Method and system for monitoring abrasion of hob cutter of shield tunneling machine in real time

Abstract

The invention provides a real-time monitoring method and system for abrasion of a hobbing cutter of a shield tunneling machine, and relates to the technical field of underground engineering construction equipment.The method comprises the steps that working pressure data, working temperature data and surface image data of a target hobbing cutter of the shield tunneling machine are collected based on a pressure sensor, a temperature sensor and a hobbing cutter camera; transmitting the data to a central processing system based on an advanced encryption standard algorithm and carrying out preprocessing operation; analyzing the data based on a long and short term memory network model, a support vector machine model and a convolutional neural network model, and generating a pressure data score, a temperature data score and an image data score; calculating a wear state score based on a fusion evaluation algorithm; the remaining service life of the hob is determined based on the Monte Carlo simulation prediction algorithm, and when the service life is lower than the preset service life lower limit, an abrasion early warning mechanism is triggered, so that the abrasion state of the hob can be accurately monitored, the maintenance cost is reduced, and the safety and economic benefits of shield construction are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of underground engineering construction equipment, and particularly to a real-time monitoring method and system for the wear of the cutter head cutters of a shield machine. Background Art

[0002] During tunnel boring construction, the wear state of the cutter head cutters of a shield machine directly affects the boring efficiency, cost control, and construction safety.

[0003] Traditional manual regular inspection methods have many drawbacks. The manual inspection method not only consumes a large amount of time and manpower, but also, due to the subjective judgment differences of inspectors, it is difficult to ensure the accuracy of the inspection results. In addition, the manual inspection method cannot achieve continuous real-time monitoring, making it difficult to promptly detect potential cutter wear problems, thereby resulting in a lag in fault response. At the same time, when the cutters are severely worn but not replaced in a timely manner, a series of problems will occur, such as a reduction in the boring efficiency of the shield machine, an increase in cutter costs, and even safety accidents, delays in the project schedule, and huge economic losses.

[0004] Therefore, it is necessary to provide a real-time monitoring method and system for the wear of the cutter head cutters of a shield machine to solve the above technical problems. Summary of the Invention

[0005] To solve the above technical problems, the present invention provides a real-time monitoring method and system for the wear of the cutter head cutters of a shield machine, which is used to solve the problems that the existing manual inspection method cannot perform real-time and accurate monitoring on the wear state of the cutters, thereby resulting in relatively high maintenance costs and poor safety and economic benefits in shield construction.

[0006] A real-time monitoring method for the wear of the cutter head cutters of a shield machine provided by the present invention includes: Based on a pressure sensor, a temperature sensor, and a cutter head camera, respectively monitor and collect the working pressure data, working temperature data, and surface image data of the cutter head cutters of the target shield machine; Based on the Advanced Encryption Standard algorithm, transmit the working pressure data, the working temperature data, and the surface image data to a central processing system and perform preprocessing operations; Based on the long short-term memory network model, support vector machine model, and convolutional neural network model built in the central processing system, respectively analyze the working pressure data, the working temperature data, and the surface image data to generate corresponding pressure data scores, temperature data scores, and image data scores; Based on a fusion evaluation algorithm, calculate a wear state score according to the pressure data score, the temperature data score, and the image data score; Based on the Monte Carlo simulation prediction algorithm, determine the remaining service life of the hob according to the wear state score, and trigger the wear warning mechanism when the remaining service life of the hob is lower than the preset lower limit of the service life.

[0007] Preferably, at the edge of the contact surface of the hob of the target shield machine, install one of the pressure sensors at intervals of a preset distance; install the temperature sensor in the reserved hole at the axis center of the hob of the target shield machine; install the hob camera on the side of the hob of the target shield machine.

[0008] Preferably, the preprocessing operation specifically includes: Perform denoising processing on the working pressure data based on the moving average filtering algorithm; Fill in the missing values in the working temperature data by using the linear interpolation method; Enhance the contrast of the surface image data by using the histogram equalization algorithm.

[0009] Preferably, analyze the working pressure data based on the convolutional neural network model to generate the pressure data score, which specifically includes: Input the working pressure data into the forget gate of the convolutional neural network model, and determine the data that needs to be retained in the working pressure data based on the forget gate. The corresponding calculation formula is as follows: In the formula, represents the output of the forget gate at the current time t; represents the weight of the forget gate at the current time t; represents the hidden state at the previous time t - 1; represents the working pressure data input at the current time t; represents the hidden state and the working pressure data to form a vector; represents the bias term of the forget gate; represents the activation function; Based on the input gate, determine the data that needs to be added to the candidate cell state in the working pressure data. The corresponding calculation formula is as follows: In the formula, represents the output of the input gate at the current time t; represents the candidate cell state at the current time t; and respectively represent the weights of the input gate and the candidate cell state; represents the hidden state at the previous time t - 1; represents the working pressure data input at the current time t; Indicates the hidden state and the working pressure data to form a vector; and respectively represent the bias terms of the input gate and the candidate cell state; represents the activation function; represents the hyperbolic tangent function; Update the cell state according to the output of the forget gate, the output of the input gate and the candidate cell state. The corresponding calculation formula is as follows: In the formula, represents the cell state at the current time t; represents the output of the forget gate at the current time t; represents the cell state at the previous time t - 1; represents the output of the input gate at the current time t; represents the candidate cell state at the current time t; According to the cell state, determine the data to be output in the hidden state through the output gate. The corresponding calculation formula is as follows: In the formula, represents the output of the output gate at the current time t; represents the activation function; represents the weight of the output gate; represents the hidden state at the previous time t - 1; represents the working pressure data input at the current time t; represents the hidden state and the working pressure data to form a vector; represents the bias term of the output gate; represents the hidden state at the current time t; represents the hyperbolic tangent function; represents the cell state at the current time t; Taking the hidden state at the current time as the predicted pressure data, the pressure data score is the ratio of the predicted pressure data to the working pressure data.

[0010] Preferably, analyze the working temperature data based on the support vector machine model to generate the temperature data score, specifically including: Extract the working temperature features corresponding to the working temperature data, and perform feature classification on the working temperature features based on the support vector machine model. The corresponding optimization problem is as follows: In the formula, It represents the weight of the $u$-th support vector machine and is used to determine the direction of the classification hyperplane; It represents the bias term of the $u$-th support vector machine and is used to determine the distance between the classification hyperplane and the origin; It represents the slack variable when the $v$-th working temperature feature is linearly inseparable in the $u$-th support vector machine; $G$ represents the regularization parameter, which is used to control the trade-off between the classification margin and classification error; $R$ represents the number of working temperature features; $\min$ represents the minimum operation; The constraint conditions corresponding to the optimization problem are as follows: In the formula, It represents the feature class label corresponding to the $v$-th working temperature feature; It represents the $v$-th working temperature feature; It represents the weight of the $u$-th support vector machine and is used to determine the direction of the classification hyperplane; It represents the bias term of the $u$-th support vector machine and is used to determine the distance between the classification hyperplane and the origin; It represents the slack variable when the $v$-th working temperature feature is linearly inseparable in the $u$-th support vector machine; $R$ represents the number of working temperature features; Based on the feature class label, the abnormal temperature feature and the corresponding abnormal temperature data are discriminated; The temperature data score is the ratio of the abnormal temperature data to the number of working temperature data.

[0011] Preferably, based on the convolutional neural network model, the surface image data is analyzed to generate the image data score, which specifically includes: Set the height of the surface image data $X$ to be $H$, the width to be $W$, and the number of channels to be $E$; the height of the convolutional kernel of the convolutional neural network model is $h$, the width is $w$, and the number is $M$; the width and height of the pooling window of the convolutional neural network model are both $k$, and the stride is $s$; Based on the convolutional kernel, feature extraction is performed on the surface image data to obtain the following convolutional image features: In the formula, It represents the value of the convolutional image feature output by the $m$-th convolutional kernel at the position ; It represents the value of the $m$-th convolutional kernel at the position ; It represents the value of the convolutional image feature at the position ; It represents the bias term corresponding to the $m$-th convolutional kernel; Based on the pooling window, max pooling processing is performed on the surface image data to obtain the following pooling image features: In the formula, represents the value of the pooled image feature output by the m-th pooling window at the position ; max represents the maximum value operation; represents the value of the convolutional image feature output by the m-th convolutional kernel at the position ; Input the pooled image feature into the fully connected layer of the convolutional neural network model, and output the wear image feature as follows: In the formula, represents the output of the l-th fully connected layer, that is, the wear image feature; l represents the number of layers of the fully connected layer; represents the activation function; represents the weight of the l-th fully connected layer; represents the output of the (l - 1)-th fully connected layer; represents the bias term of the l-th fully connected layer; The image data score is the ratio of the number of the wear image feature to the surface image feature.

[0012] Preferably, based on the fusion evaluation algorithm, calculate the wear state score according to the pressure data score, the temperature data score, and the image data score. The corresponding calculation formula is as follows: In the formula, represents the wear state score; respectively represent the pressure data score, the temperature data score, and the image data score; respectively represent the score weights corresponding to the pressure data score, the temperature data score, and the image data score.

[0013] Preferably, a real-time monitoring system for the wear of the cutter head of a shield machine, the monitoring system includes: A monitoring and acquisition module, which is used to monitor and acquire the working pressure data, the working temperature data, and the surface image data of the cutter head of the target shield machine based on a pressure sensor, a temperature sensor, and a cutter head camera respectively; An encryption and transmission module, which is used to transmit the working pressure data, the working temperature data, and the surface image data to the central processing system and perform preprocessing operations based on the advanced encryption standard algorithm; A score generation module, which is used to analyze the working pressure data, the working temperature data, and the surface image data respectively based on the long short-term memory network model, the support vector machine model, and the convolutional neural network model built in the central processing system, and generate the corresponding pressure data score, temperature data score, and image data score; A fusion evaluation module, configured to calculate a wear state score based on a fusion evaluation algorithm according to the pressure data score, the temperature data score, and the image data score; An early warning trigger module, configured to determine the remaining service life of the hob based on the Monte Carlo simulation prediction algorithm according to the wear state score, and trigger a wear early warning mechanism when the remaining service life of the hob is lower than the preset lower limit of the service life.

[0014] An electronic device, comprising a memory and a processor, wherein a computer program is stored in the memory, and when the processor runs the computer program stored in the memory, the processor executes the steps of the real-time monitoring method for the wear of the shield machine hob as described in any one of the above.

[0015] A readable storage medium, wherein a computer program is stored in the readable storage medium, and when the computer program is executed by a processor, it is used to implement the steps of the real-time monitoring method for the wear of the shield machine hob as described in any one of the above.

[0016] Compared with the related art, the real-time monitoring method and system for the wear of the shield machine hob provided by the present invention have the following beneficial effects: The present invention can respectively monitor and collect the working pressure data, the working temperature data, and the surface image data of the target shield machine hob based on a pressure sensor, a temperature sensor, and a hob camera; based on the Advanced Encryption Standard algorithm, transmit the working pressure data, the working temperature data, and the surface image data to a central processing system and perform preprocessing operations; based on the long short-term memory network model, the support vector machine model, and the convolutional neural network model built in the central processing system, respectively analyze the working pressure data, the working temperature data, and the surface image data to generate corresponding pressure data scores, temperature data scores, and image data scores; based on a fusion evaluation algorithm, calculate a wear state score according to the pressure data score, the temperature data score, and the image data score; based on the Monte Carlo simulation prediction algorithm, determine the remaining service life of the hob according to the wear state score, and trigger a wear early warning mechanism when the remaining service life of the hob is lower than the preset lower limit of the service life, so as to realize real-time and accurate monitoring of the wear state of the shield machine hob, reduce the maintenance cost, ensure the smooth progress of the project, and improve the safety and economic benefits of shield construction.

[0017] Through multi-dimensional sensor layout and multi-model collaborative analysis, the present invention can monitor the wear state of the cutter head of a shield machine with high precision and in real time, thereby enabling timely detection of minor wear changes, improving the accuracy and timeliness of wear monitoring, and avoiding equipment failure problems caused by excessive wear. Through accurate wear prediction and intelligent decision support, the present invention can make the maintenance plan more scientific and reasonable, reduce unnecessary cutter replacement and maintenance work, lower the maintenance cost, and extend the service life of the cutter. Through the real-time monitoring and warning mechanism, the present invention can effectively prevent safety accidents caused by severe wear of the cutter head, ensure the safety of construction personnel and the normal operation of construction equipment, and ensure the safety and reliability of the construction process. The present invention shortens the downtime caused by cutter failures, can quickly replace the cutter according to the maintenance suggestions and resume construction, improves the tunneling efficiency of the shield machine, speeds up the project progress, and enhances the overall economic benefits of the project. Description of the Drawings

[0018] Figure 1 It is a flowchart of a method for real-time monitoring of cutter wear of a shield machine cutter head provided by an embodiment of the present invention; Figure 2 It is a flowchart of a preprocessing operation provided by an embodiment of the present invention; Figure 3 It is a system block diagram of a system for real-time monitoring of cutter wear of a shield machine cutter head provided by an embodiment of the present invention; Figure 4 It is a schematic hardware structure diagram of an electronic device provided by an embodiment of the present invention. Detailed Embodiments

[0019] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention. Apparently, the described embodiments are only a part of the embodiments of the present invention, rather than all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

[0020] As Figure 1 shown, it is a flowchart of a method for real-time monitoring of cutter wear of a shield machine cutter head provided by an embodiment of the present invention, Figure 1The execution entity of the method shown can be a software and / or hardware device. The execution entity of this application can include but is not limited to at least one of the following: user equipment, network equipment, etc. Among them, the user equipment can include but is not limited to computers, smart phones, personal digital assistants (Personal Digital Assistant, abbreviated as: PDA), and the electronic devices mentioned above, etc. The network equipment can include but is not limited to a single network server, a server group composed of multiple network servers, or a cloud composed of a large number of computers or network servers based on cloud computing. Among them, cloud computing is a type of distributed computing, which consists of a super virtual computer composed of a group of loosely coupled computers. This embodiment does not limit this. It includes steps S1 to S5, specifically as follows: S1. Based on a pressure sensor, a temperature sensor, and a hob camera, respectively monitor and collect the working pressure data, working temperature data, and surface image data of the target shield machine hob; In tunnel engineering construction, a shield machine is a key construction equipment, and the wear state of its hob tool has a crucial impact on the project progress, cost, and safety.

[0021] In order to effectively monitor the wear condition of the shield machine hob, key data can first be obtained by relying on a pressure sensor, a temperature sensor, and a hob camera. Among them, the pressure sensor can accurately monitor the pressure borne by the hob during work and collect the working pressure data. These data reflect the force condition of the hob during the process of cutting rock and soil, and are an important basis for evaluating the wear degree of the hob. The temperature sensor is used to measure the temperature of the hob during work and obtain the working temperature data. Because the hob will generate heat during high-speed operation and cutting, the change in temperature can indirectly reflect the wear state of the hob and whether there are abnormal working conditions. The hob camera is used to collect the surface image data of the hob, and through these images, the wear condition of the hob surface can be directly observed, such as whether there are defects such as cracks and spalling.

[0022] S2. Based on the Advanced Encryption Standard algorithm, transmit the working pressure data, the working temperature data, and the surface image data to the central processing system and perform preprocessing operations; It should be noted that the Advanced Encryption Standard algorithm (AES) is a symmetric encryption algorithm, which has high-strength encryption performance. During the data transmission process, the Advanced Encryption Standard algorithm can encrypt the working pressure data, the working temperature data, and the surface image data to ensure the security of the transmitted data and prevent the data from being stolen or tampered with during the transmission process.

[0023] Then, the data in the central processing system can be preprocessed, including removing noise, data calibration, etc., to improve the quality and usability of the data.

[0024] S3. Based on the long short - term memory network model, support vector machine model, and convolutional neural network model built in the central processing system, analyze the working pressure data, working temperature data, and surface image data respectively to generate corresponding pressure data scores, temperature data scores, and image data scores. Among them, the central processing system has built - in a long short - term memory network model (LSTM), a support vector machine model (SVM), and a convolutional neural network model (CNN). The long short - term memory network model is a special recurrent neural network, which is good at processing time - series data and can effectively capture the long - term and short - term dependencies in the working pressure data. By analyzing the pressure change trend, it can generate corresponding pressure data scores to evaluate the hob wear condition reflected by the pressure data. The support vector machine model is a classification and regression model. When processing temperature data, it can judge whether the temperature state of the hob is normal according to the temperature change characteristics and historical data and generate temperature data scores. The convolutional neural network model is a neural network specifically used to process image data. It extracts and analyzes features from the surface image data through structures such as convolutional layers, pooling layers, and fully - connected layers, identifies the type and degree of defects on the hob surface, and then generates image data scores.

[0025] S4. Based on the fusion evaluation algorithm, calculate the wear state score according to the pressure data score, temperature data score, and image data score. In practical applications, in order to comprehensively evaluate the wear state of the hob, a fusion evaluation algorithm can be adopted to assign corresponding weights to the pressure data score, temperature data score, and image data score, and calculate the wear state score through weighted calculation, so as to fully consider the contribution of different data to the hob wear evaluation and improve the accuracy of wear evaluation.

[0026] S5. Based on the Monte Carlo simulation prediction algorithm, determine the remaining service life of the hob according to the wear state score. When the remaining service life of the hob is lower than the preset lower limit of the service life, trigger the wear warning mechanism.

[0027] Among them, the process of determining the remaining service life of the hob is as follows: Set the number of Monte Carlo simulations as N, and the critical wear state score of the target shield machine hob is , the wear state score at the th time step is , then the time to reach the wear limit in the nth simulation is , and the corresponding remaining service life of the hob is .

[0028] It should be noted that the Monte Carlo simulation is a method for solving problems through random sampling and statistical analysis. First, based on the Monte Carlo simulation prediction algorithm, a large number of random simulations can be used to predict the future wear of the hob, so as to determine its remaining service life. When the remaining service life of the hob is lower than the lower limit of the preset life, the wear warning mechanism is immediately triggered to remind the staff to check, repair or replace the hob in time, so as to avoid engineering accidents and economic losses caused by excessive wear of the hob.

[0029] In the specific implementation process, at the edge of the contact surface of the hob of the target shield machine, a pressure sensor is installed at each preset distance; the temperature sensor is installed in the reserved hole at the axis center of the hob of the target shield machine; the hob camera is installed on the side surface of the hob of the target shield machine.

[0030] During the process of the shield machine excavating the tunnel, in order to accurately monitor the working condition of the hob, a pressure sensor needs to be installed at the edge of the contact surface of the hob of the target shield machine at each preset distance. The preset distance refers to the distance set according to the structure, mechanical properties of the hob and past engineering experience, such as 5 - 10 cm. Through such a layout, the pressure sensor can comprehensively collect the pressure change data during the operation of the hob, and convert the pressure data received by the hob into an electrical signal, providing a basis for the force condition of the hob.

[0031] The temperature sensor is installed in the reserved hole at the axis center of the hob of the target shield machine. Since heat is generated during the high-speed rotation and cutting process of the hob, the hob can be installed at the axis center to facilitate directly measuring the core temperature of the hob. By monitoring the temperature change, it can be judged whether there is abnormal wear or overload of the hob.

[0032] Finally, a hob camera can be installed on the side surface of the hob of the target shield machine. The hob camera is a special camera device with waterproof and dustproof functions. It can take real-time pictures of the surface condition of the hob, facilitating visually observing whether there are wear signs such as cracks and spalling on the surface of the hob, and providing image support for subsequent maintenance and replacement decisions.

[0033] As Figure 2 shown, the preprocessing operation specifically includes: Denosing the working pressure data based on the moving average filtering algorithm; Filling the missing values in the working temperature data by using the linear interpolation method; Enhancing the contrast of the surface image data by using the histogram equalization algorithm.

[0034] It should be noted that, first, the working pressure data can be denoised based on the moving average filtering algorithm. The moving average filtering algorithm is a commonly used signal processing method. By averaging multiple consecutive pressure data points, it can effectively remove the noise interference in the data. For example, five consecutive pressure data points can be selected and their average value is calculated as the filtered result, which can make the pressure data curve smoother and more accurately reflect the pressure change of the hob during actual work.

[0035] Furthermore, during the data acquisition process, the temperature data may be missing due to various reasons. Through the linear interpolation method, the missing value can be estimated according to the known temperature data before and after the missing value according to the linear relationship to ensure the integrity of the temperature data.

[0036] In addition, through the histogram equalization algorithm, the pixel gray values of the surface image data can be redistributed to make the gray distribution of the image more uniform, thereby enhancing the image contrast and clearly presenting the wear characteristics on the hob surface, such as cracks, notches, etc.

[0037] Analyzing the working pressure data based on the convolutional neural network model to generate the pressure data score specifically includes: Inputting the working pressure data into the forget gate of the convolutional neural network model, and determining the data that needs to be retained in the working pressure data based on the forget gate. The corresponding calculation formula is as follows: In the formula, represents the output of the forget gate at the current time t; represents the weight of the forget gate at the current time t; represents the hidden state at the previous time t-1; represents the working pressure data input at the current time t; represents the hidden state and the working pressure data to form a vector; represents the bias term of the forget gate; represents the activation function; Determining the data that needs to be added to the candidate cell state in the working pressure data based on the input gate. The corresponding calculation formula is as follows: In the formula, represents the output of the input gate at the current time t; represents the candidate cell state at the current time t; and respectively represent the weights of the input gate and the candidate cell state; represents the hidden state at the previous time t-1; represents the working pressure data input at the current moment t; represents the hidden state and the working pressure data to form a vector; and respectively represent the bias terms of the input gate and the candidate cell state; represents the activation function; represents the hyperbolic tangent function; Update the cell state according to the output of the forget gate, the output of the input gate, and the candidate cell state. The corresponding calculation formula is as follows: In the formula, represents the cell state at the current moment t; represents the output of the forget gate at the current moment t; represents the cell state at the previous moment t - 1; represents the output of the input gate at the current moment t; represents the candidate cell state at the current moment t; Determine the data to be output in the hidden state through the output gate according to the cell state. The corresponding calculation formula is as follows: In the formula, represents the output of the output gate at the current moment t; represents the activation function; represents the weight of the output gate; represents the hidden state at the previous moment t - 1; represents the working pressure data input at the current moment t; represents the hidden state and the working pressure data to form a vector; represents the bias term of the output gate; represents the hidden state at the current moment t; represents the hyperbolic tangent function; represents the cell state at the current moment t; Taking the hidden state at the current moment as the predicted pressure data, the pressure data score is the ratio of the predicted pressure data to the working pressure data.

[0038] Among them, components such as forget gates and input gates can be used to carefully screen and process the working pressure data, accurately extract key information in the data, such as pressure change trends, abnormal fluctuations, etc., providing support for accurately evaluating the wear state of the hob later. For example, under complex geological conditions, the pressure change of the hob is relatively complex, and this method can effectively distinguish normal pressure fluctuations and abnormal pressure changes caused by increased wear.

[0039] Then, the cell state and hidden state can be dynamically updated according to the pressure data at each moment to adapt to the pressure changes of the hob in different working stages. In the initial and later stages of the shield machine tunneling, the wear degree of the hob is different, and the characteristics of the pressure data are also different. This dynamic update mechanism can reflect these changes in real time, making the pressure data scoring more in line with the actual situation.

[0040] Finally, the ratio of the predicted pressure data to the working pressure data can be used as the pressure data score to accurately measure the difference between the pressure change and the expectation. This score can visually present the working pressure condition of the hob, facilitating the staff to quickly judge whether there is a risk of abnormal wear of the hob, so as to take timely measures, such as adjusting tunneling parameters or replacing the hob, to ensure the stable operation of the shield machine and reduce the risk of shield tunneling construction.

[0041] Analyzing the working temperature data based on the support vector machine model to generate the temperature data score specifically includes: Extracting the working temperature characteristics corresponding to the working temperature data, and performing feature classification on the working temperature characteristics based on the support vector machine model. The corresponding optimization problem is as follows: In the formula, represents the weight of the u-th support vector machine, which is used to determine the direction of the classification hyperplane; represents the bias term of the u-th support vector machine, which is used to determine the distance between the classification hyperplane and the origin; represents the slack variable when the v-th working temperature characteristic is linearly inseparable in the u-th support vector machine; G represents the regularization parameter, which is used to control the trade-off between the classification margin and classification error; R represents the number of working temperature characteristics; min represents the operation of taking the minimum value; The constraint conditions corresponding to the optimization problem are as follows: In the formula, represents the feature category label corresponding to the v-th working temperature characteristic; represents the v-th working temperature characteristic; represents the weight of the u-th support vector machine, which is used to determine the direction of the classification hyperplane; represents the bias term of the u-th support vector machine, which is used to determine the distance between the classification hyperplane and the origin; It represents the relaxation variable when the v-th working temperature feature is linearly inseparable in the u-th support vector machine; R represents the number of working temperature features; Based on the feature category labels, the abnormal temperature features and the corresponding abnormal temperature data are discriminated; The temperature data score is the ratio of the number of the abnormal temperature data to the number of the working temperature data.

[0042] In practical applications, through the support vector machine model, the features of the working temperature data can be effectively extracted, and the classification hyperplane can be determined by solving the optimization problem. This enables the model to accurately classify the working temperature features and distinguish between normal and abnormal temperature features. The use of the regularization parameter can balance between the classification margin and classification errors, improving the classification accuracy and generalization ability.

[0043] By discriminating the feature category labels, the abnormal temperature features and their corresponding abnormal temperature data can be quickly determined. During the operation of the shield machine, abnormal temperatures may indicate the wear, failure or other potential problems of the hob cutters, and timely discovery of these abnormal temperature data helps to take measures in advance to avoid equipment damage or engineering accidents.

[0044] The temperature data score is the ratio of the number of the abnormal temperature data to the number of the working temperature data. This score can help the staff quickly understand the working temperature condition of the hob cutters and judge whether further inspections and maintenance are needed. At the same time, this score can also be accumulated and analyzed as historical data to provide a basis for subsequent equipment optimization and fault prediction.

[0045] Through the above methods, the accuracy and efficiency of the hob temperature monitoring of the shield machine can be improved, providing a strong guarantee for the safe and stable operation of the equipment.

[0046] Based on the convolutional neural network model, the surface image data is analyzed to generate the image data score, which specifically includes: Set the height of the surface image data X as H, the width as W, and the number of channels as E; the height of the convolution kernel of the convolutional neural network model is h, the width is w, and the number is M; the width and height of the pooling window of the convolutional neural network model are both k, and the stride is s; Based on the convolution kernel, feature extraction is performed on the surface image data to obtain the following convolution image features: In the formula, represents the value of the convolution image feature output by the m-th convolution kernel at the position ; represents the value of the m-th convolution kernel at the position ; represents the value of the convolutional image feature at the position ; represents the bias term corresponding to the m-th convolutional kernel; Performing max pooling on the surface image data based on the pooling window, the pooled image feature is obtained as follows: In the formula, represents the value of the pooled image feature output by the m-th pooling window at the position ; max represents the operation of taking the maximum value; represents the value of the convolutional image feature output by the m-th convolutional kernel at the position ; Inputting the pooled image feature into the fully connected layer of the convolutional neural network model, the worn image feature is output as follows: In the formula, represents the output of the l-th fully connected layer, that is, the worn image feature; l represents the number of layers of the fully connected layer; represents the activation function; represents the weight of the l-th fully connected layer; represents the output of the (l - 1)-th fully connected layer; represents the bias term of the l-th fully connected layer; The image data score is the ratio of the number of the worn image feature to the surface image feature.

[0047] It should be noted that through the convolutional kernels of the convolutional neural network model, multi-dimensional feature extraction can be performed on the surface image data. Different convolutional kernels can capture subtle features such as the edges and textures of the hob surface, such as wear marks like cracks and spalling on the hob surface, providing rich information for accurately evaluating the wear state of the hob.

[0048] Furthermore, through the pooling window, the dimensionality of the convolutional image feature can be reduced. While reducing the data volume, the key feature information is retained, improving the model processing efficiency, enabling the model to efficiently process data in subsequent analysis, and not losing important wear features, ensuring the accuracy of the hob wear assessment.

[0049] Finally, the pooled image feature can be input into the fully connected layer to obtain the worn image feature. The ratio of the number of the worn image feature to the surface image feature is the image data score. This score intuitively quantifies the wear condition of the hob surface, and the higher the score, the more severe the hob wear. The staff can quickly judge the wear degree of the hob based on this score, and then make decisions on repairing or replacing the hob in a timely manner to ensure the stable operation of the shield machine.

[0050] Based on the fusion evaluation algorithm, the wear state score is calculated according to the pressure data score, the temperature data score, and the image data score. The corresponding calculation formula is as follows: In the formula, represents the wear state score; respectively represent the pressure data score, the temperature data score, and the image data score; respectively represent the scoring weights corresponding to the pressure data score, the temperature data score, and the image data score.

[0051] In practical applications, the pressure data score reflects the wear trend under force, the temperature data score reflects the abnormal temperature situation, and the image data score intuitively presents the wear condition of the hob surface. The integration of the three avoids the one-sidedness of single data evaluation and can more comprehensively and accurately evaluate the wear state of the hob.

[0052] In addition, the staff can more accurately judge the actual wear degree of the hob based on this comprehensive score, make timely maintenance or replacement decisions, ensure the efficient and stable operation of the shield machine, and reduce the risk of equipment failure.

[0053] As Figure 3 shown, it is a system block diagram of a real-time monitoring system for the wear of a shield machine hob provided by an embodiment of the present invention. The monitoring system includes: A monitoring and acquisition module, which is used to monitor and acquire the working pressure data, working temperature data, and surface image data of the target shield machine hob based on a pressure sensor, a temperature sensor, and a hob camera respectively; An encryption and transmission module, which is used to transmit the working pressure data, the working temperature data, and the surface image data to a central processing system based on the Advanced Encryption Standard algorithm and perform preprocessing operations; A score generation module, which is used to analyze the working pressure data, the working temperature data, and the surface image data respectively based on the long short-term memory network model, support vector machine model, and convolutional neural network model built in the central processing system, and generate corresponding pressure data scores, temperature data scores, and image data scores; A fusion evaluation module, which is used to calculate the wear state score based on the fusion evaluation algorithm according to the pressure data score, the temperature data score, and the image data score; An early warning trigger module, which is used to determine the remaining service life of the hob based on the Monte Carlo simulation prediction algorithm according to the wear state score, and trigger a wear early warning mechanism when the remaining service life of the hob is lower than the preset lower limit of the service life.

[0054] Figure 3 The device in the shown embodiment can correspondingly be used to executeFigure 1 The steps in the method embodiments shown have similar implementation principles and technical effects, which will not be elaborated here.

[0055] An electronic device includes a memory and a processor. A computer program is stored in the memory. When the processor runs the computer program stored in the memory, the processor executes the steps of a real-time monitoring method for cutter wear of a shield machine as described in any one of the above.

[0056] As Figure 4 shown, it is a schematic hardware structure diagram of an electronic device provided by an embodiment of the present invention. The electronic device 40 includes: a processor 41, a memory 42, and a computer program; wherein The memory 42 is used to store the computer program. This memory can also be a flash memory. The computer program is, for example, an application program, a functional module, etc. that implements the above method.

[0057] The processor 41 is used to execute the computer program stored in the memory to implement each step performed by the device in the above method. For details, please refer to the relevant descriptions in the previous method embodiments.

[0058] Optionally, the memory 42 can be either independent or integrated with the processor 41.

[0059] When the memory 42 is a device independent of the processor 41, the device may further include: A bus 43 for connecting the memory 42 and the processor 41.

[0060] A readable storage medium stores a computer program. When the computer program is executed by a processor, it is used to implement the steps of a real-time monitoring method for cutter wear of a shield machine as described in any one of the above.

[0061] Among them, the readable storage medium can be a computer storage medium or a communication medium. The communication medium includes any medium that facilitates the transmission of a computer program from one place to another. The computer storage medium can be any available medium accessible by a general-purpose or special-purpose computer. For example, the readable storage medium is coupled to the processor, enabling the processor to read information from the readable storage medium and write information to the readable storage medium. Of course, the readable storage medium can also be a component of the processor. The processor and the readable storage medium can be located in an application specific integrated circuit (ASIC). Additionally, the ASIC can be located in a user device. Of course, the processor and the readable storage medium can also exist as discrete components in a communication device. The readable storage medium can be a read-only memory (ROM), a random access memory (RAM), a CD-ROM, magnetic tape, a floppy disk, and an optical data storage device, etc.

[0062] The present invention also provides a program product, which includes execution instructions stored in a readable storage medium. At least one processor of the device can read the execution instructions from the readable storage medium, and the execution of the execution instructions by at least one processor causes the device to implement the methods provided by the above various embodiments.

[0063] In the above embodiments of the device, it should be understood that the processor can be a central processing unit (CPU), and can also be other general-purpose processors, digital signal processors (DSPs), application specific integrated circuits (ASICs), etc. The general-purpose processor can be a microprocessor or the processor can also be any conventional processor, etc. The steps of the method disclosed in combination with the present invention can be directly embodied as being completed by the execution of a hardware processor, or can be completed by a combination of hardware and software modules in the processor.

[0064] Through the introduction of the above embodiments, the present invention, through the real-time monitoring method and system for the wear of shield machine hob tools, can respectively monitor and collect the working pressure data, working temperature data, and surface image data of the target shield machine hob based on pressure sensors, temperature sensors, and hob cameras; based on the Advanced Encryption Standard algorithm, transmit the working pressure data, working temperature data, and surface image data to the central processing system and perform preprocessing operations; based on the long short-term memory network model, support vector machine model, and convolutional neural network model built in the central processing system, analyze the working pressure data, working temperature data, and surface image data respectively to generate corresponding pressure data scores, temperature data scores, and image data scores; based on the fusion evaluation algorithm, calculate the wear state score according to the pressure data score, temperature data score, and image data score; based on the Monte Carlo simulation prediction algorithm, determine the remaining service life of the hob according to the wear state score. When the remaining service life of the hob is lower than the preset lower limit of the service life, trigger the wear warning mechanism, so that the wear state of the shield machine hob can be monitored in real time and accurately, reduce the maintenance cost, ensure the smooth progress of the project, and improve the safety and economic benefits of shield construction.

[0065] Through the multi-dimensional sensor layout and multi-model collaborative analysis of the present invention, the wear state of the shield machine hob can be monitored with high precision and in real time, so that small wear changes can be detected in time, improving the accuracy and timeliness of wear monitoring and avoiding equipment failures caused by excessive wear. Through accurate wear prediction and intelligent decision support of the present invention, the maintenance plan can be made more scientific and reasonable, reducing unnecessary tool replacement and maintenance work, lowering the maintenance cost, and extending the tool service life. Through the real-time monitoring and warning mechanism of the present invention, safety accidents caused by severe hob wear can be effectively prevented, ensuring the life safety of construction personnel and the normal operation of construction equipment, and ensuring the safety and reliability of the construction process. The present invention shortens the downtime caused by tool failures, can quickly replace the tool and resume construction according to the maintenance suggestions, improves the tunneling efficiency of the shield machine, speeds up the project progress, and enhances the overall economic benefits of the project.

[0066] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that they can still modify the technical solutions described in the foregoing embodiments, or perform equivalent replacements on some or all of the technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. A real-time monitoring method for the wear of the hob cutter of a shield machine, characterized in that The monitoring method includes: Based on a pressure sensor, a temperature sensor, and a hob camera, respectively monitor and collect the working pressure data, working temperature data, and surface image data of the target shield machine's hob; Based on the Advanced Encryption Standard algorithm, transmit the working pressure data, the working temperature data, and the surface image data to a central processing system and perform preprocessing operations; Based on the Long Short-Term Memory network model, Support Vector Machine model, and Convolutional Neural Network model built in the central processing system, respectively analyze the working pressure data, the working temperature data, and the surface image data to generate corresponding pressure data scores, temperature data scores, and image data scores; Based on a fusion evaluation algorithm, calculate the wear state score according to the pressure data score, the temperature data score, and the image data score; Based on the Monte Carlo simulation prediction algorithm, determine the remaining service life of the hob according to the wear state score. When the remaining service life of the hob is lower than the preset lower limit of the service life, trigger a wear warning mechanism.

2. The real-time monitoring method for the cutter wear of a shield machine hob according to claim 1, characterized in that Install one of the pressure sensors at intervals of a preset distance at the contact edge of the target shield machine's hob; install the temperature sensor in a reserved hole at the axis of the target shield machine's hob; install the hob camera on the side of the target shield machine's hob.

3. A real-time monitoring method for the wear of a cutter of a shield machine according to claim 1, characterized in that, The preprocessing operation specifically includes: Perform denoising processing on the working pressure data based on the moving average filtering algorithm; Fill in the missing values in the working temperature data using linear interpolation; Enhance the contrast of the surface image data using the histogram equalization algorithm.

4. A real-time monitoring method for the wear of a cutter of a shield machine according to claim 1, characterized in that, Based on the Convolutional Neural Network model, analyze the working pressure data to generate the pressure data score, specifically including: Input the working pressure data into the forget gate of the Convolutional Neural Network model, and determine the data that needs to be retained in the working pressure data based on the forget gate. The corresponding calculation formula is as follows: In the formula, represents the output of the forgetting gate at the current time t; represents the weight of the forgetting gate at the current time t; represents the hidden state at the previous time t - 1; represents the working pressure data input at the current time t; represents the hidden state and the working pressure data to form a vector; represents the bias term of the forgetting gate; represents the activation function; Based on the input gate, determine the data that needs to be added to the candidate cell state in the working pressure data. The corresponding calculation formula is as follows: Wherein, represents the output of the input gate at the current time t; represents the candidate cell state at the current time t; and represent the weights of the input gate and the candidate cell state respectively; represents the hidden state at the previous time t-1; represents the working pressure data input at the current time t; represents the hidden state and the working pressure data to form a vector; and represent the bias terms of the input gate and the candidate cell state respectively; represents the activation function; represents the hyperbolic tangent function; Update the cell state according to the output of the forget gate, the output of the input gate, and the candidate cell state. The corresponding calculation formula is as follows: In the formula, represents the cell state at the current time t; represents the output of the forget gate at the current time t; represents the cell state at the previous time t - 1; represents the output of the input gate at the current time t; represents the candidate cell state at the current time t; According to the cell state, determine the data that needs to be output in the hidden state through the output gate. The corresponding calculation formula is as follows: In the formula, represents the output of the output gate at the current time t; represents the activation function; represents the weight of the output gate; represents the hidden state at the previous time t - 1; represents the working pressure data input at the current time t; represents the hidden state and the working pressure data to form a vector; represents the bias term of the output gate; represents the hidden state at the current time t; represents the hyperbolic tangent function; represents the cell state at the current time t; Take the hidden state at the current moment as the predicted pressure data, and the pressure data score is the ratio of the predicted pressure data to the working pressure data.

5. A real-time monitoring method for the wear of the cutter of a shield machine according to claim 1, characterized in that Based on the Support Vector Machine model, analyze the working temperature data to generate the temperature data score, specifically including: Extract the working temperature features corresponding to the working temperature data, and perform feature classification on the working temperature features based on the Support Vector Machine model. The corresponding optimization problem is as follows: In the formula, represents the weight of the \(u\)-th support vector machine, which is used to determine the direction of the classification hyperplane; represents the bias term of the \(u\)-th support vector machine, which is used to determine the distance between the classification hyperplane and the origin; represents the slack variable when the \(v\)-th working temperature feature is linearly inseparable in the \(u\)-th support vector machine; \(G\) represents the regularization parameter, which is used to control the trade-off between the classification margin and classification error; \(R\) represents the number of working temperature features; min represents the minimum value operation; The constraint conditions corresponding to the optimization problem are as follows: Wherein, represents the feature category label corresponding to the v-th working temperature feature; represents the v-th working temperature feature; represents the weight of the u-th support vector machine, which is used to determine the direction of the classification hyperplane; represents the bias term of the u-th support vector machine, which is used to determine the distance between the classification hyperplane and the origin; represents the slack variable when the v-th working temperature feature is linearly inseparable in the u-th support vector machine; R represents the number of working temperature features; Based on the feature category label, discriminate abnormal temperature features and the corresponding abnormal temperature data; The temperature data score is the ratio of the number of abnormal temperature data to the working temperature data.

6. A real-time monitoring method for cutter wear of a shield machine hob according to claim 1, characterized in that Analyze the surface image data based on the convolutional neural network model to generate the image data score, specifically including: Set the height of the surface image data X as H, the width as W, and the number of channels as E; the height of the convolutional kernel of the convolutional neural network model is h, the width is w, and the number is M; the width and height of the pooling window of the convolutional neural network model are both k, and the stride is s; Extract features from the surface image data based on the convolution to obtain the following convolutional image features: In the formula, represents the value of the convolutional image feature output by the m-th convolutional kernel at the position ; represents the value of the m-th convolutional kernel at the position ; represents the value of the convolutional image feature at the position ; represents the bias term corresponding to the m-th convolutional kernel; Perform max-pooling processing on the surface image data based on the pooling window to obtain the following pooled image features: In the formula, represents the value of the pooled image feature output by the m-th pooling window at the position ; max represents the maximum value operation; represents the value of the convolutional image feature output by the m-th convolutional kernel at the position ; Input the pooled image features into the fully connected layer of the convolutional neural network model to output the following wear image features: In the formula, represents the output of the l-th fully connected layer, that is, the wear image feature; l represents the number of layers of the fully connected layer; represents the activation function; represents the weight of the l-th fully connected layer; represents the output of the (l - 1)-th fully connected layer; represents the bias term of the l-th fully connected layer; The image data score is the ratio of the number of the wear image features to the surface image features.

7. A real-time monitoring method for cutter wear of a shield machine hob according to claim 1, characterized in that Based on the fusion evaluation algorithm, calculate the wear state score according to the pressure data score, the temperature data score, and the image data score. The corresponding calculation formula is as follows: In the formula, represents the wear state score; respectively represent the pressure data score, the temperature data score, and the image data score; respectively represent the scoring weights corresponding to the pressure data score, the temperature data score, and the image data score.

8. A real-time monitoring system for cutter wear of a shield machine, which is applied to a real-time monitoring method for cutter wear of a shield machine as described in any one of claims 1-7, is characterized in that, The monitoring system includes: A monitoring and acquisition module, which is used to monitor and acquire the working pressure data, the working temperature data, and the surface image data of the cutter head of the target shield machine respectively based on a pressure sensor, a temperature sensor, and a hob camera; An encryption and transmission module, which is used to transmit the working pressure data, the working temperature data, and the surface image data to the central processing system based on the advanced encryption standard algorithm and perform preprocessing operations; A score generation module, which is used to analyze the working pressure data, the working temperature data, and the surface image data respectively based on the long short-term memory network model, the support vector machine model, and the convolutional neural network model built in the central processing system to generate the corresponding pressure data score, temperature data score, and image data score; A fusion evaluation module, which is used to calculate the wear state score according to the pressure data score, the temperature data score, and the image data score based on the fusion evaluation algorithm; An early warning trigger module, which is used to determine the remaining service life of the hob based on the Monte Carlo simulation prediction algorithm according to the wear state score. When the remaining service life of the hob is lower than the preset lower limit of the service life, trigger the wear early warning mechanism.

9. An electronic device, comprising a memory and a processor, wherein a computer program is stored in the memory, characterized in that, When the processor runs the computer program stored in the memory, the processor executes the steps of a real-time monitoring method for cutter head tool wear of a shield machine according to any one of claims 1-7.

10. A readable storage medium storing a computer program, characterized in that, When the computer program is executed by the processor, it is used to implement the steps of a real-time monitoring method for cutter head tool wear of a shield machine according to any one of claims 1-7.

Images