then exponentially calculated to obtain the deviation exponential attenuation term. The standard deviation of the stepper motor current fluctuation, the standard deviation of the stepper motor voltage fluctuation, and the Bluetooth power-saving readout delay exponent are calculated, and the inverse hyperbolic sine is taken and one is added to obtain the current and voltage delay joint suppression term. The Bluetooth power-saving communication status exponent is calculated and one is added to obtain the communication enhancement factor. The communication enhancement factor is calculated and divided by the current and voltage delay joint suppression term to obtain the communication adjustment coefficient. The product of the deviation exponential attenuation term and the communication adjustment coefficient is calculated to obtain the structural status reliability value.

[0016] Furthermore, the specific process for generating reliable structural state data, marking abnormal layers, and archiving fusion anomaly records based on the results of the reliable structural state analysis is as follows: Real-time comparison of the reliable structural state value and the reliable structural state threshold: When the reliable structural state value is less than the reliable structural state threshold, the main control microcontroller controls the sensor to retract one sampling layer distance via the stepper motor drive interface circuit and stops at one of the adjacent sampling layers before or after the current sampling layer; the laser switch detection interface circuit verifies the zero-return trigger state, and the Bluetooth energy-saving wireless communication circuit rereads the sensor tilt angle data, sensor displacement data, reading completion status, sensor cache timestamp, and main control reading completion timestamp corresponding to the sampling layer; simultaneously, an abnormal layer mark is generated, a fusion anomaly record is created, and the sending of the sampling layer to data upload and remote archiving is temporarily suspended; when the reliable structural state value is greater than or equal to the reliable structural state threshold, the layer data is marked as reliable structural state data, and the sensor layer confirmation depth, sensor layer confirmation tilt angle, sensor layer confirmation displacement, Bluetooth energy-saving communication status index, and laser zero-return consistency value are created and recorded in the fusion database, allowing data upload and remote archiving.

[0017] Furthermore, the specific process for uploading and remotely archiving trusted structural status data, abnormal layer markers, and fused anomaly records is as follows: trusted structural status data, abnormal layer markers, and fused anomaly records are uploaded to the remote monitoring platform via the 4G mobile communication Long Term Evolution (LTE) circuit, the 4G mobile communication subscriber identification card circuit, the 4G mobile communication power supply circuit, and the communication level conversion circuit; after receiving the data, the remote monitoring platform generates layer depth curves, tilt trend diagrams, displacement trend diagrams, abnormal layer marker diagrams, and acquisition integrity reports, and archives the trusted structural status data, abnormal layer markers, and fused anomaly records to the structural status database; during on-site maintenance, the local acquisition records are read, and abnormal data and fused anomaly records are corrected via the Universal Serial Bus (USB) parameter configuration port circuit, the USB interface power supply circuit, and the 485 communication circuit.

[0018] Furthermore, the specific process of performing multi-level power domain sleep switching and status writing storage is as follows: After the upload is completed, the main control microcontroller controls the power supply status of the stepper motor power circuit, brake control circuit, Bluetooth energy-saving wireless communication circuit, laser switch detection interface circuit, global indicator circuit, and parameter storage circuit through the main control power circuit, and performs peripheral power-off, status writing, and sleep switching; writes the abnormal level information, stepper motor current fluctuation standard deviation, stepper motor voltage fluctuation standard deviation, Bluetooth energy-saving communication status index, and Bluetooth energy-saving read latency index into the parameter storage circuit, switches the global indicator circuit to sleep prompt state, and then shuts down the sensor, stepper motor, Bluetooth energy-saving wireless communication circuit, and fourth-generation mobile communication long-term evolution circuit, so that the device enters a low-power sleep state and waits for the next round of timed wake-up execution.

[0019] The second aspect of this invention provides an adaptive acquisition and control system for structural monitoring data based on multi-source sensing, comprising: an acquisition and preprocessing module for acquiring synchronous layer data from multiple sources, performing time binding, state verification, and standardization processing of the synchronous layer data; an adaptive layer correction and sampling gating module for performing reliable analysis of sampling correction based on the synchronous layer data from multiple sources, and performing layer compensation control, sampling gating determination, and reliable sampling layer generation based on the results of the reliable analysis; a multi-source data fusion and anomaly monitoring module for performing reliable structural state analysis on the reliable sampling layers, generating reliable structural state data, marking abnormal layers, and archiving fused anomaly records based on the results of the reliable structural state analysis; and a communication and visualization archiving module for uploading and remotely archiving the reliable structural state data, abnormal layer markings, and fused anomaly records, and performing multi-level power domain sleep switching and state writing storage.

[0020] The third aspect of this invention provides an inclinometer, comprising: a main control microcontroller module, used to complete power supply, self-test, and sampling cycle startup of the inclinometer after timed wake-up, and to control the inclinometer to enter a low-power sleep state after acquisition; a stepper motor layer drive module, used to control the sensor to descend and ascend along the inclinometer tube, to perform stop sampling after reaching the sampling layer, and to perform layer compensation based on the sampling correction reliability analysis results; an encoder and laser zero-return detection module, used to record the sensor's descent height and confirm the sampling layer, and to complete the zero-return trigger detection when the sensor ascends to the initial position; a Bluetooth power-saving reading module, used to wake up and connect to the sensor, read the layer acquisition results cached by the sensor, and complete the binding of the sensor-side sampling timing and the main control-side reading timing; a reliability analysis control module, used to perform sampling correction reliability analysis and structural state reliability analysis on the layer acquisition process, and generate reliable sampling layer, reliable structural state data, abnormal layer markers, and fusion anomaly records; and a fourth-generation mobile communication and parameter storage module, used to upload the reliable structural state data, abnormal layer markers, and fusion anomaly records to a remote monitoring platform, and to write the acquisition status and anomaly records after the upload is completed.

[0021] Beneficial effects

[0022] The present invention has the following beneficial effects:

[0023] (1) This invention, through sampling correction and reliability analysis, links stepper motor drive, brake release and relock, extended dwell and rereading actions, so that the inclinometer tube layer acquisition process is transformed from fixed docking sampling to an adaptive acquisition control process with feedback correction, thereby realizing the effect of dynamic compensation of sampling layer and closed-loop control of acquisition action, effectively solving the problem that the acquisition control in the prior art depends on fixed layer spacing docking and is difficult to adaptively correct according to the sampling state.

[0024] (2) This invention incorporates encoder depth, sensor layer confirmation depth, stepper motor running fluctuation, sensor tilt angle disturbance, Bluetooth power-saving communication status and laser zero-return consistency into the sampling gating judgment, so that the reliability of the current sampling layer can be screened before entering the subsequent analysis, thereby achieving the effect of intercepting the risk of mis-sampling in advance, and effectively solving the problem that it is difficult to detect layer shift and steady-state misjudgment in time due to relying solely on encoder counting in the prior art.

[0025] (3) In this invention, abnormal layer information is generated when the sampling correction confidence value does not meet the requirements, and it is prevented from being directly used as the input of the confidence structural state data. This enables the abnormal layer data and the confidence sampling layer data to be processed separately, thereby achieving the effect of isolated archiving of abnormal sampling results. This effectively solves the problem in the prior art that data with insufficient layer correction directly enters the structural state analysis and affects the accuracy of subsequent judgments.

[0026] (4) This invention performs a reliable analysis of the structural state of a reliable sampling layer and combines the reference of adjacent layers, driving fluctuations, communication reading delay and zero return state for fusion judgment. This makes the structural state output not only dependent on the measurement value of a single layer, but also verified by combining the stability of the acquisition process. This achieves the effect of reliable output of structural state data and effectively solves the problem that monitoring results may still be directly generated when the multi-source response is inconsistent in the prior art. Attached Figure Description

[0027] Figure 1 This is a flowchart of the adaptive acquisition and control method for structural monitoring data based on multi-source sensing according to the present invention.

[0028] Figure 2 This is a structural diagram of the adaptive acquisition and control system for structural monitoring data based on multi-source sensing according to the present invention.

[0029] Figure 3 This is a circuit diagram of the BLE wireless communication of the present invention;

[0030] Figure 4 This is a graph showing the trend of the sampling correction reliability value of the present invention changing with operating conditions;

[0031] Figure 5 This is a flowchart of the structural state confidence value determination and processing of the present invention;

[0032] Figure 6 This is an example diagram of the abnormal layer marking of the present invention;

[0033] Figure 7 This is an example circuit diagram of the main control microcontroller of the present invention. Detailed Implementation

[0034] To enable those skilled in the art to better understand the technical solution, the present invention will be described in detail below with reference to embodiments. The description in this part is only exemplary and explanatory, and should not be used to limit the scope of protection of the present invention in any way.

[0035] Please see Figures 1-7This invention provides a technical solution: an adaptive acquisition and control method for structural monitoring data based on multi-source sensing, comprising: S1, acquiring multi-source sensing synchronous layer data, performing time binding, state verification, and standardization processing of the multi-source sensing synchronous layer data; S2, performing sampling correction reliability analysis based on the multi-source sensing synchronous layer data, and performing layer compensation control, sampling gating determination, and reliable sampling layer generation based on the sampling correction reliability analysis results; S3, performing structural state reliability analysis on the reliable sampling layers, and generating reliable structural state data, marking abnormal layers, and archiving fusion anomaly records based on the structural state reliability analysis results; S4, uploading and remotely archiving the reliable structural state data, abnormal layer markings, and fusion anomaly records, and performing multi-level power domain sleep switching and state writing storage.

[0036] Specifically, the process of acquiring multi-source sensor synchronous layer data and performing time binding, state verification, and standardization of the multi-source sensor synchronous layer data is as follows: During the inclinometer layer acquisition stage, the main control microcontroller controls the sensor to descend along the inclinometer tube through the stepper motor drive interface circuit, stepper motor power supply, and brake control circuit. The stepper motor drive interface circuit receives the direction control signal, pulse control signal, and enable control signal output by the main control microcontroller. The stepper motor power supply and brake control circuit provides power to the stepper motor and performs brake release and locking control. The main control microcontroller confirms adjacent sampling layers based on encoder count changes and triggers a stop sampling. When the encoder count change reaches the count change condition corresponding to the distance between adjacent sampling layers, it confirms that the sensor has reached the next sampling layer. The encoder interface circuit feeds back the encoder pulse signal to the main control microcontroller, and the laser switch detection interface circuit feeds back the zero-return trigger level to the main control microcontroller. The sensor caches the tilt angle and displacement data of the corresponding layer in memory. The Bluetooth energy-saving wireless communication circuit works with the clock circuit to synchronize the sensor clock and the main control clock. The system generates layer-level acquisition records for the same sampling round; it collects multi-source sensor synchronous layer-level data, including: sensor depth position data acquired through encoder interface and stepper motor pulse counting; sensor tilt angle data acquired through MEMS tilt module; sensor displacement data acquired through displacement measurement module; stepper motor current data acquired through stepper motor power supply and drive interface circuit; stepper motor voltage data acquired through stepper motor power supply and drive interface circuit; Bluetooth power-saving communication status data read through Bluetooth power-saving wireless communication circuit, including packet loss count, successful read count, sensor buffer timestamp, master control read completion timestamp, and read completion status; and laser switch trigger status data acquired through laser switch detection interface circuit, including zero-return trigger level and zero-return trigger timestamp. Local acquisition records are generated according to the sampling round and sampling layer, and each local acquisition record carries at least the sampling round, sampling layer, sensor buffer timestamp, and master control read completion timestamp to maintain the correspondence between the sensor-side buffer timing and the master control-side read timing.

[0037] Outlier removal and noise smoothing are performed on sensor depth, tilt, and displacement data. Specifically, abrupt changes exceeding physical continuity are eliminated by verifying the amplitude of changes in adjacent layers within a sampling cycle, while retaining valid layer change trends. Missing data and peak removal are performed on stepper motor current and voltage data using linear interpolation algorithms. Short-term missing segments are repaired using interpolation with valid sampling points before and after the missing segments, and single-point peaks are suppressed using neighborhood amplitude consistency judgment. Continuity correction and missing marker completion are performed on Bluetooth power-saving communication status data. Communication interruption segments are identified based on packet loss, successful reads, and read completion status, and missing markers are generated for incompletely read sampling layers. Sensor buffer timestamps and master control read completion timestamps are bound to the same sampling cycle to maintain a traceable correspondence between sensor-side buffer timing and master control-side read timing. Logical verification is performed on laser switch trigger status data to confirm sensor homing completion. The zero-return trigger record is retained, with the zero-return trigger level and timestamp used together as the basis for determining the validity of the zero-return, avoiding false confirmations caused by a single level jump. A zero-mean unit variance standardization algorithm is used to standardize the multi-source sensor synchronous layer data, reducing the impact of dimensional differences between different acquisition channels on the calculation of subsequent sampling correction reliability values. A maximum-minimum linear normalization algorithm is used to normalize sensor depth position data, sensor tilt angle data, sensor displacement data, stepper motor current data, stepper motor voltage data, and Bluetooth power-saving communication status data, mapping continuous data to a unified numerical range, allowing encoder depth changes, attitude changes, displacement changes, drive state changes, and communication state changes to enter the same nonlinear discrimination scale. The laser switch trigger status data is retained in binary logic to maintain a clear boundary for the zero-return trigger result in subsequent laser zero-return consistency calculations, ensuring that all subsequent formula calculation parameters are on a dimensionless calculation scale.

[0038] In this embodiment, as Figure 3The diagram shows a BLE wireless communication circuit. This circuit uses the Bluetooth power-saving communication chip U5 as its core. The VCC pin is decoupled from the ground via capacitors C51 and C52 connected in parallel. The RFIO pin is connected to an external antenna matching terminal for RF transmission and reception. The nReset pin is connected to the BLE_nReset control signal via resistor R51, allowing the main microcontroller to perform a hardware reset control on the Bluetooth power-saving communication chip. Connector J3 leads the BLE_TXD and BLE_RXD signals of U5 to the universal asynchronous transceiver interface of the main microcontroller circuit. The main microcontroller sends a read command to BLE_RXD via TXD3_2 and receives sensor data and communication status data returned by BLE_TXD via RXD3_2, forming a serial data interaction channel. The BLE_Wake pin is used by the main microcontroller to send a wake-up signal after a timed wake-up, putting the Bluetooth power-saving communication chip into working mode. During the sampling pause, the main control microcontroller reads the sensor tilt angle data and sensor displacement data cached by the sensor through the serial communication links corresponding to BLE_TXD and BLE_RXD, and synchronizes the sensor clock with the main control clock by combining the sensor cache timestamp and the main control reading completion timestamp. The number of packet loss, number of successful reads, reading completion status and number of rereads are obtained by the main control microcontroller based on the reading frame sequence number, response status and reading result. After processing by the main control microcontroller, the Bluetooth power-saving communication status index and the Bluetooth power-saving reading latency index are generated.

[0039] In this implementation plan, the inclinometer can uniformly bind sensor descent docking, encoder layer confirmation, sensor data caching, Bluetooth power-saving reading, laser zero-return feedback, and master control side time recording to the same sampling round and sampling layer during the layered acquisition process. This avoids layer correspondence confusion caused by inconsistent reading times, communication interruptions, or false zero-return triggers between different acquisition channels. At the same time, through outlier removal, missing value completion, continuity correction, logic verification, standardization, and normalization, depth position, tilt angle, displacement, drive status, communication status, and zero-return status can participate in subsequent calculations on a unified numerical scale. This forms traceable, comparable, and nonlinearly suitable multi-source sensor synchronous layer data, thereby providing a stable and reliable data foundation for reliable sampling correction analysis, reliable structural status analysis, abnormal layer marking, and remote archiving.

[0040] Specifically, the process of sampling correction and reliability analysis based on multi-source sensor synchronous layer data is as follows: Sensor depth position data is converted to encoder depth by encoder pulse counting. The main control records the encoder depth. The direction of motion is determined based on the phase sequence of encoder A-phase pulses and B-phase pulses. The cumulative effective pulse count is multiplied by the single-pulse displacement equivalent. After the laser switch triggers zeroing, the cumulative count is zeroed and calibrated. Sensor depth position data, laser switch trigger status data, and Bluetooth power-saving communication status data are used to obtain the sensor layer confirmation depth through zeroing deviation deduction, reading time alignment, and layer dwell confirmation. The zeroing trigger timestamp is used to establish the sampling depth benchmark for this round. The sensor buffer timestamp and the main control reading completion timestamp are used for matching within the same round. Within the same dwell window, the valid depth position with reading completion status is selected as the layer confirmation object. The layer dwell window is the time range within which the sensor remains sampling after reaching the current sampling layer. The time range corresponds to each 0° descent of the sensor.The sampling pause period after 5 meters lasts for 2 to 5 seconds, with the duration determined based on the time required for the sensor to complete tilt and displacement measurements, data buffer writing, sensor clock synchronization with the main control clock, and the stepper motor to come to a stop. The instantaneous change in stepper motor current data is calculated by subtracting the absolute value of the difference between adjacent sampling times. The current value at the current sampling time is then subtracted from the current value at the previous sampling time, and the absolute value is taken, and then bound accordingly according to the sampling time interval. Similarly, the instantaneous change in stepper motor voltage data is calculated by subtracting the absolute value of the difference between adjacent sampling times. The current voltage value at the current sampling time is then subtracted from the current value at the previous sampling time, and the absolute value is taken, and then bound accordingly. The voltage value at the previous sampling time is subtracted and its absolute value is taken, then recorded synchronously with the current change result at the same sampling layer. The standard deviation of the sensor tilt angle data is calculated using the standard deviation within the corresponding dwell window of the sensor layer confirmation depth. Specifically, multiple frames of tilt angle values ​​are extracted within the dwell window corresponding to the same sensor layer confirmation depth. These multiple frames of tilt angle values ​​come from the set of sampling points corresponding to the current layer dwell window, and the root mean square result of the sum of squared deviations of the mean is calculated. The sensor displacement data is obtained by extracting values ​​within the layer dwell window and performing layer correspondence verification to obtain the sensor layer confirmation displacement within the corresponding dwell window of the sensor layer confirmation depth. The displacement sequence is read internally, and valid displacement values ​​are filtered according to the reading completion status. For sensor tilt data, the sensor layer-confirmed tilt angle is obtained by extracting values ​​within the corresponding dwell window of the sensor layer confirmation depth and verifying the layer correspondence. The tilt angle sequence is read within the corresponding dwell window of the sensor layer confirmation depth, and valid tilt angle values ​​are filtered according to the correspondence between the sensor buffer timestamp and the main control reading completion timestamp. The number of lost packets and successful reads are calculated by the reading success ratio and linearly normalized by the maximum and minimum values ​​to obtain the Bluetooth power-saving communication status index. The number of lost packets and successful reads are based on the buffer data reading frame corresponding to the current sampling layer. The success rate is calculated by dividing the number of successful reads by the sum of the number of successful reads and the number of lost packets, and then mapping this success rate to a unified numerical range. The laser zero-return consistency value is obtained by verifying the zero-return trigger level and zero-return trigger timestamp through valid trigger state checks and zero-return time consistency checks. Specifically, it is determined whether the zero-return trigger level has reached a valid trigger state, and the zero-return trigger timestamp is compared with the end time of the current sampling round for time consistency. If the zero-return trigger level has reached a valid trigger state and the difference between the zero-return trigger timestamp and the end time of the current sampling round is within a preset time consistency range, the laser zero-return consistency value is set to one; otherwise, it is set to zero.

[0041] The absolute value of the difference between the encoder depth recorded by the master control and the sensor layer confirmation depth is calculated, multiplied by the sum of the instantaneous current change of the stepper motor, the instantaneous voltage change of the stepper motor, and one, to obtain the coding depth deviation weighting term. This eliminates the difference in depth offset direction by using the absolute value. Multiplying this by current and voltage disturbance factors amplifies the impact of mechanical load changes on layer deviation. Adding one ensures that the depth deviation contribution is retained even when the motor disturbance is zero. The coding depth deviation weighting term is then negatively exponentially calculated to obtain the depth deviation exponential suppression term. The negative exponential function reduces the corresponding reliable contribution when the layer deviation increases, compressing the deviation impact to a stable value range. The sum of the Bluetooth power-saving communication status index, the laser zero-return consistency value, and one is calculated to obtain the communication zero-return support term. Adding one prevents subsequent response terms from failing when the communication status or zero-return status is zero, and ensures that communication reading and zero-return confirmation jointly participate in the sampling reliable support. The inverse hyperbolic sine value of the sensor tilt angle fluctuation standard deviation is calculated and one is added to obtain the tilt angle fluctuation suppression term. The inverse hyperbolic sine function compresses the abnormal amplitude of tilt angle fluctuation, and adding one prevents the denominator from being zero. A stable suppression benchmark is established; the ratio of the hyperbolic tangent of the communication zero-return support term to the tilt fluctuation suppression term is calculated to obtain the communication zero-return nonlinear response term. The gain upper limit of the communication zero-return support term is limited by the hyperbolic tangent function, and the tilt fluctuation suppression term is used to impose an inverse constraint on the sensor attitude disturbance; the product of the depth deviation exponential suppression term and the communication zero-return nonlinear response term is calculated to obtain the sampling correction confidence value. The product structure ensures that the sampling correction confidence value is reduced when either the depth deviation constraint or the communication zero-return constraint is not satisfied. The layer deviation, communication zero-return support, and tilt fluctuation are nonlinearly compressed by exponential operation, hyperbolic tangent operation, and inverse hyperbolic sine operation, respectively, to suppress the direct amplification of the sampling correction confidence value by single mechanical disturbances, reading anomalies, and attitude fluctuation spikes. Since the encoder depth, sensor layer confirmation depth, stepper motor instantaneous current change, stepper motor instantaneous voltage change, sensor tilt fluctuation standard deviation, Bluetooth power-saving communication status index, and laser zero-return consistency are all standardized or normalized, the sampling correction confidence value is a dimensionless value. The specific calculation formula is as follows:

[0042] ;

[0043] In the formula, This represents the sampling correction confidence value, used to characterize whether the current sampling layer meets the conditions for layer correction and sampling gating acceptance; This indicates the depth of the master control record encoder, used to characterize the layer count result obtained by the master control microcontroller based on the encoder interface; This indicates the sensor layer confirmation depth, used to characterize the sampling layer depth result after zero-return verification, reading time alignment, and dwell confirmation; It represents the instantaneous change in the stepper motor current, used to characterize the effect of changes in friction on the inner wall of the inclinometer tube on the stepper motor load; It represents the instantaneous voltage change of the stepper motor, used to characterize the impact of changes in the power supply state of the stepper motor on the stability of layer execution; This represents the standard deviation of the sensor tilt angle fluctuation, used to characterize the degree of sensor attitude disturbance during the sampling dwell period; The Bluetooth Power Saving Communication Status Index is used to characterize the completeness of the Bluetooth Power Saving Wireless Communication Circuit's reading of the current sampling layer data. This represents the laser zero-return consistency value, used to characterize the degree of confirmation of the zero-return state of the current sampling by the laser switch detection interface circuit.

[0044] In this embodiment, Table 1 is an example data table for calculating the sampling correction reliability value. It details the absolute value of the deviation between the master control record encoder depth and the sensor layer confirmation depth, the sum of the instantaneous current change and the instantaneous voltage change of the stepper motor, the standard deviation of the sensor tilt angle fluctuation, the Bluetooth power-saving communication status index, the laser zero-return consistency value, and the finally calculated sampling correction reliability value under different operating conditions. This table is used to quantify the impact of each input parameter on the reliability of layer correction. Specifically: for the baseline operating condition, the absolute value of the deviation between the master control record encoder depth and the sensor layer confirmation depth is 0.02, the sum of the instantaneous current change and the instantaneous voltage change of the stepper motor is 0.15, the standard deviation of the sensor tilt angle fluctuation is 0.12, the Bluetooth power-saving communication status index is 0.96, the laser zero-return consistency value is 1, and the sampling correction reliability value is 0.971; for the depth deviation increase operating condition, the absolute value of the deviation between the master control record encoder depth and the sensor layer confirmation depth is 0.15, the sum of the instantaneous current change and the instantaneous voltage change of the stepper motor is 0.15, the standard deviation of the sensor tilt angle fluctuation is 0.12, the Bluetooth power-saving communication status index is 0.96, the laser zero-return consistency value is 1, and the sampling correction reliability value is 0.971. The sum is 0.15, the standard deviation of sensor tilt angle fluctuation is 0.12, the Bluetooth power-saving communication status index is 0.96, the laser zero-return consistency is 1, and the sampling correction confidence value is 0.776; the absolute value of the deviation between the master control record encoder depth and the sensor layer confirmation depth corresponding to the motor disturbance increase condition is 0.02, the sum of the instantaneous current change and the instantaneous voltage change of the stepper motor is 0.80, the standard deviation of sensor tilt angle fluctuation is 0.12, the Bluetooth power-saving communication status index is 0.96, the laser zero-return consistency is 1, and the sampling correction confidence value is 0.926; tilt angle The absolute value of the deviation between the encoder depth and the sensor layer confirmation depth recorded by the main control system under the condition of increased fluctuation is 0.02, the sum of the instantaneous current change and the instantaneous voltage change of the stepper motor is 0.15, the standard deviation of the sensor tilt angle fluctuation is 0.55, the Bluetooth power-saving communication status index is 0.96, the laser zero-return consistency value is 1, and the sampling correction confidence value is 0.915. The absolute value of the deviation between the encoder depth and the sensor layer confirmation depth recorded by the main control system under the condition of communication deterioration and zero-return failure is 0.02, the sum of the instantaneous current change and the instantaneous voltage change of the stepper motor is 0.15, the standard deviation of the sensor tilt angle fluctuation is 0.55, the Bluetooth power-saving communication status index is 0.96, the laser zero-return consistency value is 1, and the sampling correction confidence value is 0.915. The sum of these values ​​is 0.15, the standard deviation of sensor tilt fluctuation is 0.12, the Bluetooth power-saving communication status index is 0.25, the laser zero-return consistency is 0, and the sampling correction confidence value is 0.703. Under the multi-factor deterioration superposition condition, the absolute value of the deviation between the master control record encoder depth and the sensor layer confirmation depth is 0.15, the sum of the instantaneous current change and the instantaneous voltage change of the stepper motor is 0.80, the standard deviation of sensor tilt fluctuation is 0.55, the Bluetooth power-saving communication status index is 0.25, the laser zero-return consistency is 0, and the sampling correction confidence value is 0.329.

[0045]

[0046] like Figure 4The graph shows the trend of sampling correction reliability values ​​as a function of operating conditions. The vertical axis represents the sampling correction reliability value, and each point on the horizontal axis corresponds to a different operating condition label. The dots represent the actual sampling correction reliability values ​​for each operating condition, and the dots are connected by broken lines. The arrows above the dots indicate the corresponding operating condition type. Table 1 shows that the attenuation of sampling correction reliability values ​​varies significantly under different operating conditions. The highest sampling correction reliability value (0.971) is found in the baseline operating condition, indicating that the layer level correction reliability is optimal when depth deviation is small, motor disturbance is low, tilt angle is stable, communication is complete, and zero-return confirmation is effective. The sampling correction reliability value drops to 0.776 under the condition of increased depth deviation, showing that a single increase in the deviation between the encoder depth and the sensor-confirmed depth can cause a significant decrease in reliability. The lowest sampling correction reliability value (0.329) is found under the condition of multiple factors causing deterioration, indicating that the layer level correction reliability is unacceptable when multiple factors such as depth deviation, motor disturbance, tilt angle fluctuation, communication deterioration, and zero-return failure are combined. Overall, the line in the trend chart of the sampling correction confidence value changes with the operating condition gradually slopes downward from the baseline operating condition to the multi-factor deterioration superposition operating condition. The text labels and arrows above each data point point to the corresponding circles, which intuitively reflect the progressive compression effect of the gradual superposition of various deterioration factors on the sampling correction confidence value. It can serve as the basis for selecting the sampling correction confidence threshold and triggering the stratification compensation control command in the adaptive stratification correction and sampling gating module.

[0047] This implementation scheme transforms the sampling layer positioning from single encoder counting to multi-source constraint confirmation positioning, enabling the master controller to record encoder depth, sensor layer confirmation depth, drive disturbance status, attitude fluctuation status, communication readout status, and homing verification status to jointly participate in the sampling reliability determination. This process can reduce the impact of inclinometer inner wall friction changes, stepper motor power supply fluctuations, sensor tilt angle disturbances, Bluetooth power-saving communication delays, and homing trigger deviations on the layer positioning results, allowing the sampling correction reliability value to more accurately reflect whether the current sampling layer meets the conditions for subsequent structural state reliability analysis. By applying nonlinear joint constraints to layer deviation, mechanical disturbances, communication status, and homing consistency, the targeting of layer compensation control and sampling gating determination can be improved, reducing the risk of mis-sampling, layer drift, and invalid sampling data entering the subsequent analysis process.

[0048] Specifically, the process of performing layer compensation control, sampling gating determination, and generating reliable sampling layers based on the sampling correction reliability analysis results is as follows: Real-time comparison of the sampling correction reliability value and the sampling correction reliability threshold:

[0049] When the sampling correction confidence value is less than the sampling correction confidence threshold, it is determined that there is a risk of layer shift or steady-state misjudgment at the current sampling layer. The main control microcontroller issues a layer compensation control command through the stepper motor power supply and drive interface circuit. The layer compensation control command includes the compensation direction, the number of compensation pulses, and the dwell time after compensation. It is used to control the sensor to reposition itself in the neighborhood of the current sampling layer. The stepper motor power supply and brake control circuit control the brake release and relock, so that the sensor can release the mechanical lock before compensation movement and maintain the layer dwell state after compensation. At the current sampling layer, the system performs extended dwell, rereads the multi-source sensor synchronous layer data, and reconfirms the laser switch return to zero status. The reread multi-source sensor synchronous layer data includes at least sensor depth position data, sensor tilt angle data, sensor displacement data, Bluetooth power-saving communication status data, and laser switch trigger status data. This is used to verify whether the layer shift is caused by mechanical execution error or communication reading abnormality. If the sampling correction confidence value is still lower after reprocessing, the system will take further action. If the sampled layer is less than the sampling correction confidence threshold, based on the comparison between the reprocessed sampling correction confidence value and the sampling correction confidence threshold, and considering the calculation items that caused the sampling correction confidence value to fail to meet the standard, such as the master control record encoder depth, sensor layer confirmation depth, stepper motor instantaneous current change, stepper motor instantaneous voltage change, sensor tilt angle fluctuation standard deviation, Bluetooth power-saving communication status index, and laser zero-return consistency, abnormal layer markers are generated and abnormal layer information is created. The master control record encoder depth, sensor layer confirmation depth, stepper motor instantaneous current change, stepper motor instantaneous voltage change, sensor tilt angle fluctuation standard deviation, Bluetooth power-saving communication status index, and laser zero-return consistency are created and archived in the correction anomaly database to retain the key calculation items and execution status that caused the sampling correction confidence value to fail to meet the standard. No reliable sampling layer marker is generated, and the current sampling layer is not archived as a reliable recording unit in the structural state database, nor is it used as reliable structural state data input. Only the abnormal layer information is carried into the structural state reliability analysis.

[0050] When the sampling correction confidence value is greater than or equal to the sampling correction confidence threshold, the current sampling layer is determined to meet the layer correction and layer confirmation sampling acceptance conditions, and is marked as a reliable sampling layer. The sensor layer confirmation depth, sensor layer confirmation tilt angle, sensor layer confirmation displacement, stepper motor current data, stepper motor voltage data, Bluetooth energy-saving communication status index, and laser zero-return consistency value of the sampling layer are created and archived into the structural status database, so that the spatial position, sensor response, drive status, and communication status of the sampling layer form the same recording unit and enter the structural status confidence analysis.

[0051] This implementation scheme achieves layer compensation control and sampling gating driven by the sampling correction confidence value. This allows sampling layers that do not meet the sampling correction confidence threshold to undergo closed-loop processing involving compensation direction, number of compensation pulses, post-compensation dwell time, and zero-return status verification before deciding whether to proceed to subsequent structural state confidence analysis. This process records the causes of anomalies such as layer offset, mechanical execution error, communication readout abnormalities, and zero-return inconsistencies, forming traceable correction anomaly data and preventing insufficiently corrected layers from being directly used as reliable sampling layers. For sampling layers that meet the sampling correction confidence threshold, sensor layer confirmation depth, sensor layer confirmation tilt angle, sensor layer confirmation displacement, drive status, communication status, and zero-return status are bound to the same recording unit, improving the data integrity, layer consistency, and decision reliability of subsequent structural state confidence analysis.

[0052] Specifically, the process of performing structural state reliability analysis on reliable sampling layers is as follows: The sliding median of the sensor layer confirmation depths of adjacent layers within the reliable sampling layer is calculated to obtain reference values ​​for sensor layer confirmation depths. The adjacent layer windows are determined according to the layer order formed by sensor descent within the same sampling cycle. Layer records within the same sampling cycle that are located before and after the current sampling layer and have been marked as reliable sampling layers are selected, and the median of the corresponding sensor layer confirmation depths is taken after sorting. The sliding median of the sensor layer confirmation tilt angles of adjacent layers within the reliable sampling layer is calculated to obtain reference values ​​for sensor layer confirmation tilt angles. Sensor layer confirmation tilt angles are extracted within the same adjacent layer window, arranged in numerical order, and the median is taken. The sliding median of the sensor layer confirmation displacements of adjacent layers within the reliable sampling layer is calculated to obtain reference values ​​for sensor layer confirmation displacements. Sensor layer confirmation displacements are extracted within the same adjacent layer window. The median was taken after arranging the data in numerical order. The standard deviation of stepper motor current fluctuation was calculated using the standard deviation of adjacent layer windows for the stepper motor current data. Stepper motor current data corresponding to the current sampling layer and adjacent reliable sampling layers were extracted, and the root mean square of the mean deviation of the sum of squares was calculated. The standard deviation of stepper motor voltage fluctuation was calculated using the standard deviation of adjacent layer windows for the stepper motor voltage data. Stepper motor voltage data corresponding to the current sampling layer and adjacent reliable sampling layers were extracted, where adjacent reliable sampling layers are the layer records marked as reliable sampling layers within the adjacent layer window. The root mean square of the mean deviation of the sum of squares was calculated. The Bluetooth power-saving read latency index was obtained by calculating and normalizing the time difference between the sensor buffer timestamp and the master control read completion timestamp. The read latency was obtained by subtracting the sensor buffer timestamp from the master control read completion timestamp, and then linearly normalized to a unified numerical range using the maximum and minimum values.

[0053] The differences between the sensor-confirmed depth and its reference value, the differences between the sensor-confirmed tilt angle and its reference value, and the differences between the sensor-confirmed displacement and its reference value are calculated. These differences are then squared, summed, and the square root is taken to obtain the Euclidean quantity of the multi-source sensing deviation. Squaring eliminates the difference in deviation direction, and taking the square root maintains the correspondence between the deviation quantity and the original difference scale, thus forming a unified distance metric for depth, tilt, and displacement deviations. The laser zero-return consistency value is calculated and incremented by one to obtain the zero-return normalization denominator. Incrementing by one avoids the denominator being zero and ensures that the zero-return confirmation result provides a stable basis for the deviation attenuation process. The adjustment process involves: calculating the Euclidean quantity of multi-source sensor deviation and dividing it by the denominator of the zero-return adjustment to obtain the zero-return adjustment deviation. This division structure allows the laser zero-return consistency quantity to exert a suppressive constraint on the Euclidean quantity of multi-source sensor deviation. The zero-return adjustment deviation is then negatively exponentially calculated to obtain a deviation exponential attenuation term. This negative exponential function reduces the reliable contribution of increasing layer deviation, dip angle deviation, and displacement deviation, compressing the deviation impact to a finite numerical range. Finally, the standard deviations of stepper motor current fluctuation and voltage fluctuation are calculated, along with the Bluetooth power-saving readout delay exponent. Taking an inverse hyperbolic sine and adding one yields a joint current and voltage delay suppression term, which is then summed to reflect the driving effect. The superimposed disturbances of fluctuations and communication delays are compressed by an inverse hyperbolic sine function, and the denominator is avoided by adding one. The Bluetooth power-saving communication state index is calculated and then incremented by one to obtain the communication enhancement factor. This increment ensures that the communication state index still has basic computational support even when it is zero, and allows communication integrity to participate in the adjustment of the structural state reliability value. The communication enhancement factor is calculated and divided by the current-voltage-delay joint suppression term to obtain the communication adjustment coefficient. This coefficient, through a fractional structure, provides positive support for the reliable output from the communication readout state, while creating negative constraints on the reliable output from current fluctuations, voltage fluctuations, and readout delays. Finally, the deviation index attenuation term and the communication adjustment coefficient are calculated. The product of these factors yields the structural state reliability value. This product structure ensures that the structural state reliability value decreases if either the multi-source deviation constraint or the communication drive constraint is not satisfied. Since the sensor layer confirmation depth, sensor layer confirmation tilt angle, sensor layer confirmation displacement, reference values ​​for sensor layer confirmation depth, sensor layer confirmation tilt angle, and sensor layer confirmation displacement, as well as the standard deviation of stepper motor current fluctuation, the standard deviation of stepper motor voltage fluctuation, the Bluetooth power-saving readout latency index, the Bluetooth power-saving communication state index, and the laser zero-return consistency value are all standardized or normalized, the structural state reliability value is a dimensionless value. The specific calculation formula is as follows:

[0054] ;

[0055] In the formula, This represents the structural state confidence value, used to characterize whether the current sampling layer meets the confidence structural state data output conditions; This indicates the sensor-confirmed layer depth, used to characterize the current sampling layer depth result after layer confirmation; This indicates the sensor-confirmed tilt angle, used to characterize the attitude response result corresponding to the current sampling layer; This indicates the sensor-confirmed displacement at the current sampling layer, used to characterize the displacement response at that layer. This indicates the sensor layer confirmed depth reference value, which is used to provide a depth comparison benchmark between adjacent reliable sampling layers; This indicates the dip angle reference value confirmed by the sensor layer, which is used to provide a dip angle comparison benchmark for adjacent reliable sampling layers; This indicates the sensor layer displacement reference value, which is used to provide a displacement comparison benchmark for adjacent reliable sampling layers; It represents the standard deviation of stepper motor current fluctuation, used to characterize the load fluctuation of stepper motors within adjacent layer windows; It represents the standard deviation of stepper motor voltage fluctuation, used to characterize the power supply fluctuation of stepper motors within adjacent layer windows; This represents the Bluetooth power-saving read latency index, used to characterize the read latency state between the sensor cache timestamp and the master control read completion timestamp; The Bluetooth power-saving communication status index is used to characterize the completeness of the Bluetooth power-saving wireless communication circuit's reading of the current sampling layer data; This represents the laser zero-return consistency quantity, used to characterize the consistency of the laser switch detection interface circuit in confirming the zero-return state of this round of sampling.

[0056] This implementation scheme achieves the construction of neighborhood references and the reliable discrimination of structural states between reliable sampling layers. This means that the current sampling layer no longer relies solely on single-layer data for output judgment, but instead combines adjacent reliable sampling layers to form a reference for depth, tilt, and displacement. This process can uniformly measure the deviation of the current sampling layer from the neighboring structural state using multi-source sensor deviation Euclidean quantities, and constrain and correct the deviation results using laser zero-return consistency, stepper motor current fluctuation standard deviation, stepper motor voltage fluctuation standard deviation, Bluetooth power-saving readout latency index, and Bluetooth power-saving communication status index. By comprehensively judging layer deviation, attitude deviation, displacement deviation, drive fluctuation, and communication readout status using reliable structural state values, the reliability of abnormal layer identification can be improved, and misjudgments caused by local communication delays, mechanical disturbances, or single-layer sampling anomalies can be reduced. This provides a stable basis for the generation of reliable structural state data, the marking of abnormal layers, and the archiving of fused anomaly records.

[0057] Specifically, the process of generating reliable structural state data, marking abnormal layers, and archiving fused abnormal records based on the results of the structural state reliability analysis is as follows: Figure 5The diagram shows the flowchart for determining and processing the structural state confidence value, which compares the structural state confidence value and the structural state confidence threshold in real time.

[0058] When the structural state confidence value is less than the structural state confidence threshold, it is determined that there is inconsistency in multi-source fusion or that the abnormal monitoring does not meet the output conditions at the current sampling level. The main control microcontroller controls the sensor to retreat by one sampling layer distance through the stepper motor drive interface circuit and stop at one of the adjacent sampling layers before and after the current sampling level. Here, one sampling layer distance is the distance between two adjacent stopping sampling positions within the same sampling cycle, and is determined by the corresponding change in encoder pulse count. The adjacent sampling layers before and after refer to the sampling layers located one position before and one position after the current sampling level according to the descending sampling order of the sensor in the inclinometer tube, so that the sensor still stops at the predefined layer sampling position after retreating. The zero-trigger state is verified by the laser switch detection interface circuit and the Bluetooth power-saving wireless communication circuit is used to re-verify the zero-trigger state. The system reads sensor tilt angle data, sensor displacement data, reading completion status, sensor cache timestamp, and main control reading completion timestamp corresponding to the sampling layer. This is used to verify whether there are Bluetooth power-saving reading delays, incomplete layer data readings, or timestamp mismatches at the sampling layer. Simultaneously, it generates an abnormal layer marker and stores the sensor layer confirmation depth, sensor layer confirmation tilt angle, sensor layer confirmation displacement, stepper motor current fluctuation standard deviation, stepper motor voltage fluctuation standard deviation, Bluetooth power-saving reading latency index, Bluetooth power-saving communication status index, and laser zeroing consistency value in the fusion anomaly record database. A fusion anomaly record is created and bound to the current sampling round, the current sampling layer, and the abnormal layer marker for archiving. The data upload and remote archiving of the sampling layer are temporarily suspended.

[0059] When the structural state confidence value is greater than or equal to the structural state confidence threshold, the current sampling layer is determined to meet the structural state confidence output condition. The layer data is marked as reliable structural state data. The sensor layer confirmation depth, sensor layer confirmation tilt angle, sensor layer confirmation displacement, Bluetooth energy-saving communication state index and laser zero-return consistency value are created and recorded in the fusion database, and data upload and remote archiving are allowed.

[0060] This implementation scheme achieves a trusted output offloading mechanism driven by the trusted structural state value. This allows sampling layers that do not meet the trusted structural state threshold to first undergo a neighborhood backtracking, zero-return verification, and Bluetooth power-saving rereading process. This prevents layer anomalies caused by communication latency, incomplete readings, or timestamp mismatches from being directly output as structural state anomalies. This process binds sensor layer confirmation depth, sensor layer confirmation tilt angle, sensor layer confirmation displacement, driving fluctuation status, communication status, and zero-return status to generate a fused anomaly record, forming a traceable link from sampling layer and sampling round to the cause of the anomaly. For sampling layers that meet the trusted structural state threshold, the layer data can be marked as trusted structural state data and entered into data upload and remote archiving, improving the reliability, completeness, and subsequent trend analysis reliability of the data received by the remote monitoring platform.

[0061] Specifically, the process of uploading and remotely archiving trusted structural state data, abnormal layer markers, and fusion anomaly records is as follows: Trusted structural state data, abnormal layer markers, and fusion anomaly records are uploaded to the remote monitoring platform via the 4G mobile communication Long Term Evolution (LTE) circuit, the 4G mobile communication subscriber identification card circuit, the 4G mobile communication power supply circuit, and the communication level conversion circuit. The 4G mobile communication power supply circuit provides independent power to the communication module, and the communication level conversion circuit matches the serial port level of the main control microcontroller with the interface level of the communication module, ensuring that the uploaded data can be stably transmitted according to the agreed protocol frames. Before uploading, the main control microcontroller checks the sensor layer confirmation depth integrity, Bluetooth power-saving communication status index, Bluetooth power-saving read latency index, laser zero-return consistency, stepper motor current fluctuation standard deviation, and stepper motor voltage fluctuation standard deviation for each sampling layer. It then generates a pre-upload verification state based on the above check results. This pre-upload verification state is used to distinguish between trusted structural state data, abnormal layer markers, and fusion anomalies. The uploaded data is categorized as follows: After receiving the data, the remote monitoring platform generates layered depth curves, dip trend maps, displacement trend maps, abnormal layer marker maps, and a data acquisition integrity report. The layered depth curves are generated according to the sensor-confirmed layer depths. The dip trend maps and displacement trend maps are generated using the sampling round and sampling layer as indices, respectively. The abnormal layer marker map is used to display the layer locations where the structural state confidence value does not meet the structural state confidence threshold. The credible structural state data, abnormal layer markers, and fused abnormal records are archived to the structural state database. During archiving, the sampling round, sampling layer, upload time, and device number are used as retrieval indexes to provide a data foundation for long-term multi-round monitoring and trend analysis. During on-site maintenance, the local acquisition records, correction abnormal data, and fused abnormal records are read through the universal serial bus parameter configuration port circuit, universal serial bus interface power circuit, and 485 communication circuit. This is used to retrieve acquisition process data, layer correction process data, and structural state confidence analysis process data when remote communication is unavailable or during on-site verification.

[0062] In this embodiment, as Figure 6 The diagram shows an example of anomaly stratum marking. The horizontal axis represents the sensor-confirmed stratum depth, covering the range from the initial position to the maximum measurement depth. The vertical axis distinguishes between normal and abnormal strata and has no actual dimensions. Each sampling stratum in the diagram is represented by a marker point: Strata marked with blue dots correspond to structural state confidence values ​​greater than or equal to the structural state confidence threshold, indicating that this stratum has been determined to be a reliable structural state stratum after multi-source data fusion, and its sensor-confirmed stratum depth, sensor-confirmed tilt angle, and sensor-confirmed displacement data have been successfully uploaded and archived remotely. Strata marked with red crosses correspond to structural state confidence values ​​less than the structural state confidence threshold, indicating that there is inconsistency or anomaly in multi-source fusion at this stratum. The platform generates anomaly stratum markers here based on the anomaly judgment results of the third module. The location of the anomaly stratum markers directly reflects the depth range within the inclinometer tube where suspicious data exists, providing intuitive depth positioning basis for on-site verification and subsequent trend analysis.

[0063] This implementation scheme achieves the classified uploading, verification transmission, and indexing and archiving of trusted structural status data, abnormal layer markers, and fused abnormal records from the field acquisition terminal to the remote monitoring platform. This eliminates reliance solely on communication link connectivity during data upload; instead, it incorporates pre-upload verification based on sensor layer depth integrity confirmation, Bluetooth power-saving communication status, Bluetooth power-saving read latency, laser zero-return consistency, and stepper motor fluctuation status. This process ensures that the data received by the remote monitoring platform has a clear sampling round, sampling layer, and abnormal type source, and is visualized through curves, marker graphs, and acquisition integrity reports, improving the query efficiency, trend analysis continuity, and abnormal traceability capabilities of long-term, multi-round monitoring data. Simultaneously, the field maintenance interface can read local acquisition records, correct abnormal data, and fuse abnormal records when remote communication is unavailable or requires verification, enhancing the convenience of field operation and maintenance and fault location.

[0064] Specifically, the process of performing multi-level power domain sleep switching and state writing storage is as follows: After uploading, the main control microcontroller controls the power supply status of the stepper motor power circuit, brake control circuit, Bluetooth energy-saving wireless communication circuit, laser switch detection interface circuit, global indicator circuit, and parameter storage circuit through the main control power circuit. It executes peripheral power-off, state writing, and sleep switching. According to the power domain hierarchical shutdown strategy, it first cuts off the power supply to the high-power drive circuit, then maintains the power supply to the digital logic circuit, completing state writing storage during the power-down process. It writes the abnormal level information, stepper motor current fluctuation standard deviation, stepper motor voltage fluctuation standard deviation, Bluetooth energy-saving communication status index, and Bluetooth energy-saving read latency index into the parameter storage circuit, maintaining the write operation through the backup power domain, with sector alignment... The data is written to the non-volatile memory area of ​​the parameter storage circuit. After writing, a readback verification is performed to ensure that the data is intact and not lost during low-voltage power loss. The global indicator circuit is switched to sleep mode, and the indicator lights are driven to flash in a breathing pattern with a low duty cycle and long intervals. This indicates that the device has entered sleep mode without waking up the main controller. Then, the sensor power supply enable, stepper motor drive enable, Bluetooth energy-saving wireless communication circuit enable, and 4G mobile communication LTE circuit enable levels are pulled low sequentially through the IO ports. At the same time, the power supply of the corresponding analog circuit bias source and RF front-end power amplifier is turned off, cutting off all power islands that are not necessary for standby. Then, the sensor, stepper motor, Bluetooth energy-saving wireless communication circuit, and 4G mobile communication LTE circuit are turned off, so that the device enters a low-power sleep state and waits for the next round of timed wake-up execution.

[0065] In this embodiment, as Figure 7 The diagram shown is an example of a main control microcontroller circuit. The main control microcontroller U1 forms the control core through power decoupling, a reset circuit, and an external crystal oscillator. Its universal asynchronous transceiver interface connects to a Bluetooth energy-saving wireless communication circuit and a fourth-generation mobile communication long-term evolution circuit. Its universal input / output interfaces connect to a stepper motor drive interface circuit, a laser switch detection interface circuit, and an encoder interface circuit, respectively. During layered data acquisition, U1 obtains the encoder depth based on the encoder's A-phase pulse, B-phase pulse, and Z-phase index signals. It reads sensor tilt angle data, sensor displacement data, and corresponding timestamps through the Bluetooth energy-saving wireless communication circuit, and generates the sensor layer confirmation depth by combining this with the laser switch's zero-return trigger state. When the sampling correction confidence value is less than the sampling correction confidence threshold, U1 performs layer compensation through the stepper motor drive interface circuit and completes brake release and relocking through the stepper motor power supply and brake control circuit.

[0066] This implementation scheme achieves status protection, tiered power-off, and low-power standby management after data upload is completed. This ensures that before entering sleep mode, the device reliably writes abnormal layer information, drive fluctuation status, and communication status, and then systematically shuts down high-power execution and communication components. This process avoids the problems of lost abnormal records, untraceable communication status, or lack of reference data for the next round of sampling caused by direct power loss of peripherals. Furthermore, the parameter storage circuit retains the previous round of acquisition and control results, providing continuous data for the next round of layer compensation control, reliable structural status analysis, and on-site maintenance verification. Simultaneously, through a global indicator circuit and multi-level power domain sleep switching, the device's operating status remains identifiable while reducing average operating power consumption, improving endurance and operational reliability in long-term field deployment scenarios.

[0067] like Figure 2 As shown, the second aspect of the present invention provides an adaptive acquisition and control system for structural monitoring data based on multi-source sensing, comprising: an acquisition and preprocessing module for acquiring multi-source sensing synchronous stratigraphic data, performing time binding, state verification, and standardization processing of the multi-source sensing synchronous stratigraphic data, so that the depth position, tilt angle, displacement, driving state, communication state, and homing state under the same sampling cycle form a traceable data foundation; and an adaptive stratigraphic correction and sampling gating module for performing sampling correction reliability analysis based on multi-source sensing synchronous stratigraphic data, performing stratigraphic compensation control, sampling gating determination, and reliable sampling stratigraphic generation based on the sampling correction reliability analysis results, so that sampling results with stratigraphic offset or unstable docking are processed. Before subsequent analysis, compensation and verification are performed; the multi-source data fusion and anomaly monitoring module is used to perform structural state credibility analysis on credible sampling layers, generate credible structural state data, mark abnormal layers, and archive fusion anomaly records based on the results of the structural state credibility analysis, so that layer data with inconsistent multi-source responses are distinguished and recorded, avoiding their direct output as credible structural state data; the communication and visualization archiving module is used to upload and remotely archive credible structural state data, abnormal layer markings, and fusion anomaly records, and perform multi-level power domain sleep switching and status writing storage, so that the acquisition results, anomaly causes, and device sleep status form a continuous archiving link, which is convenient for remote display and subsequent verification.

[0068] In this implementation scheme, through the coordinated setup of the acquisition and preprocessing module, the adaptive stratification correction and sampling gating module, the multi-source data fusion and anomaly monitoring module, and the communication and visualization archiving module, the depth position, tilt angle, displacement, driving status, communication status, and zero-return status of the sensor during the stratified acquisition process of the inclinometer tube can be synchronously bound and reliably verified in the same sampling round. When there is a shift in the sampling stratum, unstable docking, abnormal communication reading, or inconsistent zero-return, stratum compensation, extended dwell time, and rereading can be performed first before deciding whether to generate a reliable sampling stratum. When there is inconsistency among multi-source responses, abnormal stratum markers and fusion anomaly records can be generated to prevent abnormal stratum data from being directly uploaded and archived as reliable structural status data. Ultimately, the stratified acquisition, anomaly identification, remote uploading, and low-power sleep of structural monitoring data form a continuous closed loop, thereby improving the reliability of stratum correspondence, traceability of anomaly records, and stability of remote monitoring of the stratified acquisition data of the inclinometer tube.

[0069] The third aspect of this invention provides an inclinometer, comprising: a main control microcontroller module, used to complete the inclinometer power supply, self-test, and sampling cycle start-up after timed wake-up, coordinate the execution timing of layered driving, synchronous reading, reliability analysis, and communication uploading according to the acquisition sequence, and control the inclinometer to enter a low-power sleep state after acquisition, so that the single-cycle acquisition process and the next round of timed wake-up process are continuously connected; a stepper motor layered driving module, used to control the sensor to descend and ascend along the inclinometer tube, and cooperate with brake release and locking control to maintain the sensor's stopping stability at the sampling layer, perform dwell sampling after reaching the sampling layer, and perform layer compensation based on the sampling correction reliability analysis results, so that the layer offset can be repositioned and verified during the sampling stage; an encoder and laser zero-return detection module, used to record the sensor's descent height and confirm the sampling layer, provide hardware detection basis for the main control to record the encoder depth, sensor layer confirmation depth, and laser zero-return consistency, and complete the zero-return trigger detection when the sensor ascends to the initial position, so that the sampling depth benchmark of this round can be calibrated after the ascent; The tooth-energy-saving reading module is used to wake up and connect to the sensor, read the layered acquisition results cached by the sensor, and retain the reading completion status and reading timing information during the reading process. It also completes the binding of the sensor-side sampling timing and the main control-side reading timing, so that the layer acquisition records can be traced according to the same sampling round. The reliability analysis and control module is used to perform sampling correction reliability analysis and structural state reliability analysis on the layered acquisition process. It links the layer compensation results, sampling gating results and structural state output results for judgment, and generates reliable sampling layers, reliable structural state data, abnormal layer markers and fusion anomaly records, so that data with insufficient layer correction or inconsistent multi-source responses can be distinguished and processed. The fourth-generation mobile communication and parameter storage module is used to upload reliable structural state data, abnormal layer markers and fusion anomaly records to the remote monitoring platform, and maintain the correspondence between the uploaded data and the sampling round, sampling layer and anomaly record. After the upload is completed, it writes the acquisition status and anomaly record, so that the inclinometer's on-site acquisition process, remote display results and local trace data form a closed loop archive.

[0070] In this implementation scheme, the inclinometer integrates timed wake-up, layered acquisition, layer position correction, reliability determination, remote upload, and low-power sleep mode into a single execution carrier. This allows the sensor to continuously complete its descent docking, homing, buffer reading, and data upload within the inclinometer tube, all within the same sampling cycle. Layered driving by stepper motors, layer position confirmation by encoders, laser homing detection, and Bluetooth energy-saving reading enhance the stability of sampling layer positioning, sampling timing binding, and data reading processes. Reliability analysis of sampling correction and structural state compensates, marks, and diverts data exhibiting layer position shifts, unstable docking, communication reading anomalies, and inconsistent multi-source responses, preventing abnormal layer position data from being directly output as reliable structural state data. Remote upload and local tracking are achieved through fourth-generation mobile communication and a parameter storage module, enabling the inclinometer to achieve reliable layered acquisition, traceable abnormal processes, reliable structural state output, and stable long-term low-power operation.

[0071] It should be noted that, in this document, the terms "comprising," "including," and any other variations are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Specific examples have been used in this document to illustrate the principles and implementation methods of the present invention. These examples are merely for the purpose of helping to understand the method and core ideas of the present invention. The above descriptions are only preferred embodiments of the present invention. It should be pointed out that, due to the limitations of written expression and the objective existence of infinite specific structures, those skilled in the art can make several improvements, modifications, or variations without departing from the principles of the present invention, and can also combine the above technical features in an appropriate manner. These improvements, modifications, variations, or combinations, or the direct application of the concept and technical solution of the present invention to other situations without modification, should all be considered within the scope of protection of the present invention.

Claims

1. An adaptive acquisition and control method for structural monitoring data based on multi-source sensing, characterized in that, Includes the following steps: S1, collect multi-source sensor synchronization layer data, and perform time binding, status verification and standardization processing of multi-source sensor synchronization layer data; S2, based on multi-source sensor synchronous layer data, performs sampling correction reliability analysis, and performs layer compensation control, sampling gating determination and reliable sampling layer generation based on the sampling correction reliability analysis results; S3, perform structural state credibility analysis on the credible sampling layer, and generate credible structural state data, mark abnormal layers, and archive fusion anomaly records based on the results of the structural state credibility analysis. S4 performs data upload and remote archiving of trusted structure status data, abnormal level markers, and fused abnormal records, and executes multi-level power domain hibernation switching and status write storage.

2. The adaptive acquisition and control method for structural monitoring data based on multi-source sensing according to claim 1, characterized in that: The specific process of collecting multi-source sensor synchronization layer data, performing time binding, status verification, and standardization of the multi-source sensor synchronization layer data is as follows: During the inclinometer tube layer acquisition stage, the main control microcontroller controls the sensor to move down the inclinometer tube through the stepper motor drive interface circuit, stepper motor power supply and brake control circuit. The main control microcontroller confirms the adjacent sampling layer position based on the encoder count change and triggers stop sampling. Multi-source sensor synchronous layer data is collected, including: sensor depth and position data acquired through the encoder interface and stepper motor pulse counting; sensor tilt angle data, sensor displacement data, stepper motor current data, and stepper motor voltage data; Bluetooth power-saving communication status data read through the Bluetooth power-saving wireless communication circuit, including packet loss count, successful read count, sensor buffer timestamp, master control read completion timestamp, and read completion status; and laser switch trigger status data acquired through the laser switch detection interface circuit, including zero-return trigger level and zero-return trigger timestamp. Local acquisition records are generated according to the sampling round and sampling layer. Outlier removal and noise smoothing are performed on sensor depth position data, sensor tilt angle data, and sensor displacement data; missing data completion and peak removal are performed on stepper motor current data and stepper motor voltage data using a linear interpolation algorithm. The Bluetooth power-saving communication status data is corrected and missing markers are filled in; the laser switch trigger status data is logically verified to confirm that the sensor has returned to zero and to retain the zero-return trigger record; the multi-source sensor synchronization layer data is standardized using the zero-mean unit variance standardization algorithm; the sensor depth position data, sensor tilt angle data, sensor displacement data, stepper motor current data, stepper motor voltage data, and Bluetooth power-saving communication status data are normalized using the maximum and minimum value linear normalization algorithm, and the binary logic state of the laser switch trigger status data is retained.

3. The adaptive acquisition and control method for structural monitoring data based on multi-source sensing according to claim 1, characterized in that: The specific process of performing sampling correction reliability analysis based on multi-source sensor synchronous layer data is as follows: The following calculations are performed: The sensor depth position data is converted to encoder pulse count to obtain the main control recorded encoder depth; the sensor layer confirmation depth is obtained from sensor depth position data, laser switch trigger status data, and Bluetooth power-saving communication status data through zero-return deviation deduction, read time alignment, and layer dwell confirmation; the stepper motor current data is calculated from the absolute value of the difference between adjacent sampling times to obtain the instantaneous current change of the stepper motor; the stepper motor voltage data is calculated from the absolute value of the difference between adjacent sampling times to obtain the instantaneous voltage change of the stepper motor; the sensor tilt angle data is calculated from the standard deviation within the dwell window corresponding to the sensor layer confirmation depth to obtain the sensor tilt angle fluctuation standard deviation; the sensor displacement data is obtained from the value extraction within the dwell window and layer correspondence verification to obtain the sensor layer confirmation displacement; the sensor tilt angle data is obtained from the value extraction within the dwell window and layer correspondence verification to obtain the sensor layer confirmation tilt angle; the number of lost packets and the number of successful reads are calculated from the successful read ratio and linearly normalized by the maximum and minimum values ​​to obtain the Bluetooth power-saving communication status index; the zero-return trigger level and zero-return trigger timestamp are obtained from the effective trigger status verification and zero-return time consistency verification to obtain the laser zero-return consistency quantity. Calculate the absolute value of the difference between the encoder depth recorded by the main control and the sensor layer confirmation depth, multiply it by the sum of the instantaneous change in stepper motor current, the instantaneous change in stepper motor voltage, and one, to obtain the encoding depth deviation weighting term; after negativeing ​​the encoding depth deviation weighting term, perform an exponential operation to obtain the depth deviation exponential suppression term; Calculate the sum of the Bluetooth power-saving communication status index, the laser zero-return consistency value, and one to obtain the communication zero-return support term; The tilt fluctuation suppression term is obtained by adding one to the inverse hyperbolic sine value of the standard deviation of the sensor tilt fluctuation. The ratio of the hyperbolic tangent of the communication zero-return support term to the tilt fluctuation suppression term is calculated to obtain the communication zero-return nonlinear response term; The product of the depth deviation index suppression term and the communication return-to-zero nonlinear response term is calculated to obtain the sampling correction confidence value.

4. The adaptive acquisition and control method for structural monitoring data based on multi-source sensing according to claim 1, characterized in that: The specific process of performing layer compensation control, sampling gating determination, and reliable sampling layer generation based on the sampling correction reliability analysis results is as follows: Real-time comparison of sampling correction confidence value and sampling correction confidence threshold: When the sampling correction confidence value is less than the sampling correction confidence threshold, the main control microcontroller issues a layer compensation control command through the stepper motor power supply and drive interface circuit, and controls the brake release and relock through the stepper motor power supply and brake control circuit. At the current sampling layer, it performs extended dwell, rereads multi-source sensor synchronous layer data, and reconfirms the laser switch return to zero status. If the sampling correction confidence value is still less than the sampling correction confidence threshold after reprocessing, abnormal layer is marked and abnormal layer information is generated. The main control records encoder depth, sensor layer confirmation depth, stepper motor instantaneous current change, stepper motor instantaneous voltage change, sensor tilt angle fluctuation standard deviation, Bluetooth energy-saving communication status index, and laser return to zero consistency value, and creates and archives them into the correction anomaly database. These data are not used as input for reliable structural state data, but only carry abnormal layer information into the structural state confidence analysis. When the sampling correction confidence value is greater than or equal to the sampling correction confidence threshold, it is marked as a reliable sampling layer. The sensor layer confirmation depth, sensor layer confirmation tilt angle, sensor layer confirmation displacement, stepper motor current data, stepper motor voltage data, Bluetooth energy-saving communication status index and laser zero-return consistency value of the sampling layer are created and archived into the structural status database and then enter the structural status confidence analysis.

5. The adaptive acquisition and control method for structural monitoring data based on multi-source sensing according to claim 1, characterized in that: The specific process for performing structural state credibility analysis on the credible sampling layer is as follows: The sliding median is used to calculate the sensor layer confirmation depth, sensor layer confirmation dip angle, and sensor layer confirmation displacement of adjacent layers of the reliable sampling layer to obtain reference values ​​for sensor layer confirmation depth, sensor layer confirmation dip angle, and sensor layer confirmation displacement. The standard deviation of stepper motor current fluctuation is obtained by calculating the standard deviation of adjacent layer windows for stepper motor current data; the standard deviation of stepper motor voltage fluctuation is obtained by calculating the standard deviation of adjacent layer windows for stepper motor voltage data; the Bluetooth power-saving reading latency index is obtained by calculating and normalizing the time difference between the sensor buffer timestamp and the main control reading completion timestamp. Calculate the differences between the sensor layer confirmed depth and the sensor layer confirmed depth reference value, the differences between the sensor layer confirmed tilt angle and the sensor layer confirmed tilt angle reference value, and the differences between the sensor layer confirmed displacement and the sensor layer confirmed displacement reference value. Square each of these differences, sum them, and then take the square root to obtain the Euclidean quantity of the multi-source sensor deviation. Calculate the laser zero-return consistency quantity and add one to obtain the zero-return denominator. Calculate the multi-source sensor deviation Euclidean quantity and divide it by the zero-return denominator to obtain the zero-return adjustment deviation. Negate the zero-return adjustment deviation and perform an exponential operation to obtain the deviation exponential decay term. Calculate the sum of the standard deviation of stepper motor current fluctuation, the standard deviation of stepper motor voltage fluctuation, and the Bluetooth power-saving readout delay index, take the inverse hyperbolic sine and add one to obtain the current and voltage delay joint suppression term; Calculate the Bluetooth power-saving communication state index plus one to obtain the communication enhancement factor; calculate the communication enhancement factor and divide it by the current, voltage and time delay joint suppression term to obtain the communication adjustment coefficient; calculate the product of the deviation index attenuation term and the communication adjustment coefficient to obtain the structural state confidence value.

6. The adaptive acquisition and control method for structural monitoring data based on multi-source sensing according to claim 1, characterized in that: The specific process of generating reliable structural state data, marking abnormal layers, and archiving fused abnormal records based on the results of the reliable structural state analysis is as follows: Real-time comparison of structural state confidence value and structural state confidence threshold: When the structural state confidence value is less than the structural state confidence threshold, the main control microcontroller controls the sensor to retreat one sampling layer distance through the stepper motor drive interface circuit and stops at one of the adjacent sampling layers before or after the current sampling layer; the laser switch detection interface circuit verifies the zero-trigger state, and the Bluetooth energy-saving wireless communication circuit rereads the sensor tilt angle data, sensor displacement data, reading completion status, sensor buffer timestamp, and main control reading completion timestamp corresponding to the sampling layer; at the same time, an abnormal layer mark is generated, a fusion anomaly record is created, and the sending of the sampling layer into data upload and remote archiving is temporarily suspended. When the structural state confidence value is greater than or equal to the structural state confidence threshold, the layer data is marked as reliable structural state data. The sensor layer confirmation depth, sensor layer confirmation tilt angle, sensor layer confirmation displacement, Bluetooth power-saving communication state index, and laser zero-return consistency value are created and recorded in the fusion database, and data upload and remote archiving are allowed.

7. The adaptive acquisition and control method for structural monitoring data based on multi-source sensing according to claim 1, characterized in that: The specific process for uploading and remotely archiving trusted structural state data, anomaly level markers, and fused anomaly records is as follows: Trusted structural status data, abnormal stratum markers, and fusion anomaly records are uploaded to the remote monitoring platform via the 4G mobile communication Long Term Evolution (LTE) circuit, 4G mobile communication subscriber identification card circuit, 4G mobile communication power supply circuit, and communication level conversion circuit. After receiving the data, the remote monitoring platform generates stratification depth curves, dip angle trend maps, displacement trend maps, abnormal stratum marker maps, and acquisition integrity reports, and archives the trusted structural status data, abnormal stratum markers, and fusion anomaly records into the structural status database. During on-site maintenance, the local acquisition records are read, and abnormal data and fusion anomaly records are corrected via the Universal Serial Bus (USB) parameter configuration port circuit, USB interface power supply circuit, and 485 communication circuit.

8. The adaptive acquisition and control method for structural monitoring data based on multi-source sensing according to claim 1, characterized in that: The specific process of performing multi-level power domain hibernation switching and state writing to storage is as follows: After the upload is complete, the main control microcontroller controls the power supply status of the stepper motor power circuit, brake control circuit, Bluetooth energy-saving wireless communication circuit, laser switch detection interface circuit, global indicator circuit, and parameter storage circuit through the main control power circuit. It performs peripheral power-off, status writing, and sleep switching. It writes the abnormal level information, stepper motor current fluctuation standard deviation, stepper motor voltage fluctuation standard deviation, Bluetooth energy-saving communication status index, and Bluetooth energy-saving read latency index into the parameter storage circuit. It switches the global indicator circuit to sleep prompt state, and then shuts down the sensor, stepper motor, Bluetooth energy-saving wireless communication circuit, and 4G mobile communication LTE circuit, so that the device enters a low-power sleep state and waits for the next round of timed wake-up execution.

9. A structure monitoring data adaptive acquisition and control system based on multi-source sensing, employing the structure monitoring data adaptive acquisition and control method based on multi-source sensing as described in any one of claims 1-8, characterized in that, include: The acquisition and preprocessing module is used to acquire multi-source sensor synchronous layer data, and to perform time binding, status verification and standardization processing of multi-source sensor synchronous layer data; The adaptive tomographic correction and sampling gating module is used to perform sampling correction reliability analysis based on multi-source sensor synchronous tomographic data, and to perform tomographic compensation control, sampling gating determination and reliable sampling tomographic generation based on the sampling correction reliability analysis results. The multi-source data fusion and anomaly monitoring module is used to perform structural state credibility analysis on credible sampling layers, generate credible structural state data, mark abnormal layers, and archive fusion anomaly records based on the structural state credibility analysis results. The communication and visualization archiving module is used to upload and remotely archive trusted structure status data, anomaly level markers, and fused anomaly records, and to perform multi-level power domain hibernation switching and status writing to storage.

10. An inclinometer, employing the adaptive acquisition and control method for structural monitoring data based on multi-source sensing as described in any one of claims 1-8, characterized in that, include: The main control microcontroller module is used to power on the inclinometer, perform self-test, start the sampling cycle after a timed wake-up, and control the inclinometer to enter a low-power sleep state after the data acquisition is completed. The stepper motor layer drive module is used to control the sensor to descend and ascend along the inclinometer tube, perform stop sampling after reaching the sampling layer, and perform layer compensation based on the sampling correction confidence analysis results; The encoder and laser zero-return detection module are used to record the sensor's descent height and confirm the sampling layer position, and complete the zero-return trigger detection when the sensor rises back to the initial position; The Bluetooth power-saving read module is used to wake up and connect to the sensor, read the layered acquisition results cached by the sensor, and complete the binding of the sensor-side sampling timing and the master control-side reading timing. The credibility analysis and control module is used to perform sampling correction credibility analysis and structural state credibility analysis on the hierarchical acquisition process, and generate credible sampling layers, credible structural state data, abnormal layer markers and fusion anomaly records. The fourth-generation mobile communication and parameter storage module is used to upload trusted structure status data, abnormal layer markers, and fused abnormal records to the remote monitoring platform, and write the acquisition status and abnormal records after the upload is completed.

Images