Drill bit state quick judgment method for tunnel excavation of three-arm drill jumbo
Through the rapid discrimination method of drill bit status based on vibration monitoring, the cepspectral coefficient method and BP neural network model are used to solve the subjectivity and inaccuracy of the traditional drill bit status monitoring method, and high accuracy and high efficiency drill bit status recognition are achieved, reducing maintenance costs.
Patent Information
- Application Number
- CN202411664943.4
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-11-20
- Publication Date
- 2025-03-11
AI Technical Summary
Traditional drill bit status monitoring methods rely on manual inspection, which is time-consuming and labor-intensive and is easily affected by subjective factors, resulting in inaccurate judgment and difficult to effectively identify the working status of the drill bit.
The drill bit state rapid discrimination method based on vibration monitoring is adopted. By installing a vibration monitoring system, vibration data is obtained and processed, feature signals are extracted using the cepspectral coefficient method, and a BP neural network model is constructed for state discrimination.
It improves the accuracy and efficiency of drill bit status recognition, supports real-time monitoring and feedback, reduces maintenance costs, and extends the service life of the equipment.
Smart Images
Figure CN119593737A_ABST
Abstract
Description
Technical Field
[0001] The invention relates to a method for quickly judging the drill bit state of a three-arm rock drilling trolley for tunnel excavation, and is applicable to the field of tunnel excavation. Background Art
[0002] In modern tunnel excavation operations, the three-arm rock drilling rig is an important construction equipment, and the working state of its drill bit directly affects the construction efficiency and safety. The drill bit may be in a variety of states during operation, such as normal operation, wear, drill jamming, deflection, etc. Timely identification of these states is of great significance for preventing equipment failures, reducing maintenance costs and improving construction efficiency.
[0003] Traditional drill bit status monitoring methods mainly rely on manual inspection and empirical judgment, which is not only time-consuming and labor-intensive, but also easily affected by subjective factors, resulting in inaccurate judgment. With the development of sensor technology and data processing technology, status recognition methods based on vibration monitoring have gradually become a research hotspot. Vibration signals can reflect the operating status of equipment. By analyzing vibration data, the working status of the drill bit can be effectively identified.
[0004] In the vibration monitoring system, high-quality vibration signals can be obtained through reasonable installation and data collection. Data processing and analysis are key steps to achieve state recognition, which usually include data preprocessing, feature extraction and model building. As an effective feature extraction method, the cepstral coefficient method can extract useful feature information from the vibration signal and provide support for subsequent state discrimination.
[0005] In summary, the method of rapid drill bit status identification based on vibration monitoring, combined with modern data processing technology and machine learning algorithms, can effectively improve the accuracy and efficiency of drill bit status identification and provide reliable technical support for tunnel excavation operations. This method can not only reduce labor costs, but also improve construction safety, and has broad application prospects. Summary of the invention
[0006] The purpose of the present invention is to propose a method for quickly judging the drill bit status of a three-arm rock drilling rig tunnel excavation based on vibration data recognition, which can detect equipment abnormalities in time through data monitoring and real-time monitoring to ensure the smooth progress of construction. The purpose of the present invention can be achieved by adopting the following technical solutions:
[0007] A method for quickly judging the drill bit status of a three-arm rock drilling rig tunnel excavation, comprising the following steps:
[0008] S101, Installation of vibration monitoring system;
[0009] The installation of the vibration monitoring system includes three vibration pickups, a data transmission line and a data collection and processing system. The vibration pickups can be used to obtain vibration data. The vibration data is transmitted to the data collection and processing system through the data transmission line. The data processing and collection system is built into the driving computer of the three-arm rock drilling rig;
[0010] The installation of the vibration monitoring system includes installing the three vibration pickups on the rock drills of the three-arm rock drilling rig respectively, and the installation positions are required to be completely consistent;
[0011] S102, obtaining vibration monitoring data;
[0012] A vibration pickup is installed at a fixed position near the three rock drills of the three-arm rock drilling rig, and when the three-arm rock drilling rig performs tunnel excavation operations, vibration data of a fixed time period is obtained at a certain time interval, and transmitted to the data collection and processing system through the data transmission line;
[0013] S103, data processing and analysis;
[0014] The data processing and analysis includes the data collection and processing system performing data processing and analysis on the vibration data after receiving the vibration data;
[0015] The data processing and analysis includes preprocessing n vibration data samples included in the vibration data by the data processing;
[0016] The data processing and analysis includes extracting corresponding characteristic signal data from each vibration data sample using the cepstral coefficient method, and extracting 24 corresponding characteristic signal data from each vibration data sample;
[0017] The data processing and analysis also includes expert evaluation of the drill bit status when the n vibration data samples are obtained to obtain labels, wherein the labels are 1 to 5. When the expert evaluates that the drill bit is in a normal working state, the label assigned to the vibration sample is 1; when the expert evaluates that the drill bit is in a wear state, the label assigned to the vibration sample is 2; when the expert evaluates that the drill bit is in a stuck state, the label assigned to the vibration sample is 3; when the expert evaluates that the drill bit is in a deflected state, the label assigned to the vibration sample is 4; when the expert evaluates that the drill bit is in other abnormal states, the label assigned to the vibration sample is 5;
[0018] The data processing and analysis includes obtaining characteristic signal data and labels for each vibration data sample and constructing model training data, wherein the model training data has 25 dimensions, the first dimension is the label identification, and the next 24 dimensions are the characteristic signal data;
[0019] S104, constructing a BP neural network model;
[0020] The constructing of the BP neural network model includes constructing a BP neural network model, wherein the BP neural network structure adopts 24-20-5, which means that the BP neural network input layer has 24 nodes, the hidden layer has 20 nodes, and the output layer has 5 nodes;
[0021] S105, model training;
[0022] The model training includes using the BP neural network to carry out model training, and the model training steps are: a) normalizing the model training data, b) dividing the model training data into training set data and test set data according to a certain ratio, c) setting the expected output value of each group of model training data according to the first dimension label identification of the model training data, when the label is 1, the expected output value is
[10000] , when the label is 2, the expected output value is
[01000] , when the label is 3, the expected output value is
[00100] , when the label is 4, the expected output value is
[00010] , when the label is 5, the expected output value is
[00001] , d) initializing the BP neural network, using the training set data to train the BP neural network, and adjusting the weight and threshold of the network according to the network prediction error during the training process, e) using the test set data to evaluate and optimize the BP neural network, the basis of the evaluation is the classification error of the BP neural network, and the optimization is to optimize the parameters of the BP neural network according to the evaluation results, and f) finally obtaining a BP neural network model whose classification error meets the requirements;
[0023] S106, model application;
[0024] The model application includes using the BP neural network model whose classification error meets the requirements to apply it to the drill bit status identification work of actual three-arm rock drilling rig tunnel excavation.
[0025] Furthermore, in the above S103, the preprocessing includes denoising the vibration data samples, and the denoising is achieved by filtering.
[0026] Furthermore, in the above S105, the normalization process adopts the maximum-minimum method.
[0027] Furthermore, in the above S105, the standard of the BP neural network model that the classification error meets the requirements is that the recognition accuracy of each category label is greater than or equal to 0.85.
[0028] The beneficial effects of the present invention are:
[0029] The beneficial effects of the present invention are: (1) improving the accuracy of drill bit status identification; by adopting a combination of vibration monitoring and BP neural network model, the drill bit status can be accurately judged through precise analysis of vibration data. Through feature extraction and deep learning model training, it is possible to effectively distinguish between normal status, wear, stuck drill, deflection and other states, thereby significantly improving the accuracy of drill bit status identification; (2) real-time monitoring and feedback; the vibration monitoring system can obtain the vibration data of the drill bit in real time, and process and analyze it through the driving computer to support real-time monitoring. In this way, the operator can get feedback on the drill bit status in real time, so as to take maintenance or adjustment measures in time to reduce the risk of failure. (3) reducing maintenance costs; through real-time monitoring and accurate status identification, potential problems of the drill bit can be discovered in advance, and maintenance and repair can be carried out in time, thereby avoiding the occurrence of serious failures. Preventive maintenance helps to reduce the overall maintenance cost of equipment and extend the service life of equipment. BRIEF DESCRIPTION OF THE DRAWINGS
[0030] Figure 1 It is a flow chart of the method for quickly judging the drill bit status of a three-arm rock drilling rig tunnel excavation according to the present invention;
[0031] Figure 2 This is a relationship diagram between prediction error and number of hidden layer nodes of an embodiment of a method for quickly judging drill bit status during tunnel excavation using a three-arm drilling rig of the present invention. DETAILED DESCRIPTION
[0032] The specific embodiments of the present invention are described in detail below in conjunction with the accompanying drawings; it should be understood that the specific embodiments given here are only used to illustrate and explain the present invention and cannot be used to limit the present invention.
[0033] The following is a specific embodiment of a method for quickly judging the drill bit status of a three-arm rock drilling rig in tunnel excavation.
[0034] The specific basis project of this embodiment is: the total length of the C2 line of the Eighth Division of the General Manager's Office of the East Coast Railway Project of China Communications Construction Corporation Malaysia is 3595m, the import mileage is CH000+606, the export mileage is CH004+201, the line changes from the import end to CH002+400 with a slope of 3‰ and a slope of 6‰, the maximum burial depth is 335m, the minimum burial depth is 5m, the tunnel is designed as a composite lining structure, and the standard line spacing is 4.2m. The design of the Sengwendan No. 1 tunnel is an inclined shaft auxiliary construction and rescue channel with a length of 261m, the clear section size of the inclined shaft is 5m*8m, the longitudinal slope of A1-1CH0+261-A1-1CH0+052 is 10%, and the longitudinal slope of A1-1CH0+052-A1-1CH0+000 is 1.6%. It adopts a trackless double-lane design and an anchor sprayed lining + molded lining structure. The main design technical standards of the Sengwendan No. 1 tunnel are shown in Table 1.
[0035] Table 1 Main design technical standards of Shuangwendan No. 1 tunnel
[0036] Tunnel Name Serendah Tunnel 1 Railway Grade National Railway Class I Track Type Ballasted track gauge 1435mm (standard track gauge) Number of main lines Double Line Tunnel length 3595m Maximum slope 5‰ Minimum curve radius 1000m Clearance section on rail surface Height 8.15m, width 11.2m Clearance area on track surface 76.63m2 Traction type Electric traction Occlusion type Automatic Block
[0037] Reference Figure 1 In this embodiment, a method for quickly judging the drill bit status of a three-arm rock drilling trolley for tunnel excavation includes the following steps:
[0038] S101, Installation of vibration monitoring system;
[0039] The installation of the vibration monitoring system includes three vibration pickups, a data transmission line and a data collection and processing system. The vibration pickups can be used to obtain vibration data. The vibration data is transmitted to the data collection and processing system through the data transmission line. The data processing and collection system is built into the driving computer of the three-arm rock drilling rig;
[0040] The installation of the vibration monitoring system includes installing the three vibration pickups on the rock drills of the three-arm rock drilling rig respectively, requiring the installation positions to be completely consistent.
[0041] In this embodiment, the vibration pickup used is a vibration sensor, and the index parameters of the vibration sensor are shown in Table 2.
[0042] Table 2 Index parameters of vibration sensor
[0043]
[0044]
[0045] S102, obtaining vibration monitoring data;
[0046] A vibration pickup is installed at a fixed position near the three rock drills of the three-arm rock drilling rig, and when the three-arm rock drilling rig performs tunnel excavation operations, vibration data of a fixed time period is obtained at a certain time interval, and transmitted to the data collection and processing system through the data transmission line;
[0047] S103, data processing and analysis;
[0048] The data processing and analysis includes the data collection and processing system performing data processing and analysis on the vibration data after receiving the vibration data;
[0049] The data processing and analysis includes preprocessing n vibration data samples included in the vibration data by the data processing;
[0050] The data processing and analysis includes extracting corresponding characteristic signal data from each vibration data sample using the cepstral coefficient method, and extracting 24 corresponding characteristic signal data from each vibration data sample;
[0051] The data processing and analysis also includes expert evaluation of the drill bit status when the n vibration data samples are obtained to obtain labels, wherein the labels are 1 to 5. When the expert evaluates that the drill bit is in a normal working state, the label assigned to the vibration sample is 1; when the expert evaluates that the drill bit is in a wear state, the label assigned to the vibration sample is 2; when the expert evaluates that the drill bit is in a stuck state, the label assigned to the vibration sample is 3; when the expert evaluates that the drill bit is in a deflected state, the label assigned to the vibration sample is 4; when the expert evaluates that the drill bit is in other abnormal states, the label assigned to the vibration sample is 5;
[0052] The data processing and analysis includes obtaining characteristic signal data and labels for each vibration data sample, and constructing model training data, wherein the model training data is 25 dimensions, the first dimension is the label identifier, and the next 24 dimensions are the characteristic signal data; further, in the above S103, the preprocessing includes denoising the vibration data samples, and the denoising is achieved by filtering.
[0053] S104, constructing a BP neural network model;
[0054] The constructing of the BP neural network model includes constructing a BP neural network model, wherein the BP neural network structure adopts 24-20-5, which means that the BP neural network input layer has 24 nodes, the hidden layer has 20 nodes, and the output layer has 5 nodes;
[0055] In this embodiment, in order to determine the reasonable number of hidden layer nodes, we set the hidden layer to 15 to 30 nodes respectively, and observe the relationship between the number of hidden layer nodes and the classification error rate. Figure 2 As shown, through Figure 2 It can be seen that when the number of hidden layer nodes is 20, the classification error rate is the lowest, so the number of hidden layer stages is selected as 20.
[0056] S105, model training;
[0057] The model training includes using the BP neural network to carry out model training, and the model training steps are: a) normalizing the model training data, b) dividing the model training data into training set data and test set data according to a certain ratio, c) setting the expected output value of each group of model training data according to the first dimension label identification of the model training data, when the label is 1, the expected output value is
[10000] , when the label is 2, the expected output value is
[01000] , when the label is 3, the expected output value is
[00100] , when the label is 4, the expected output value is
[00010] , when the label is 5, the expected output value is
[00001] , d) initializing the BP neural network, using the training set data to train the BP neural network, and adjusting the weight and threshold of the network according to the network prediction error during the training process, e) using the test set data to evaluate and optimize the BP neural network, the basis of the evaluation is the classification error of the BP neural network, and the optimization is to optimize the parameters of the BP neural network according to the evaluation results, and f) finally obtaining a BP neural network model whose classification error meets the requirements;
[0058] Furthermore, in the above S105, the normalization process adopts the maximum-minimum method;
[0059] Furthermore, in the above S105, the standard of the BP neural network model that the classification error meets the requirements is that the recognition accuracy of each category label is greater than or equal to 0.85.
[0060] In this embodiment, the recognition accuracy rate for each tag is finally shown in Table 3, and the recognition accuracy rate meets the requirement of being greater than or equal to 0.85.
[0061] Table 3 BP neural network classification accuracy table
[0062] Category Label 1 2 3 4 5 Recognition accuracy 0.8621 0.9211 0.9012 0.8666 0.9021
[0063] S106, model application;
[0064] The model application includes using the BP neural network model whose classification error meets the requirements to apply it to the drill bit status identification work of actual three-arm rock drilling rig tunnel excavation.
[0065] In the above embodiments, the present invention discloses a method for quickly judging the state of a drill bit for tunnel excavation using a three-arm drilling rig; including the installation of a vibration system, the acquisition of vibration monitoring data, the processing and analysis of data, the construction of a BP neural network model, model training and model application; the method for quickly judging the state of a drill bit for tunnel excavation using a three-arm drilling rig proposed in the present invention can accurately judge the state of the drill bit and make the judgment in a timely manner, which is conducive to further improving the efficiency of tunnel excavation work using a three-arm drilling rig, and can be widely used in the field of tunnel excavation.
[0066] The above description is a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principle of the present invention should be included in the protection scope of the present invention.
Claims
1. A method for quickly judging the drill bit status of a three-arm drilling rig tunnel excavation, characterized in that The following steps are involved: Installation of S101 vibration monitoring system; The vibration monitoring system includes three vibration pickups, a data transmission line and a data collection and processing system. The vibration pickups are used to obtain vibration data and transmit the data to the data collection and processing system through the data transmission line. The data processing and collection system is built into the driving computer of the three-arm rock drilling rig. At the same time, the three vibration pickups are respectively installed on the rock drill of the three-arm rock drilling rig, and the installation positions must be completely consistent. S102 Acquisition of vibration monitoring data; A vibration pickup installed at a fixed position near the rock drill of the three-arm rock drilling rig, when the three-arm rock drilling rig performs tunnel excavation operations, obtains vibration data of a fixed time period at a certain time interval, and transmits the data to the data collection and processing system through the data transmission line; S103 data processing and analysis; The data processing and analysis includes the data collection and processing system performing data processing and analysis on the vibration data after receiving the vibration data; The data processing and analysis includes preprocessing n vibration data samples included in the vibration data by the data processing; The data processing and analysis includes extracting corresponding characteristic signal data from each vibration data sample using the cepstral coefficient method, and extracting 24 corresponding characteristic signal data from each vibration data sample; The data processing and analysis also includes expert evaluation of the drill bit status when the n vibration data samples are obtained to obtain labels, wherein the labels are 1 to 5. When the expert evaluates that the drill bit is in a normal working state, the label assigned to the vibration sample is 1; when the expert evaluates that the drill bit is in a wear state, the label assigned to the vibration sample is 2; when the expert evaluates that the drill bit is in a stuck state, the label assigned to the vibration sample is 3; when the expert evaluates that the drill bit is in a deflected state, the label assigned to the vibration sample is 4; when the expert evaluates that the drill bit is in other abnormal states, the label assigned to the vibration sample is 5; The data processing and analysis includes obtaining characteristic signal data and labels for each vibration data sample and constructing model training data, wherein the model training data has 25 dimensions, the first dimension is the label identification, and the next 24 dimensions are the characteristic signal data; S104 builds a BP neural network model; The constructing of the BP neural network model includes constructing a BP neural network model, wherein the BP neural network structure adopts 24-20-5, which means that the BP neural network input layer has 24 nodes, the hidden layer has 20 nodes, and the output layer has 5 nodes; S105 model training; The model training includes using the BP neural network to carry out model training, and the model training steps are: a) normalizing the model training data; b) dividing the model training data into training set data and test set data according to a certain ratio; c) according to the first dimension label identification of the model training data, setting the expected output value of each group of model training data, when the label is 1, the expected output value is [10000], when the label is 2, the expected output value is [01000], when the label is 3, the expected output value is [00100], when the label is 4, the expected output value is [00010], when the label is 5, the expected output value is [00001]; d) Initialize the BP neural network, use the training set data to train the BP neural network, and adjust the network weights and thresholds according to the network prediction error during the training process; e) using the test set data to evaluate and optimize the BP neural network, wherein the evaluation is based on the classification error of the BP neural network, and the optimization is to optimize the parameters of the BP neural network according to the evaluation results; f) Finally, a BP neural network model with classification error meeting the requirements is obtained; S106 Model Application; The model application includes using the BP neural network model whose classification error meets the requirements to apply it to the drill bit status identification work of actual three-arm rock drilling rig tunnel excavation.
2. The method for quickly judging the drill bit status of a three-arm drilling rig tunnel excavation according to claim 1 is characterized in that: In the step S103, the preprocessing includes denoising the vibration data samples, and the denoising is achieved by filtering.
3. The method for quickly judging the drill bit status of a three-arm drilling rig tunnel excavation according to claim 1, characterized in that: In the step S105, the normalization process adopts the maximum-minimum method.
4. The method for quickly judging the drill bit status of a three-arm drilling rig tunnel excavation according to claim 1, characterized in that: In step S105, the standard of the BP neural network model that the classification error meets the requirements is that the recognition accuracy of each category label is greater than or equal to 0.85.