Method and device for detecting, controlling and controlling compaction quality of highway subgrade

Abstract

The invention provides a highway subgrade compaction quality detection management and control method and device, and relates to the technical field of subgrade management and control, and the method comprises the steps: obtaining multi-source subgrade data; performing edge preprocessing on the multi-source roadbed data to obtain edge roadbed data; performing dynamic resource allocation on the edge roadbed data based on a block chain consensus mechanism and a 5G network slice to obtain a standard data stream; performing anomaly judgment on the standard data stream according to a self-adaptive threshold mechanism and an anomaly scoring model to obtain an anomaly judgment result; predicting the fault probability of the equipment based on the anomaly judgment result and a generalized regression neural network to obtain a fault probability prediction value; and triggering a three-level management and control decision mechanism according to the fault probability predicted value, and executing roadbed detection and equipment maintenance decisions to obtain a management and control result. The problem that early warning information cannot be generated based on identification of abnormal data is solved.

Description

Technical Field

[0001] This invention relates to the field of roadbed management technology, and more specifically, to a method and device for detecting and managing the compaction quality of highway roadbeds. Background Technology

[0002] In the field of roadbed management technology, manual sampling inspection is commonly used to assess roadbed quality. Inspectors, following established sampling rules, conduct a series of tasks including sample collection, data recording, and preliminary analysis at different locations along the roadbed. However, in actual inspections, the professional skills of inspectors vary, and their understanding and execution of inspection standards and operating procedures differ. Some inexperienced inspectors may fail to select representative samples as required during sample collection, or make clerical errors or omissions during data recording. Furthermore, manual sampling inspection can only obtain localized data, making it difficult to achieve effective coverage of the entire road section. It also lacks the function of automatically identifying abnormal data, resulting in the inability to generate early warning information based on the identification of abnormal data.

[0003] Therefore, there is an urgent need for a method and device for detecting and controlling the compaction quality of highway subgrade, which solves the problem of not being able to generate early warning information based on the identification of abnormal data. Summary of the Invention

[0004] The purpose of this invention is to provide a method and apparatus for detecting and controlling the compaction quality of highway subgrade, so as to improve the above-mentioned problems. To achieve the above objective, the technical solution adopted by this invention is as follows:

[0005] Firstly, this application provides a method for testing and controlling the compaction quality of highway subgrade, including:

[0006] Acquire multi-source roadbed data, which includes vibration sensor data, environmental factors, and ground-penetrating radar data;

[0007] Edge preprocessing is performed on the multi-source roadbed data to obtain edge roadbed data;

[0008] Based on the blockchain consensus mechanism and 5G network slicing, the edge roadbed data is dynamically allocated to obtain a standard data stream;

[0009] Anomaly detection is performed on the standard data stream based on an adaptive threshold mechanism and anomaly scoring model to obtain anomaly detection results;

[0010] Based on the anomaly detection results and the generalized regression neural network to predict the failure probability of the device, the predicted failure probability value is obtained.

[0011] The three-level control decision-making mechanism is triggered based on the predicted failure probability value to execute roadbed inspection and equipment maintenance decisions, thereby obtaining the control results.

[0012] Secondly, this application also provides a highway subgrade compaction quality testing and control device, comprising:

[0013] The acquisition module is used to acquire multi-source roadbed data, which includes vibration sensor data, environmental factors, and ground-penetrating radar data.

[0014] The preprocessing module is used to perform edge preprocessing on the multi-source roadbed data to obtain edge roadbed data;

[0015] The allocation module is used to dynamically allocate resources to the edge roadbed data based on the blockchain consensus mechanism and 5G network slicing to obtain a standard data stream.

[0016] The judgment module is used to judge the anomalies of the standard data stream according to the adaptive threshold mechanism and the anomaly scoring model, and obtain the anomaly judgment result.

[0017] The prediction module is used to obtain a predicted fault probability value based on the anomaly judgment result and the generalized regression neural network to predict the fault probability of the device.

[0018] The execution module is used to trigger a three-level control decision mechanism based on the predicted fault probability value, execute roadbed detection and equipment maintenance decisions, and obtain control results.

[0019] The beneficial effects of this invention are as follows:

[0020] This invention ensures efficient and reliable data transmission through blockchain consensus mechanisms and dynamic resource allocation via 5G network slicing, thereby guaranteeing data security and integrity and improving transmission efficiency. During data transmission, an adaptive threshold mechanism and an anomaly scoring model are used to monitor and identify anomalies in the standard data stream in real time, promptly detecting anomalies in roadbed compaction and providing early warnings. Simultaneously, a generalized regression neural network is used to predict equipment failure probabilities, accurately predicting potential faults and scheduling maintenance in advance to avoid inspection interruptions or roadbed compaction quality issues. A three-level control decision-making mechanism triggers corresponding decisions based on the predicted failure probability values, executing roadbed inspections and equipment maintenance to achieve refined control of roadbed compaction quality. In summary, this invention solves the problem of being unable to generate early warning information based on the identification of anomaly data by ensuring data transmission, accurately identifying and warning of abnormal data, and predicting and maintaining equipment failures in advance.

[0021] Other features and advantages of the invention will be set forth in the following description, and will be apparent in part from the description, or may be learned by practicing embodiments of the invention. The objects and other advantages of the invention may be realized and obtained by means of the structures particularly pointed out in the written description, claims, and drawings. Attached Figure Description

[0022] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings used in the embodiments will be briefly introduced below. It should be understood that the following drawings only show some embodiments of the present invention and should not be regarded as a limitation on the scope. For those skilled in the art, other related drawings can be obtained based on these drawings without creative effort.

[0023] Figure 1 This is a schematic diagram of the process for detecting and controlling the compaction quality of highway subgrade as described in this embodiment of the invention;

[0024] Figure 2 This is a schematic diagram of the highway subgrade compaction quality testing and control equipment described in an embodiment of the present invention.

[0025] The markings in the diagram are: 800, Highway subgrade compaction quality testing and control equipment; 801, Processor; 802, Memory; 803, Multimedia component; 804, I / O interface; 805, Communication component. Detailed Implementation

[0026] 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 with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. The components of the embodiments of the present invention described and shown in the accompanying drawings can generally be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of the present invention provided in the accompanying drawings is not intended to limit the scope of the claimed invention, but merely to illustrate selected embodiments of the invention. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without inventive effort are within the scope of protection of the present invention.

[0027] It should be noted that similar reference numerals and letters in the following figures indicate similar items; therefore, once an item is defined in one figure, it does not need to be further defined and explained in subsequent figures. Furthermore, in the description of this invention, terms such as "first," "second," etc., are used only to distinguish descriptions and should not be construed as indicating or implying relative importance.

[0028] Example 1:

[0029] This embodiment provides a method for detecting and controlling the compaction quality of highway subgrade.

[0030] See Figure 1 The figure shows that the method includes steps S1 to S6, including:

[0031] S1: Acquire multi-source roadbed data, which includes vibration sensor data, environmental factors, and ground-penetrating radar data;

[0032] In this step, the vibration sensor data is used to monitor the roadbed surface by a road roller and a testing vehicle to obtain real-time monitoring vibration frequency, amplitude, and acceleration, which are used to reflect the compaction energy transfer and soil response; the ground-penetrating radar data is used to detect the geological structure of the roadbed (such as soil type and density distribution) and assist in the dynamic deployment of sensors.

[0033] It also includes the compaction machinery trajectory, which is monitored by a GPS / laser scanner and used to generate a three-dimensional terrain model of the roadbed, and to identify key locations such as compaction overlap areas and edge areas.

[0034] S2: Perform edge preprocessing on the multi-source roadbed data to obtain edge roadbed data;

[0035] To clarify the specific method for acquiring edge roadbed data, step S2 includes S21 to S25, specifically:

[0036] S21: Perform wavelet transform on the multi-source roadbed data to obtain wavelet roadbed data;

[0037] In this step, the wavelet transform analyzes the data in the time and frequency domains to extract features at different scales.

[0038] S22: Perform empirical mode decomposition on the wavelet roadbed data, and filter the effective intrinsic mode function components through a preset kurtosis coefficient to obtain noise-reduced roadbed data;

[0039] In this step, the empirical mode decomposition adaptively decomposes the signal based on its characteristics to identify and correct sensor faults or environmental interference data.

[0040] S23: Based on a long short-term memory network, feature extraction is performed on the noise-reduced roadbed data to obtain key features;

[0041] In this step, the Long Short-Term Memory (LSTM) network solves the problem that traditional feature extraction methods cannot effectively handle long-term dependencies in time series data.

[0042] S24: Perform feature weighting calculation on the key features according to the edge decision tree algorithm to generate a local compaction score;

[0043] In this step, the key features are trained according to the edge decision tree algorithm, the importance score of each feature is extracted, the weight of each feature is calculated according to the feature weight formula, the key features are multiplied by their corresponding weights and summed to generate a local compaction score.

[0044] S25: Mark the local compaction score on the noise reduction subgrade data to generate edge subgrade data.

[0045] In this step, the local compaction score is time-aligned and associated with the noise-reduced subgrade data, and marked on the noise-reduced subgrade data to form edge subgrade data with compaction score.

[0046] S3: Based on the blockchain consensus mechanism and 5G network slicing, the edge roadbed data is dynamically allocated to obtain a standard data stream;

[0047] In this step, the blockchain consensus mechanism is used to ensure the security and integrity of data, and the dynamic resource allocation of the 5G network slice dynamically adjusts network resources according to data transmission needs to improve data transmission efficiency.

[0048] To clarify the specific method for acquiring the standard data stream, step S3 includes S31 to S35, specifically:

[0049] S31: Based on the blockchain consensus mechanism, perform hash calculation on the edge roadbed data, and input the calculated hash value into the blockchain for protection processing to form a data fingerprint chain;

[0050] In this step, the edge data is hashed based on a blockchain consensus mechanism. The calculated hash values ​​are then linked sequentially into a chain structure for protection (the hash value of the previous data packet serves as the input for the next data packet), forming an immutable data fingerprint chain. Simultaneously, header and trailer information are added to each data packet to make it identifiable and routable, ensuring the integrity and security of the data encapsulation.

[0051] S32: Dynamically allocate resources to the edge roadbed data based on the 5G network slice to obtain the slice resource allocation result;

[0052] In this step, based on 5G network slicing technology, and considering the real-time needs, priorities, real-time monitoring results, and algorithm optimization of edge roadbed data, the allocation of network, computing, and storage resources is dynamically adjusted. This ensures high reliability of data transmission and avoids transmission failures due to insufficient resources or unreasonable allocation.

[0053] S33: Based on the slice resource allocation result, the data fingerprint chain is encrypted and transmitted to obtain the transmission status result;

[0054] In this step, based on the slice resource allocation result, the data fingerprint chain is encrypted using an encryption algorithm. The encrypted data is transmitted through the allocated slice resources, and the transmission status is recorded in real time to form a transmission status result.

[0055] The transmission status includes transmission delay, packet loss rate, and encryption integrity.

[0056] S34: Based on the blockchain consensus mechanism, the slice is reconfigured, and the backup slice is activated according to the transmission status result to obtain the slice configuration scheme;

[0057] In this step, based on the blockchain consensus mechanism, the load, performance, and fault status of the slices are monitored in real time. When a slice is overloaded, fails, or its performance degrades and cannot meet the demand, the slice resources are automatically or semi-automatically reconfigured, and a standby slice is activated to take over the resources or functions of the primary slice, ensuring the continuity and stability of network services. This ultimately results in a new slice configuration scheme.

[0058] S35: Adjust the transmission process according to the transmission status record and the slice configuration scheme to generate a standard data stream.

[0059] In this step, the transmission strategy is dynamically adjusted based on the transmission status record and sliced ​​resources. Data transmission paths, retransmission mechanisms, and resource allocation are optimized to ensure data transmission stability. The adjusted data is then integrated to generate a standard data stream.

[0060] S4: Based on the adaptive threshold mechanism and anomaly scoring model, perform anomaly judgment on the standard data stream to obtain anomaly judgment results;

[0061] In this step, the adaptive threshold mechanism automatically adjusts the threshold based on the dynamic characteristics of the standard data stream, and the anomaly scoring model scores the standard data stream. When the score exceeds the set threshold, anomalies occurring during the roadbed compaction process are detected in a timely manner, and early warning information is generated.

[0062] To clarify the specific method for obtaining the anomaly judgment result, step S4 includes S41 to S45, specifically:

[0063] S41: The standard data stream is dynamically segmented according to the adaptive threshold mechanism and multimodal features to obtain a data sliding window;

[0064] In this step, the mean and standard deviation of the standard data stream are calculated based on the adaptive threshold mechanism to initialize the adaptive threshold. Changes in multimodal characteristics (abrupt vibration amplitudes or significant changes in environmental factors) are used as the basis for segmentation, and the size of the sliding window is dynamically adjusted to obtain the data sliding window. This ensures that the data within each window have similar characteristics.

[0065] S42: Calculate the standard deviation based on the distribution change of the data sliding window, and obtain the anomaly threshold by comparing the standard deviation with the preset baseline threshold;

[0066] In this step, the standard deviation expression is:

[0067] (1)

[0068] In the above formula (1), In time The dynamic standard deviation at time t, For window size, For the first in the sliding window Data points, In time The window mean at any given time.

[0069] S43: Extract time-series features from the data sliding window based on the multi-source roadbed data, and calculate the comprehensive anomaly score by combining the weights of the time-series features and environmental factors;

[0070] To clarify the specific method for obtaining the comprehensive anomaly score, step S43 includes steps S431 to S434, which are as follows:

[0071] S431: Based on the data rate, equipment moving speed and environmental factors in the multi-source roadbed data, the window size is dynamically adjusted to obtain a multi-dimensional dynamic window adaptive algorithm;

[0072] This step addresses the problem of insufficient adaptability of traditional fixed window size methods under different environments and data changes.

[0073] S432: Based on the multi-dimensional dynamic window adaptive algorithm, the vibration time series features are extracted from the data sliding window to obtain the time series features;

[0074] S433: Based on a machine learning model, time-series features and environmental factors are dynamically learned to obtain environmental factor weights;

[0075] In this step, the time-series features and environmental factors are trained based on a machine learning model, the importance of each feature is extracted, and the weights of environmental factors are calculated based on the importance of the features.

[0076] S434: The comprehensive anomaly score is obtained by weighted summation based on the time-series characteristics and environmental factor weights.

[0077] In this step, the expression for the comprehensive anomaly score is:

[0078] (2)

[0079] In the above formula (2), In time The overall anomaly score at any given time. The total number of features, For the first The weights of each feature For the first The sample at the th Values ​​on each feature In order to be in Time of the first The window mean of each feature, In order to be in Time of the first The dynamic standard deviation of each feature.

[0080] S44: Based on adaptive moment estimation and Nesterov momentum training, anomaly scoring model is used, and online reinforcement learning algorithm is combined to dynamically adjust the anomaly threshold to obtain optimized anomaly results;

[0081] To clarify the specific method for obtaining abnormal results, step S44 includes steps S441 to S443, specifically:

[0082] S441: Based on the combination of adaptive moment estimation and Nesterov momentum, the anomaly scoring model is trained to obtain an optimized anomaly scoring model;

[0083] In this step, the parameters of the anomaly scoring model are initialized based on adaptive moment estimation and Nesterov momentum. The hyperparameters of the Adam optimizer are then set. In each iteration, the gradient of the loss function is calculated and the model parameters are updated. The iteration is repeated until the model converges, resulting in the optimized anomaly scoring model.

[0084] S442: Dynamically adjust the parameters of the optimized anomaly scoring model based on the cumulative historical gradient and adaptive learning rate in the optimized anomaly scoring model to obtain the optimized model parameters;

[0085] In each iteration, the current gradient is calculated and the accumulated historical gradients are updated. The adjusted learning rate is then calculated and the model parameters are updated using the adjusted learning rate.

[0086] S443: Based on optimized model parameters, the anomaly threshold is dynamically adjusted through online reinforcement learning algorithms and environmental changes to obtain optimized anomaly results.

[0087] In this step, the parameters of the Q-learning algorithm are initialized based on the optimized model parameters. In each iteration, environmental changes are detected, actions are selected and rewards are calculated based on the environmental changes, and the Q value is updated. The anomaly threshold is adjusted based on the updated Q value, and the anomaly scoring model is evaluated using the optimized anomaly threshold to generate optimized anomaly results.

[0088] S45: Based on the optimized anomaly results and the comprehensive anomaly score, an anomaly judgment result is generated.

[0089] In this step, the judgment is made based on the optimized anomaly results and the comprehensive anomaly score, which improves the accuracy of anomaly judgment.

[0090] S5: Based on the anomaly judgment result and the generalized regression neural network to predict the failure probability of the device, the failure probability prediction value is obtained;

[0091] In this step, the anomaly detection results and the generalized regression neural network accurately predict the probability of equipment failure. By analyzing equipment operating data, the neural network model predicts potential equipment failures in advance.

[0092] To clarify the specific method for obtaining the fault probability prediction value, step S5 includes S51 to S56, specifically:

[0093] S51: Obtain historical trends;

[0094] S52: Based on the anomaly detection result, feature extraction is performed to obtain a feature vector;

[0095] S53: Construct a fault prediction model for the generalized regression neural network based on the feature vectors;

[0096] In this step, a fault prediction model is constructed using a generalized regression neural network (GRNN) based on feature vectors and historical fault data as the training set.

[0097] S54: Based on the historical trend and the feature vector, calculate the changes in decision variables to obtain an improved predictive dynamic multi-objective evolutionary algorithm;

[0098] In this step, the improved predictive dynamic multi-objective evolutionary algorithm is a dynamic multi-objective optimization algorithm based on neural network learning of environmental change patterns. The improved predictive dynamic multi-objective evolutionary algorithm optimizes the multi-objective problem by dynamically adjusting decision variables.

[0099] S55: Based on the deep coupling of the fault prediction model and the improved prediction dynamic multi-objective evolutionary algorithm, the optimal decision scheme is obtained by optimizing the preset equipment operating parameters.

[0100] In this step, the fault prediction model is combined with the optimization algorithm to improve the overall performance of the system. By optimizing the equipment operating parameters, the optimal decision scheme is generated to reduce the equipment failure rate.

[0101] S56: Based on the optimal decision scheme, predict the failure probability of the equipment and obtain the failure probability prediction value.

[0102] S6: Trigger the three-level control decision mechanism based on the predicted fault probability value, execute roadbed inspection and equipment maintenance decisions, and obtain control results.

[0103] In this step, if the compaction degree in the predicted failure probability value is lower than the threshold or if the equipment experiences a sudden failure, a three-level response is immediately triggered. The three-level response includes a red alert, automatic adjustment, and expert system intervention.

[0104] Red Alert: A notification is sent to the construction terminal and monitoring platform, prompting immediate shutdown and inspection. Automatic Adjustment: Edge nodes temporarily increase sensor sampling rates and activate backup equipment. Expert System Intervention: Remote experts guide on-site adjustments via an AR visualization interface.

[0105] Example 2:

[0106] This embodiment provides a highway subgrade compaction quality testing and control device, the device comprising:

[0107] The acquisition module is used to acquire multi-source roadbed data, which includes vibration sensor data, environmental factors, and ground-penetrating radar data.

[0108] The preprocessing module is used to perform edge preprocessing on the multi-source roadbed data to obtain edge roadbed data;

[0109] The allocation module is used to dynamically allocate resources to the edge roadbed data based on the blockchain consensus mechanism and 5G network slicing to obtain a standard data stream.

[0110] To clarify the specific methods for obtaining the allocation module, the following are included:

[0111] The computing unit is used to perform hash calculations on the edge roadbed data based on the blockchain consensus mechanism, and input the calculated hash value into the blockchain for protection processing to form a data fingerprint chain;

[0112] The allocation unit is used to dynamically allocate resources to the edge roadbed data according to the 5G network slice, and obtain the slice resource allocation result;

[0113] A transmission unit is used to encrypt and transmit the data fingerprint chain based on the slice resource allocation result, and obtain a transmission status result.

[0114] The configuration unit is used to reconfigure the slice based on the blockchain consensus mechanism and activate the standby slice according to the transmission status result to obtain the slice configuration scheme.

[0115] The adjustment unit is used to adjust the transmission process according to the transmission status record and the slice configuration scheme to generate a standard data stream.

[0116] The judgment module is used to judge the anomalies of the standard data stream according to the adaptive threshold mechanism and the anomaly scoring model, and obtain the anomaly judgment result.

[0117] To clearly determine the specific method of obtaining the module, the following are included:

[0118] The segmentation unit is used to dynamically segment the standard data stream according to the adaptive threshold mechanism and multimodal features to obtain a data sliding window;

[0119] The comparison unit is used to calculate the standard deviation based on the distribution change of the data sliding window, and obtain the anomaly threshold by comparing the standard deviation with a preset baseline threshold;

[0120] The extraction unit is used to extract time-series features from the data sliding window based on the multi-source roadbed data, and calculate the comprehensive anomaly score by combining the weights of the time-series features and environmental factors.

[0121] To clarify the specific methods for obtaining the extraction units, the following are included:

[0122] The first adjustment subunit is used to dynamically adjust the window size based on the data rate, equipment moving speed and environmental factors in the multi-source roadbed data, to obtain a multi-dimensional dynamic window adaptive algorithm.

[0123] Extraction sub-units are used to extract vibration time-series features from the data sliding window based on the multi-dimensional dynamic window adaptive algorithm to obtain time-series features;

[0124] The learning sub-unit is used to dynamically learn temporal features and environmental factors based on a machine learning model, and obtain the environmental factor weights.

[0125] The calculation subunit is used to perform weighted summation based on the time-series characteristics and environmental factor weights to obtain a comprehensive anomaly score.

[0126] The training unit is used to train the anomaly scoring model based on adaptive moment estimation and Nesterov momentum, and dynamically adjust the anomaly threshold by combining online reinforcement learning algorithm to obtain optimized anomaly results;

[0127] To clarify the specific methods for obtaining training units, the following are included:

[0128] The training subunit is used to train the anomaly scoring model based on the combination of adaptive moment estimation and Nesterov momentum, so as to obtain an optimized anomaly scoring model.

[0129] The second adjustment subunit is used to dynamically adjust the parameters of the optimized anomaly scoring model based on the accumulated historical gradient and adaptive adjustment learning rate in the optimized anomaly scoring model, so as to obtain the optimized model parameters.

[0130] The third adjustment subunit is used to dynamically adjust the anomaly threshold based on the optimized model parameters through online reinforcement learning algorithms and environmental changes, thereby obtaining optimized anomaly results.

[0131] The judgment unit is used to make a judgment based on the optimized anomaly result and the comprehensive anomaly score, and generate anomaly judgment result.

[0132] The prediction module is used to obtain a predicted fault probability value based on the anomaly judgment result and the generalized regression neural network to predict the fault probability of the device.

[0133] The execution module is used to trigger a three-level control decision mechanism based on the predicted fault probability value, execute roadbed detection and equipment maintenance decisions, and obtain control results.

[0134] It should be noted that the specific manner in which each module performs its operation in the apparatus described in the above embodiments has been described in detail in the embodiments of the method, and will not be elaborated here.

[0135] Example 3:

[0136] Corresponding to the above method embodiments, this embodiment also provides a highway subgrade compaction quality testing and control device. The highway subgrade compaction quality testing and control device described below and the highway subgrade compaction quality testing and control method described above can be referred to in correspondence.

[0137] Figure 2 This is a block diagram illustrating a highway subgrade compaction quality testing and control device 800 according to an exemplary embodiment. Figure 2 As shown, the highway subgrade compaction quality testing and control device 800 may include: a processor 801 and a memory 802. The highway subgrade compaction quality testing and control device 800 may also include one or more of the following: a multimedia component 803, an I / O interface 804, and a communication component 805.

[0138] The processor 801 controls the overall operation of the highway subgrade compaction quality testing and control equipment 800 to complete all or part of the steps in the aforementioned highway subgrade compaction quality testing and control method. The memory 802 stores various types of data to support the operation of the highway subgrade compaction quality testing and control equipment 800. This data may include, for example, instructions for any application or method operating on the highway subgrade compaction quality testing and control equipment 800, as well as application-related data such as contact data, sent and received messages, images, audio, video, etc. The memory 802 can be implemented using any type of volatile or non-volatile storage device or a combination thereof, such as Static Random Access Memory (SRAM), Electrically Erasable Programmable Read-Only Memory (EEPROM), Erasable Programmable Read-Only Memory (EPROM), Programmable Read-Only Memory (PROM), Read-Only Memory (ROM), magnetic storage, flash memory, magnetic disk, or optical disk. The multimedia component 803 may include a screen and an audio component. The screen may be, for example, a touchscreen, and the audio component is used to output and / or input audio signals. For example, the audio component may include a microphone for receiving external audio signals. The received audio signals may be further stored in the memory 802 or transmitted via the communication component 805. The audio component also includes at least one speaker for outputting audio signals. I / O interface 804 provides an interface between processor 801 and other interface modules, such as keyboards, mice, and buttons. These buttons can be virtual or physical. Communication component 805 is used for wired or wireless communication between the highway subgrade compaction quality inspection and control equipment 800 and other devices. Wireless communication includes, for example, Wi-Fi, Bluetooth, Near Field Communication (NFC), 2G, 3G, or 4G, or a combination thereof. Therefore, the corresponding communication component 805 may include a Wi-Fi module, a Bluetooth module, and an NFC module.

[0139] In an exemplary embodiment, the highway subgrade compaction quality inspection and control device 800 may be implemented by one or more application-specific integrated circuits (ASICs), digital signal processors (DSPs), digital signal processing devices (DSPDs), programmable logic devices (PLDs), field-programmable gate arrays (FPGAs), controllers, microcontrollers, microprocessors, or other electronic components to perform the highway subgrade compaction quality inspection and control method described above.

[0140] Example 4:

[0141] Corresponding to the above method embodiments, this embodiment also provides a medium. The medium described below can be referred to in conjunction with the highway subgrade compaction quality detection and control method described above.

[0142] A medium storing a computer program, which, when executed by a processor, implements the steps of the highway subgrade compaction quality detection and control method described in the above method embodiments.

[0143] The medium can specifically be any medium capable of storing program code, such as a USB flash drive, external hard drive, read-only memory (ROM), random access memory (RAM), magnetic disk, or optical disk.

[0144] The above description is merely a preferred embodiment of the present invention and is not intended to limit the invention. Various modifications and variations can be made to the present invention by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

[0145] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the present invention should be included 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 method for detecting and controlling the compaction quality of highway subgrade, characterized in that, include: Acquire multi-source roadbed data, which includes vibration sensor data, environmental factors, and ground-penetrating radar data; Edge preprocessing is performed on the multi-source roadbed data to obtain edge roadbed data; Based on the blockchain consensus mechanism and 5G network slicing, the edge roadbed data is dynamically allocated to obtain a standard data stream; Anomaly detection is performed on the standard data stream based on an adaptive threshold mechanism and anomaly scoring model to obtain anomaly detection results; Based on the anomaly detection results and the generalized regression neural network to predict the failure probability of the device, the predicted failure probability value is obtained. The three-level control decision-making mechanism is triggered based on the predicted failure probability value to execute roadbed inspection and equipment maintenance decisions, thereby obtaining the control results.

2. The method for detecting and controlling the compaction quality of highway subgrade according to claim 1, characterized in that, based on a blockchain consensus mechanism and 5G network slicing, dynamic resource allocation is performed on the edge subgrade data to obtain a standard data stream, including: The edge roadbed data is hashed based on the blockchain consensus mechanism, and the hash value is input into the blockchain for protection processing to form a data fingerprint chain. Dynamic resource allocation is performed on the edge roadbed data based on 5G network slicing to obtain slice resource allocation results; Based on the slice resource allocation result, the data fingerprint chain is encrypted and transmitted to obtain the transmission status result; Based on the blockchain consensus mechanism, the slices are reconfigured, and the backup slices are activated according to the transmission status results to obtain the slice configuration scheme. The transmission process is adjusted based on the transmission status record and the slice configuration scheme to generate a standard data stream.

3. The method for detecting and controlling the compaction quality of highway subgrade according to claim 1, characterized in that, Anomaly detection is performed on the standard data stream based on an adaptive threshold mechanism and an anomaly scoring model to obtain anomaly detection results, including: The standard data stream is dynamically segmented based on an adaptive threshold mechanism and multimodal features to obtain a data sliding window; The standard deviation is calculated based on the distribution change of the data sliding window, and the anomaly threshold is obtained by comparing the standard deviation with the preset baseline threshold. Based on the multi-source roadbed data, time-series features are extracted from the data sliding window, and a comprehensive anomaly score is calculated by combining the weights of the time-series features and environmental factors. Based on adaptive moment estimation and Nesterov momentum training anomaly scoring model, combined with online reinforcement learning algorithm to dynamically adjust anomaly threshold, optimized anomaly results are obtained; An anomaly judgment result is generated based on the optimized anomaly result and the comprehensive anomaly score.

4. The method for detecting and controlling the compaction quality of highway subgrade according to claim 3, characterized in that, based on the multi-source subgrade data, time-series features are extracted from the data sliding window, and a comprehensive anomaly score is calculated by combining the weights of the time-series features and environmental factors, including: Based on the data rate, equipment moving speed and environmental factors in the multi-source roadbed data, the window size is dynamically adjusted to obtain a multi-dimensional dynamic window adaptive algorithm. Based on the multi-dimensional dynamic window adaptive algorithm, vibration time-series features are extracted from the data sliding window to obtain time-series features; The environmental factor weights are obtained by dynamically learning time-series features and environmental factors based on machine learning models. The comprehensive anomaly score is obtained by weighted summation based on the time-series characteristics and environmental factor weights.

5. The method for detecting and controlling the compaction quality of highway subgrade according to claim 3, characterized in that, based on adaptive moment estimation and Nesterov momentum training anomaly scoring model, and combined with online reinforcement learning algorithm, the anomaly threshold is dynamically adjusted to obtain optimized anomaly results, including: An optimized anomaly scoring model is obtained by training the anomaly scoring model by combining adaptive moment estimation and Nesterov momentum. The parameters of the optimized anomaly scoring model are obtained by dynamically adjusting the parameters of the optimized anomaly scoring model based on the cumulative historical gradient and the adaptive learning rate. Based on optimized model parameters, the anomaly threshold is dynamically adjusted through online reinforcement learning algorithms and environmental changes to obtain optimized anomaly results.

6. A device for detecting and controlling the compaction quality of highway subgrade, characterized in that, include: The acquisition module is used to acquire multi-source roadbed data, which includes vibration sensor data, environmental factors, and ground-penetrating radar data. The preprocessing module is used to perform edge preprocessing on the multi-source roadbed data to obtain edge roadbed data; The allocation module is used to dynamically allocate resources to the edge roadbed data based on the blockchain consensus mechanism and 5G network slicing to obtain a standard data stream. The judgment module is used to judge the anomalies of the standard data stream according to the adaptive threshold mechanism and the anomaly scoring model, and obtain the anomaly judgment result. The prediction module is used to obtain a predicted fault probability value based on the anomaly judgment result and the generalized regression neural network to predict the fault probability of the device. The execution module is used to trigger a three-level control decision mechanism based on the predicted fault probability value, execute roadbed detection and equipment maintenance decisions, and obtain control results.

7. The highway subgrade compaction quality detection and control device according to claim 6, characterized in that the distribution module comprises: The computing unit is used to perform hash calculations on the edge roadbed data based on the blockchain consensus mechanism, and input the calculated hash value into the blockchain for protection processing to form a data fingerprint chain. The allocation unit is used to dynamically allocate resources to the edge roadbed data according to the 5G network slice, and obtain the slice resource allocation result; A transmission unit is used to encrypt and transmit the data fingerprint chain based on the slice resource allocation result, and obtain a transmission status result. The configuration unit is used to reconfigure the slice based on the blockchain consensus mechanism and activate the standby slice according to the transmission status result to obtain the slice configuration scheme. The adjustment unit is used to adjust the transmission process according to the transmission status record and the slice configuration scheme to generate a standard data stream.

8. The highway subgrade compaction quality detection and control device according to claim 6, characterized in that the judgment module includes: The segmentation unit is used to dynamically segment the standard data stream according to the adaptive threshold mechanism and multimodal features to obtain a data sliding window; The comparison unit is used to calculate the standard deviation based on the distribution change of the data sliding window, and obtain the anomaly threshold by comparing the standard deviation with a preset baseline threshold; The extraction unit is used to extract time-series features from the data sliding window based on the multi-source roadbed data, and calculate the comprehensive anomaly score by combining the weights of the time-series features and environmental factors. The training unit is used to train the anomaly scoring model based on adaptive moment estimation and Nesterov momentum, and dynamically adjust the anomaly threshold by combining online reinforcement learning algorithm to obtain optimized anomaly results; The judgment unit is used to make a judgment based on the optimized anomaly result and the comprehensive anomaly score, and generate anomaly judgment result.

9. The highway subgrade compaction quality detection and control device according to claim 8, characterized in that the extraction unit comprises: The first adjustment subunit is used to dynamically adjust the window size based on the data rate, equipment moving speed and environmental factors in the multi-source roadbed data, to obtain a multi-dimensional dynamic window adaptive algorithm. The extraction sub-unit is used to extract vibration time-series features from the data sliding window based on the multi-dimensional dynamic window adaptive algorithm to obtain time-series features; The learning sub-unit is used to dynamically learn temporal features and environmental factors based on a machine learning model, and obtain the environmental factor weights. The calculation subunit is used to perform weighted summation based on the time-series characteristics and environmental factor weights to obtain a comprehensive anomaly score.

10. The highway subgrade compaction quality detection and control device according to claim 6, characterized in that, Training units include: The training subunit is used to train the anomaly scoring model based on the combination of adaptive moment estimation and Nesterov momentum, so as to obtain an optimized anomaly scoring model. The second adjustment subunit is used to dynamically adjust the parameters of the optimized anomaly scoring model based on the accumulated historical gradient and adaptive adjustment learning rate in the optimized anomaly scoring model, so as to obtain the optimized model parameters. The third adjustment subunit is used to dynamically adjust the anomaly threshold based on the optimized model parameters through online reinforcement learning algorithms and environmental changes, thereby obtaining optimized anomaly results.

Images