A multi-source information collaborative intelligent compaction monitoring system based on dynamic compensation

Abstract

The present invention is a multi-source information collaborative intelligent compaction monitoring system based on dynamic compensation, comprising a data acquisition module, a displacement acquisition module, a data processing module, a shear wave velocity acquisition module, and a soil compaction value prediction model construction module. The data processing module extracts intelligent compaction indicators as well as the soil pressure peak value and the acceleration peak value inside the soil body. The displacement acquisition module obtains soil displacement data. The shear wave velocity acquisition module corrects the coordinates of the bending element sensor based on the soil displacement data and then calculates the shear wave velocity. The soil compaction degree is obtained through an on-site sand filling method experiment. A neural network model is constructed using the soil pressure peak value, the acceleration peak value, the lateral and vertical shear wave velocities, the soil displacement data, and the intelligent compaction indicator as input to achieve accurate prediction of the compaction degree. The present invention significantly improves the accuracy, real-time performance, and applicability of compaction quality evaluation.

Description

Technical Field

[0001] The present invention relates to the technical field of intelligent road construction, and in particular to a multi-source information collaborative intelligent compaction monitoring system based on dynamic compensation. Background Art

[0002] In modern road construction, soil compaction quality is crucial to the stability and durability of pavement projects. Traditional compaction quality assessment methods often rely on field tests, which are time-consuming and prone to localized inhomogeneities and human error. These methods fail to fully and comprehensively reflect the soil's state during the compaction process. In recent years, with the rapid development of artificial intelligence and sensing technologies, intelligent compaction quality assessment methods have gradually become mainstream. Intelligent compaction technology utilizes sensors mounted on the roller to collect the roller's vibration signals in real time. Combined with GPS positioning systems, it enables real-time monitoring of the rolling process and continuous assessment of compaction quality. However, most current research focuses on roller vibration signals, which are insufficient to fully reflect the soil's compaction state. Furthermore, there is a lack of an intelligent monitoring system that combines internal soil physical properties (such as acceleration, displacement, earth pressure, and shear wave velocity) with external roller measurement signals for comprehensive and coordinated assessment.

[0003] During the construction process, measuring the soil pressure is relatively simple, and data can usually be directly obtained by burying soil pressure sensors. However, there is a key problem: as the compaction process proceeds, the sensor will settle and displace along with the soil, causing the actual depth position corresponding to the measured data to change. Directly measuring the settlement of soil at different depths is more difficult, especially without destroying the soil structure, which is difficult to achieve with traditional methods. In addition, the existing technology calculates the settlement outside the soil by loading multiple displacement sensors on the roller, and the monitoring accuracy of this method needs to be improved. When obtaining the shear wave velocity through the bending element method, the accurate calculation of the distance between the transmitting end and the receiving end is difficult to determine, which puts higher requirements on the combination of the physical properties inside the soil and the compaction performance for more accurate intelligent monitoring. Summary of the Invention

[0004] The purpose of the present invention is to provide a multi-source information collaborative intelligent compaction monitoring system based on dynamic compensation to solve the problems raised in the above background technology.

[0005] In order to achieve the above objects, the technical solution adopted by the present invention is:

[0006] In a first aspect of the present invention, a multi-source information collaborative intelligent compaction monitoring system based on dynamic compensation is provided, the system comprising:

[0007] a data acquisition module, the data acquisition module comprising a composite sensor disposed within the compacted soil and an intelligent compaction device mounted on the vibratory roller, the composite sensor comprising an earth pressure sensor, an acceleration sensor, and a bend element sensor, the composite sensors being classified into two categories: a transmitting end sensor and a receiving end sensor, depending on whether the bend element sensor is a transmitting end or a receiving end; the earth pressure sensor being used to measure earth pressure signals during the compaction process, the acceleration sensor being used to measure acceleration signals within the soil during the compaction process, and the bend element sensor being used to measure shear wave signals of the soil during the compaction process; and the intelligent compaction device being used to collect vibration wheel acceleration signals, roller parameters, and position information throughout the entire compaction process;

[0008] A data processing module, comprising a filtering unit and an extraction unit, for processing data collected by the composite sensor and the intelligent compaction equipment during the compaction process, obtaining a filtered vibration wheel acceleration signal, a filtered earth pressure signal, and a filtered soil internal acceleration signal through the filtering unit, and extracting an intelligent compaction index, a peak earth pressure value within the soil, and a peak-to-peak acceleration value within the soil through the extraction unit;

[0009] a displacement acquisition module, the displacement acquisition module being electrically connected to the data processing module and configured to perform an integration operation on the acceleration based on the filtered acceleration signal inside the soil to obtain soil displacement data;

[0010] A shear wave velocity acquisition module is used to acquire soil shear wave velocity data, including transverse shear wave velocity and vertical shear wave velocity, based on the shear wave signal collected by the bending element sensor and the soil displacement data obtained by the displacement acquisition module;

[0011] The soil compaction value prediction model construction module is based on a neural network model, with dynamic response characteristic values ​​and intelligent compaction indicators as inputs of the neural network model, and compaction values ​​as outputs of the neural network model. The neural network model is optimized using an intelligent optimization algorithm, and a soil compaction value prediction model is obtained through training. The dynamic response characteristic values ​​include the peak soil pressure inside the soil, the peak-to-peak acceleration inside the soil, the transverse shear wave velocity, the vertical shear wave velocity, and the soil displacement data.

[0012] Furthermore, the composite sensor is arranged inside the compacted soil as follows:

[0013] Use an electric drill to vertically drill a hole in the loose layer of the road compaction construction site. After digging to the depth set in the test plan, place the composite sensor at the bottom of the pit and backfill. During the burial process, ensure that the soil particles in contact with the composite sensor are fine, with the Z axis of the composite sensor pointing vertically upward and the X axis parallel to the direction of travel of the roller.

[0014] In a transverse arrangement, the transmitter and receiver sensors are placed on the left and right sides of the compacted road, aligned transversely, and buried at the same depth within the loose layer to measure the transverse shear wave signals of the soil. The horizontal distance between the transmitter and receiver sensors in the transverse arrangement is determined by the width of the roller's vibratory wheel.

[0015] When arranged vertically, the transmitting end sensor and the receiving end sensor are arranged vertically in the loose pavement in the middle and / or at the edge of the compacted road. The transmitting end sensor and the receiving end sensor are located on the surface and bottom of the loose pavement respectively, and are used to measure the vertical shear wave signal of the soil.

[0016] Along the direction of the lane, a group of composite sensors is arranged horizontally or vertically every △L meters in the soil under compaction. Each group of composite sensors consists of a transmitting end sensor and a receiving end sensor, which are distributed at intervals horizontally and vertically to simultaneously collect the horizontal and vertical shear wave signals of the soil.

[0017] Furthermore, the compaction degree at the location of each group of composite sensors is obtained, and the soil pressure peak value, acceleration peak value, lateral shear wave velocity, vertical shear wave velocity, soil displacement data and intelligent compaction index obtained after each compaction, as well as the corresponding compaction degree, are used as training samples to train the neural network model.

[0018] Furthermore, the diameter of the drill hole is 5-10 cm, and the thickness of the loose layer is 20-30 cm.

[0019] Furthermore, the processing process of the filtering unit is:

[0020] Taking the signal peak as the center point, 4-second signal segments are taken before and after the center point, and the 8-second signal segments are digitally filtered using a Hamming window.

[0021] Furthermore, the process of integrating the acceleration to obtain soil displacement data is:

[0022] The filtered soil internal acceleration signal is zeroed using the average value of the signal's data over the previous two seconds as a benchmark. The first and last times the soil internal acceleration signal exceeds the threshold acceleration after the zeroing operation are defined as the vibration start time t1 and vibration end time t2, respectively. The threshold acceleration is determined based on the average peak value of the soil internal acceleration signal under no load.

[0023] The acceleration signal inside the soil after the zeroing operation is integrated to obtain the velocity signal; the velocity signal before t1 is kept as the integrated velocity data without correction, and the velocity signal from t2 to the end of the entire record is fitted by the least squares method to obtain the fitting straight line;

[0024] Connect the point corresponding to time t1 on the velocity signal and the point corresponding to time t2 on the fitting line as the trend line of the vibration stage, and subtract the trend line of the vibration stage from the velocity signal to obtain the corrected velocity signal;

[0025] The corrected velocity signal is integrated in the time domain to obtain the soil displacement signal, and the steady-state value from t2 to the end of the entire record is extracted as the soil displacement after compaction.

[0026] Furthermore, the process of obtaining the soil shear wave velocity data is:

[0027] Each composite sensor consists of three earth pressure sensors, three acceleration sensors and a bending element sensor. The three earth pressure sensors and acceleration sensors are arranged in three directions respectively. The data of the acceleration sensors in the three directions can be used to obtain the soil displacement data in the three directions respectively.

[0028] The initial buried positions of the transmitting and receiving sensors are recorded in the form of three-dimensional coordinates. After each composite sensor obtains the soil displacement in three directions, the three-dimensional coordinates of the composite sensor are corrected according to the soil displacement in the three directions. The distance between the transmitting and receiving sensors in each group of composite sensors is then calculated based on their respective corrected three-dimensional coordinates, and the shear wave velocity is calculated based on the distance.

[0029] In a second aspect of the present invention, an electronic device is further provided. The electronic device includes at least one processor; and a memory communicatively connected to the at least one processor; the memory storing instructions executable by the at least one processor, the instructions being executed by the at least one processor to enable the at least one processor to execute the system of the first aspect.

[0030] Compared with the prior art, the present invention has the following beneficial effects:

[0031] The system, based on soil pressure and acceleration signals collected from within the soil, fuses multi-source data with soil characteristic parameters to construct a nonlinear mapping relationship, enabling accurate prediction of compaction. The proposed soil compaction prediction model is adaptive to changes in soil quality and construction conditions, dynamically assessing compaction in real time and providing feedback to operators to guide them in adjusting compaction parameters, thereby enabling intelligent monitoring and closed-loop control of the compaction process.

[0032] The present invention significantly improves the accuracy, real-timeness and applicability of compaction quality evaluation, and provides efficient and reliable technical support for construction quality control of roads, railways and other projects. BRIEF DESCRIPTION OF THE DRAWINGS

[0033] Figure 1Schematic diagram of the structure of the multi-source information collaborative intelligent compaction monitoring system based on dynamic compensation in the present invention;

[0034] Figure 2 Schematic diagram of the arrangement of the composite sensor during the compaction process of the present invention;

[0035] Figure 3 Schematic diagram of displacement signal with trend term in time domain integration;

[0036] Figure 4 It is a segmented schematic diagram of the acceleration signal;

[0037] Figure 5 Schematic diagram of the displacement signal obtained by integrating the acceleration signal. DETAILED DESCRIPTION

[0038] The following description is intended to disclose the present invention so that those skilled in the art can implement the present invention, but is not intended to limit the scope of protection of this application. The embodiments described below are for illustrative purposes only, and those skilled in the art may conceive of other obvious variations.

[0039] Example 1:

[0040] Please refer to Figure 1 As shown, the multi-source information collaborative intelligent compaction monitoring system 100 based on dynamic compensation in this embodiment includes:

[0041] Data acquisition module 110, data acquisition module 110 includes a composite sensor arranged inside the compacted soil and an intelligent compaction device installed on the vibratory roller, the composite sensor includes an earth pressure sensor, an acceleration sensor and a bending element sensor, the earth pressure sensor is used to measure the earth pressure signal of the soil during compaction, the acceleration sensor is used to measure the acceleration signal inside the soil during compaction, the bending element sensor is used to measure the shear wave signal of the soil during the compaction process, and the intelligent compaction device is used to collect the vibration wheel acceleration signal, roller parameters, position information, etc. during the entire compaction process.

[0042] The data processing module 120 includes a filtering unit and an extraction unit. The data processing module is used to process the data collected by the composite sensor and the intelligent compaction equipment during the compaction process, obtain the filtered vibration wheel acceleration signal, the filtered soil pressure signal and the filtered soil internal acceleration signal through the filtering unit, and extract the intelligent compaction index and the soil pressure peak value and the acceleration peak-to-peak value inside the soil through the extraction unit.

[0043] The displacement acquisition module 130 is electrically connected to the data processing module 120 and is used to integrate the acceleration based on the filtered acceleration signal inside the soil to obtain soil displacement data;

[0044] The shear wave velocity acquisition module 140 is electrically connected to the displacement acquisition module 130 and is used to obtain soil shear wave velocity data, including transverse shear wave velocity and vertical shear wave velocity, based on the shear wave signal collected by the bending element sensor and the soil displacement data obtained by the displacement acquisition module.

[0045] The soil compaction value prediction model construction module 150 selects soil near the composite sensor during the on-site compaction process and uses the sand filling method to measure the compaction. Based on the neural network model, the dynamic response characteristic value and the intelligent compaction index are used as the input of the neural network model, and the compaction value is used as the output of the neural network model. The neural network model is optimized by the intelligent optimization algorithm, and a soil compaction value prediction model is obtained through training. The soil compaction value prediction model has a high accuracy prediction effect; the dynamic response characteristic value includes the soil pressure peak value inside the soil body, the acceleration peak-to-peak value inside the soil body, the transverse shear wave velocity, the vertical shear wave velocity and the soil displacement data.

[0046] Those skilled in the art will appreciate that by placing composite sensors (soil pressure sensors, acceleration sensors, and bending element sensors) within the compacted soil and installing intelligent compaction equipment on a vibratory roller, the present invention enables the system to simultaneously collect soil pressure signals, internal soil acceleration signals, shear wave signals, as well as the roller's vibratory wheel acceleration signals, roller parameters, and position information. The filtering unit in the data processing module effectively removes noise and interference from the collected signals, improving signal purity. Furthermore, the extraction unit extracts key information from the signals processed by the filtering unit, such as intelligent compaction indicators, peak-to-peak acceleration values ​​within the soil, and peak soil pressure values ​​within the soil, facilitating subsequent analysis. The shear wave velocity acquisition module uses soil displacement data to correct the spacing between the receiving and transmitting sensors, and simultaneously calculates the shear wave velocity based on the shear wave signal. This parameter reflects the mechanical properties of the soil and the compaction effect.

[0047] Example 2:

[0048] In this embodiment, the soil compaction value prediction model construction module adopts a neural network model and introduces the dragonfly algorithm to implement it. It can fully utilize the learning ability of the neural network and the optimization ability of the dragonfly algorithm to obtain a highly accurate soil compaction value prediction model, thereby improving the model's prediction accuracy. The highly accurate soil compaction value prediction model can provide scientific guidance for construction. According to the predicted compaction value, it can help construction personnel to adjust compaction parameters in a timely manner, such as the number of compaction times and vibration frequency, to ensure that the soil compaction meets the design requirements. At the same time, the model also helps to improve the quality of the project and reduce quality problems caused by insufficient or excessive compaction.

[0049] The dragonfly algorithm is:

[0050] ;

[0051] ;

[0052] ;

[0053] ;

[0054] ;

[0055]

[0056] Where, 、 、 、 、 They represent the displacement distances generated by the i-th dragonfly individual in its collision avoidance, pairing, aggregation, predation and enemy avoidance behaviors; and Represent the positions of the i-th and j-th dragonfly individuals, Y represents the number of dragonflies adjacent to the i-th dragonfly individual, Indicates the location of food. represents the location of the natural enemy, α, b, c, d, and e represent the weights of dragonfly group behavior, represents the inertia weight, t is the current iteration number, express Generation population update step size; That is, the step vector at the tth iteration.

[0057] The neural network model and the introduction of the dragonfly algorithm specifically include the following steps:

[0058] Initialize the neural network model, which includes an input layer, an output layer, and a hidden layer. The dynamic response eigenvalue and the intelligent compaction index are used as inputs of the input layer, and the output layer outputs the predicted compaction value. The dynamic response eigenvalues ​​include the soil pressure peak value inside the soil, the acceleration peak value inside the soil, the transverse shear wave velocity, the vertical shear wave velocity, and the soil displacement data (soil displacement data in three directions). When a value is missing, it is supplemented with 0.

[0059] Select the initial value of the dragonfly population behavior weight, determine the dragonfly population size N and the number of iterations T;

[0060] Arrange the neural network model weights w and thresholds θ in an orderly manner to form a row vector (w, θ), which is used as the position X of the dragonfly individual. Set the weight and threshold ranges, and randomly initialize the position of the dragonfly individual according to the weight and threshold ranges.

[0061] Calculate the individual dragonfly fitness value and record the current optimal solution;

[0062] Select mean square error as the fitness function and update the location of the food and the location of natural enemies , and the dragonfly population behavior weight is updated according to the above formula of the dragonfly algorithm. 、 、 、 、 ;

[0063] Update the dragonfly step vector , if the number of iterations t>T, then retain the connection weight w and threshold θ, otherwise t=t+1, return to calculate the individual fitness value of the dragonfly;

[0064] The weight w and threshold θ corresponding to the optimal solution are used as the initial connection weight and threshold of the neural network. The neural network model is then trained using training samples to obtain a highly accurate soil compaction value prediction model.

[0065] Example 3:

[0066] Please refer to Figure 2 As shown in FIG, arranging a composite sensor inside the compacted soil includes the following steps:

[0067] When embedding composite sensors within the soil, the arrangement of the receiving and transmitting sensors has a direct impact on the shear wave velocity measurement results. The loose paving thickness of on-site road compaction construction is generally 20-30 cm, taking a loose paving thickness of 30 cm as an example.

[0068] Use a small diameter (5-10 cm) electric drill to drill vertically into the loose soil to the specified depth set in the test plan. Place the composite sensor at the bottom of the pit and backfill with small-particle soil. The composite sensor's Z axis is vertically upward, and the X axis is parallel to the direction of the roller's travel.

[0069] In a horizontal arrangement, the transmitter and receiver sensors are placed on the left and right sides of the compacted road, aligned transversely, to measure the soil's transverse shear wave signals. Both sensors are buried at the same depth (e.g., 10 cm), meaning 20 cm from the surface of the compacted layer. The horizontal spacing between the transmitter and receiver sensors is determined by the width of the roller's vibratory drum.

[0070] In a vertical arrangement, a composite sensor set is placed vertically at a specific location on the road (in the middle or at the edge). Each composite sensor consists of a transmitter and a receiver, located at the upper (surface) and lower (e.g., 30 cm deep) ends of the loose pavement, with a vertical spacing equal to the pavement thickness. These sensors measure vertical shear wave signals in the soil. To ensure accurate data, the soil in contact with the sensors must be fine-grained. Coarse particles (such as gravel) can lead to uneven contact stress distribution, resulting in higher sensor readings due to localized point loads.

[0071] Example 4:

[0072] In this embodiment, the data collected by the composite sensor and the intelligent compaction device during the compaction process is filtered. The specific process is:

[0073] The amount of raw signal data collected during the test is enormous, requiring selective processing. With the signal peak as the center point, 4-second signal segments are taken before and after the center point. These 8-second segments are digitally filtered using a Hamming window. These signals, including the vibration wheel acceleration signal, the earth pressure signal, and the soil internal acceleration signal, are processed by a filtering unit to obtain the filtered vibration wheel acceleration signal, the filtered earth pressure signal, and the filtered soil internal acceleration signal, respectively.

[0074] These 8 seconds of data cover the entire compaction process and reduce the storage and computing burden.

[0075] Feature parameters are extracted from the filtered vibration wheel acceleration signal, the filtered soil internal acceleration signal, and the filtered soil pressure signal. The feature value extracted from the filtered vibration wheel acceleration signal is the intelligent compaction index CMV, the feature value extracted from the filtered soil internal acceleration signal is the peak-to-peak value of the acceleration inside the soil, and the feature value extracted from the filtered soil pressure signal is the peak soil pressure inside the soil.

[0076] Example 5:

[0077] In this embodiment, the process of integrating the acceleration to obtain soil displacement data is as follows:

[0078] S1. The filtered soil internal acceleration signal is reset to zero using the average value of the data 2 seconds before the signal as a reference to eliminate the initial offset of the instrument.

[0079] S2, define the time when the acceleration signal inside the soil exceeds the threshold acceleration for the first time and the last time after the zeroing operation as the vibration start time t1 and the vibration end time t2 respectively; Figure 4 As shown, t1=about 1.2s, t2=7s.

[0080] The threshold acceleration can be determined as 0.15 m / s² based on experience. The threshold acceleration is determined based on the average peak value of the acceleration signal under no-load conditions. Exceeding this value indicates the influence of external loads, i.e., the roller is acting on the buried location of the sensor. In this embodiment, 0.15 m / s² is defined as the peak value under no-load conditions, i.e., the threshold acceleration.

[0081] S3, integrating the acceleration signal inside the soil after the zeroing operation to obtain a velocity signal, which shows a nonlinear trend due to baseline drift;

[0082] S4, keeping the integrated speed data of the speed signal before t1 without making any correction, and performing least square fitting on the speed signal from t2 to the end of the entire record to obtain a fitting straight line;

[0083] S5. Finally, draw a straight line connecting the point corresponding to time t1 on the velocity signal and the point corresponding to time t2 on the fitting line. Use this straight line as the trend line of the vibration stage. Subtract the trend line of the vibration stage from the velocity signal obtained in step S3 to obtain a corrected velocity signal, thereby eliminating baseline drift.

[0084] S6. Perform time domain integration on the corrected velocity signal to obtain the soil displacement signal, and extract the steady-state value from t2 to the end of the entire record as the soil displacement after compaction. Figure 5 As shown, t2=7s. After t2, the displacement curve gradually stabilizes, and the value when the displacement value does not change is taken as the steady-state value, that is, the steady-state displacement.

[0085] The acceleration signal is integrated once to obtain the velocity signal, and then integrated twice to obtain the displacement signal. If the conventional acceleration is integrated twice to obtain the displacement signal, the error will be amplified during the integration process due to the accumulation of DC components and noise. At this time, the displacement signal contains a trend term, such as Figure 3 shown.

[0086] Example 6:

[0087] In this embodiment, the process of obtaining soil shear wave velocity data is:

[0088] Each composite sensor consists of three earth pressure sensors, three acceleration sensors and a bending element sensor. The three earth pressure sensors and acceleration sensors are arranged in the X, Y, and Z directions respectively. The data of the acceleration sensors in the three directions can be used to obtain the soil displacement data in the three directions respectively.

[0089] The initial buried positions of the transmitting and receiving sensors are recorded in the form of three-dimensional coordinates (X, Y, Z). After each composite sensor obtains the soil displacement in the three directions, the three-dimensional coordinates of the composite sensor are corrected according to the soil displacement in the three directions. The distance between the transmitting and receiving sensors in each group of composite sensors is then calculated based on their respective corrected three-dimensional coordinates, and the shear wave velocity is then calculated based on the distance.

[0090] In this example, X represents the coordinate in the direction of the roller's travel, Y represents the coordinate perpendicular to the roller's travel, and Z represents the depth, expressed in meters. For example, when a pair of composite sensors A and B are arranged horizontally, the coordinates of their initial burial positions are A (0, 0, 0.3) and B (0, 2.2, 0.3), respectively. 2.2 meters is the horizontal distance between sensors A and B, determined by the vibrating wheel width. Z represents the burial depth, determined by the depth at the time of burial. X is 0 at the starting point of paving.

[0091] When a set of composite sensors A and B is arranged vertically, the coordinates of their initial locations are A (0, 0, 0) and B (0, 0, 0.3), where X is the coordinate in the direction of the roller's travel, Y is the coordinate perpendicular to the roller's travel, and Z is the depth in meters. 0.3m is the vertical distance between sensors A and B, determined by the thickness of the loose soil layer. At the starting point of the installation, both X and Y are 0.

[0092] A is the transmitter sensor, B is the receiver sensor, and the distance between the transmitter and receiver sensors is calculated based on the coordinates of the transmitter and receiver sensors corrected by the soil displacement in three directions at each compaction moment.

[0093] The distance between the transmitter sensor and the receiver sensor is calculated as follows:

[0094]

[0095]

[0096]

[0097]

[0098] Where, is the corrected coordinate of the receiving end sensor B on the X axis, is the corrected coordinate of the transmitter sensor A on the X axis, is the corrected coordinate of the receiving end sensor on the Y axis, is the corrected coordinate of the transmitter sensor on the Y axis, is the corrected coordinate of the receiving end sensor on the Z axis, is the corrected coordinate of the transmitter sensor on the Z axis, and L is the distance between the transmitter sensor and the receiver sensor; is the coordinate difference between the receiving sensor and the transmitting sensor on the X axis, is the coordinate difference between the receiving sensor and the transmitting sensor on the Y axis, is the coordinate difference between the receiving sensor and the transmitting sensor on the Z axis.

[0099] The first arrival point of the shear wave (i.e., the moment when the signal first significantly deviates from the baseline) is identified from the signal received by the receiving sensor. The time difference is directly read by comparing the time domain signals from the transmitting and receiving sensors. The transverse and vertical shear wave velocities of the soil are calculated by dividing the time difference by the distance between the transmitting and receiving sensors.

[0100] In this embodiment, soil displacement data in three directions is obtained by using data from acceleration sensors in three directions. The soil displacement data is directional, and when performing coordinate correction, only the initial three-dimensional coordinate values ​​need to be uniformly added with the corresponding soil displacement data.

[0101] In a second aspect of the present invention, an electronic device is provided. The electronic device includes at least one processor; and a memory communicatively connected to the at least one processor; the memory storing instructions executable by the at least one processor, the instructions being executed by the at least one processor to enable the at least one processor to execute the system according to the first aspect of the present invention.

[0102] The above shows and describes the basic principles, main features, and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The above embodiments and descriptions merely illustrate the principles of the present invention. Various changes and modifications may be made to the present invention without departing from the spirit and scope of the present invention. Such changes and modifications are intended to fall within the scope of the present invention. The scope of protection claimed by the present invention is defined by the appended claims and their equivalents.

[0103] Any matters not described in the present invention are applicable to the prior art.

Claims

1. A multi-source information collaborative intelligent compaction monitoring system based on dynamic compensation, characterized in that: The system comprises: a data acquisition module, the data acquisition module comprising a composite sensor disposed within the compacted soil and an intelligent compaction device mounted on the vibratory roller, the composite sensor comprising an earth pressure sensor, an acceleration sensor, and a bend element sensor, the composite sensors being classified into two categories: a transmitting end sensor and a receiving end sensor, depending on whether the bend element sensor is a transmitting end or a receiving end; the earth pressure sensor being used to measure earth pressure signals during the compaction process, the acceleration sensor being used to measure acceleration signals within the soil during the compaction process, and the bend element sensor being used to measure shear wave signals of the soil during the compaction process; and the intelligent compaction device being used to collect vibration wheel acceleration signals, roller parameters, and position information throughout the entire compaction process; The arrangement of composite sensors inside the compacted soil is as follows: Use an electric drill to vertically drill a hole in the loose layer of the road compaction construction site. After digging to the depth set in the test plan, place the composite sensor at the bottom of the pit and backfill. During the burial process, ensure that the soil particles in contact with the composite sensor are fine, with the Z axis of the composite sensor pointing vertically upward and the X axis parallel to the direction of travel of the roller. In a transverse arrangement, the transmitter and receiver sensors are placed on the left and right sides of the compacted road, aligned transversely, and buried at the same depth within the loose layer to measure the transverse shear wave signals of the soil. The horizontal distance between the transmitter and receiver sensors in the transverse arrangement is determined by the width of the roller's vibratory wheel. When arranged vertically, the transmitting end sensor and the receiving end sensor are arranged vertically in the loose pavement in the middle and / or at the edge of the compacted road. The transmitting end sensor and the receiving end sensor are located on the surface and bottom of the loose pavement respectively, and are used to measure the vertical shear wave signal of the soil. Along the direction of the lane, a group of composite sensors are arranged horizontally or vertically every △L meters in the soil under compaction. Each group of composite sensors consists of a transmitting end sensor and a receiving end sensor, which are spaced horizontally and vertically to simultaneously collect the horizontal and vertical shear wave signals of the soil. A data processing module, comprising a filtering unit and an extraction unit, for processing data collected by the composite sensor and the intelligent compaction equipment during the compaction process, obtaining a filtered vibration wheel acceleration signal, a filtered earth pressure signal, and a filtered soil internal acceleration signal through the filtering unit, and extracting an intelligent compaction index, a peak earth pressure value within the soil, and a peak-to-peak acceleration value within the soil through the extraction unit; The processing process of the filtering unit is: Taking the signal peak as the center point, take 4-second signal segments before and after the center point, and use the Hamming window to perform digital filtering on the 8-second signal segments. a displacement acquisition module, the displacement acquisition module being electrically connected to the data processing module and configured to perform an integration operation on the acceleration based on the filtered acceleration signal inside the soil to obtain soil displacement data; The process of integrating the acceleration to obtain soil displacement data is as follows: The filtered soil internal acceleration signal is zeroed using the average value of the signal's data over the previous two seconds as a benchmark. The first and last times the soil internal acceleration signal exceeds the threshold acceleration after the zeroing operation are defined as the vibration start time t1 and vibration end time t2, respectively. The threshold acceleration is determined based on the average peak value of the soil internal acceleration signal under no load. The acceleration signal inside the soil after the zeroing operation is integrated to obtain the velocity signal; the velocity signal before t1 is kept as the integrated velocity data without correction, and the velocity signal from t2 to the end of the entire record is fitted by the least squares method to obtain the fitting straight line; Connect the point corresponding to time t1 on the velocity signal and the point corresponding to time t2 on the fitting line as the trend line of the vibration stage, and subtract the trend line of the vibration stage from the velocity signal to obtain the corrected velocity signal; Perform time domain integration on the corrected velocity signal to obtain the soil displacement signal, and extract the steady-state value from t2 to the end of the entire record as the soil displacement after compaction; A shear wave velocity acquisition module is used to acquire soil shear wave velocity data, including transverse shear wave velocity and vertical shear wave velocity, based on the shear wave signal collected by the bending element sensor and the soil displacement data obtained by the displacement acquisition module; The process of obtaining the soil shear wave velocity data is as follows: Each composite sensor consists of three earth pressure sensors, three acceleration sensors and a bending element sensor. The three earth pressure sensors and acceleration sensors are arranged in three directions respectively. The data of the acceleration sensors in the three directions can be used to obtain the soil displacement data in the three directions respectively. The initial buried positions of the transmitter and receiver sensors are recorded in the form of three-dimensional coordinates. After each composite sensor obtains soil displacement in three directions, the three-dimensional coordinates of the composite sensor are corrected according to the soil displacement in the three directions. The distance between the transmitter and receiver sensors in each composite sensor group is then calculated based on their respective corrected three-dimensional coordinates, and the shear wave velocity is calculated based on the distance. The soil compaction value prediction model construction module is based on a neural network model, with dynamic response characteristic values ​​and intelligent compaction indicators as inputs of the neural network model, and compaction values ​​as outputs of the neural network model. The neural network model is optimized using an intelligent optimization algorithm, and a soil compaction value prediction model is obtained through training. The dynamic response characteristic values ​​include the peak soil pressure inside the soil, the peak-to-peak acceleration inside the soil, the transverse shear wave velocity, the vertical shear wave velocity, and the soil displacement data.

2. The multi-source information collaborative intelligent compaction monitoring system based on dynamic compensation according to claim 1 is characterized in that: The compaction degree at the location of each group of composite sensors is obtained, and the soil pressure peak value, acceleration peak value, lateral shear wave velocity, vertical shear wave velocity, soil displacement data and intelligent compaction index obtained after each compaction, as well as the corresponding compaction degree, are used as training samples to train the neural network model.

3. The multi-source information collaborative intelligent compaction monitoring system based on dynamic compensation according to claim 1 is characterized in that: The diameter of the drill hole is 5-10 cm, and the thickness of the loose layer is 20-30 cm.

4. The multi-source information collaborative intelligent compaction monitoring system based on dynamic compensation according to claim 1 is characterized in that: The intelligent optimization algorithm is the dragonfly algorithm.

Images