Road culvert water leakage monitoring system and method based on wireless sensor network

Abstract

The invention relates to the technical field of permeability testing, in particular to a road culvert water leakage monitoring system and method based on a wireless sensor network. Firstly, particle leakage data collected by a sensor in different seasons are obtained and processed, and particle leakage time sequence features are obtained; particle leakage change rate characteristics are obtained according to the particle leakage time sequence characteristics, and a particle leakage monitoring area is determined according to the particle leakage change rate characteristics; setting the position of the sensor network according to the particle leakage change rate characteristics; acquiring vehicle load data and vibration data of the ground and the culvert in the monitoring time, establishing a temperature-vibration compensation item, obtaining a leakage monitoring curve, and analyzing the short-time permeability of particles; sampling to obtain suspended particle concentration of a sensor monitoring position; and a vehicle load-particle leakage rate correlation model is constructed, vibration data, particle short-time permeability, particle volume short-time change rate and other data are input, particle leakage indexes are output, and whether leakage exists or not is judged.

Description

Technical Field

[0001] The present invention relates to the technical field of permeability testing, and specifically to a highway culvert leakage monitoring system and method based on a wireless sensor network. Background Technique

[0002] Leakage monitoring is a key link in ensuring the safety of transportation infrastructure. The core lies in accurately evaluating the permeability, pore volume, and particle loss characteristics of porous media. Traditional methods rely on laboratory tests, which require destructive sampling and take more than 24 hours, and cannot capture the dynamic changes at the culvert site, such as the fluctuation of the particle leakage rate caused by vehicle vibration. Existing on-site monitoring technologies can only obtain a single parameter, and it is difficult to quantify the coupling relationship between water permeability and soil particle loss rate, resulting in insufficient leakage warning accuracy.

[0003] Therefore, a highway culvert leakage monitoring system and method based on a wireless sensor network are proposed. Summary of the Invention

[0004] The purpose of the present invention is to provide a highway culvert leakage monitoring system and method based on a wireless sensor network. First, by obtaining and processing the particle leakage data collected by sensors in different seasons, the particle leakage time series characteristics are obtained; according to the particle leakage time series characteristics, the particle leakage change rate characteristics are obtained, and according to the particle leakage change rate characteristics, the particle leakage monitoring area is determined; according to the particle leakage change rate characteristics, the positions of the sensor network are set; the vehicle load data, the vibration data of the ground and the culvert during the monitoring time are obtained, a temperature-vibration compensation term is established, a leakage monitoring curve is obtained, and the short-term particle permeability is analyzed; the suspended particle concentration at the sensor monitoring position is sampled; a vehicle load-particle leakage rate correlation model is constructed, and data such as vibration data, short-term particle permeability, and short-term change rate of particle volume are input, and a particle leakage index is output to determine whether there is leakage.

[0005] To achieve the above purpose, the present invention provides the following technical solutions:

[0006] A highway culvert leakage monitoring method based on a wireless sensor network, including:

[0007] Obtaining and processing the particle leakage data collected by sensors in different seasons to obtain particle leakage time series characteristics; obtaining particle leakage change rate characteristics according to the particle leakage time series characteristics, and determining the particle leakage monitoring area according to the particle leakage change rate characteristics;

[0008] Further, the specific steps for obtaining the particle leakage monitoring area are as follows: The particle leakage data includes pore water pressure data, temperature data, joint displacement data, and leakage flow rate data; the particle leakage data is processed through Fourier transform, Hodrick-Prescott filter, and Bayesian change point analysis to obtain three types of particle leakage time series characteristics, including periodic characteristics, trend characteristics, and mutation characteristics; the daily, weekly, monthly, and quarterly change rate characteristics of the three types of particle leakage time series characteristics are constructed to obtain the particle leakage change rate characteristics; the change rate characteristics are processed to output the particle leakage monitoring area.

[0009] Obtain the location of the particle leakage monitoring area and set the location of the sensor network according to the particle leakage change rate characteristics;

[0010] Further, the specific steps for setting the location of the sensor network are as follows: Perform clustering analysis on the particle leakage time series characteristics of the historical monitoring points in the particle leakage monitoring area to obtain the sensor placement points and arrange the sensor network.

[0011] Obtain the vehicle load data, ground and culvert vibration data during the monitoring time, establish a temperature-vibration compensation term, obtain the leakage monitoring curve, and analyze the short-term particle permeability; sample the suspended particle concentration at the sensor monitoring location;

[0012] Further, obtain the vehicle load, traffic flow, highway width, road surface vibration, and culvert vibration data during the monitoring time, and establish a leakage data collection index. The formula is:

[0013]

[0014] where LR represents the leakage data collection index, Q represents the vehicle load within the time window, N represents the traffic flow within the time window, W represents the highway width, E road represents the road surface vibration frequency, E culvert represents the culvert vibration frequency, E0 represents the reference vibration frequency, and γ represents the vibration enhancement coefficient.

[0015] Further, if the leakage data collection index is always greater than the threshold value, continuously collect the leakage data during this time period, use the temperature-vibration compensation term to correct the seepage flow rate, eliminate the interference of environmental temperature and mechanical vibration on the piezometric sensor, obtain the true seepage flow rate, and construct the leakage monitoring curve during the collection time according to the true seepage flow rate.

[0016] Further, the steps for obtaining the short-term particle permeability, short-term particle volume change rate, and short-term pore area change rate include:

[0017] Calculate the mean value of the derivative of the leakage monitoring curve during the monitoring time to obtain the short-term particle permeability K;

[0018] Sample the pore volume within the monitoring time and subtract it from the initial pore volume to obtain the short-term change rate V of the particle volume; sample the pore area within the monitoring time and subtract it from the initial pore area to obtain the short-term change rate S of the pore area.

[0019] Construct a vehicle load - particle leakage rate correlation model, input vibration data, short-term particle permeability, short-term change rate V of the particle volume, short-term change rate S of the pore area, and suspended particle concentration, and output the particle leakage index.

[0020] Furthermore, if the particle leakage index exceeds the threshold, mark the particle leakage monitoring area as a leakage point for early warning.

[0021] Furthermore, the particle leakage index includes a vibration - leakage coupling term, a pore structure instability term, a hydraulic - leakage synergy term, and a joint deformation term, and the calculation formula is:

[0022]

[0023] Among them, R represents the particle leakage index, A rms represents the maximum vibration eigenvalue within the acquisition time, A0 represents the average vibration eigenvalue within the historical monitoring time, K represents the short-term particle permeability, K0 represents the initial permeability, that is, the inherent permeability of the soil under the condition of no load on the road, exp() represents the exponential function, V represents the short-term change rate of the particle volume, S represents the short-term change rate of the pore area, ε represents the pore structure stability coefficient, P pore represents the suspended particle concentration, P crit represents the average value of the historical suspended particle concentration, K′ represents the absolute value of the maximum permeability mutation, K th represents the permeability mutation threshold, that is, the critical permeability at which the pore seepage changes from linear to non-linear, tanh() represents the hyperbolic tangent function, and C represents the degree of joint deformation.

[0024] The present invention also provides a highway culvert leakage monitoring system based on a wireless sensor network, including:

[0025] A historical leakage data processing module, including a leakage feature extraction unit and a particle leakage monitoring area identification unit, where the leakage feature extraction unit obtains and processes the particle leakage data collected by historical sensors to obtain the particle leakage time series features; obtains the particle leakage change rate features according to the particle leakage time series features; the particle leakage monitoring area identification unit determines the particle leakage monitoring area according to the particle leakage change rate features;

[0026] A sensor arrangement module, used to obtain the location of the particle leakage monitoring area and arrange a sensor network according to the particle leakage change rate features;

[0027] The leakage data acquisition module includes an acquisition trigger unit, a data acquisition unit, and a data analysis unit. The acquisition trigger unit obtains leakage vibration data during the monitoring time and establishes leakage data acquisition indicators. If it is greater than the threshold, data acquisition is performed. The data acquisition unit performs data acquisition to obtain a leakage monitoring curve. The data analysis unit processes the acquired data to obtain the processed leakage data.

[0028] The leakage monitoring module constructs a vehicle load - particle leakage rate correlation model, inputs vibration data, short - term particle permeability, short - term change rate of particle volume, short - term change rate of pore area, and suspended particle concentration, and outputs particle leakage indicators.

[0029] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0030] 1. By extracting and analyzing the data characteristics of the particle leakage data collected at different times, subtly and crucially changing in the historical leakage process can be accurately captured at multiple levels and scales, finding out the joint areas that may be abnormal, improving the sensitivity and accuracy of monitoring; using historical data and clustering analysis algorithms, the optimal sensor placement points can be determined, and the sensor network can be reasonably arranged to achieve high - density monitoring in abnormal joint areas, ensuring monitoring accuracy and coverage.

[0031] 2. By comprehensively constructing leakage data acquisition indicators with vehicle load, traffic flow, highway width, road surface, and culvert vibration data, the leakage risk under actual road conditions can be fully reflected; through the processing of the temperature - vibration compensation term, the interference of environmental temperature fluctuations and mechanical vibrations on the signal of the osmotic pressure sensor can be eliminated, so as to obtain the true seepage flow calculated based on Darcy's law, and thus generate a detailed monitoring curve; through further processing of the acquired data, data support is provided for judging the leakage state inside the culvert.

[0032] 3. By constructing particle leakage indicators with vibration - leakage coupling terms, pore structure instability terms, hydraulic - leakage synergy terms, and joint deformation terms, considering the interference of environmental and other factors, accurate monitoring and dynamic early warning of the culvert leakage phenomenon caused by vehicle load are realized, providing strong technical support for the safe operation and maintenance decision - making of highway culverts. Description of the Drawings

[0033] Figure 1 It is a flow chart of a highway culvert leakage and water seepage monitoring method based on a wireless sensor network provided by an embodiment of the present invention.

[0034] Figure 2 It is a flow chart of calculating new particle leakage indicators provided by an embodiment of the present invention.

[0035] Figure 3Schematic diagram of the structure of a highway culvert leakage monitoring system based on a wireless sensor network provided by an embodiment of the present invention. Detailed implementation manners

[0036] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

[0037] Embodiment 1:

[0038] A certain company introduced a highway culvert leakage monitoring method based on a wireless sensor network provided by the present invention to improve the accuracy and efficiency of highway culvert leakage monitoring. The method flow is as Figure 1 shown, and the specific implementation manners are as follows:

[0039] Obtain and process the particle leakage data of different seasons collected by historical sensors to obtain the particle leakage time series characteristics; obtain the particle leakage change rate characteristics according to the particle leakage time series characteristics, and determine the particle leakage monitoring area according to the particle leakage change rate characteristics;

[0040] Furthermore, the specific steps for obtaining the particle leakage monitoring area include:

[0041] The particle leakage data includes pore water pressure data, temperature data, joint displacement data, and leakage flow data. Table 1 shows some particle leakage data;

[0042] Table 1. Some particle leakage data

[0043] Timestamp Pore water pressure (kPa) Temperature (°C) Joint displacement Leakage flow rate (ml) 2023-12-10 08:00:00 35.2 -5.3 0.02 80 2023-12-10 12:00:00 62.7 -2.1 0.08 152 2023-12-10 16:00:00 38.9 -2.8 -0.05 94 2023-12-10 20:00:00 28.4 -8.3 0.09 59

[0044] Process the particle leakage data through Fourier transform, use the Hodrick-Prescott filter to separate the long-term trend, and identify sudden jump points based on Bayesian change point analysis (BCP) to obtain three types of particle leakage time series characteristics, including periodic characteristics, trend characteristics, and mutation characteristics; construct the daily, weekly, monthly, and quarterly change rate characteristics of the three types of particle leakage time series characteristics to obtain the particle leakage change rate characteristics, and finally process the particle leakage change rate characteristics to output the particle leakage monitoring area.

[0045] Further, first, preprocess the characteristics of pore water pressure data, temperature data, joint displacement data, and leakage flow data through Fourier transform, Hodrick-Prescott filter, and Bayesian change point analysis to obtain periodic time series characteristics, trend time series characteristics, and mutation time series characteristics, forming a feature matrix with a structure of 4×3;

[0046] Further, use convolutional neural network, recurrent neural network, and transformer mechanism to construct daily, weekly, monthly, and quarterly change rate characteristics of the above three types of key time series characteristics, and finally output the number of the particle leakage monitoring area through a feedforward neural network.

[0047] First, separate the three time series characteristics of period, trend, and mutation, and then construct daily, weekly, monthly, and quarterly change rate characteristics respectively, which can comprehensively capture the dynamic changes of the leakage process from different time scales, thereby improving the accuracy and timeliness of anomaly detection.

[0048] Obtain the location of the particle leakage monitoring area and arrange the sensor network according to the particle leakage change rate characteristics;

[0049] Further, the specific steps for setting the location of the sensor network include: obtaining the data of historical monitoring points and performing cluster analysis on the particle leakage time series characteristics of the monitoring points in the particle leakage monitoring area to obtain the sensor placement points and arrange the sensor network.

[0050] Further, in the above process, a matrix with a structure of 4×3 is obtained. Obtain all the matrix data of the monitoring points in the particle leakage monitoring area, and perform cluster analysis through the K-means clustering algorithm to obtain several cluster centers. Then, set the positions of the cluster centers as the sensor placement points and arrange the sensor network.

[0051] Through cluster analysis of the particle leakage time series characteristics of historical monitoring points, the most representative feature patterns in the area can be extracted; the cluster centers obtained by using K-means clustering can represent the key points in the monitoring area, thus making the arrangement of sensors more scientific and accurate, and ensuring that the data collection covers the most typical and sensitive positions in the area.

[0052] Obtain the leakage vibration data during the monitoring time and establish leakage data collection indicators. If it is greater than the threshold, data collection is performed to obtain the leakage monitoring curve, and the leakage data at the sensor monitoring position is sampled to obtain the suspended particle concentration;

[0053] Table 2. Highway-related data

[0054]

[0055] Furthermore, the construction of leakage data acquisition indicators includes: leakage vibration data including vehicle load, traffic flow, highway width, road surface vibration, and culvert vibration data. Obtain the vehicle load, traffic flow, highway width, road surface vibration, and culvert vibration data within the monitoring time, and establish leakage data acquisition indicators. The formula is:

[0056]

[0057] Among them, LR represents the leakage data acquisition indicator, Q represents the vehicle load within the time window, N represents the traffic flow within the time window, W represents the highway width, E road represents the road surface vibration frequency, E culvert represents the culvert vibration frequency, E0 represents the reference vibration frequency, an empirical threshold, the average value monitored continuously for 24 hours under the static load state of the highway, γ represents the vibration enhancement coefficient, indicating the amplification effect of the vibration frequency on the leakage risk, which can be calibrated through soil dynamic tests.

[0058] Furthermore, sensors are also installed on the highway to monitor vehicle load, traffic flow, vibration data, etc. Table 2 shows the relevant data on the highway monitored at some time points;

[0059] By constructing a leakage data acquisition indicator that comprehensively considers vehicle load, traffic flow, highway width, road surface vibration, and culvert vibration data, data acquisition can be automatically triggered by the threshold, effectively improving the accuracy, real-time performance, and intelligent level of culvert leakage monitoring, thus providing a solid technical guarantee for the structural safety management and early warning of culverts.

[0060] Obtain the vehicle load data, ground and culvert vibration data during the acquisition time, establish a temperature-vibration compensation term, obtain the leakage monitoring curve, and analyze the short-term permeability of particles; sample the suspended particle concentration at the sensor monitoring position;

[0061] Furthermore, the steps to obtain the leakage monitoring curve include:

[0062] If the leakage data acquisition indicator is always greater than the threshold, enter the acquisition time. During this period, continuously collect leakage data, use the temperature-vibration compensation term to correct the seepage flow rate, eliminate the interference of environmental temperature and mechanical vibration on the piezometric sensor, and obtain the true seepage flow rate. The formula is:

[0063]

[0064] Among them, Q c (t) represents the true seepage flow rate at time t, Q raw(t) represents the measured seepage flow rate obtained according to Darcy's law, α represents the temperature sensitivity coefficient, which can be calibrated in the laboratory, ΔT(t) represents the offset of the temperature at the monitoring location relative to the calibrated temperature, β represents the coupling coefficient of vibration to the sensor sensitivity, calibrated according to actual measurement, a rms (t) represents the root mean square value of the vibration acceleration;

[0065] Construct a leakage monitoring curve during the acquisition time based on the true seepage flow rate.

[0066] By introducing a temperature-vibration compensation term, the interference of ambient temperature changes and mechanical vibration on the osmotic pressure sensor can be effectively eliminated, thereby obtaining the true seepage flow rate obtained according to Darcy's law. This ensures that the collected data is more real, reduces errors, and improves the overall accuracy of the monitoring system.

[0067] Further, the steps to obtain the short-term permeability of particles, the short-term change rate of particle volume, and the short-term change rate of pore area include:

[0068] Calculate the mean value of the derivative of the leakage monitoring curve during the monitoring time to obtain the short-term permeability K of the particles;

[0069] Sample the pore volume during the monitoring time and subtract it from the initial pore volume to obtain the short-term change rate V of the particle volume; sample the pore area during the monitoring time and subtract it from the initial pore area to obtain the short-term change rate S of the pore area.

[0070] Further, the specific steps to obtain the suspended particle concentration include:

[0071] Step S1: Use a dedicated turbidity or laser scattering sensor at the sensor monitoring location to collect the optical signal in the leakage water in real time. This signal is closely related to the concentration of suspended particles in the water;

[0072] Step S2: Filter, amplify, and denoise the collected original optical or electrical signal to eliminate the influence of environmental interference (such as light changes, temperature fluctuations, etc.) and ensure the accuracy of subsequent data conversion;

[0073] Step S3: Use the standard curve pre-calibrated in the laboratory to convert the processed sensor signal into the suspended particle concentration value; during the calibration process, the change in sensor sensitivity caused by temperature or other environmental factors can be considered for compensation.

[0074] Through the extraction and processing of the above data, the accuracy and reliability of data measurement are improved, and the leakage state in the culvert and the dynamic changes of the pore structure can be comprehensively and accurately reflected, providing real-time data support and scientific decision-making basis for the safety monitoring of the culvert.

[0075] Construct a vehicle load - particle leakage rate correlation model, input features such as vibration data, short - term particle permeability, short - term change rate of particle volume, short - term change rate of pore area, suspended particle concentration, etc., and output particle leakage indicators.

[0076] Furthermore, if the particle leakage indicator exceeds the threshold, the particle leakage monitoring area is marked as a leakage point for early warning.

[0077] Furthermore, the vehicle load - pore leakage rate correlation model is to explore the correlation between vehicle data, highway data, and leakage data. Its input data includes features such as vibration data, short - term particle permeability, short - term change rate of particle volume, short - term change rate of pore area, suspended particle concentration, etc. It constructs a vibration - leakage coupling term, a pore structure instability term, a hydraulic - leakage synergy term, and a joint deformation term, and further obtains the particle leakage indicator. The calculation formula is:

[0078]

[0079] Among them, R represents the particle leakage indicator, A rms represents the maximum vibration eigenvalue during the acquisition time, A0 represents the average vibration eigenvalue during the historical monitoring time, K represents the short - term particle permeability, K0 represents the initial permeability, that is, the inherent permeability of the soil under the condition of no vehicle load on the highway, exp() represents the exponential function, V represents the short - term change rate of particle volume, S represents the short - term change rate of pore area, ε represents the pore structure stability coefficient, an empirical parameter related to the soil strength, and the smaller the value, the more unstable the structure. P pore represents the suspended particle concentration, P crit represents the average value of the historical suspended particle concentration, K′ represents the absolute value of the maximum permeability mutation, K th represents the permeability mutation threshold, which is the critical permeability marking the transition of pore seepage from linear to non - linear, tanh() represents the hyperbolic tangent function, and C represents the degree of joint deformation. As shown in Table 3, the particle leakage indicators for some acquisition time periods in the third monitoring area are all greater than the threshold of 1. Through the particle leakage indicators of multiple acquisition time periods, it can be judged that there is indeed a leakage risk in the third monitoring area.

[0080] Table 3. Particle Leakage Indicators

[0081] Collection time period Particle leakage index From 2024-8-23 12:54:42 to 12:59:01 1.23 From 2024-8-23 13:24:52 to 13:29:26 1.20 From 2024-8-23 14:23:24 to 14:35:48 1.39

[0082] By continuously monitoring and analyzing the particle leakage indicators, taking into account factors such as vibration caused by vehicle load, pore structure changes, hydraulic conditions, and joint deformation, it can more comprehensively reflect the particle leakage behavior under the actual highway conditions and improve the accuracy of the assessment.

[0083] A method for monitoring leakage of highway culverts based on a wireless sensor network provided by the present invention first processes historical data and extracts features, fully utilizes the historical data to mine areas that may have leakage risks, and performs real-time data collection and analysis; then, by introducing a temperature-vibration compensation term, the interference of ambient temperature and the vibration of the highway and culvert on the sensor measurement data is effectively eliminated, and the accuracy of the leakage monitoring data is significantly improved; at the same time, a particle leakage index is constructed by using key influencing factors such as pore structure instability, hydraulic-leakage synergy, and joint deformation, further enhancing the ability to identify leakage risks and providing strong support for the maintenance and safety management of highway culverts.

[0084] Embodiment 2:

[0085] The present invention also provides a system for monitoring leakage of highway culverts based on a wireless sensor network. The system structure is as Figure 3 shown, and the specific implementation method is as follows:

[0086] The historical leakage data processing module includes a leakage feature extraction unit and a particle leakage monitoring area identification unit. The leakage feature extraction unit obtains the particle leakage data collected by historical sensors and processes it to obtain the time series features of particle leakage; the change rate feature of particle leakage is obtained according to the time series features of particle leakage; the particle leakage monitoring area identification unit determines the particle leakage monitoring area according to the change rate feature of particle leakage.

[0087] Furthermore, the specific steps for obtaining the particle leakage monitoring area include: the particle leakage data includes pore water pressure data, temperature data, joint displacement data, and leakage flow data; the particle leakage data is processed by Fourier transform, Hodrick-Prescott filter, and Bayesian change point analysis to obtain three types of time series features of particle leakage, including periodic features, trend features, and mutation features; the daily, weekly, monthly, and quarterly change rate features of the three types of time series features of particle leakage are constructed to obtain the change rate feature of particle leakage; the change rate feature is processed to output the particle leakage monitoring area.

[0088] The sensor arrangement module is used to obtain the positions of the particle leakage monitoring areas and arrange a sensor network according to the change rate feature of particle leakage.

[0089] Furthermore, the specific steps for setting the positions of the sensor network include: performing cluster analysis on the time series features of particle leakage at the historical monitoring points in the particle leakage monitoring area to obtain the sensor placement points and arrange the sensor network.

[0090] Leakage data acquisition module, including an acquisition trigger unit, a data acquisition unit, and a data analysis unit. The acquisition trigger unit obtains leakage vibration data during the monitoring time and establishes leakage data acquisition indicators. If it is greater than the threshold, data acquisition is performed. The data acquisition unit performs data acquisition to obtain a leakage monitoring curve. The data analysis unit processes the acquired data to obtain processed leakage data.

[0091] Furthermore, the processed leakage data includes the true seepage flow rate, leakage monitoring curve, short-term particle permeability, short-term change rate of particle volume, short-term change rate of pore area, etc.

[0092] Furthermore, constructing the leakage data acquisition indicators includes: obtaining vehicle load, traffic flow, highway width, road surface vibration, and culvert vibration data during the monitoring time, and establishing leakage data acquisition indicators. The formula is:

[0093]

[0094] where LR represents the leakage data acquisition indicator, Q represents the vehicle load within the time window, N represents the traffic flow within the time window, W represents the highway width, E road represents the road surface vibration frequency, E culvert represents the culvert vibration frequency, E0 represents the reference vibration frequency, and γ represents the vibration enhancement coefficient.

[0095] Furthermore, if the leakage data acquisition indicator is always greater than the threshold, leakage data is continuously acquired during this time period. The seepage flow rate is corrected using the temperature-vibration compensation term to eliminate the interference of environmental temperature and mechanical vibration on the piezometric sensor, and the true seepage flow rate is obtained. A leakage monitoring curve during the acquisition time is constructed based on the true seepage flow rate.

[0096] Furthermore, the data analysis unit also includes: calculating the mean value of the derivative of the leakage monitoring curve during the monitoring time to obtain the short-term particle permeability K;

[0097] Sampling the pore volume during the monitoring time and taking the difference with the initial pore volume to obtain the short-term change rate of particle volume V; sampling the pore area during the monitoring time and taking the difference with the initial pore area to obtain the short-term change rate of pore area S.

[0098] Leakage monitoring module, constructing a vehicle load-particle leakage rate correlation model, inputting vibration data, short-term particle permeability, short-term change rate of particle volume, short-term change rate of pore area, suspended particle concentration, and outputting particle leakage indicators.

[0099] Furthermore, if the particle leakage indicator exceeds the threshold, the particle leakage monitoring area is marked as a leakage point for early warning.

[0100] Furthermore, the particle leakage index includes a vibration-leakage coupling term, a pore structure instability term, a hydraulic-leakage synergy term, and a joint deformation term. In this embodiment, a rain correction term is further introduced to construct the particle leakage index. The structure is as shown in Figure 2 and the calculation formula is:

[0101]

[0102] where R represents the particle leakage index, A rms represents the maximum vibration eigenvalue during the acquisition time, A0 represents the mean vibration eigenvalue during the historical monitoring time, K represents the short-term particle permeability, K0 represents the initial permeability, that is, the inherent permeability of the soil under the condition of no load on the road, exp() represents the exponential function, V represents the short-term change rate of the particle volume, S represents the short-term change rate of the pore area, ε represents the pore structure stability coefficient, an empirical parameter related to the soil strength, and the smaller the value, the easier the structure is to be unstable, P pore represents the suspended particle concentration, P crit represents the mean value of the historical suspended particle concentration, K′ represents the absolute value of the maximum permeability mutation, K th represents the permeability mutation threshold, which is the critical permeability indicating the transition of pore seepage from linear to nonlinear, tanh() represents the hyperbolic tangent function, C represents the degree of joint deformation; Y represents the rainfall per unit time, D represents the continuous rainfall time, S0 represents the initial soil saturation, H(t - t rain ) represents the rainfall lag function, the Heaviside step function, indicating the delay effect of rainfall on leakage, t rain represents the rainfall start time, and t represents the acquisition end time. As shown in Table 4, the results calculated by the new particle leakage index are all less than the threshold value of 1.2, so there is no leakage risk at this position.

[0103] Table 4. Particle Leakage Index

[0104] Collection time period Particle leakage index From 2024-12-23 9:24:45 to 9:26:02 0.92 From 2024-12-23 9:56:23 to 10:03:18 0.87 From 2024-12-23 10:14:29 to 10:18:41 0.72

[0105] Although the embodiments of the present invention have been shown and described, those of ordinary skill in the art can understand that various changes, modifications, substitutions, and variations can be made to these embodiments without departing from the principles and spirit of the present invention. The scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A method for monitoring leakage of highway culverts based on a wireless sensor network, characterized in that, Including: Obtain and process the particulate leakage data collected by sensors in different seasons to obtain the particulate leakage time series characteristics; obtain the particulate leakage change rate characteristics based on the particulate leakage time series characteristics, and determine the particulate leakage monitoring area according to the particulate leakage change rate characteristics; Obtain the location of the particulate leakage monitoring area and set the location of the sensor network according to the particulate leakage change rate characteristics; Obtain the vehicle load data, ground and culvert vibration data during the acquisition time, establish a temperature-vibration compensation term, obtain the leakage monitoring curve, and analyze the short-term particulate permeability; sample the suspended particulate concentration at the sensor monitoring location; Construct a vehicle load-particulate leakage rate correlation model, input the vibration data, short-term particulate permeability, short-term particulate volume change rate, short-term pore area change rate, and suspended particulate concentration, and output the particulate leakage index.

2. The method for monitoring leakage of highway culverts based on a wireless sensor network according to claim 1, wherein The specific steps for obtaining the particulate leakage monitoring area include: the particulate leakage data includes pore water pressure data, temperature data, joint displacement data, and leakage flow data; process the particulate leakage data through Fourier transform, Hodrick-Prescott filter, and Bayesian change point analysis to obtain three types of particulate leakage time series characteristics, including periodic characteristics, trend characteristics, and mutation characteristics; construct the daily, weekly, monthly, and quarterly change rate characteristics of the three types of particulate leakage time series characteristics to obtain the particulate leakage change rate characteristics; process the particulate leakage change rate characteristics and output the particulate leakage monitoring area.

3. A method for monitoring leakage of highway culverts based on a wireless sensor network according to claim 1, characterized in that, The specific steps for setting the location of the sensor network include: perform cluster analysis on the particulate leakage time series characteristics of the historical monitoring points in the particulate leakage monitoring area to obtain the sensor placement points and arrange the sensor network.

4. The method for monitoring leakage of highway culverts based on a wireless sensor network according to claim 1, characterized in that, Obtain the vehicle load, traffic flow, highway width, road surface vibration, and culvert vibration data during the monitoring time, and establish a leakage data acquisition index, the formula is: Among them, LR represents the leakage data acquisition index, Q represents the vehicle load within the time window, N represents the traffic flow within the time window, W represents the highway width, and E road represents the pavement vibration frequency, E culvert represents the culvert vibration frequency, E0 represents the reference vibration frequency, and γ represents the vibration enhancement coefficient.

5. The method for monitoring leakage of highway culverts based on wireless sensor network according to claim 4, characterized in that, If the leakage data acquisition index is always greater than the threshold, continuously collect the leakage data during this time period, use the temperature-vibration compensation term to correct the seepage flow rate, eliminate the interference of environmental temperature and mechanical vibration on the piezometric sensor, obtain the true seepage flow rate, and construct the leakage monitoring curve during the acquisition time according to the true seepage flow rate.

6. The method for monitoring leakage of highway culverts based on a wireless sensor network according to claim 1, characterized in that, The steps for obtaining the short-term particulate permeability, short-term particulate volume change rate, and short-term pore area change rate include: Calculate the mean value of the derivative of the leakage monitoring curve during the monitoring time to obtain the short-term particulate permeability K; Sample the pore volume during the monitoring time and perform a difference operation with the initial pore volume to obtain the short-term particulate volume change rate V; sample the pore area during the monitoring time and perform a difference operation with the initial pore area to obtain the short-term pore area change rate S.

7. A method for monitoring leakage of highway culverts based on a wireless sensor network according to claim 1, characterized in that, If the particulate leakage index exceeds the threshold, mark the particulate leakage monitoring area as a leakage point for early warning.

8. A method for monitoring leakage of highway culverts based on a wireless sensor network according to claim 1, characterized in that, The particulate leakage index includes a vibration-leakage coupling term, a pore structure instability term, a hydraulic-leakage synergy term, and a joint deformation term, and the calculation formula is: Among them, R represents the particle leakage index, A rms represents the maximum vibration eigenvalue during the acquisition time, A0 represents the mean value of the vibration characteristics during the historical monitoring time, K represents the short-term particle permeability, K0 represents the initial permeability, that is, the inherent permeability of the soil under the condition of no load on the road, exp() represents the exponential function, V represents the short-term change rate of the particle volume, S represents the short-term change rate of the pore area, ε represents the pore structure stability coefficient, P pore represents the suspended particle concentration, P crit represents the mean value of the historical suspended particle concentration, K′ represents the absolute value of the maximum permeability mutation, K th represents the permeability mutation threshold, that is, the critical permeability at which the pore seepage changes from linear to nonlinear, tanh() represents the hyperbolic tangent function, and C represents the degree of joint deformation.

9. A highway culvert leakage monitoring system based on a wireless sensor network, characterized in that, Including: The historical leakage data processing module includes a leakage feature extraction unit and a particulate leakage monitoring area identification unit. The leakage feature extraction unit acquires and processes the particulate leakage data collected by historical sensors to obtain the particulate leakage time series features; obtains the particulate leakage change rate features based on the particulate leakage time series features; and the particulate leakage monitoring area identification unit determines the particulate leakage monitoring area according to the particulate leakage change rate features. The sensor arrangement module is used to obtain the location of the particulate leakage monitoring area and arrange a sensor network according to the particulate leakage change rate features. The leakage data acquisition module includes an acquisition trigger unit, a data acquisition unit, and a data analysis unit. The acquisition trigger unit acquires the leakage vibration data during the monitoring time and establishes a leakage data acquisition index. If it is greater than the threshold, data acquisition is performed; the data acquisition unit performs data acquisition to obtain a leakage monitoring curve; and the data analysis unit processes the acquired data to obtain the processed leakage data. The leakage monitoring module constructs a vehicle load - particulate leakage rate correlation model, and inputs the vibration data, particulate short - term permeability, particulate volume short - term change rate, pore area short - term change rate, and suspended particle concentration, and outputs the particulate leakage index.

Images