Intelligent acquisition and analysis monitoring method and system for urban tube well safety

Abstract

The invention discloses an intelligent acquisition and analysis monitoring method and system for urban tube well safety, and relates to the technical field of urban tube well monitoring and control, and the method comprises the steps: obtaining the vibration data of a well lid; inputting the vibration data into the MSCNN model to obtain abnormal vibration data; the MSCNN model is a pre-training model which takes the vibration data as input and takes the abnormal vibration data as output and is used for identifying the abnormal vibration data in the vibration data; inputting the abnormal vibration data into a VGG-CNN model for damage identification and classification to obtain a damage category prediction result of the well lid; the VGG-CNN model is a pre-training model which takes the abnormal vibration data as input and the damage category prediction result as output and is used for predicting the health state of the well lid according to the abnormal vibration data. According to the invention, accurate damage identification and classification and health state evaluation of the well lid can be realized, and the safety of pedestrians and vehicles around the well lid of the urban tube well is ensured.

Description

Technical Field

[0001] The present application relates to the field of urban pipe well monitoring and control technology, and in particular to an intelligent collection and analysis monitoring method and system for urban pipe well safety. Background Art

[0002] At present, traditional urban manhole monitoring methods usually rely on regular manual inspections and some simple sensor monitoring methods, such as basic vibration sensors, three-axis acceleration sensors, and temperature and humidity sensors. These monitoring methods can provide monitoring and protection to a certain extent, such as identifying whether the manhole cover has been illegally moved or detecting obvious physical damage to the manhole cover. However, such methods are usually only discovered after the problem has become obvious or has caused major consequences. It is impossible to predict the extent of damage to the manhole cover in advance and it is impossible to accurately assess the health status of the manhole cover. For example, small cracks, slight deformations, or slight looseness, if these subtle changes are not quickly identified and handled, may evolve into more serious problems, and even cause huge safety hazards to pedestrians and vehicles around the manhole cover. Summary of the invention

[0003] The purpose of this application is to provide an intelligent collection and analysis monitoring method and system for urban manhole safety, which can realize accurate damage identification and classification and health status assessment of manhole covers, and ensure the safety of pedestrians and vehicles around urban manhole covers.

[0004] To achieve the above objectives, this application provides the following solutions.

[0005] In a first aspect, the present application provides an intelligent data collection, analysis and monitoring method for urban pipe well safety, and the intelligent data collection, analysis and monitoring method for urban pipe well safety includes the following steps.

[0006] Get the vibration data of the manhole cover.

[0007] The vibration data is input into the MSCNN model to obtain abnormal vibration data; the MSCNN model is a pre-trained model that takes the vibration data as input and the abnormal vibration data as output, and is used to identify the abnormal vibration data in the vibration data; the abnormal vibration data is data corresponding to the abnormal vibration state in the vibration data.

[0008] The abnormal vibration data is input into the VGG-CNN model for damage identification and classification to obtain a damage category prediction result of the manhole cover; the VGG-CNN model is a pre-trained model that takes the abnormal vibration data as input and the damage category prediction result as output, and is used to predict the health status of the manhole cover based on the abnormal vibration data; the damage category prediction result is the probability value corresponding to each damage category of the manhole cover.

[0009] Optionally, obtaining vibration data of the manhole cover specifically includes the following steps.

[0010] An acceleration sensor is installed at the center of the back side of the manhole cover, and the acceleration sensor is used to measure the vibration data of the manhole cover in the vertical direction at a predetermined frequency.

[0011] The manhole cover is monitored by using the acceleration sensor to collect the vibration data.

[0012] Optionally, after the step of acquiring the vibration data of the manhole cover, the intelligent collection and analysis monitoring method for urban manhole safety further includes the following steps.

[0013] The vibration data is preprocessed by data cleaning, data normalization and data noise reduction to obtain preprocessed vibration data; the preprocessed vibration data is used as the vibration data to be input into the MSCNN model to obtain the abnormal vibration data.

[0014] Optionally, the vibration data is preprocessed by data cleaning, data normalization and data noise reduction to obtain preprocessed vibration data, which specifically includes the following steps.

[0015] Abnormal values ​​in the vibration data are removed, and missing values ​​in the vibration data are filled to obtain vibration data after data cleaning.

[0016] The vibration data after data cleaning is normalized to obtain the vibration data after data normalization.

[0017] A wavelet denoising algorithm is used to perform data denoising on the vibration data after the data normalization to obtain the preprocessed vibration data.

[0018] Optionally, after the step of inputting the abnormal vibration data into the VGG-CNN model for damage identification and classification to obtain the damage category prediction result of the manhole cover, the intelligent collection and analysis monitoring method for urban manhole safety also includes the following steps.

[0019] The health status level of the manhole cover is determined according to the damage category prediction result.

[0020] Optionally, the health status level of the manhole cover includes a slight damage level, a medium damage level and a severe damage level in order from small to large damage degrees.

[0021] Optionally, the damage categories corresponding to the slight damage level include tiny cracks, slight deformation and slight loosening; the damage categories corresponding to the medium damage level include medium cracks, medium deformation and medium loosening; the damage categories corresponding to the severe damage level include severe cracks, severe deformation and severe loosening.

[0022] Optionally, after the step of determining the health status level of the manhole cover according to the damage category prediction result, the intelligent collection and analysis monitoring method for urban manhole safety further includes the following steps.

[0023] For the manhole covers with serious cracks, serious deformation and / or serious looseness, the location of the manhole cover is obtained, and the health status level and location of the manhole cover are reported, so that the inspection staff can repair the manhole cover in time.

[0024] Optionally, the sampling frequency of the acceleration sensor is 2500 Hz.

[0025] In the second aspect, the present application proposes an intelligent collection and analysis monitoring system for urban pipe well safety, including: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the intelligent collection and analysis monitoring method for urban pipe well safety described in the first aspect.

[0026] According to the specific embodiments provided in this application, this application discloses the following technical effects.

[0027] The present application provides an intelligent collection and analysis monitoring method and system for urban pipe well safety, which combines the MSCNN model and the VGG-CNN model and applies them to the manhole cover monitoring scenario of the urban pipe well. First, the MSCNN model is used to accurately identify abnormal vibration data based on vibration data, and the abnormal vibration data output by the MSCNN model is used as the input of the VGG-CNN model. Then, by inputting the abnormal vibration data into the VGG-CNN model, the VGG-CNN model is used to perform damage identification and classification and health status assessment on the manhole cover, and accurate and reliable damage category prediction results can be obtained, thereby achieving accurate damage identification and classification of the manhole cover, which is beneficial to the effective management of the urban pipe well cover and ensuring the safety of pedestrians and vehicles around the urban pipe well cover. BRIEF DESCRIPTION OF THE DRAWINGS

[0028] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0029] Figure 1 A flowchart of an intelligent collection and analysis monitoring method for urban pipe well safety provided in one embodiment of the present application.

[0030] Figure 2 A schematic diagram of a process for identifying abnormal vibration data provided in an embodiment of the present application.

[0031] Figure 3 A schematic diagram of a preprocessing process provided in an embodiment of the present application.

[0032] Figure 4 A schematic diagram of the structure of the MSCNN model provided in one embodiment of the present application.

[0033] Figure 5 A schematic diagram of the convolution process of the MSCNN model provided in one embodiment of the present application.

[0034] Figure 6 A schematic diagram of the parameter optimization process of the VGG-CNN model provided in one embodiment of the present application.

[0035] Figure 7 A schematic diagram of the structure of a VGG-CNN model provided in one embodiment of the present application. DETAILED DESCRIPTION

[0036] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0037] In order to make the above-mentioned objects, features and advantages of the present application more obvious and easy to understand, the present application is further described in detail below in conjunction with the accompanying drawings and specific implementation methods.

[0038] like Figure 1 As shown, this embodiment proposes an intelligent collection and analysis monitoring method for urban pipe well safety, which mainly combines the MSCNN (Multi-Scale Convolutional Neural Networks) model and the VGG-CNN (Visual Geometry Group-Convolutional Neural Network) model and applies it to the manhole cover monitoring scenario of urban pipe wells, aiming to achieve accurate damage identification and classification and health status assessment of manhole covers, thereby ensuring the safety of pedestrians and vehicles around urban pipe well covers. The method includes the following steps.

[0039] Step S1, obtaining vibration data of the manhole cover.

[0040] Step S2: input the vibration data into the MSCNN model to obtain abnormal vibration data.

[0041] Among them, the MSCNN model is a pre-trained model that takes the vibration data as input and the abnormal vibration data as output, and is used to identify the abnormal vibration data in the vibration data; the abnormal vibration data is data corresponding to the abnormal vibration state in the vibration data.

[0042] Step S3: input the abnormal vibration data into the VGG-CNN model for damage identification and classification to obtain a damage category prediction result of the manhole cover.

[0043] Among them, the VGG-CNN model is a pre-trained model that takes the abnormal vibration data as input and the damage category prediction result as output, and is used to predict the health status of the manhole cover based on the abnormal vibration data; the damage category prediction result is the probability value corresponding to each damage category of the manhole cover.

[0044] In this embodiment, step S1 obtains vibration data of the manhole cover, which specifically includes the following steps.

[0045] Step S11, installing an acceleration sensor at the center position of the back side of the manhole cover, wherein the acceleration sensor is used to measure the acceleration signal of the manhole cover in the vertical direction at a predetermined frequency as vibration data.

[0046] In this embodiment, an acceleration sensor of model ADXL1001 is used, which is installed at the center of the back of the manhole cover to measure the acceleration signal in the vertical direction of the manhole cover as vibration data at a predetermined frequency, so as to realize real-time collection of vibration data of the manhole cover. The sampling frequency of the acceleration sensor in this embodiment is 2500Hz.

[0047] Step S12: monitor the manhole cover using the acceleration sensor to collect the vibration data.

[0048] In this embodiment, after the step of obtaining the vibration data of the manhole cover in step S1, the intelligent collection and analysis monitoring method for urban manhole safety further includes the following steps.

[0049] The vibration data is preprocessed by data cleaning, data normalization and data noise reduction to obtain preprocessed vibration data, wherein the preprocessed vibration data is used as the vibration data to be input into the MSCNN model to obtain the abnormal vibration data.

[0050] In this embodiment, the vibration data is preprocessed by data cleaning, data normalization and data noise reduction to obtain the preprocessed vibration data, which specifically includes the following steps.

[0051] (1) Removing abnormal values ​​from the vibration data and filling in missing values ​​from the vibration data to obtain vibration data after data cleaning.

[0052] (2) Normalizing the vibration data after data cleaning to obtain normalized vibration data.

[0053] (3) Using a wavelet denoising algorithm, the vibration data after the data normalization is subjected to data denoising to obtain the preprocessed vibration data.

[0054] This embodiment uses an acceleration sensor to collect vibration data of the manhole cover in real time at a predetermined frequency, and then performs pre-processing such as data cleaning, data normalization, and data noise reduction on the vibration data. Figure 2 As shown in the figure, the pre-processed vibration data is input into the microcontroller of the manhole cover terminal, and the vibration data is analyzed by using the preset MSCNN model in the microcontroller to determine whether the vibration data is abnormal vibration data. If so, the abnormal vibration data is written into the SD card, and then fed back to the microcontroller, and the abnormal vibration data is transmitted to the server through the 4G module, otherwise it is directly discarded. The abnormal vibration data uploaded to the server is input into the preset VGG-CNN model to identify and classify the abnormal vibration data and evaluate the health status, so as to obtain the health status of the manhole cover and determine the health status level of the manhole cover. When the health status level of the manhole cover is judged to be severely damaged, that is, when the health status level of the manhole cover is at a dangerous level, the server reports the health status level and the position of the manhole cover to the inspection staff in real time, so that the inspection staff can repair the severely damaged manhole cover in time, and notify the manhole cover terminal to send an alarm to remind pedestrians and vehicles to pass carefully. In addition, when there is no one to inspect, the digital intelligent management and control of "perception-diagnosis-alarm" of urban manholes is realized.

[0055] After obtaining the vibration data, this embodiment first performs preprocessing such as data cleaning, data normalization, and data noise reduction on the vibration data. The acceleration sensor outputs an analog signal, and abnormal values ​​and missing values ​​are inevitable when ADC sampling is used. According to the basic parameters of the acceleration sensor, it can be determined that when the voltage exceeds the power supply voltage V DD When the outlier is found, the outlier should be removed and filled with the average of the five surrounding sample values. The same principle is used for missing values, which is to fill the missing value with the average of the five surrounding sample values. The expression of data normalization is as follows.

[0056] .

[0057] Among them, X norm represents the normalization result of vibration data, X represents vibration data, X max Indicates the maximum value of the vibration data, X min Indicates the minimum value of vibration data.

[0058] Since there is often construction and underground facility operation around urban pipe wells, low-frequency or high-frequency vibration noise may be generated. Therefore, it is necessary to perform noise reduction processing on the collected vibration signal sequence. This embodiment uses a wavelet noise reduction algorithm to perform noise reduction processing. The steps of the wavelet noise reduction algorithm include: (1) Determine the number of decomposition layers. The preprocessed vibration signal is decomposed into multiple resolution levels through wavelet transform to obtain a series of wavelet coefficients in different frequency bands. The low-frequency component usually represents the overall trend of the signal, while the high-frequency component contains detail information and noise. (2) Determine the threshold standard. A soft threshold noise reduction algorithm is used to set a filter threshold, and based on the filter threshold, the wavelet coefficients of different frequency bands are quantized using a threshold function, and the wavelet coefficients less than the filter threshold are removed, and the wavelet coefficients greater than or equal to the filter threshold are kept unchanged or reduced to achieve multi-scale wavelet coefficient denoising. (3) Reconstruct the signal by inverse wavelet transform. Through inverse wavelet transform, the above-processed wavelet coefficients are reconstructed into a signal to obtain a denoised vibration signal, thereby achieving noise reduction processing of the vibration signal. The preprocessing process is as follows: Figure 3 As shown. After investigation, it was found that abnormal vibration states such as tiny cracks, slight deformations and slight looseness of the manhole cover correspond to low-frequency signals of 0~1000Hz. In this embodiment, the sampling frequency is set to 2500Hz, and the Mallet algorithm is used to implement discrete wavelet transform. There is no need to further distinguish components with a frequency less than 20Hz, so the resolution level of the multi-scale decomposition of the wavelet transform in this embodiment is 5 layers, and a series of wavelet coefficients in different frequency bands are obtained after decomposition. The soft threshold denoising algorithm is then used to process the wavelet coefficients, and its threshold estimation method uses an unbiased risk estimation filter threshold.

[0059] In this embodiment, after the step of inputting the abnormal vibration data into the VGG-CNN model for damage identification and classification to obtain the damage category prediction result of the manhole cover in step S3, the intelligent collection and analysis monitoring method for urban manhole safety also includes the following steps.

[0060] Step S4: Determine the health status level of the manhole cover according to the damage category prediction result.

[0061] In this embodiment, the health status level of the manhole cover includes slight damage level, medium damage level and severe damage level in order from small to large damage degree. Among them, the damage categories corresponding to the slight damage level include tiny cracks, slight deformation and slight looseness. The damage categories corresponding to the medium damage level include medium cracks, medium deformation and medium looseness. The damage categories corresponding to the severe damage level include severe cracks, severe deformation and severe looseness.

[0062] In this embodiment, after the step of determining the health status level of the manhole cover according to the damage category prediction result in step S4, the intelligent collection and analysis monitoring method for urban manhole safety also includes the following steps.

[0063] Step S5: For the manhole cover with severe cracks, severe deformation and / or severe looseness, obtain the location of the manhole cover, and report the health status level and location of the manhole cover so that the inspection staff can repair the manhole cover in time.

[0064] Figure 4 The structure of the MSCNN model is shown. The MSCNN model includes an input layer, a convolutional layer, a Concat layer, a Maxpool layer, and an output layer. This structure can extract input layer data in parallel and extract feature information layer by layer. The network has only 8 layers, which reduces the complexity of the calculation and improves the real-time performance of the calculation. The input of the MSCNN model is vibration data, specifically preprocessed vibration data, which is in the form of a vibration signal sequence. Therefore, the input layer is a preprocessed vibration signal sequence with a size of 2500. The main feature extraction part has 4 convolutional layers, which are 1 1 convolutional layer, 1 2 convolutional layers, 1 3 convolutional layers and 1 5 convolutional layers, the activation function is the Relu function, and the feature extraction of the input data adopts a parallel mode, which can retain the features of the input data as much as possible. The convolution process of the MSCNN model is as follows Figure 5As shown in the figure, each convolution layer uses serial feature extraction, which can ensure the depth of feature information extracted by the convolution kernel of each convolution layer. The Concat layer concatenates the features of abnormal vibration signals (cracks, deformations, and looseness) extracted by each convolution layer. Bias is an additional input to neurons, which is used to adjust the activation level of neurons. In the Concat layer, the role of bias is to provide a fixed offset to help neurons better adapt to different input data, thereby improving the generalization ability of the MSCNN model. Since the features extracted from the input layer by each convolution layer have a large amount of repeated information, and the abnormal vibration signal to be identified has the characteristic of large energy, the maximum pooling layer can well identify this feature and avoid more computational waste. The MSCNN model is a binary classification model. The output layer uses the sigmoid function, and the loss function uses the binary cross entropy loss function. The output result is the probability value of the current vibration data being abnormal vibration data.

[0065] Before using the MSCNN model, the MSCNN model needs to be trained and its parameters optimized. First, 2400 sets of vibration signal sequences are collected on healthy manhole covers and abnormal manhole covers through the manhole cover monitor, including 1800 sets of abnormal vibration signal sequences and 600 sets of normal vibration signal sequences. It should be noted that the 1800 sets of abnormal vibration signal sequences collected are from manhole covers in 9 situations: small cracks, slight deformation, slight looseness, medium cracks, medium deformation, medium looseness, severe cracks, severe deformation and severe looseness, 200 sets each, to facilitate the subsequent use of the VGG-CNN model for more detailed diagnosis; then the vibration signal sequence is annotated to form a data set, that is, the normal vibration signal sequence is annotated as 0, abnormal vibration signal sequence is marked as 1, and then the data set is divided into 1680 training sets, 480 test sets, and 240 validation sets according to the ratio of 7:2:1; the hyperparameters of the MSCNN model are set, the learning rate η is 0.01, the batch is set to 32, the number of training epochs is 300, and the model accuracy is 92.99% after 300 rounds of training; the classification threshold is determined, and a series of abnormal vibration candidate probability thresholds t=0.1, 0.2, ..., 0.9 are set. 100 groups of validation set samples are input into the MSCNN model to obtain the predicted probability of each sample by the MSCNN model p , for each threshold t , the predicted probability is calculated according to the following formula p Convert to category labels , including two classification labels 0 and 1, where 1 is the positive class, indicating abnormal vibration data; 0 is the negative class, indicating normal vibration data.

[0066] .

[0067] Then, the evaluation index is calculated. In this embodiment, the F1 score (F1 Score) is used as the evaluation index, and the calculation formula is as follows.

[0068] .

[0069] Among them, F1 is the F1 score, Precision is the proportion of samples predicted as positive that are actually positive, and Recall is the proportion of all positive samples that are successfully predicted as positive. t =0.7 is the best classification threshold.

[0070] In this embodiment, before using the VGG-CNN model, the model parameters of the VGG-CNN model need to be trained. First, the original vibration data uses 1800 sets of abnormal vibration signal sequences, and the abnormal vibration signal sequences are re-labeled. For example, the vibration signal sequence of small cracks is labeled as 1, the vibration signal sequence of slight deformation is labeled as 2, the vibration signal sequence of slight looseness is labeled as 3, the vibration signal sequence of medium cracks is labeled as 4, the vibration signal sequence of medium deformation is labeled as 5, the vibration signal sequence of medium looseness is labeled as 6, the vibration signal sequence of severe cracks is labeled as 7, the vibration signal sequence of severe deformation is labeled as 8, and the vibration signal sequence of severe looseness is labeled as 9. Then, the VGG-CNN model is established, and the hyperparameters of the model training are set. The batch value is 32, the epoch value is 400, the learning rate η is set to 0.01, and the random inactivation rate dropout value is set to 0.1 to prevent overfitting; finally, the model training of the VGG-CNN model is performed, and the accuracy of the VGG-CNN model reaches 94% after 400 rounds of training. The parameter optimization process of the VGG-CNN model is as follows Figure 6 shown.

[0071] In this embodiment, when establishing the VGG-CNN model, first, the abnormal vibration data is preprocessed to convert the one-dimensional vector into a two-dimensional matrix; then, features are extracted through three convolutional layers, namely, convolutional layer C1, convolutional layer C2, and convolutional layer C3, and down-sampled through two maximum pooling layers, namely, maximum pooling layer P1 and maximum pooling layer P2; then, features are integrated using two fully connected layers, namely, fully connected layer F1 and fully connected layer F2, and finally classified through a Softmax layer. The structure of the VGG-CNN model is shown in Figure 1. Figure 7 The input data of the input layer is the abnormal vibration data received by the server from the manhole cover terminal, and its size is 2500. Before the abnormal vibration data is input to the convolution layer C1, the abnormal vibration data (1 2500) is transformed into a one-dimensional vector of (50 50) two-dimensional vector; convolution layer C1 uses 16 (5 5) convolution kernel to the input layer data to form (16 50 50) tensor; the output tensor of the convolution layer C1 is then passed through a tensor of size (2 2) The maximum pooling layer P1 obtains (16 25 25) tensor; the convolution layer C2 then calculates the output tensor of the maximum pooling layer P1, using 32 (3 3) Convolution kernel performs convolution, and the output tensor size is (32 25 25). The convolution layer C3 then calculates the output tensor using 64 (3 3) The convolution kernel performs convolution, and the output tensor size is (64 25 25). The maximum pooling layer P2 performs maximum pooling calculation on the output tensor of the convolution layer C3. The size of the maximum pooling layer P2 is (64 5 5), the output tensor size is (64 5 5). The fully connected layer F1 contains 256 neurons. The vector length of the flattened input tensor is 1600, so its weight matrix size is (256 1600), the output tensor size is (1 256); The principle of the fully connected layer F2 is the same as that of the fully connected layer F1, and the final output tensor is (1 9). Finally, the Softmax layer is used for classification to obtain the probability values ​​of the following nine situations: tiny cracks, slight deformation, slight looseness, medium cracks, medium deformation, medium looseness, severe cracks, severe deformation, and severe looseness.

[0072] In this embodiment, the health status level of the manhole cover includes slight damage level, medium damage level and severe damage level in order from small to large according to the degree of damage. Among them, the damage categories corresponding to the slight damage level include tiny cracks, slight deformation and slight looseness; the damage categories corresponding to the medium damage level include medium cracks, medium deformation and medium looseness; the damage categories corresponding to the severe damage level include severe cracks, severe deformation and severe looseness. In this embodiment, the damage category corresponding to the maximum probability value is taken as the health status of the manhole cover corresponding to the vibration signal sequence. For example, assuming that the probability value corresponding to medium cracks in a manhole cover predicted by the VGG-CNN model is the largest, reaching 97%, then it can be determined that the health status of the manhole cover is medium cracks, and the health status level of the manhole cover is medium damage level.

[0073] In this embodiment, when the VGG-CNN model infers that the manhole cover corresponding to the vibration signal sequence is at a serious damage level (including serious cracks, serious deformation and / or serious looseness), it is reported, and other situations are recorded, which can be written to the FLASH memory inside the STM32 chip, and reported regularly when the 4G module is idle.

[0074] At present, the traditional urban pipe well monitoring method has a huge amount of collected vibration signal data and needs to be transmitted to the cloud platform over long distances, which increases the burden of transmitting information and the storage pressure of the cloud platform. Among these data, most of them are useless noise data, which causes a huge waste of storage resources of the cloud platform. This embodiment is based on neural network technology, by constructing MSCNN model and VGG-CNN model, and training and learning MSCNN model and VGG-CNN model, it can better improve the accuracy of identifying abnormal vibration data, effectively filter out invalid vibration data, and send valid abnormal vibration data to the server side, reduce storage pressure, save storage space, and solve the current problems of high storage pressure and huge waste of storage resources.

[0075] At present, due to the limited computing power of the manhole cover terminal, it is necessary to analyze the collected abnormal vibration data on the server side to identify tiny cracks, slight deformations or loose displacements, and evaluate the health status of the manhole cover. This embodiment combines the MSCNN model and the VGG-CNN model, first identifies abnormal vibration data in the vibration data through the MSCNN model, and then uses the VGG-CNN model to identify and classify the abnormal vibration data and evaluate the health status, which can achieve accurate identification of abnormal vibration data and accurate evaluation of the health status of the manhole cover, solving the problem that the current manhole cover terminal controller has limited computing power and cannot identify and classify abnormal vibration data and evaluate the health status.

[0076] In an exemplary embodiment, an intelligent collection and analysis monitoring system for urban pipe well safety is also provided, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the intelligent collection and analysis monitoring method for urban pipe well safety.

[0077] The technical features of the above embodiments may be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0078] This specification uses specific examples to illustrate the principles and implementation methods of this application. The above examples are only used to help understand the method and core ideas of this application. At the same time, for those skilled in the art, according to the ideas of this application, there will be changes in the specific implementation methods and application scope. In summary, the content of this specification should not be understood as limiting this application.

Claims

1. An intelligent collection and analysis monitoring method for urban pipe well safety, characterized in that: The intelligent collection and analysis monitoring method for urban pipe well safety includes: Get the vibration data of the manhole cover; The vibration data is input into the MSCNN model to obtain abnormal vibration data; the MSCNN model is a pre-trained model that takes the vibration data as input and the abnormal vibration data as output and is used to identify the abnormal vibration data in the vibration data; the abnormal vibration data is data corresponding to the abnormal vibration state in the vibration data; The abnormal vibration data is input into the VGG-CNN model for damage identification and classification to obtain a damage category prediction result of the manhole cover; the VGG-CNN model is a pre-trained model that takes the abnormal vibration data as input and the damage category prediction result as output, and is used to predict the health status of the manhole cover based on the abnormal vibration data; the damage category prediction result is the probability value corresponding to each damage category of the manhole cover.

2. The intelligent collection and analysis monitoring method for urban pipe well safety according to claim 1 is characterized in that: Obtain vibration data of manhole covers, including: An acceleration sensor is installed at the center of the back of the manhole cover, and the acceleration sensor is used to measure the vibration data of the manhole cover in the vertical direction at a predetermined frequency; The manhole cover is monitored by using the acceleration sensor to collect the vibration data.

3. The intelligent collection and analysis monitoring method for urban pipe well safety according to claim 1 is characterized in that: After the step of obtaining the vibration data of the manhole cover, the intelligent collection and analysis monitoring method for urban manhole safety further includes: The vibration data is preprocessed by data cleaning, data normalization and data noise reduction to obtain preprocessed vibration data; the preprocessed vibration data is used as the vibration data to be input into the MSCNN model to obtain the abnormal vibration data.

4. The intelligent collection and analysis monitoring method for urban pipe well safety according to claim 3 is characterized in that: The vibration data is preprocessed by data cleaning, data normalization and data noise reduction to obtain preprocessed vibration data, specifically including: Removing abnormal values ​​from the vibration data and filling in missing values ​​from the vibration data to obtain vibration data after data cleaning; Normalizing the vibration data after data cleaning to obtain normalized vibration data; A wavelet denoising algorithm is used to perform data denoising on the vibration data after the data normalization to obtain the preprocessed vibration data.

5. The intelligent collection and analysis monitoring method for urban pipe well safety according to claim 1 is characterized in that: After the step of inputting the abnormal vibration data into the VGG-CNN model for damage identification and classification to obtain the damage category prediction result of the manhole cover, the intelligent collection and analysis monitoring method for urban manhole safety also includes: The health status level of the manhole cover is determined according to the damage category prediction result.

6. The intelligent collection and analysis monitoring method for urban pipe well safety according to claim 5 is characterized in that: The health status levels of the manhole cover include a slight damage level, a medium damage level and a severe damage level in order from small to large damage degrees.

7. The intelligent data collection and analysis monitoring method for urban pipe well safety according to claim 6 is characterized in that: The damage categories corresponding to the minor damage level include tiny cracks, minor deformation and minor loosening; the damage categories corresponding to the medium damage level include medium cracks, medium deformation and medium loosening; the damage categories corresponding to the severe damage level include severe cracks, severe deformation and severe loosening.

8. The intelligent data collection and analysis monitoring method for urban pipe well safety according to claim 7 is characterized in that: After the step of determining the health status level of the manhole cover according to the damage category prediction result, the intelligent collection and analysis monitoring method for urban manhole safety also includes: For the manhole covers with serious cracks, serious deformation and / or serious looseness, the location of the manhole cover is obtained, and the health status level and location of the manhole cover are reported, so that the inspection staff can repair the manhole cover in time.

9. The intelligent collection and analysis monitoring method for urban pipe well safety according to claim 2 is characterized in that: The sampling frequency of the acceleration sensor is 2500 Hz.

10. An intelligent data collection and analysis monitoring system for urban pipe well safety, comprising: A memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the intelligent collection, analysis, and monitoring method for urban pipe well safety as described in any one of claims 1 to 9.

Images