Real-time remote monitoring method and system for foundation pit support inclination based on wireless inclinometer

Abstract

This invention discloses a method and system for real-time remote monitoring of foundation pit support tilt based on wireless inclinometers, relating to the fields of smart construction site safety monitoring and geotechnical engineering technology. It aims to solve the current technical problems of data distortion and low early warning accuracy caused by periodic vibration interference. The method involves first aggregating the raw tilt data collected by each wireless inclinometer node to an edge computing gateway, and then synchronously adjusting the sampling frequency and wireless data upload frequency of the wireless inclinometer nodes; the higher the deformation rate, the greater the sampling and upload density. This invention utilizes a complete technical system that coordinates edge-side spatiotemporal domain joint filtering correction, adaptive control of sampling and upload linked to tilt change rate, and a multi-factor dynamic early warning threshold model. This system comprehensively solves the core pain points of data distortion, power consumption and real-time performance imbalance, and insufficient early warning accuracy under strong vibration environments, significantly improving the reliability and safety management efficiency of foundation pit support monitoring in special scenarios.

Description

Technical Field

[0001] This invention relates to the field of smart construction site safety monitoring and geotechnical engineering technology, and more specifically, to a method and system for real-time remote monitoring of foundation pit support tilt based on a wireless inclinometer. Background Technology

[0002] In the field of smart construction site safety monitoring, tilt monitoring of foundation pit support based on wireless inclinometers has become the mainstream technical solution, replacing traditional manual reading and wired sensor monitoring methods. Existing technologies typically employ an architecture of wireless sensor nodes and a cloud monitoring platform. Inclinometers collect tilt data of the support structure, which is then uploaded to the cloud via a wireless communication network. After processing the data using a general filtering algorithm, risk warnings are issued by comparing it with fixed allowable values ​​specified in the standards. Some optimized solutions introduce a variable sampling frequency mechanism, adjusting the sampling frequency according to the deformation threshold to balance the monitoring frequency and equipment battery life, which can basically meet the daily monitoring needs of foundation pits in ordinary sites.

[0003] However, in high-risk and special scenarios such as deep foundation pits adjacent to operating rail transit lines, existing technologies have significant limitations: general filtering algorithms cannot effectively isolate the interference of periodic surface vibrations generated by train operation, resulting in the submergence of real deformation signals, distortion of monitoring data, and a high rate of false alarms; simple threshold-triggered variable sampling strategies are easily triggered by vibration interference, making it impossible to truly achieve a balance between power consumption and real-time monitoring; warning thresholds mostly use fixed standard values ​​and are not dynamically adapted to the control requirements of surrounding highly sensitive protected objects; at the same time, the cloud-based centralized processing architecture suffers from high transmission latency and complete failure of the warning function when the network is down, and there is a lack of maintenance-free online correction methods for the zero-point drift of sensors during long-term operation. In view of this, we propose a method and system for real-time remote monitoring of foundation pit support tilt based on a wireless inclinometer. Summary of the Invention

[0004] The purpose of this invention is to overcome the shortcomings of the prior art, adapt to practical needs, and provide a method and system for real-time remote monitoring of the tilt of foundation pit support based on a wireless inclinometer, so as to solve the current technical problems of distorted monitoring data and low accuracy of early warning caused by periodic vibration interference.

[0005] To solve the above-mentioned technical problems, the present invention provides the following technical solution: a method for real-time remote monitoring of the tilt of foundation pit support based on a wireless inclinometer, comprising: The raw tilt data collected by each wireless tilt sensor node is first aggregated to the edge computing gateway, where spatiotemporal joint filtering correction is performed: first, the reference period and amplitude characteristics of the periodic vibration of the surrounding environment are identified through a time-domain sliding window, and a standard vibration interference template is generated. Then, combined with the spatial position constraints of multiple measuring points within the same support unit, the standard vibration interference template is adapted for point amplitude attenuation and propagation phase difference. The adapted periodic vibration interference component is canceled from the raw tilt data, and the real tilt signal of the support structure is extracted. The tilt rate of change per unit time is calculated based on the real tilt signal after filtering and correction. Two-level rate thresholds are preset to divide the three working modes into low-power sleep, regular monitoring and emergency encryption. According to the level of tilt rate of change, the sampling frequency and wireless data upload frequency of the wireless tilt sensing node are adjusted synchronously. The higher the deformation rate, the greater the sampling and upload density. The tilt angle data after edge preprocessing is uploaded to the cloud monitoring platform. The cloud uses the allowable tilt value of the support structure design as the benchmark threshold, and introduces the risk level weighting coefficient of the surrounding protected objects, the excavation depth correction coefficient and the deformation trend correction coefficient to construct a multi-factor dynamic early warning threshold model. Combined with the tilt angle data of multiple points in the entire foundation pit, a global risk classification assessment is carried out, multi-level early warning instructions are generated and remote early warning information is output to multiple terminals.

[0006] Preferably, in the spatiotemporal joint filtering correction, the calculation of the true tilt signal satisfies the following relationship: ; In the formula, For the first The actual tilt signal at each measuring point This is the original dip angle data for this measuring point. For standard vibration interference template, This is the spatial amplitude correction coefficient for the measuring point, and it is inversely proportional to the square root of the horizontal distance from the measuring point to the vibration source. The phase difference is the vibration propagation to the measuring point and is calculated from the distance from the measuring point to the vibration source and the shear wave velocity of the soil.

[0007] Preferably, the switching of the three working modes is equipped with a delay anti-shake mechanism: the working mode switching is triggered only after the tilt angle change rate continuously crosses the threshold for a preset time, so as to avoid frequent switching between sampling and uploading modes caused by periodic vibration fluctuations.

[0008] Preferably, in the multi-factor dynamic early warning threshold model, the early warning thresholds at each level... Satisfying the relation: ; In the formula, The allowable inclination benchmark value is designed for the support structure. Risk level coefficient for surrounding protected objects, This is a correction factor for excavation depth, and the threshold is gradually tightened as the excavation depth increases. It is a deformation trend correction coefficient that tightens the threshold as the deformation rate increases.

[0009] Preferably, the edge computing gateway periodically performs online self-calibration of zero-point drift using the spatial geometric constraints of multiple wireless tilt sensing nodes within the same support unit: based on the tilt difference between adjacent measuring points in the initial calibration state, the cumulative zero-point drift of each node is calculated. It also performs online correction of the current monitoring value, and the calibration process does not interrupt the normal monitoring process.

[0010] A real-time remote monitoring system for the tilt of foundation pit support based on a wireless inclinometer includes multiple wireless inclinometer nodes deployed at various monitoring points on the foundation pit support structure, at least one edge computing gateway, and a cloud monitoring platform. The wireless tilt sensing node has a built-in tilt acquisition unit, node control unit, wireless transmission unit and power management unit, which are used to acquire raw tilt data and adjust the sampling and uploading parameters according to the control command; The edge computing gateway is used to aggregate data from wireless tilt sensor nodes within its jurisdiction, perform spatiotemporal filtering correction, tilt change rate calculation and preliminary risk assessment, and send mode control commands to the sensor nodes. The cloud-based monitoring platform is used to receive preprocessed monitoring data, run a multi-factor dynamic early warning threshold model to complete global risk assessment, and output remote early warning information from multiple terminals.

[0011] Preferably, the power management unit of the wireless tilt sensing node is linked with the node control unit to dynamically adjust the power supply strategy according to the current working mode: in low-power sleep mode, unnecessary peripheral modules are turned off, and in emergency encryption mode, full power supply is provided. The power management unit supports automatic switching between solar power and built-in battery power.

[0012] Preferably, the edge computing gateway has a built-in local early warning module and an offline data caching unit. In the event of a network interruption with the cloud, it can independently complete data filtering and correction and local audible and visual early warning. After the network is restored, it can automatically re-transmit the offline cached data to the cloud monitoring platform.

[0013] Preferably, the cloud monitoring platform has a built-in three-dimensional visualization module, which can spatially map and display the tilt data and early warning information of each monitoring point in conjunction with the foundation pit BIM model, and supports multi-channel remote access and early warning information push from the web and mobile terminals.

[0014] A method for real-time remote monitoring of foundation pit support tilt based on a wireless inclinometer is proposed. This method is applied to the scenario of monitoring the tilt of deep foundation pit support adjacent to an operating rail transit line. The periodic vibration interference of the surrounding environment is the periodic vibration of the ground surface generated by the operation of rail transit trains.

[0015] Compared with the prior art, the beneficial effects of the present invention are: 1. This invention integrates edge spatiotemporal domain joint filtering, rate adaptive sampling control, and a multi-factor dynamic early warning model to form an integrated technical system. Spatiotemporal domain filtering, combined with temporal vibration feature recognition and spatial measurement point vibration propagation attenuation adaptation, can accurately filter out periodic surface vibration noise from subway trains and restore the true tilt deformation of the support. Based on the true deformation rate, the sampling upload frequency is controlled in stages. The sampling frequency is reduced to extend the endurance during the stable stage of the foundation pit, and the monitoring is automatically intensified to capture sudden changes when deformation accelerates. The multi-factor dynamic threshold, combined with the surrounding protection level, excavation depth, and deformation rate, dynamically adjusts the early warning standard to meet the stringent control requirements of highly sensitive foundation pits next to subways, and simultaneously solves the core problems of data distortion, power consumption and real-time performance contradictions, and false alarms.

[0016] 2. This invention also incorporates a delay-based anti-jitter mechanism into the three-level operating mode switching logic. Mode switching is only triggered when the tilt angle change rate continuously crosses a threshold for a preset duration. This effectively filters interference from periodic vibration fluctuations and instantaneous spikes, avoiding frequent erroneous switching between sampling and uploading modes and ensuring stable system operation. This mechanism makes the adaptive power consumption control logic more closely match the actual deformation trend of the support structure, reducing unnecessary mode switching overhead and power loss, and further enhancing the accuracy and practicality of the variable sampling strategy.

[0017] 3. This invention also constructs a two-tier collaborative processing architecture consisting of wireless sensor nodes, edge computing gateways, and a cloud monitoring platform. The edge computing gateway performs data filtering and correction, rate calculation, and preliminary risk assessment locally, directly triggering on-site audible and visual early warnings. The cloud platform is responsible for global risk assessment, model iteration, and remote early warning dissemination to multiple terminals. This architecture not only reduces the pressure on wireless transmission bandwidth and controls local early warning response latency to the second level, but also maintains complete local monitoring and early warning functions by relying on the edge gateway during network interruptions, eliminating early warning blind spots caused by network outages and achieving complementary advantages between rapid local response and global cloud assessment.

[0018] 4. This invention also addresses the shortcomings of the aforementioned low-power operation mode and edge-side computing power limitations, where sensors are prone to cumulative zero-point drift and gradual attenuation of monitoring accuracy during long-term operation, through an online self-calibration mechanism for zero-point drift with multi-node spatial constraints within the same support unit. This mechanism uses the tilt angle difference of each measuring point in the initial calibration state as a benchmark, calculates the cumulative drift of each node using the geometrical position correlation of multiple nodes, and completes online correction. The calibration process does not require interruption of normal monitoring or manual on-site calibration. While retaining the advantages of edge autonomy and low power consumption, it effectively controls long-term monitoring data deviation and improves the accuracy and reliability of continuous system operation. Attached Figure Description

[0019] Figure 1 This is a flowchart of the overall monitoring method of the present invention; Figure 2 This is a diagram showing the state switching of the three-level working mode of the present invention; Figure 3 This is a logic diagram of the multi-factor dynamic early warning threshold model of the present invention. Detailed Implementation

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

[0021] It should be noted that the technical solutions described in the embodiments of this application are particularly applicable to monitoring scenarios of deep foundation pit projects located adjacent to operating rail transit lines, main roads, and other areas with significant periodic environmental vibration sources, but are also equally applicable to tilt monitoring of other conventional foundation pit supports. The terms "first," "second," etc., used in the specification and claims of this application are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence.

[0022] Before elaborating on the technical solution of this application, some key terms involved in this application will be explained for ease of understanding.

[0023] Wireless tilt sensor node: refers to an integrated sensor device that is deployed at monitoring points on the foundation pit support structure and integrates tilt measurement, data acquisition, wireless communication and power management functions.

[0024] Edge computing gateway: A computing device deployed at the network edge of a monitoring site, close to the data source, responsible for aggregating sensor node data and performing real-time preprocessing, local decision-making, and command issuance.

[0025] Spatiotemporal joint filtering correction: This application proposes a signal processing method that extracts periodic vibration features in the time domain and combines them with the geometric constraints of multiple measurement points in the spatial domain to compensate for and cancel interference in the original signal.

[0026] Multi-factor dynamic early warning threshold model: Unlike fixed thresholds, this is a cloud-based dynamic early warning threshold calculation model that comprehensively considers multiple influencing factors such as support design allowable values, risk levels of surrounding protected objects, excavation depth, and real-time deformation trends.

[0027] Zero-point drift online self-calibration: refers to a mechanism that automatically calculates and corrects the zero-point offset of sensors caused by long-term operation without interrupting the monitoring process during normal system monitoring operation, using the spatial geometric relationship of multiple sensor nodes within the same support unit.

[0028] like Figure 1 As shown in the figure, this application provides a method for real-time remote monitoring of the tilt of foundation pit support based on a wireless inclinometer. The method includes: Step S101: The raw tilt data collected by each wireless tilt sensor node is first aggregated to the edge computing gateway, and spatiotemporal joint filtering correction is performed at the edge. Step S102: Calculate the tilt angle change rate per unit time based on the real tilt signal after filtering and correction, and pre-set two-level rate thresholds to divide the three working modes into low-power sleep, regular monitoring and emergency encryption. Synchronously adjust the sampling frequency and wireless data upload frequency of the wireless tilt angle sensing node. In step S103, the tilt angle data after edge-side preprocessing is uploaded to the cloud monitoring platform. The cloud constructs a multi-factor dynamic early warning threshold model to conduct global risk classification assessment, generates multi-level early warning instructions, and outputs remote early warning information to multiple terminals.

[0029] In this embodiment, step S101 performs spatiotemporal domain joint filtering correction through an edge computing gateway, combining the periodic vibration pattern recognition in the time domain with propagation attenuation and phase delay in the spatial domain. This eliminates interference from periodic vibrations in the surrounding environment, such as train operation, on the tilt angle data, providing a high-fidelity real tilt signal for all subsequent analyses. Step S102 monitors the rate of change of the real tilt signal to achieve real-time dynamic perception of the safety status of the foundation pit support structure. Based on this, the sampling and data transmission strategies of the sensor nodes are adaptively adjusted, significantly saving power consumption under stable deformation conditions and immediately increasing monitoring density under emergency conditions of accelerated deformation, achieving an optimal balance between power consumption and real-time performance. Step S103 delivers the preprocessed high-quality data to a powerful cloud monitoring platform, where a multi-factor dynamic early warning threshold model is run. This overcomes the limitations of traditional fixed threshold alarms, combining multi-source influencing factors and multi-point data from the entire foundation pit to complete a precise risk assessment at the global level, ultimately achieving multi-terminal, second-level push of multi-level early warning information.

[0030] Taking a specific application scenario as an example, in a deep foundation pit project adjacent to an operating subway line, multiple wireless tilt sensor nodes are deployed at different depths of the foundation pit support piles. When a subway train passes by, the resulting periodic vibrations are superimposed on the raw tilt data collected by each wireless tilt sensor node. This data is first aggregated to the on-site edge computing gateway through step S101. The processing program in the edge computing gateway first performs a sliding window analysis in the time domain to identify the vibration characteristics caused by the subway operation with fixed frequency and relatively stable amplitude, generating a "standard vibration interference template". Subsequently, the program calculates the amplitude attenuation and propagation phase delay of the standard vibration template based on the horizontal distance of each wireless tilt sensor node from the subway tunnel (vibration source) and the site soil conditions, thereby customizing a unique disturbance component for each measuring point and accurately subtracting it from the raw data.

[0031] Next, step S102 uses the purified real tilt signal to calculate the tilt change rate of the support pile in real time. If the deformation rate is extremely low under normal conditions, the edge computing gateway will issue an instruction to put the wireless tilt sensor node into a low-power sleep mode and reduce the data upload frequency; once the deformation rate is detected to be continuously rising and crossing a higher rate threshold, the edge computing gateway will immediately instruct the relevant wireless tilt sensor node to switch to emergency encryption mode, while increasing the sampling and upload frequency.

[0032] Finally, the data, after edge-side filtering and mode adjustment, is sent to the cloud monitoring platform via step S103. The cloud platform utilizes a dynamic early warning threshold model that includes three factors: the surrounding subway tunnel (a high-risk protected object), the current excavation depth of the foundation pit, and the real-time deformation acceleration trend. The calculated dynamic threshold is comprehensively analyzed with the data from all measuring points in the entire foundation pit. Once the tilt deformation of a certain area is found to exceed the dynamically set safety threshold, a yellow or red warning is immediately generated, and alarm information is simultaneously pushed to the mobile phones of the project manager and safety officer, as well as to the on-site audible and visual alarms.

[0033] In one embodiment, for the spatiotemporal joint filtering correction in step S101, the calculation of the true tilt signal satisfies the following relationship: ; In the formula, For the first The actual tilt signal at each measuring point This is the original dip angle data for this measuring point. For standard vibration interference template, This is the spatial amplitude correction coefficient for the measuring point, and it is inversely proportional to the square root of the horizontal distance from the measuring point to the vibration source. The phase difference is the vibration propagation to the measuring point and is calculated from the distance from the measuring point to the vibration source and the shear wave velocity of the soil.

[0034] In this embodiment, the formula precisely defines how to extract the original dip angle data. The true tilt signal was restored in the middle. The core issue lies in the fact that periodic vibrations are in a changing state during propagation; their intensity decreases with increasing distance, and they arrive at different locations at different times. Therefore, it is necessary to perform measurements at each measuring point. It accurately adapts to the actual vibration interference it senses. Among these, the spatial amplitude correction coefficient... Based on the geometric diffusion principle of waves, the characteristic of vibrational energy attenuating with the square root of the distance is considered; phase difference This introduces site soil dynamics parameters from the field of geotechnical engineering, using the distance between the measuring point and the vibration source divided by the shear wave velocity to simulate the time difference of vibration reaching different measuring points.

[0035] As a specific implementation method, in scenarios adjacent to rail transit, the vibration source can be specifically defined as the outer contour center of the left and right tunnels of the subway, while the site soil shear wave velocity is obtained through on-site wave velocity testing or geological survey reports. The digital signal processor (DSP) inside the edge computing gateway first performs time-domain autocorrelation or cross-correlation analysis on the aggregated multi-channel data to pinpoint the periodic dominant frequency and phase caused by the train, in order to establish... Subsequently, using the pre-stored coordinates of each measuring point and the tunnel coordinates, the coordinates of each measuring point were calculated. corresponding and The above formula is iteratively calculated to remove vibration interference point by point in real time, thereby obtaining a pure and true tilt signal of the support structure.

[0036] In one embodiment, such as Figure 2 As shown, for the process of adjusting the sampling frequency and upload frequency in step S102, the switching of the three working modes is equipped with a delay anti-jitter mechanism: the working mode switching is only triggered after the tilt angle change rate continuously crosses the threshold for a preset duration, avoiding frequent switching of sampling and upload modes caused by periodic vibration fluctuations. The process of dynamically adjusting the sampling frequency and upload frequency based on the tilt angle change rate is specifically as follows: the tilt angle change rate per unit time is calculated based on the actual tilt signal. The system is divided into three working modes based on two preset rate thresholds, with varying sampling frequencies. The wireless data upload frequency increases with the increase in speed level. With sampling frequency According to a fixed proportional coefficient Linked matching, satisfying the following relational expression: ; In the formula, the proportionality coefficient Values Different proportional coefficients should be set for different working modes. .

[0037] In this embodiment, a time-delay anti-shake mechanism is used to solve the problem of false triggering of working modes caused by occasional or instantaneous vibration spikes. Only when the deformation trend continuously and stably exceeds a certain threshold is it determined that the risk level has indeed changed substantially, thereby ensuring the stability and reliability of the system operation.

[0038] Furthermore, sampling frequency With upload frequency The interrelationships were precisely modeled. Proportionality coefficient. Essentially, this reflects the compression ratio or selective upload strategy of the data processing. In low-power sleep mode, The value can be set relatively small, such as 0.1, so that the sensor node only uploads a very small number of representative statistical feature data; while in emergency encryption mode, The value can be set to a large value, such as 0.8. The sensor node will quickly and densely upload almost all the sampling point data to ensure that the cloud can capture every deformation detail.

[0039] Preferably, in one example, the sampling frequency set for the low-power sleep mode is... The monitoring rate is 1 time / minute, with the regular monitoring mode at 1 time / 10 seconds and the emergency encryption mode at 10 times / second. The two-level rate thresholds are set at 0.1 mm / day and 1 mm / day respectively, based on the foundation pit design safety level and engineering experience. The edge computing gateway will calculate based on real-time data... Within the given range, the system searches a preset mode mapping table, generates and sends control commands to the corresponding wireless tilt sensor node, and adjusts the sampling trigger clock of its internal analog-to-digital converter (ADC) and the transmission interval of the communication module.

[0040] In one embodiment, such as Figure 3 As shown, for the multi-factor dynamic early warning threshold model in step S103, the early warning thresholds at each level are... Satisfying the relation: ; In the formula, The allowable inclination benchmark value is designed for the support structure. Risk level coefficient for surrounding protected objects, This is a correction factor for excavation depth, and the threshold is gradually tightened as the excavation depth increases. It is a deformation trend correction coefficient that tightens the threshold as the deformation rate increases.

[0041] In this embodiment, multiple independent influencing factors are organically coupled through multiplication to achieve dynamic calibration of the early warning threshold. Among them, The severity of deformation control in the surrounding environment of the foundation pit is quantified. For example, the protection level required for subway tunnels is higher than that for ordinary municipal pipelines, and the coefficient value will be smaller, thus sounding the alarm in advance. This reflects the spatial effect of foundation pit excavation. The greater the depth, the stronger the soil unloading, the lower the allowable deformation of the support structure, and the threshold should be tightened accordingly. This introduces a risk differential over time. Even if the current total deformation is not large, if it shows a continuous accelerating trend, it indicates that destruction is imminent. Therefore, introducing this coefficient can help detect risks in advance.

[0042] By multiplying the above three dynamic factors by the design baseline value, the final early warning threshold is generated. It is a dynamic value that adapts and adjusts itself in response to changes in the environment, construction progress, and structural response.

[0043] As a specific implementation method, in the multi-factor dynamic early warning threshold model deployed on the cloud monitoring platform, It can be directly imported from engineering drawings and design files. The protection requirements for different levels of external operation impact can be graded and quantified, for example, 0.6 for special protection objects and 0.8 for first-level protection objects. It can be constructed as a decreasing function of the ratio of the current excavation depth to the final depth of the foundation pit. It can be constructed as a deformation rate This is a continuous correction function for the variable. The cloud platform uses these preset rules to automatically refresh the warning thresholds for all monitoring points in the entire foundation pit every calculation cycle, and compares them with real-time monitoring data to achieve differentiated and dynamic risk diagnosis.

[0044] In one embodiment, the edge computing gateway periodically performs online self-calibration of zero-point drift using the spatial geometric constraints of multiple wireless tilt sensing nodes within the same support unit: based on the tilt difference between adjacent measuring points in the initial calibration state, the cumulative zero-point drift of each node is calculated. The system performs online calibration of the current monitoring value without interrupting the normal monitoring process until the final monitoring value is obtained. The following relation is satisfied: ; ; In the formula, This refers to the total number of measuring points within the same support unit. , Measurement points under initial calibration conditions , The reference value for the tilt angle.

[0045] In this embodiment, the online self-calibration mechanism cleverly utilizes the mechanical principle that the relative tilt angle difference between measuring points within the same area should remain relatively stable in the short term as a constraint. If a signal from a certain node is detected... Other nodes If the relative relationship shows a general and consistent shift, it is more likely due to zero-point drift of the node's own sensor rather than a local structural abrupt change. This method, by comparing each measurement point pairwise and averaging the results, accurately separates the cumulative drift error of each node. The algorithm is simple and suitable for efficient operation on resource-constrained devices such as edge computing gateways.

[0046] As a specific implementation method, the triggering time for the zero-drift online self-calibration process can be set to a period of relatively stable monitoring data, such as early morning each day, triggered by a timed task built into the edge computing gateway. The initial reference value upon which the calibration calculation relies... , After the system is built and has been running stably for a period of time, manual calibration or automatic locking through statistical analysis can be performed and the data stored in the edge computing gateway's memory. The entire calibration process runs in the background and does not affect the real-time processing flow of steps S101 and S102. The final monitored value after calibration... It will be used as a data source for subsequent rate calculations and cloud analysis, thereby ensuring that the monitoring system can maintain long-term monitoring accuracy over a continuous operating cycle of several months or even several years.

[0047] As can be seen from the above embodiments, the monitoring method provided in this application overcomes many shortcomings of existing technologies, such as data distortion and low early warning accuracy in complex vibration environments. Through spatiotemporal domain joint filtering at the edge, high-precision monitoring is achieved under periodic disturbances such as train vibration; through an adaptive variable sampling and variable transmission mechanism, the device's endurance is extended while ensuring real-time performance; through a two-level collaborative architecture between the edge and the cloud, local early warning response latency is reduced, and early warning blind spots caused by network outages are eliminated; through multi-node spatial constraint self-calibration, the data deviation of the sensor is reduced over long-term operation. Overall, this application has achieved significant progress in improving monitoring accuracy, balancing power consumption and real-time performance, enhancing the intelligence level of early warning, and improving the long-term reliability of the system.

[0048] This application also provides a real-time remote monitoring system for the tilt of foundation pit support based on a wireless inclinometer, including multiple wireless inclinometer nodes deployed at various monitoring points of the foundation pit support structure, at least one edge computing gateway, and a cloud monitoring platform.

[0049] This wireless tilt sensor node integrates a tilt acquisition unit, a node control unit, a wireless transmission unit, and a power management unit. It is used to acquire raw tilt data and adjust the sampling and uploading parameters according to control commands.

[0050] This edge computing gateway is used to aggregate data from all wireless tilt sensor nodes within its jurisdiction, perform spatiotemporal filtering and correction, tilt change rate calculation and preliminary risk assessment, and issue mode control commands to the sensor nodes.

[0051] This cloud-based monitoring platform is used to receive pre-processed monitoring data, run a multi-factor dynamic early warning threshold model to complete a global risk assessment, and output remote early warning information from multiple terminals.

[0052] In this system, the wireless tilt sensor node, edge computing gateway, and cloud monitoring platform serve as the physical execution carriers for steps S101 to S103 in the above method embodiments, together constituting a fully functional remote monitoring system. The wireless tilt sensor node is responsible for front-end sensing and data acquisition; the edge computing gateway is responsible for data aggregation, purification, and local control decisions at the edge; and the cloud monitoring platform is responsible for global intelligent analysis and early warning dissemination.

[0053] In one embodiment, the power management unit of the wireless tilt sensing node is linked with the node control unit to dynamically adjust the power supply strategy according to the current operating mode: in low-power sleep mode, unnecessary peripheral modules are shut down; in emergency encryption mode, full power supply is provided. The power management unit supports automatic switching between solar power and built-in battery power. This refined power management design maximizes the matching of the three operating modes in the monitoring method, achieving deep power consumption optimization at the hardware-level clock and peripheral power supply levels. For example, in sleep mode, only the real-time clock and watchdog timer are maintained, while in emergency mode, the high-precision sampling front-end and high-speed communication interface are activated. The dual-mode automatic switching power supply strategy further enhances the system's energy autonomy during long-term field operation.

[0054] In one embodiment, the edge computing gateway integrates a local early warning module and an offline data caching unit. During network outages, it independently performs data filtering and correction, and generates local audible and visual alarms. Once the network is restored, it automatically re-uploads the offline cached data to the cloud monitoring platform. This design endows the system with robust autonomous capabilities during network outages. The local early warning module runs a streamlined local risk assessment algorithm consistent with cloud logic, directly triggering the audible and visual alarms connected to the gateway based on the data processed at the edge. The offline data caching unit employs a cyclic overwrite storage strategy to ensure that critical monitoring data is not lost during network outages.

[0055] In one embodiment, the cloud-based monitoring platform incorporates a built-in 3D visualization module. This module combines the foundation pit's BIM model to spatially map and display the tilt data and early warning information of each monitoring point. It also supports multi-channel remote access and early warning information push notifications via web and mobile devices. The 3D visualization module significantly enhances the intuitiveness and efficiency of human-computer interaction with the monitoring data. By integrating the monitoring data with the foundation pit's 3D information model, managers can intuitively view the entire foundation pit's data status on the platform. Risky points will be highlighted on the model, and clicking on them allows tracing their historical data curves. The multi-channel push notification method ensures that early warning information is delivered to all relevant parties in a timely manner.

[0056] This application also provides a method for real-time remote monitoring of foundation pit support tilt based on a wireless inclinometer. This method is applied to the scenario of monitoring the tilt of deep foundation pit support adjacent to an operating rail transit line, where the periodic vibration interference in the surrounding environment is the periodic vibration of the ground surface generated by the operation of rail transit trains. In this application scenario, the spatiotemporal domain joint filtering correction in step S101 mentioned in all the preceding embodiments of this application, which needs to identify and cancel the reference period and amplitude characteristics, is clearly defined as the characteristics of the periodic vibration of the ground surface generated by the operation of rail transit trains. This makes the application of the method more targeted, and its effect is fully realized in this high-risk scenario. For example, it can more accurately cancel the strong interference caused by subway operation and ensure the safety of deep foundation pit construction next to an operating tunnel.

[0057] The embodiments disclosed in this invention are preferred embodiments, but are not limited thereto. Those skilled in the art can easily understand the spirit of this invention based on the above embodiments and make different extensions and variations, but as long as they do not depart from the spirit of this invention, they are all within the protection scope of this invention.

Claims

1. A method for real-time remote monitoring of the tilt of foundation pit support based on a wireless inclinometer, characterized in that, include: The raw tilt data collected by each wireless tilt sensor node is first aggregated to the edge computing gateway, where spatiotemporal joint filtering correction is performed at the edge. The tilt rate of change per unit time is calculated based on the real tilt signal after filtering and correction. Two-level rate thresholds are preset to divide the three working modes into low-power sleep, regular monitoring and emergency encryption. According to the level of tilt rate of change, the sampling frequency and wireless data upload frequency of the wireless tilt sensing node are adjusted synchronously. The higher the deformation rate, the greater the sampling and upload density. The tilt angle data after edge preprocessing is uploaded to the cloud monitoring platform. The cloud uses the allowable tilt value of the support structure design as the benchmark threshold, and introduces the risk level weighting coefficient of the surrounding protected objects, the excavation depth correction coefficient and the deformation trend correction coefficient to construct a multi-factor dynamic early warning threshold model. Combined with the tilt angle data of multiple points in the entire foundation pit, a global risk classification assessment is carried out, multi-level early warning instructions are generated and remote early warning information is output to multiple terminals.

2. The method for real-time remote monitoring of foundation pit support tilt based on a wireless inclinometer according to claim 1, characterized in that, The specific steps of performing spatiotemporal joint filtering correction on the edge side are as follows: First, the reference period and amplitude characteristics of the periodic vibration of the surrounding environment are identified through a time-domain sliding window to generate a standard vibration interference template. Then, combined with the spatial position constraint relationship of multiple measuring points within the same support unit, the standard vibration interference template is adapted for point amplitude attenuation and propagation phase difference. The adapted periodic vibration interference component is canceled from the original tilt angle data to extract the true tilt signal of the support structure. In the spatiotemporal joint filtering correction, the calculation of the true tilt signal satisfies the following relationship: ； In the formula, For the first The actual tilt signal at each measuring point This is the original dip angle data for this measuring point. For standard vibration interference template, This is the spatial amplitude correction coefficient for the measuring point, and it is inversely proportional to the square root of the horizontal distance from the measuring point to the vibration source. The phase difference is the vibration propagation to the measuring point and is calculated from the distance from the measuring point to the vibration source and the shear wave velocity of the soil.

3. The method for real-time remote monitoring of foundation pit support tilt based on a wireless inclinometer according to claim 1, characterized in that, The three-level working mode switching configuration delay anti-shake mechanism: the mode switching can only be triggered when the tilt angle change rate continuously crosses the threshold and remains for a preset time, to prevent frequent erroneous switching of sampling and uploading modes caused by periodic vibration fluctuations.

4. The method for real-time remote monitoring of foundation pit support tilt based on a wireless inclinometer according to claim 1, characterized in that, In the multi-factor dynamic early warning threshold model, the early warning thresholds at each level Satisfying the relation: ； In the formula, The allowable inclination benchmark value is designed for the support structure. Risk level coefficient for surrounding protected objects, This is a correction factor for excavation depth, and the threshold is gradually tightened as the excavation depth increases. It is a deformation trend correction coefficient that tightens the threshold as the deformation rate increases.

5. A method for real-time remote monitoring of foundation pit support tilt based on a wireless inclinometer according to claim 1, characterized in that, The edge computing gateway periodically performs online self-calibration of zero-point drift using the spatial geometric constraints of multiple wireless tilt sensor nodes within the same support unit: based on the tilt difference between adjacent measuring points in the initial calibration state, the cumulative zero-point drift of each node is calculated. It also performs online correction of the current monitoring value, and the calibration process does not interrupt the normal monitoring process.

6. A real-time remote monitoring system for the tilt of foundation pit support based on a wireless inclinometer, characterized in that, It includes multiple wireless tilt sensor nodes deployed at various monitoring points on the foundation pit support structure, at least one edge computing gateway, and a cloud monitoring platform; The wireless tilt sensing node has a built-in tilt acquisition unit, node control unit, wireless transmission unit and power management unit, which are used to acquire raw tilt data and adjust the sampling and uploading parameters according to the control command; The edge computing gateway is used to aggregate data from wireless tilt sensor nodes within its jurisdiction, perform spatiotemporal filtering correction, tilt change rate calculation and preliminary risk assessment, and send mode control commands to the sensor nodes. The cloud-based monitoring platform is used to receive preprocessed monitoring data, run a multi-factor dynamic early warning threshold model to complete global risk assessment, and output remote early warning information from multiple terminals.

7. A real-time remote monitoring system for the tilt of foundation pit support based on a wireless inclinometer, as described in claim 6, is characterized in that, The power management unit of the wireless tilt sensing node is linked with the node control unit to dynamically adjust the power supply strategy according to the current working mode: in low-power sleep mode, unnecessary peripheral modules are turned off, and in emergency encryption mode, full power supply is provided. The power management unit supports automatic switching between solar power and built-in battery power.

8. A real-time remote monitoring system for the tilt of foundation pit support based on a wireless inclinometer, as described in claim 6, is characterized in that, The edge computing gateway has a built-in local early warning module and an offline data caching unit. It can independently complete data filtering and correction and local sound and light early warning in the event of a network interruption with the cloud. After the network is restored, it automatically re-transmits the offline cached data to the cloud monitoring platform.

9. A real-time remote monitoring system for the tilt of foundation pit support based on a wireless inclinometer, as described in claim 6, is characterized in that, The cloud-based monitoring platform has a built-in 3D visualization module, which can spatially map and display the tilt data and early warning information of each monitoring point in conjunction with the foundation pit BIM model. It also supports multi-channel remote access and early warning information push from web and mobile terminals.

10. A method for real-time remote monitoring of the tilt of foundation pit support based on a wireless inclinometer, characterized in that, This method is applied to the scenario of monitoring the tilt of deep foundation pit support adjacent to an operating rail transit line, where the periodic vibration interference of the surrounding environment is the periodic vibration of the ground surface generated by the operation of rail transit trains.

Images