A self-early warning intelligent geogrid system and method based on multi-parameter fusion

Abstract

This invention discloses a self-early warning intelligent geogrid system and method based on multi-parameter fusion, comprising: a visual early warning module integrated on the load-bearing substrate, which changes color in response to environmental stimuli; a multi-parameter sensing module deployed on the load-bearing substrate in conjunction with the visual early warning module, which acquires strain, humidity, and temperature electrical signals reflecting the subgrade condition through a micro-sensor array; a data processing and communication module connected to the multi-parameter sensing module, which collects and fuses the electrical signals and transmits the extracted feature information to the cloud via a wireless protocol; and an analysis and early warning module connected to the data processing and communication module, which performs multi-dimensional correlation analysis and lesion assessment on the feature information through a preset algorithm model and outputs graded early warning signals. This invention enables early perception, multi-dimensional diagnosis, and intuitive visual early warning of hidden defects within geotechnical structures, overcoming the technical shortcomings of existing monitoring methods, such as high cost, strong lag, and limited information.

Description

Technical Field

[0001] This invention relates to the field of geotechnical structure monitoring technology, and in particular to a self-early warning intelligent geogrid system and method based on multi-parameter fusion. Background Technology

[0002] Currently, the main methods for monitoring hidden defects such as voids and settlement within geotechnical structures like roadbeds are traditional point sensors, fiber optic sensors, and ground-penetrating radar. With increasing monitoring demands, the industry's technological trend is moving towards intelligent integration, such as embedding electronic components like radio frequency identification (RFID) into geogrids to identify defects by sensing component breakage or failure caused by geogrid stretching.

[0003] However, existing monitoring technologies still face severe technical bottlenecks and market challenges in practical applications. Traditional monitoring methods generally suffer from objective problems such as high construction costs, limited monitoring coverage, overly specialized data interpretation, and difficulty in achieving real-time and intuitive early warnings. Emerging embedded radio frequency monitoring technologies exhibit a clear passive response characteristic, typically only generating signal loss when the damage has progressed to the point of causing significant grid deformation and electronic component breakage, resulting in severe delays in early warning. Furthermore, such technologies provide overly simplistic information dimensions, only achieving a binary determination of the presence or absence of a signal, failing to effectively identify the specific causes of the damage (such as water seepage or freeze-thaw cycles), the rate of damage development, and the precise geographical location.

[0004] In terms of practical inspection efficiency and environmental adaptability, existing solutions heavily rely on specialized equipment such as vehicle-mounted or handheld RF scanners, making it impossible to achieve rapid, large-area, and intuitive identification using conventional equipment such as the naked eye or drones. Furthermore, RF signals exhibit poor stability in complex soil environments, easily affected by environmental factors such as soil moisture content and metal components, impacting monitoring accuracy. Therefore, the market urgently needs an intelligent geotechnical monitoring technology capable of early identification, intuitive display, multi-dimensional data perception, and autonomous early warning to fill the gaps in real-time performance, multi-dimensional diagnostics, and convenient inspection capabilities of existing technologies. Summary of the Invention

[0005] In view of this, the purpose of this invention is to provide a self-early warning intelligent geogrid system and method based on multi-parameter fusion, which can realize early perception, multi-dimensional diagnosis and intuitive visual early warning of hidden defects inside geostructures, and overcome the technical defects of existing monitoring methods such as high cost, strong lag and single information.

[0006] In a first aspect, embodiments of the present invention provide a self-early warning intelligent geogrid system based on multi-parameter fusion, which includes a visual early warning module, a multi-parameter sensing module, a data processing and communication module, and an analysis and early warning module.

[0007] The visual warning module is integrated into the stress-bearing substrate and changes color in response to environmental stimuli.

[0008] The multi-parameter sensing module and the visual early warning module are deployed together on the stress-bearing substrate, and the strain, humidity and temperature electrical signals reflecting the condition of the roadbed are acquired through a micro-sensor array.

[0009] The data processing and communication module is connected to the multi-parameter sensing module, collects and fuses electrical signals, and transmits the extracted feature information to the cloud via a wireless protocol.

[0010] The analysis and early warning module is connected to the data processing and communication module. It performs multi-dimensional correlation analysis and lesion assessment on the feature information through a preset algorithm model and outputs a graded early warning signal.

[0011] The analysis and early warning module is equipped with a multi-parameter fusion risk assessor, which calculates the local disease risk index based on a weighted normalization function, combines a spatiotemporal anomaly detection algorithm, determines the propagation probability of the abnormal area based on the continuous temporal fluctuation pattern and spatial connectivity, adjusts the weight coefficients of the multi-parameter fusion risk assessor, and outputs the adjusted local disease risk index.

[0012] The analysis and early warning module further inputs the adjusted local disease risk index as the core feature vector into the disease type identification model to determine the disease category.

[0013] By utilizing edge computing hierarchical early warning logic, the disease category and its corresponding risk index are determined based on fuzzy logic reasoning, and hierarchical early warning signals corresponding to different degrees of danger are output.

[0014] In conjunction with the first aspect, embodiments of the present invention provide a first possible implementation of the first aspect, wherein,

[0015] The visual warning module uses microencapsulated smart color-changing material, which is coated or embedded in the rib intersection nodes and / or longitudinal ribs of the grid.

[0016] The microencapsulated smart color-changing material includes at least one of stress-induced color-changing materials, wet-induced color-changing materials, and thermochromic materials.

[0017] When the strain generated by the stress-bearing substrate sensed by the multi-parameter sensing module exceeds a preset threshold, the stress-induced color-changing material undergoes a reversible or irreversible change from one color to another.

[0018] When the humidity of the surrounding soil sensed by the multi-parameter sensing module reaches the response threshold, the wet-induced color-changing material undergoes a reversible or irreversible change from one color to another.

[0019] When the multi-parameter sensing module detects an abnormal temperature in the freeze-thaw cycle zone or in a localized area, the thermochromic material undergoes a reversible or irreversible change from one color to another.

[0020] In conjunction with the first aspect, embodiments of the present invention provide a second possible implementation of the first aspect, wherein,

[0021] The multi-parameter sensing module employs a miniature sensor array, which is printed or embedded on the surface of the grid using flexible printed electronics technology.

[0022] The micro-sensor array includes at least one of a flexible resistance strain gauge, a humidity sensor, and a temperature sensor.

[0023] The flexible resistance strain gauge monitors the distributed strain at various parts of the grid.

[0024] The humidity sensor monitors the volumetric moisture content of the soil surrounding the grid.

[0025] The temperature sensor monitors the local temperature of various parts of the grille.

[0026] In conjunction with the first aspect, embodiments of the present invention provide a third possible implementation of the first aspect, wherein,

[0027] The data processing and communication module acquires the electrical signal of the multi-parameter sensing module, fits the time series of each parameter using a piecewise linear compression algorithm, and extracts feature inflection point data to reduce the data size.

[0028] The feature inflection point data is encapsulated to obtain fused feature information.

[0029] The fused feature information is sent to the analysis and early warning module via LoRa, NB-IoT, or Bluetooth wireless communication protocols.

[0030] In conjunction with the first aspect, this embodiment of the invention provides a fourth possible implementation of the first aspect, wherein the weighted normalization function used by the analysis and early warning module when calculating the local disease risk index is:

[0031] , If the value exceeds a preset threshold, such as 0.6, an alert will be triggered.

[0032] in, For normalized strain values, .

[0033] Normalized humidity value .

[0034] For temperature anomaly, .

[0035] This is a local disease risk index. , and These are the weighting coefficients. The strain value at the current moment. The humidity value at the current moment. The temperature value at the current moment. The initial strain reference value, This is the initial humidity baseline value. This is the initial temperature reference value. The strain threshold, Humidity threshold This represents the temperature change threshold.

[0036] In conjunction with the first aspect, embodiments of the present invention provide a fifth possible implementation of the first aspect, wherein,

[0037] The spatiotemporal anomaly detection algorithm performs sliding window statistics on each sensing node. If the current sampled value exceeds the sum of the sliding mean and the preset standard deviation, it is determined to be a node anomaly. The calculation formula is as follows: ,in, This is the current sampled value. It is the moving average. For the sliding standard deviation, The determination coefficient.

[0038] If multiple consecutive nodes in a space simultaneously malfunction, the regional anomaly event is determined by the spatial propagation probability, which is calculated using the following formula: ,in, For spatial propagation probability, The distance from the anomaly center. For threshold distance, denoted as the propagation coefficient.

[0039] The analysis and early warning module utilizes the spatial propagation probability. Based on the extent of damage and coverage, the weighting coefficients in the multi-parameter fusion risk assessor are adaptively adjusted to correct and output a local disease risk index that reflects the degree of local disease development.

[0040] In conjunction with the first aspect, embodiments of the present invention provide a sixth possible implementation of the first aspect, wherein,

[0041] The disease type identification model receives a feature vector constructed from the adjusted risk index and the corresponding sensor parameters. .

[0042] A lightweight neural network algorithm is used to identify and output corresponding disease type labels, which include at least one of the following: normal state, cavitation and settlement, abnormal seepage, freeze-thaw cycle, and compound disease.

[0043] The disease type identification model is deployed in a microcontroller using TensorFlow Lite Micro.

[0044] In conjunction with the first aspect, embodiments of the present invention provide a seventh possible implementation of the first aspect, wherein,

[0045] The hierarchical early warning logic inputs the disease type label and the associated risk index into the fuzzy inference engine, and outputs an early warning level from 0 to 3 according to the preset fuzzy rule set.

[0046] Among them, low strain and normal parameters correspond to level 0, while high strain accompanied by high humidity and abnormal temperature corresponds to the highest level, level 3.

[0047] Secondly, embodiments of the present invention also provide an intelligent geogrid monitoring system, which includes an inspection unit and a cloud monitoring platform.

[0048] The inspection unit uses a drone or inspection vehicle equipped with a high-resolution camera. The high-resolution camera has a resolution of no less than 20 million pixels and generates a visual distribution map of the disease by identifying the color change areas of the visual warning module.

[0049] The cloud-based monitoring platform receives wirelessly transmitted early warning signals and spatially matches them with the visual distribution map of the disease.

[0050] The intelligent geogrid monitoring system monitors the self-early warning intelligent geogrid system based on multi-parameter fusion as described above.

[0051] Thirdly, embodiments of the present invention also provide a method for monitoring subgrade defects using a self-early warning intelligent geogrid system based on multi-parameter fusion, comprising:

[0052] By simultaneously sensing the stress and environmental parameters of the roadbed through the embedded intelligent geogrid, a preliminary intuitive warning is provided through the visual early warning module.

[0053] By combining the risk index calculated by the analysis and early warning module with the disease type, the location, severity and cause of the disease are displayed on the cloud monitoring platform through a GIS map.

[0054] Based on the output graded early warning signals and the color-changing area images captured by the drone, maintenance personnel formulate corresponding maintenance decisions.

[0055] The self-early warning intelligent geogrid system based on multi-parameter fusion is the self-early warning intelligent geogrid system based on multi-parameter fusion as described above.

[0056] The beneficial effects of the embodiments of the present invention are:

[0057] This invention achieves an earlier and more intuitive early warning mechanism. Through intelligent color-changing materials, it generates a color change at the initial stage of minor deformation of the grid or abnormal environmental parameters. Its response point is significantly earlier than the passive mode of traditional radio frequency components that fail only due to breakage. Furthermore, it allows for rapid inspection through visual inspection or conventional equipment without the need for specialized scanning equipment. This invention greatly broadens the dimensions of monitoring information, enabling the simultaneous acquisition of multi-dimensional information such as mechanical deformation, hydrology, and temperature. This elevates the monitoring process from simply detecting anomalies to a preliminary diagnosis of the causes and types of damage, providing a scientific basis for precise operation and maintenance.

[0058] In terms of practical application effectiveness, this invention boasts extremely high inspection efficiency and coverage, supporting large-scale UAV visual inspections. It achieves rapid location of areas by automatically identifying color-changing hotspots and performs in-depth analysis based on wirelessly transmitted data, significantly improving the automation level of inspections. Simultaneously, this invention exhibits excellent construction compatibility and durability. The flexible electronics and coating technologies employed have minimal impact on the original reinforcement performance of the geogrid, and the construction method is fully compatible with traditional geogrids, ensuring long-term reliable operation of the monitoring system in complex geotechnical environments.

[0059] This invention boasts significant advantages in intelligence and low power consumption. Through local data fusion and edge computing capabilities, it achieves graded early warning of roadbed defects and targeted data processing and reporting, effectively reducing system power consumption and data traffic requirements. This highly integrated self-early warning system not only simplifies the monitoring process and reduces long-term maintenance costs, but also greatly enhances the safety and reliability of roadbed structure monitoring through dual-channel visual and data perception. Detailed Implementation

[0060] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. The components of the embodiments of the present invention described and shown herein can generally be arranged and designed in various different configurations.

[0061] This invention provides a self-early warning intelligent geogrid system based on multi-parameter fusion, comprising a visual early warning module, a multi-parameter sensing module, a data processing and communication module, and an analysis and early warning module. The visual early warning module is integrated onto the load-bearing substrate and changes color in response to environmental stimuli. The multi-parameter sensing module is deployed in conjunction with the visual early warning module on the load-bearing substrate, acquiring strain, humidity, and temperature electrical signals reflecting the subgrade condition through a micro-sensor array. The data processing and communication module is connected to the multi-parameter sensing module, collects and fuses the electrical signals, and transmits the extracted feature information to the cloud via a wireless protocol. The analysis and early warning module is connected to the data processing and communication module, performing multi-dimensional analysis of the feature information using a preset algorithm model. The system performs correlation analysis and lesion assessment to output graded early warning signals. The analysis and early warning module deploys a multi-parameter fusion risk assessor, which calculates a local disease risk index based on a weighted normalization function. Combined with a spatiotemporal anomaly detection algorithm, it determines the propagation probability of abnormal areas based on continuous temporal fluctuation patterns and spatial connectivity, adjusts the weight coefficients of the multi-parameter fusion risk assessor, and outputs the adjusted local disease risk index. The analysis and early warning module further inputs the adjusted local disease risk index as a core feature vector into a disease type identification model to determine the disease category. Using edge computing graded early warning logic, it determines the disease category and its corresponding risk index based on fuzzy logic reasoning, and outputs graded early warning signals corresponding to different levels of severity.

[0062] The visual warning module employs microencapsulated smart color-changing materials, coated or embedded in the cross nodes and / or longitudinal ribs of the grid. The microencapsulated smart color-changing materials include at least one of stress-induced color-changing materials, wet-induced color-changing materials, and thermochromic materials. When the strain generated by the stressed substrate sensed by the multi-parameter sensing module exceeds a preset threshold (which can be set to 0.5% or 1%), the stress-induced color-changing material undergoes a reversible or irreversible change from one color, such as green, to another color, such as red or orange. When the humidity of the surrounding soil sensed by the multi-parameter sensing module reaches a response threshold due to pipe leakage, the wet-induced color-changing material undergoes a reversible or irreversible change from one color, such as green, to another color, such as red or orange. When the temperature of the freeze-thaw cycle zone or local temperature sensed by the multi-parameter sensing module is abnormal, the thermochromic material undergoes a reversible or irreversible change from one color, such as green, to another color, such as red or orange.

[0063] The multi-parameter sensing module employs a micro-sensor array, which is printed or embedded on the surface of the grid using flexible printed electronics technology. The micro-sensor array includes at least one of a flexible resistance strain gauge, a humidity sensor, and a temperature sensor. The flexible resistance strain gauge monitors the distributed strain at various parts of the grid. The humidity sensor is a capacitive or resistive humidity sensor that monitors the volumetric water content of the soil surrounding the grid. The temperature sensor is a digital temperature sensor that monitors the local temperature at various parts of the grid.

[0064] The data processing and communication module acquires the electrical signal of the multi-parameter sensing module, fits the time series of each parameter using a piecewise linear compression algorithm, extracts feature inflection point data to reduce the data size, encapsulates the feature inflection point data to obtain fused feature information, and sends the fused feature information to the analysis and early warning module via LoRa, NB-IoT or Bluetooth wireless communication protocols.

[0065] Among them, a piecewise linear compression algorithm is used to fit the time series, storing only the inflection points, achieving a compression rate of over 80%, significantly reducing wireless transmission power consumption, and supporting cloud-based reconstruction of the original trend.

[0066] The weighted normalization function used by the analysis and early warning module when calculating the local disease risk index is as follows: , If the value exceeds a preset threshold, such as 0.6, an alert will be triggered; among which, For normalized strain values, ; Normalized humidity value ; For temperature anomaly, ; This is a local disease risk index. , and These are the weighting coefficients. The strain value at the current moment. The humidity value at the current moment. The temperature value at the current moment. The initial strain reference value, This is the initial humidity baseline value. This is the initial temperature reference value. The strain threshold, Humidity threshold This represents the temperature change threshold.

[0067] The spatiotemporal anomaly detection algorithm performs sliding window statistics on each sensing node. If the current sampled value exceeds the sum of the sliding mean and the preset standard deviation, it is determined to be a node anomaly. The calculation formula is as follows: ,in, This is the current sampled value. It is the moving average. For the sliding standard deviation, The determination coefficient is used; if multiple consecutive nodes in the space are simultaneously abnormal, the regional abnormal event is determined by the spatial propagation probability, and the formula for calculating the spatial propagation probability is as follows: ,in, For spatial propagation probability, The distance from the anomaly center. For threshold distance, The propagation coefficient is used by the analysis and early warning module; the spatial propagation probability is used as the propagation coefficient. Based on the extent of damage and coverage, the weighting coefficients in the multi-parameter fusion risk assessor are adaptively adjusted to correct and output a local disease risk index that reflects the degree of local disease development.

[0068] The disease type identification model receives a feature vector constructed from the adjusted risk index and the corresponding sensor parameters. The system uses a lightweight neural network algorithm to identify and output corresponding disease type labels, which include at least one of the following: normal state, cavitation and settlement, abnormal seepage, freeze-thaw cycle, and compound disease. The disease type identification model is deployed on a microcontroller using TensorFlow Lite Micro.

[0069] The graded early warning logic inputs the disease type label and associated risk index into the fuzzy inference engine, and outputs early warning levels from 0 to 3 according to a preset fuzzy rule set; where low strain and normal parameters correspond to level 0, and high strain accompanied by high humidity and abnormal temperature corresponds to the highest level, level 3. Examples of fuzzy rules are shown in Table 1.

[0070] Table 1

[0071] Low normal normal Level 0 middle normal normal Level 1 middle high normal Level 2 high high abnormal Level 3

[0072] This invention also provides an intelligent geogrid monitoring system, comprising an inspection unit and a cloud monitoring platform. The inspection unit uses a drone or inspection vehicle equipped with a high-resolution camera, the high-resolution camera having a resolution of no less than 20 million pixels, and generates a visual distribution map of defects by identifying color change areas in the visual early warning module. The cloud monitoring platform receives wirelessly transmitted early warning signals and spatially matches them with the visual distribution map of defects. The intelligent geogrid monitoring system monitors the self-early warning intelligent geogrid system based on multi-parameter fusion as described above.

[0073] This invention also provides a method for monitoring roadbed defects using a self-early warning intelligent geogrid system based on multi-parameter fusion. The method includes: synchronously sensing roadbed stress and environmental parameters through the embedded intelligent geogrid; providing preliminary intuitive warnings through a visual early warning module; combining the risk index and defect type calculated by the analysis and early warning module; displaying the defect location, severity, and causes on a cloud-based monitoring platform using a GIS map; and, based on the output graded early warning signals and images of color-changing areas taken by drones, having maintenance personnel formulate corresponding maintenance decisions. The self-early warning intelligent geogrid system based on multi-parameter fusion is the self-early warning intelligent geogrid system based on multi-parameter fusion described above.

[0074] The core of this invention lies in the synergistic integration of intelligent color-changing materials and a flexible multi-parameter sensing system to construct an intelligent geogrid system with dual-channel perception capabilities of vision and data. To make the objectives, technical solutions, and advantages of this invention clearer, the following detailed description is provided in conjunction with specific embodiments.

[0075] Example 1: Used for highway subgrade monitoring

[0076] This embodiment focuses on long-term monitoring of the stability of the roadbed after compaction and potential voids and settlement defects.

[0077] (1) Product preparation

[0078] Polyester grid was selected as the load-bearing substrate. Visual warning modules containing stress-induced color-changing microcapsules were integrated at the nodes of the substrate using a printing process, with a strain threshold of 0.8% for color change. A multi-parameter sensing module (including a flexible resistance strain gauge, humidity sensor, and digital temperature sensor) was integrated every 1 meter along the longitudinal ribs of the load-bearing substrate. Every 10 sensing nodes were connected to a LoRa communication relay module via wires, serving as a data processing and communication module responsible for data aggregation and transmission.

[0079] (2) On-site deployment

[0080] After the compaction process is completed in the newly constructed highway subgrade, the prepared intelligent geogrid is laid flat on the top surface of the subgrade. After the laying is completed, the road base and surface layers are laid on top of it according to conventional processes.

[0081] (3) Operational monitoring

[0082] Visual inspection: During the operation and maintenance phase, drones equipped with high-resolution cameras are used regularly (e.g., monthly) to conduct aerial photography of the entire line, and the AI ​​image recognition system is used to automatically screen for areas with abnormal colors and generate a visual distribution map of the defects.

[0083] Data Analysis: The analysis and early warning module receives feature information hourly via a cloud platform. It uses a piecewise linear compression algorithm to process the electrical signal to reduce its size and then runs a weighted normalization function to calculate the local disease risk index. .

[0084] Linked early warning: Once a significant change in the color of the grid in a certain area is detected and the risk index is high... If the threshold is exceeded (e.g., 0.6), the system will automatically issue a Level 3 (urgent) or Level 2 (serious) warning through fuzzy logic reasoning to guide ground verification.

[0085] Example 2: Used for monitoring frost damage in railway subgrade

[0086] This embodiment uses the dynamic changes in temperature and humidity to conduct specialized monitoring of the freeze-thaw cycle of railway subgrade in high-altitude and cold regions.

[0087] (1) Product preparation

[0088] A visual warning module containing thermochromic material is integrated into a stress-bearing substrate, with its color-changing temperature threshold set near 0°C. Simultaneously, a multi-parameter sensing module, including temperature and humidity sensors, is integrated at preset intervals.

[0089] (2) On-site deployment

[0090] The smart grid was laid in the antifreeze layer of the railway subgrade according to the design specifications.

[0091] (3) Operational monitoring

[0092] Visual assessment: During winter operation, maintenance personnel can visually identify the freezing range and depth of different areas of the roadbed through drone visual inspections.

[0093] Risk Analysis: The analysis and early warning module analyzes continuous temporal fluctuation patterns using a spatiotemporal anomaly detection algorithm to determine the spatial propagation probability of anomaly areas. :

[0094] ,in, The distance from the anomaly center. For threshold distance, denoted as the propagation coefficient.

[0095] Disease identification: The system calculates the risk index based on measured data. The data is then used as the core feature vector and input into the disease type identification model to output freeze-thaw cycle or compound disease labels, thereby providing early warning of the risk of local uneven settlement caused by freeze-thaw settlement.

[0096] The embodiments of the present invention aim to protect a self-early warning intelligent geogrid system and method based on multi-parameter fusion, which has the following effects:

[0097] 1. By integrating microencapsulated smart color-changing materials onto the grid substrate, the problem of signal lag in traditional radio frequency monitoring technology, which only generates signals after the grid has undergone significant deformation and fracture, has been solved. This enables intuitive and real-time visual early warning through the naked eye or conventional equipment in the early stages of disease development.

[0098] 2. By deploying a multi-parameter sensor array for strain, humidity, and temperature using flexible printed electronics technology, the problem of limited information dimensions and inability to determine the causes of diseases in existing monitoring methods has been solved. This has enabled a leap from single anomaly detection to preliminary diagnosis covering causes such as water seepage, settlement, and freeze-thaw cycles.

[0099] 3. By leveraging the efficient collaboration between drone visual inspection and wireless sensor data platform, the problems of low efficiency, limited coverage, and reliance on expensive specialized equipment in manual inspection have been solved, enabling rapid location and accurate assessment of hotspots of roadbed defects over large areas.

[0100] 4. By embedding an edge computing model and a piecewise linear compression algorithm into the microcontroller, the problems of high power consumption and large data traffic in continuous monitoring data transmission are solved, and the localized calculation of the disease risk index and the targeted and accurate reporting of hierarchical early warning information are realized.

[0101] 5. By integrating the flexible sensing unit with the grid substrate, the problem of the original reinforcement performance of the grid that may be weakened after the monitoring element is embedded is solved, and the monitoring system is highly compatible with traditional construction technology and can operate stably for a long time in complex soil environments.

[0102] The computer program product of the roadbed defect monitoring method based on a multi-parameter fusion self-early warning intelligent geogrid system provided in this embodiment of the invention includes a computer-readable storage medium storing program code. The instructions included in the program code can be used to execute the methods in the preceding method embodiments. For specific implementation, please refer to the method embodiments, which will not be repeated here.

[0103] Specifically, the storage medium can be a general-purpose storage medium, such as a portable disk or hard disk. When the computer program on the storage medium is run, it can execute the roadbed disease monitoring method based on the multi-parameter fusion self-early warning intelligent geogrid system, thereby enabling early perception, multi-dimensional diagnosis and intuitive visual early warning of hidden diseases inside the geostructure, overcoming the technical defects of existing monitoring methods such as high cost, strong lag and single information.

[0104] If the aforementioned functions are implemented as software functional units and sold or used as independent products, they can be stored in a processor-executable, non-volatile, computer-readable storage medium. Based on this understanding, the technical solution of this invention, essentially, or the part that contributes to the prior art, or a portion of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0105] Finally, it should be noted that the above-described embodiments are merely specific implementations of the present invention, used to illustrate the technical solutions of the present invention, and not to limit it. The scope of protection of the present invention is not limited thereto. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that any person skilled in the art can still modify or easily conceive of changes to the technical solutions described in the foregoing embodiments within the technical scope disclosed in the present invention, or make equivalent substitutions for some of the technical features; and these modifications, changes, or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention, and should all be covered within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A self-early warning intelligent geogrid system based on multi-parameter fusion, characterized in that, It includes a visual warning module, a multi-parameter sensing module, a data processing and communication module, and an analysis and warning module; The visual warning module is integrated on the stress-bearing substrate and changes color in response to environmental stimuli. The multi-parameter sensing module and the visual early warning module are deployed together on the stress-bearing substrate, and the strain, humidity and temperature electrical signals reflecting the roadbed condition are acquired through a micro sensor array. The data processing and communication module is connected to the multi-parameter sensing module, collects and fuses electrical signals, and transmits the extracted feature information to the cloud via a wireless protocol. The analysis and early warning module is connected to the data processing and communication module. It performs multi-dimensional correlation analysis and lesion assessment on the feature information through a preset algorithm model and outputs a graded early warning signal. The analysis and early warning module is equipped with a multi-parameter fusion risk assessor, which calculates the local disease risk index based on the weighted normalization function, combines the spatiotemporal anomaly detection algorithm, determines the propagation probability of the abnormal area based on the continuous temporal fluctuation pattern and spatial connectivity, adjusts the weight coefficient of the multi-parameter fusion risk assessor, and outputs the adjusted local disease risk index. The analysis and early warning module further inputs the adjusted local disease risk index as a core feature vector into the disease type identification model to determine the disease category; By utilizing edge computing hierarchical early warning logic, the disease category and its corresponding risk index are determined based on fuzzy logic reasoning, and hierarchical early warning signals corresponding to different degrees of danger are output.

2. The self-early warning intelligent geogrid system based on multi-parameter fusion according to claim 1, characterized in that, The visual warning module uses microencapsulated smart color-changing material, which is coated or embedded in the rib intersection nodes and / or longitudinal ribs of the grid. The microencapsulated smart color-changing material includes at least one of stress-induced color-changing materials, wet-induced color-changing materials, and thermochromic materials. When the strain generated by the stress-bearing substrate sensed by the multi-parameter sensing module exceeds a preset threshold, the stress-induced color-changing material undergoes a reversible or irreversible change from one color to another. When the humidity of the surrounding soil sensed by the multi-parameter sensing module reaches the response threshold, the wet-induced color-changing material undergoes a reversible or irreversible change from one color to another. When the multi-parameter sensing module detects an abnormal temperature in the freeze-thaw cycle zone or in a localized area, the thermochromic material undergoes a reversible or irreversible change from one color to another.

3. The self-early warning intelligent geogrid system based on multi-parameter fusion according to claim 1, characterized in that, The multi-parameter sensing module employs a miniature sensor array, which is printed or embedded on the surface of the grid using flexible printed electronics technology. The micro-sensor array includes at least one of a flexible resistance strain gauge, a humidity sensor, and a temperature sensor; The flexible resistance strain gauge monitors the distributed strain at various parts of the grid; The humidity sensor monitors the volumetric moisture content of the soil surrounding the grid. The temperature sensor monitors the local temperature of various parts of the grille.

4. The self-early warning intelligent geogrid system based on multi-parameter fusion according to claim 1, characterized in that, The data processing and communication module acquires the electrical signal of the multi-parameter sensing module, fits the time series of each parameter using a piecewise linear compression algorithm, and extracts feature inflection point data. The feature inflection point data is encapsulated to obtain fused feature information; The fused feature information is sent to the analysis and early warning module via LoRa, NB-IoT, or Bluetooth wireless communication protocols.

5. The self-early warning intelligent geogrid system based on multi-parameter fusion according to claim 1, characterized in that, The weighted normalization function used by the analysis and early warning module when calculating the local disease risk index is: ； in, For normalized strain values, ; Normalized humidity value ; For temperature anomaly, ; This is a local disease risk index. , and These are the weighting coefficients. The strain value at the current moment. The humidity value at the current moment. The temperature value at the current moment. The initial strain reference value, This is the initial humidity baseline value. This is the initial temperature reference value. The strain threshold, Humidity threshold This represents the temperature change threshold.

6. The self-early warning intelligent geogrid system based on multi-parameter fusion according to claim 1, characterized in that, The spatiotemporal anomaly detection algorithm performs sliding window statistics on each sensing node. If the current sampled value exceeds the sum of the sliding mean and the preset standard deviation, it is determined to be a node anomaly. The calculation formula is as follows: ,in, This is the current sampled value. It is the moving average. For the sliding standard deviation, The determination coefficient; If multiple consecutive nodes in a space simultaneously malfunction, the regional anomaly event is determined by the spatial propagation probability, which is calculated using the following formula: ,in, For spatial propagation probability, The distance from the anomaly center. For threshold distance, The propagation coefficient; The analysis and early warning module utilizes the spatial propagation probability. Based on the extent of damage and coverage, the weighting coefficients in the multi-parameter fusion risk assessor are adaptively adjusted to correct and output a local disease risk index that reflects the degree of local disease development.

7. The self-early warning intelligent geogrid system based on multi-parameter fusion according to claim 1, characterized in that, The disease type identification model receives a feature vector constructed from the adjusted risk index and the corresponding sensor parameters. ; A lightweight neural network algorithm is used to identify and output corresponding disease type labels, which include at least one of the following: normal state, cavitation and settlement, abnormal seepage, freeze-thaw cycle, and compound disease. The disease type identification model is deployed in a microcontroller using TensorFlow Lite Micro.

8. The self-early warning intelligent geogrid system based on multi-parameter fusion according to claim 1, characterized in that, The hierarchical early warning logic inputs the disease type label and the associated risk index into the fuzzy inference engine, and outputs an early warning level from 0 to 3 according to the preset fuzzy rule set; Among them, low strain and normal parameters correspond to level 0, while high strain accompanied by high humidity and abnormal temperature corresponds to the highest level, level 3.

9. An intelligent geogrid monitoring system, characterized in that, Includes inspection units and a cloud-based monitoring platform; The inspection unit uses a drone or inspection vehicle equipped with a high-resolution camera, the resolution of which is no less than 20 million pixels, to generate a visual distribution map of the disease by recognizing the color change areas of the visual warning module. The cloud-based monitoring platform receives wirelessly transmitted early warning signals and spatially matches them with the visual distribution map of the disease. The intelligent geogrid monitoring system monitors the self-early warning intelligent geogrid system based on multi-parameter fusion as described in any one of claims 1-8.

10. A method for monitoring subgrade defects using a self-early warning intelligent geogrid system based on multi-parameter fusion, characterized in that, include: By simultaneously sensing the stress and environmental parameters of the roadbed through the embedded intelligent geogrid, and providing preliminary intuitive warnings through the visual early warning module; By combining the risk index calculated by the analysis and early warning module with the disease type, the location, severity and cause of the disease are displayed on the cloud monitoring platform using a GIS map; Based on the output graded early warning signals and the color-changing area images captured by the drone, the operation and maintenance personnel formulate corresponding maintenance decisions; The self-early warning intelligent geogrid system based on multi-parameter fusion is the self-early warning intelligent geogrid system based on multi-parameter fusion as described in any one of claims 1-8.

Images