Dynamic Detection Method for Coal Mine Explosion-Proof Equipment Based on Laser Doppler and Pressure Data

Abstract

This application relates to the field of intelligent detection technology, and discloses a dynamic detection method for coal mine explosion-proof equipment based on laser Doppler and pressure data. The method includes: setting a measurement area and a scanning scheme according to the geometric shape, working environment and measurement requirements of the explosion-proof housing, arranging pressure sensors and laser Doppler vibrometers at measurement points, conducting explosion experiments and collecting laser Doppler data and pressure data; extracting features from the collected data to obtain vibration feature vectors and pressure feature vectors respectively, inputting the feature vectors into a deep learning model for fusion to generate fused feature vectors, and finally inputting the fused feature vectors into a comprehensive evaluation model to output the performance score of the explosion-proof housing and compare it with a threshold to determine its explosion-proof performance. This application can perform high-precision and non-contact dynamic response measurement and performance evaluation on coal mine explosion-proof equipment under the action of explosion shock, improving the efficiency and accuracy of dynamic detection of explosion-proof equipment.

Description

Technical Field

[0001] This application relates to the field of intelligent detection technology, and particularly to a dynamic detection method for coal mine explosion-proof equipment based on laser Doppler and pressure data. Background Art

[0002] Coal accounts for more than 90% of fossil energy resources and is the most stable, economical, and self-guaranteed energy source. The geological conditions for coal mining are complex, and the content of flammable and explosive gases is high. Disasters such as mechanical and electrical accidents and gas explosions caused by faults in explosion-proof equipment are particularly prominent. Therefore, the safety of coal mine explosion-proof equipment is directly related to the smooth progress of coal mine safety production work. It is necessary to improve the explosion-proof performance evaluation ability of coal mine explosion-proof equipment to avoid safety accidents caused by gas explosions.

[0003] The current measurement and evaluation system for coal mine explosion-proof equipment is relatively basic, lacking systematic research on the evolution theory of coal mine explosion-proof equipment and the testing methods of main performance parameters. The existing theories and technologies cannot effectively solve the problems of intelligent testing and safety evaluation of coal mine explosion-proof equipment in China. In traditional detection, the strength of the explosion-proof enclosure is mainly verified through pressure tests. However, the mechanical changes of the coal mine explosion-proof enclosure under the action of gas explosion are a complex process, and the mechanical changes inside the enclosure cannot be revealed and effectively evaluated by simply analyzing the pressure. Moreover, if other contact measurement methods are considered, there are problems such as sensor installation and sensor detachment caused by violent explosions, and none of the above can achieve good evaluation of the coal mine explosion-proof enclosure.

[0004] The introduction of laser Doppler measurement technology can effectively overcome these problems. In addition to avoiding the influence of explosion conditions, installation and other factors of traditional contact measurement methods, it has the advantages of high precision and fast response. Importantly, it can measure the dynamic parameters of the explosion-proof enclosure to reveal the internal laws. Therefore, the fusion of laser Doppler and pressure data can effectively solve the testing problems of coal mine explosion-proof equipment, and the performance of the explosion-proof enclosure can be effectively evaluated by combining data processing methods. Summary of the Invention

[0005] This application provides a dynamic detection method for coal mine explosion-proof equipment based on laser Doppler and pressure data, which can measure the surface dynamic response of the explosion-proof enclosure under explosion shock in real time and accurately, and improve the efficiency and accuracy of dynamic detection of coal mine explosion-proof equipment.

[0006] In a first aspect, this application provides a dynamic detection method for coal mine explosion-proof equipment based on laser Doppler and pressure data, and the method includes:

[0007] Step 1: Set the measurement area and scanning scheme according to the geometric shape, working environment and area to be measured of the explosion-proof enclosure, where the scanning scheme includes the number of measurement points, distribution positions and scanning paths;

[0008] Step 2: Set pressure sensors at each measurement point, and arrange and install a laser Doppler vibrometer based on the measurement area and scanning scheme;

[0009] Step 3: Conduct an explosion experiment, and simultaneously collect the laser Doppler data and pressure data at each measurement point;

[0010] Step 4: Extract features from the laser Doppler data to obtain vibration feature vectors, and extract features from the pressure data to obtain pressure feature vectors;

[0011] Step 5: Input the vibration feature vectors and pressure feature vectors into a deep learning model for fusion to obtain fusion feature vectors;

[0012] Step 6: Input the fusion feature vectors into a comprehensive evaluation model, output the performance score of the explosion-proof housing, and compare the performance score with a set threshold to determine the explosion-proof performance of the explosion-proof housing.

[0013] Combined with the first aspect, in the first implementation manner of the first aspect of the present application, in step 1, the generation method of the scanning scheme is as follows:

[0014] Step 11: Analyze the geometric shape and force characteristics of the explosion-proof housing, and design the first measurement point distribution method and the second measurement point distribution method based on the analysis results;

[0015] Step 12: Based on the layout methods corresponding to each measurement point distribution method, arrange virtual measurement points in the measurement area;

[0016] Step 13: Identify overlapping measurement points, delete the overlapping measurement points from the virtual measurement point set corresponding to the first measurement point distribution method to obtain the first virtual measurement point set;

[0017] Step 14: Generate a first scanning path based on the first virtual measurement point set, generate a second scanning path based on the second virtual measurement point set, and fuse the first scanning path and the second scanning path to generate a third scanning path, where the second virtual measurement point set is the virtual measurement point set corresponding to the second measurement point distribution method;

[0018] Step 15: Conduct an explosion simulation experiment, collect the vibration data at each virtual measurement point, analyze the vibration data, adjust the number and distribution of virtual measurement points according to the analysis results, and generate a new scanning path based on the adjusted results. Repeat step 15. After multiple experimental verifications and adjustments, determine the number, distribution positions, and scanning path of the measurement points.

[0019] In combination with the first aspect, in the second implementation manner of the first aspect of the present application, the first measurement point distribution method is global scanning, and the second measurement point distribution method is local scanning. Among them, the area of the local scanning includes the central area, the edge area, and the structural connection area, and the area shape is circular or rectangular;

[0020] The layout methods corresponding to the global scanning and the central area are evenly spaced layouts, and the layout methods corresponding to the edge area and the structural connection area are dense layouts.

[0021] In combination with the first aspect, in the third implementation manner of the first aspect of the present application, before step 4, it includes: performing time synchronization and space synchronization processing on the laser Doppler data and the pressure data.

[0022] In combination with the first aspect, in the fourth implementation manner of the first aspect of the present application, the laser Doppler data includes vibration displacement, vibration velocity, and vibration acceleration. In step 4, the vibration feature vector obtained by feature extraction of the laser Doppler data is:

[0023] Analyze the laser Doppler data to obtain the first feature data of the laser Doppler data in the time dimension, and combine all the first feature data to generate a time-domain feature vector. Among them, the first feature data includes the maximum amplitude, root mean square amplitude, standard deviation of vibration velocity, and mean value of vibration velocity of the laser Doppler data;

[0024] After performing Fourier transform on the laser Doppler data and then analyzing it, obtain the second feature data of the laser Doppler data in the frequency dimension, and combine all the second feature data to generate a frequency-domain feature vector. Among them, the second feature data includes the main frequency, maximum value of power spectral density, and frequency bandwidth;

[0025] Combine the time-domain feature vector and the frequency-domain feature vector to generate a vibration feature vector.

[0026] In combination with the first aspect, in the fifth implementation manner of the first aspect of the present application, in step 4, the pressure feature vector obtained by feature extraction of the pressure data is:

[0027] Analyze the pressure data to obtain the maximum pressure, root mean square pressure, pressure change rate, and pressure pulse width, and combine the maximum pressure, root mean square pressure, pressure change rate, and pressure pulse width to generate a pressure feature vector.

[0028] In combination with the first aspect, in the sixth implementation manner of the first aspect of the present application, step 5 includes:

[0029] Perform deep feature extraction on the vibration feature vector and the pressure feature vector respectively through a neural network to obtain the first deep feature corresponding to the vibration feature vector and the second deep feature corresponding to the pressure feature vector;

[0030] The first deep feature and the second deep feature are concatenated or weighted and fused to generate a fused feature vector.

[0031] Combined with the first aspect, in the seventh implementation manner of the first aspect of the present application, in step 6, a mechanical response analysis is performed on the explosion-proof housing:

[0032] The vibration displacement of each measurement point is obtained from the laser Doppler data, and the shock wave intensity of each measurement point is obtained from the pressure data. Based on the vibration displacement and the shock wave intensity, the dynamic displacement distribution on the surface of the explosion-proof housing is calculated by a numerical analysis method;

[0033] Through a frequency-domain analysis method, combined with the laser Doppler data and the pressure data, the response characteristics of the explosion-proof housing at different frequencies are obtained;

[0034] A mechanical model is selected based on the material properties, structural form, and experimental conditions of the explosion-proof housing. The pressure data is used to drive the mechanical model to obtain the mechanical response of the mechanical model. Subsequently, the laser Doppler data is used to verify and correct the mechanical model to determine the change trends of stress and strain, and a stress-strain model is constructed.

[0035] Combined with the first aspect, in the eighth implementation manner of the first aspect of the present application, the laser Doppler vibrometer is configured with a galvanometer control system. During the explosion experiment, the irradiation angle and position of the laser head are automatically adjusted to control the movement of the laser beam according to the scanning path.

[0036] Combined with the first aspect, in the ninth implementation manner of the first aspect of the present application, the laser Doppler data and the pressure data are denoised and calibrated before step 4.

[0037] Compared with the prior art, the beneficial effects of the technical solution of the present application are at least as follows:

[0038] 1. By setting multiple measurement points and scanning paths, and combining global scanning and local scanning, all key parts of the explosion-proof housing can be comprehensively covered, ensuring the comprehensiveness and accuracy of the detection.

[0039] 2. Dynamic parameters such as displacement, velocity, and acceleration on the surface of the explosion-proof housing can be accurately obtained. The fusion with the pressure data can provide effective experimental data for explosion-proof performance evaluation and design. By extracting features from the laser Doppler data and the pressure data and inputting them into a deep learning model for fusion, a more comprehensive fused feature vector can be obtained. This data fusion method can comprehensively consider information such as vibration and pressure, providing a more accurate explosion-proof performance evaluation.

[0040] 3. By performing a mechanical response analysis on the explosion-proof housing, combining the mechanical model and experimental data, calculating the dynamic displacement distribution on its surface and the responses at different frequencies, and constructing a stress-strain model, it is possible to deeply understand the dynamic behavior of the explosion-proof housing under explosion shock, which helps to identify the weak links of the explosion-proof housing and make targeted improvements. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] To more clearly illustrate the technical solutions of the embodiments of the present invention, the following will briefly introduce the drawings required for the description of the embodiments. Obviously, the drawings in the following description are some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can also be obtained based on these drawings.

[0042] Figure 1 FIG. is a schematic diagram of an embodiment of the dynamic detection method for coal mine explosion-proof equipment based on laser Doppler and pressure data in the embodiments of the present application;

[0043] Figure 2 FIG. is a schematic diagram of another embodiment of the dynamic detection method for coal mine explosion-proof equipment based on laser Doppler and pressure data in the embodiments of the present application;

[0044] Figure 3 FIG. is a schematic diagram of an embodiment of the global scan in the embodiments of the present application;

[0045] Figure 4 FIG. is a schematic diagram of an embodiment of the local scan in the embodiments of the present application. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0046] The embodiments of the present application provide a dynamic detection method for coal mine explosion-proof equipment based on laser Doppler and pressure data. The terms "first", "second", "third", "fourth", etc. (if any) in the specification, claims and drawings of the present application are used to distinguish similar objects and do not necessarily describe a specific order or sequence. It should be understood that such data can be interchanged under appropriate circumstances so that the embodiments described here can be implemented in an order other than that illustrated or described here. In addition, the terms "comprising" or "having" and any variation thereof are intended to cover non-exclusive inclusion. For example, a process, method, system, product or device that includes a series of steps or units does not necessarily have to be limited to those steps or units clearly listed, but may include other steps or units not clearly listed or inherent to these processes, methods, products or devices.

[0047] For ease of understanding, the following describes the specific process of the embodiments of the present application. Please refer to Figure 1 An embodiment of the dynamic detection method for coal mine explosion-proof equipment based on laser Doppler and pressure data in the embodiments of the present application includes:

[0048] Step 1: Set the measurement area and scanning scheme according to the geometric shape, working environment, and area to be measured of the explosion-proof housing. Among them, the scanning scheme includes the number of measurement points, distribution positions, and scanning paths.

[0049] Specifically, before Step 1, select a suitable laser Doppler vibrometer and data acquisition system to ensure that the power, frequency range, accuracy, etc. of the sensor meet the dynamic response measurement requirements of coal mine explosion-proof equipment. At the same time, to prevent the explosion-proof equipment from affecting the measurement data due to vibration or external force during the measurement process, it is necessary to ensure that the equipment is firmly fixed on the test platform, and use appropriate supports or fixtures to firmly fix the explosion-proof equipment to avoid any possible movement or shaking.

[0050] According to the geometric shape, working environment, and area to be measured of the explosion-proof housing, select and plan the measurement area. Usually, the outer surface of the explosion-proof housing is selected, especially those parts that are subjected to greater stress or more obvious vibration, such as the central area, edge area, and structural connection parts. These parts may be strongly affected by external factors such as explosion shock waves and pressure waves. Generally, when an explosion occurs in the cavity, the explosion shock wave will spread spherically around the ignition point. Ideally, it is centered ignition and evenly stressed in all directions. Therefore, the side surface of the explosion-proof housing is selected as the measurement area.

[0051] Step 2: Set pressure sensors at each measurement point, and arrange and install a laser Doppler vibrometer based on the measurement area and scanning scheme.

[0052] Specifically, set pressure sensors at each measurement point. The arrangement of the sensors needs to ensure that they can accurately sense the pressure change of the explosion shock wave. Usually, they need to be installed on the surface or inside of the explosion-proof housing, and the specific position is determined according to the measurement point distribution method determined in Step 1. Before installation, calibrate the pressure sensors to ensure their measurement accuracy and reliability. The calibration process usually includes steps such as zero calibration, range calibration, and linearity calibration.

[0053] Specifically, after determining the measurement area and scanning scheme, fix the laser head of the laser Doppler in an appropriate position in front of the test area of the explosion-proof equipment to ensure that the laser beam can accurately irradiate the surface of the measurement area, and set the optimal measurement distance and angle to ensure that the laser beam irradiates the measurement area vertically or approximately vertically. Before installation, perform necessary calibration and debugging on the laser Doppler instrument to ensure the normal operation of the equipment. Through comparison tests with known standard signals, confirm that the measurement system can accurately capture surface vibration information.

[0054] Step 3: Conduct an explosion experiment while collecting laser Doppler data and pressure data at each measurement point.

[0055] Specifically, before conducting the explosion experiment, sufficient preparatory work needs to be carried out, including checking whether the connections of all equipment are firm, whether the sensors are working properly, whether the scanning path is correct, etc. At the same time, confirm whether all safety measures have been implemented, including safety distances, protective equipment, emergency plans, etc., to ensure the safety of experimental personnel and equipment. According to the experimental design, set an appropriate explosion source. The selection and setting of the explosion source need to consider factors such as explosion equivalent, explosion location, explosion direction, etc., to ensure that the explosion shock wave can effectively act on the explosion-proof housing.

[0056] During the explosion experiment, the laser Doppler vibrometer obtains the vibration characteristic data (i.e., laser Doppler data) of the explosion-proof housing by measuring vibration displacement, vibration velocity, and vibration acceleration, and the pressure sensor measures the pressure change of the explosion shock wave in real time (i.e., pressure data).

[0057] Step 4: Extract features from the laser Doppler data to obtain a vibration feature vector, and extract features from the pressure data to obtain a pressure feature vector.

[0058] What the laser Doppler measures is the vibration information on the surface of the explosion-proof housing. The vibration data of all measurement points are collected and integrated to form a complete vibration response data set. Further, features are extracted from this vibration response data set to obtain a vibration feature vector; the pressure sensor provides local pressure change data on the surface of the explosion-proof housing. The pressure data of all measurement points are collected and integrated to form a complete pressure change data set. Further, features are extracted from this pressure change data set to obtain a pressure feature vector, providing high-quality basic data for subsequent deep learning and comprehensive evaluation.

[0059] Step 5: Input the vibration feature vector and the pressure feature vector into the deep learning model for fusion to obtain a fused feature vector.

[0060] Specifically, the vibration feature vector and the pressure feature vector can be fused by concatenation fusion or weighted fusion. Concatenation fusion simply concatenates the vibration feature vector and the pressure feature vector to generate a new feature vector. This method is simple and direct, but it may ignore the correlation between different features; weighted fusion weights the vibration feature vector and the pressure feature vector according to the contribution degree of different features to generate a new feature vector. This method can better reflect the relative importance of different features.

[0061] Step 6: Input the fused feature vector into the comprehensive evaluation model, output the performance score of the explosion-proof housing, and compare the performance score with the set threshold to determine the explosion-proof performance of the explosion-proof housing.

[0062] Specifically, the fused features (including time-domain, frequency-domain features, and pressure features) are input into a deep learning model (such as a multi-layer perceptron), and the performance of the explosion-proof housing is determined based on the output performance evaluation results.

[0063] Construct a comprehensive evaluation model (such as a neural network model) based on a physical model and a machine learning model, and input the fused feature vector into the comprehensive evaluation model to output a performance score S perf , , where is the evaluation function corresponding to the comprehensive evaluation model, and F deep is the fused feature vector.

[0064] Set a threshold , and determine whether the explosion-proof housing meets the design requirements according to the scoring result. For example, if , it is considered that the explosion-proof performance of the explosion-proof housing is qualified, otherwise it is unqualified.

[0065] Figure 2 This is another schematic diagram of the dynamic detection method for coal mine explosion-proof equipment based on laser Doppler and pressure data in the embodiments of the present application.

[0066] In a specific embodiment, in step 1, the method for generating the scanning scheme is as follows:

[0067] Step 11: Analyze the geometric shape and force characteristics of the explosion-proof housing, and design the first measurement point distribution method and the second measurement point distribution method based on the analysis results.

[0068] Step 12: Based on the layout method corresponding to each measurement point distribution method, arrange virtual measurement points in the measurement area.

[0069] Step 13: Identify overlapping measurement points, delete the overlapping measurement points from the set of virtual measurement points corresponding to the first measurement point distribution method, and obtain the first set of virtual measurement points.

[0070] Step 14: Generate a first scanning path based on the first set of virtual measurement points, generate a second scanning path based on the second set of virtual measurement points, and fuse the first scanning path and the second scanning path to generate a third scanning path, where the second set of virtual measurement points is the set of virtual measurement points corresponding to the second measurement point distribution method.

[0071] Step 15: Conduct an explosion simulation experiment, collect the vibration data at each virtual measurement point, analyze the vibration data, adjust the number and distribution of virtual measurement points according to the analysis results, and generate a new scanning path based on the adjusted results. Repeat step 15. After multiple experimental verifications and adjustments, determine the number of measurement points, the distribution positions, and the scanning path.

[0072] In a specific embodiment, the distribution pattern of the first measurement points is global scanning, and the distribution pattern of the second measurement points is local scanning. Among them, the area of local scanning includes the central area, the edge area, and the structural connection area, and the shape of the area is circular or rectangular;

[0073] The layout patterns corresponding to global scanning and the central area are evenly spaced layouts, and the layout patterns corresponding to the edge area and the structural connection area are dense layouts.

[0074] The number, distribution pattern, and scanning path of virtual measurement points are determined through multiple experiments to ensure the accuracy and efficiency of scanning. The distribution pattern can be divided into global scanning and local scanning. Global scanning is for the entire measurement area. Therefore, within the measurement area where the propagation is on the surface of the explosion-proof housing, several virtual measurement points are planned. Then, the optimal number of measurement points is selected according to different measurement point schemes. The schemes adopt evenly spaced intervals and dense arrangements according to the vibration-sensitive areas. As Figure 3 shown, the measurement points in global scanning are evenly distributed within the measurement area according to a certain grid spacing. The areas of local scanning are the central area and the edge area, and the shapes of the areas are two types: circular and rectangular areas. For the central area, which is generally the main part of the explosion-proof housing, the force is evenly distributed, and the vibration transmission is relatively obvious. An evenly spaced method is adopted, and it is laid out according to circular and rectangular grids. Since this area is relatively symmetrical, the even layout helps to obtain the overall vibration information. In the edge area, the vibration intensity of the outer edge area of the explosion-proof housing may be relatively large, especially when affected by external forces, the edge effect is more significant. The measurement points in this area can be arranged in a dense layout, that is, the measurement point density is increased in the vibration-sensitive area. Multiple measurement points can be set at equal intervals along a ring, especially more measurement points are set at the structural connection parts. As Figure 4 shown, the local scanning is laid out according to a rectangular grid, and the measurement points are evenly distributed.

[0075] According to the experimental results, the number and distribution pattern of virtual measurement points are optimized. For example: if the vibration information in a certain area is not obvious, the number of measurement points in this area can be appropriately reduced; if the vibration intensity in a certain area is relatively large or the change is complex, the measurement point density in this area can be increased. By combining global scanning and local scanning, the optimal number and distribution pattern of measurement points are determined to ensure that the measurement area can be fully covered and the vibration information of key parts can be captured focusedly, ensuring the comprehensiveness and accuracy of measurement data, and providing reliable data support for subsequent explosion-proof performance evaluation.

[0076] In a specific embodiment, before step 4, it includes: performing time synchronization and space synchronization processing on the laser Doppler data and the pressure data.

[0077] Since the sampling frequencies and sampling times of the laser Doppler vibrometer and the pressure sensor may be different, time synchronization processing is performed on these data to ensure that the data collected by different sensors are exactly aligned in time. The data can be time-aligned by interpolation methods. For example, methods such as linear interpolation and spline interpolation are used to interpolate the pressure data to the sampling time points of the laser Doppler vibrometer to ensure that the two types of data correspond at the same time step.

[0078] The vibration data provided by the laser Doppler vibrometer is usually the scanned area data, while the pressure data is collected by sensors at different positions. Therefore, the pressure data and vibration data at the same position must be used to ensure that the spatial distributions of the pressure data and the vibration data are consistent, which can be achieved by accurately measuring the relative positions between the sensors and performing coordinate transformation.

[0079] In a specific embodiment, the laser Doppler data includes vibration displacement, vibration velocity, and vibration acceleration. In step 4, the vibration feature vector obtained by feature extraction of the laser Doppler data is:

[0080] Analyze the laser Doppler data to obtain the first feature data of the laser Doppler data in the time dimension, and combine all the first feature data to generate a time-domain feature vector. Among them, the first feature data includes the maximum amplitude, root mean square amplitude, standard deviation of vibration velocity, and mean value of vibration velocity of the laser Doppler data;

[0081] After performing Fourier transform on the laser Doppler data and then analyzing it, obtain the second feature data of the laser Doppler data in the frequency dimension, and combine all the second feature data to generate a frequency-domain feature vector. Among them, the second feature data includes the main frequency, maximum value of power spectral density, and frequency bandwidth;

[0082] Combine the time-domain feature vector and the frequency-domain feature vector to generate a vibration feature vector.

[0083] Specifically, extract vibration features from the time domain and the frequency domain. The time-domain features are:

[0084] (1) Maximum amplitude : Reflects the maximum amplitude of vibration, indicating the maximum deformation of the structure at a certain instant, , where is the vibration velocity.

[0085] (2) Root mean square amplitude (RMS) : Used to describe the energy or amplitude size of vibration, , where is the time length of the vibration signal, is the velocity signal.

[0086] (3) Standard deviation : represents the volatility of the vibration signal, , where, is the mean value of the vibration signal.

[0087] The time-domain feature vector is: , where, is the maximum amplitude of the vibration, is the root mean square amplitude of the vibration, is the standard deviation of the vibration signal, is the mean value of the vibration signal.

[0088] The frequency-domain features are:

[0089] (1) The main frequency : Identifies the main frequency component of the vibration signal, usually corresponding to the modal frequency of the explosion-proof housing, , where, is the Fourier transform of the vibration signal.

[0090] (2) Power spectral density (PSD): , where, is the Fourier transform of the vibration signal.

[0091] (3) Modal frequency: By analyzing the spectrum, the natural modal frequency of the explosion-proof housing can be extracted, reflecting the natural response of the housing under excitation.

[0092] The frequency-domain feature vector is: , where, is the main frequency of the signal, is the maximum value of the power spectral density, is the frequency bandwidth (which can be obtained from spectrum analysis).

[0093] After combining the time-domain and frequency-domain features, a feature vector is obtained, which represents the response characteristics of the explosion-proof housing in terms of vibration.

[0094] In a specific embodiment, in step 4, the pressure feature vector obtained by feature extraction of the pressure data is:

[0095] Analyze the pressure data to obtain the maximum pressure, root mean square pressure, pressure change rate, and pressure pulse width, and combine the maximum pressure, root mean square pressure, pressure change rate, and pressure pulse width to generate the pressure feature vector.

[0096] Specifically, the pressure features include:

[0097] (1) Maximum pressure : Reflects the limit value of the local pressure and is directly related to the maximum degree of force on the housing, , where, is a pressure signal.

[0098] (2)Root Mean Square Pressure (RMS) : Represents the average intensity of the pressure, .

[0099] (3)Rate of change of pressure : Reflects the speed of pressure change, usually used to evaluate the degree of pressure fluctuation under excitation, .

[0100] (4)Pressure pulse width W P : By calculating the pulse width of the pressure wave, evaluate the duration of the pressure fluctuation. By setting a threshold (such as 10% of the signal maximum value), calculate the duration of the pressure signal exceeding this threshold, ;

[0101] Where:

[0102] ;

[0103] ;

[0104] ;

[0105] is the maximum value of the pressure signal;

[0106] The pressure feature vector is: , where, is the maximum value of the pressure, is the root mean square value of the pressure, is the rate of change of the pressure signal, is the pulse width of the pressure signal.

[0107] In a specific embodiment, step 5 includes:

[0108] Deep feature extraction is performed on the vibration feature vector and the pressure feature vector respectively through a neural network to obtain the first deep feature corresponding to the vibration feature vector and the second deep feature corresponding to the pressure feature vector;

[0109] The first deep feature and the second deep feature are concatenated or weighted and fused to generate a fused feature vector.

[0110] Deep features are extracted through a neural network, and the features of the laser Doppler data and the pressure data are input into a deep learning model for fusion. Specifically, two branch networks are designed, and the laser Doppler feature vector and the pressure feature vector , each branch network extracts deep features through a multi-layer fully connected network:

[0111]

[0112]

[0113] The two sets of deep features are fused by concatenation or weighted fusion to obtain the fused feature vector F deep :

[0114] , where is the first deep feature corresponding to the laser Doppler data, is the second deep feature corresponding to the pressure data.

[0115] In a specific embodiment, in step 6, a mechanical response analysis is performed on the explosion-proof housing:

[0116] The vibration displacement of each measurement point is obtained from the laser Doppler data, and the shock wave intensity of each measurement point is obtained from the pressure data. Based on the vibration displacement and the shock wave intensity, the dynamic displacement distribution on the surface of the explosion-proof housing is calculated by a numerical analysis method;

[0117] Through a frequency domain analysis method, combining the laser Doppler data and the pressure data, the response characteristics of the explosion-proof housing at different frequencies are obtained;

[0118] Based on the material properties, structural form, and experimental conditions of the explosion-proof housing, a mechanical model is selected. The pressure data is used to drive the mechanical model to obtain the mechanical response of the mechanical model. Subsequently, the laser Doppler data is used to verify and correct the mechanical model to determine the change trend of stress and strain, and a stress-strain model is constructed.

[0119] Specifically, according to the vibration displacement obtained from the laser Doppler data and the shock wave intensity in the pressure data, the dynamic displacement distribution on the surface of the explosion-proof housing is calculated by a numerical analysis method. These displacement data can reveal the deformation of the surface of the explosion-proof housing under the action of the explosion wave, especially the local maximum displacement and stress concentration areas.

[0120] Using frequency domain analysis, combining the vibration data and the pressure data, the response of the housing at different frequencies can be analyzed. Especially in the specific frequency band of the explosion wave, the displacement of the housing may vibrate significantly. By fusing the vibration data and the pressure data, the areas on the surface of the explosion-proof housing where the maximum displacement occurs can be identified. These areas are usually the key parts that are strongly impacted by the explosion wave and may be potential areas for cracks, damage, or fatigue.

[0121] The fused data can help establish a more accurate stress-strain model. The stress field and strain field distributions can be calculated on the surface and inside of the explosion-proof housing. The explosion wave information in the pressure data is used to drive the mechanical model of the housing, while the vibration data is used to determine the changing trends of stress and strain. By combining the analysis of vibration data and pressure data, the stress concentration areas can be accurately identified. Especially under a strong explosion shock, large stress concentrations may occur in local areas of the explosion-proof housing, and these areas are the parts of the housing most prone to structural failure or fatigue.

[0122] Through a detailed mechanical response analysis process, the laser Doppler data and pressure data are organically combined to deeply analyze the dynamic behavior of the explosion-proof housing under explosion shock, and a corresponding mechanical model is constructed. This analysis method can not only provide detailed mechanical performance parameters of the explosion-proof housing, but also provide a scientific basis for the design and optimization of explosion-proof equipment, thereby effectively improving the safety and reliability of coal mine explosion-proof equipment.

[0123] In a specific embodiment, the laser Doppler vibrometer is configured with a galvanometer control system. During the explosion experiment, the irradiation angle and position of the laser head are automatically adjusted, and the laser beam is controlled to move according to the scanning path.

[0124] Specifically, the laser head automatically adjusts the irradiation angle and position of the laser head through the galvanometer control system. The galvanometer controls the movement of the laser beam according to the scanning path, quickly responds and accurately projects the laser beam to the position of the predetermined measurement point, realizes accurate non-contact measurement at each virtual measurement point, and real-time collects vibration data. This method only needs to perform a single pressure resistance test on the explosion-proof housing to obtain the vibration data of all points on one side. The vibration data of all measurement points is collected and integrated into the data acquisition system to form a complete vibration response data set.

[0125] In a specific embodiment, the laser Doppler data and pressure data are denoised and calibrated before step 4.

[0126] Specifically, due to factors such as vibration and electromagnetic interference in the experimental environment, the laser Doppler data may contain noise, and filtering techniques (such as low-pass filtering, band-pass filtering, wavelet denoising, etc.) need to be used to remove the noise and retain the useful vibration signals. The pressure data may also be affected by environmental noise and requires corresponding denoising processing. Similar filtering techniques can be used to remove the noise and retain the real pressure change signals.

[0127] The laser Doppler data is calibrated to ensure the accuracy of the measured vibration displacement, velocity, and acceleration. The calibration process may include zero calibration, sensitivity calibration, and linearity calibration, etc. The pressure data is calibrated to ensure the accuracy of the measured pressure value. The calibration process may include zero calibration, range calibration, and linearity calibration, etc.

[0128] Those skilled in the art can clearly understand that for the convenience and conciseness of description, the specific working processes of the above-described systems, systems, and units can refer to the corresponding processes in the foregoing method embodiments, and will not be elaborated herein.

[0129] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application, in essence, or the part that contributes to the prior art, or all or part of this technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to enable a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present application. The foregoing storage medium includes: various media such as USB flash drives, mobile hard disks, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical discs that can store program codes.

[0130] As described above, the above embodiments are only used to illustrate the technical solutions of the present application, rather than to limit them; although the present application 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 recorded in the foregoing embodiments, or perform equivalent replacements for some of the technical features; and these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the various embodiments of the present application.

Claims

1. A dynamic detection method for explosion-proof equipment in coal mines based on laser Doppler and pressure data, characterized in that The method includes: Step 1: Set the measurement area and scanning scheme according to the geometric shape, working environment, and area to be measured of the explosion-proof housing. Among them, the scanning scheme includes the number of measurement points, distribution positions, and scanning paths of the measurement points; Step 2: Set pressure sensors at each measurement point, and arrange and install a laser Doppler vibrometer based on the measurement area and the scanning scheme; Step 3: Conduct an explosion experiment, and simultaneously collect the laser Doppler data and pressure data at each measurement point; Step 4: Extract features from the laser Doppler data to obtain a vibration feature vector, and extract features from the pressure data to obtain a pressure feature vector; Step 5: Input the vibration feature vector and the pressure feature vector into a deep learning model for fusion to obtain a fusion feature vector; Step 6: Input the fusion feature vector into a comprehensive evaluation model, output the performance score of the explosion-proof housing, compare the performance score with a set threshold, and determine the explosion-proof performance of the explosion-proof housing; In Step 6, it also includes performing a mechanical response analysis on the explosion-proof housing: Obtain the vibration displacement of each measurement point from the laser Doppler data, and obtain the shock wave intensity of each measurement point from the pressure data. Based on the vibration displacement and the shock wave intensity, calculate the dynamic displacement distribution on the surface of the explosion-proof housing by means of numerical analysis methods; By means of frequency domain analysis methods, combine the laser Doppler data and the pressure data to obtain the response characteristics of the explosion-proof housing at different frequencies; Select a mechanical model based on the material characteristics, structural form, and experimental conditions of the explosion-proof housing, use the pressure data to drive the mechanical model to obtain the mechanical response of the mechanical model, and then use the laser Doppler data to verify and correct the mechanical model to determine the change trend of stress and strain, and construct a stress-strain model.

2. The dynamic detection method for coal mine explosion-proof equipment based on laser Doppler and pressure data according to claim 1, characterized in that, In Step 1, the generation method of the scanning scheme is: Step 11: Analyze the geometric shape and force characteristics of the explosion-proof housing, and design the first measurement point distribution method and the second measurement point distribution method based on the analysis results; Step 12: Based on the layout method corresponding to each measurement point distribution method, arrange virtual measurement points in the measurement area; Step 13: Identify overlapping measurement points, delete the overlapping measurement points from the set of virtual measurement points corresponding to the first measurement point distribution method, and obtain the first set of virtual measurement points; Step 14: Generate a first scanning path based on the first set of virtual measurement points, generate a second scanning path based on the second set of virtual measurement points, and fuse the first scanning path and the second scanning path to generate a third scanning path, where the second set of virtual measurement points is the set of virtual measurement points corresponding to the second measurement point distribution method; Step 15: Conduct an explosion simulation experiment, collect vibration data at each virtual measurement point, analyze the vibration data, adjust the number and distribution of the virtual measurement points according to the analysis results, generate a new scanning path based on the adjusted results, repeat Step 15, and after multiple experimental verifications and adjustments, determine the number of the measurement points, the distribution positions, and the scanning path.

3. The dynamic detection method for coal mine explosion-proof equipment based on laser Doppler and pressure data according to claim 2, characterized in that, The distribution mode of the first measurement points is global scanning, and the distribution mode of the second measurement points is local scanning. Among them, the area of the local scanning includes the central area, the edge area, and the structural connection area, and the area shape is circular or rectangular; The layout modes corresponding to the global scanning and the central area are evenly spaced layouts, and the layout modes corresponding to the edge area and the structural connection area are dense layouts.

4. The dynamic detection method of coal mine explosion-proof equipment based on laser Doppler and pressure data according to claim 1, characterized in that, Before Step 4: Perform time synchronization and spatial synchronization processing on the laser Doppler data and the pressure data.

5. The dynamic detection method for coal mine explosion-proof equipment based on laser Doppler and pressure data according to claim 1, characterized in that, The laser Doppler data includes vibration displacement, vibration velocity, and vibration acceleration. In Step 4, the extraction of vibration feature vectors from the laser Doppler data is as follows: Analyze the laser Doppler data to obtain first feature data of the laser Doppler data in the time dimension, and combine all the first feature data to generate a time-domain feature vector. Among them, the first feature data includes the maximum amplitude of the laser Doppler data, the root mean square amplitude, the standard deviation of the vibration velocity, and the mean value of the vibration velocity; Analyze the laser Doppler data after Fourier transform to obtain second feature data of the laser Doppler data in the frequency dimension, and combine all the second feature data to generate a frequency-domain feature vector. Among them, the second feature data includes the main frequency, the maximum value of the power spectral density, and the frequency bandwidth; Combine the time-domain feature vector and the frequency-domain feature vector to generate the vibration feature vector.

6. The dynamic detection method of coal mine explosion-proof equipment based on laser Doppler and pressure data according to claim 1, characterized in that, In Step 4, the extraction of pressure feature vectors from the pressure data is as follows: Analyze the pressure data to obtain the maximum pressure, the root mean square pressure, the pressure change rate, and the pressure pulse width, and combine the maximum pressure, the root mean square pressure, the pressure change rate, and the pressure pulse width to generate the pressure feature vector.

7. The dynamic detection method of coal mine explosion-proof equipment based on laser Doppler and pressure data according to claim 1, wherein, Step 5 includes: Deeply extract features from the vibration feature vector and the pressure feature vector respectively through a neural network to obtain a first deep feature corresponding to the vibration feature vector and a second deep feature corresponding to the pressure feature vector; Stitch or weighted fuse the first deep feature and the second deep feature to generate the fused feature vector.

8. The dynamic detection method for coal mine explosion-proof equipment based on laser Doppler and pressure data according to claim 1, characterized in that, The laser Doppler vibrometer is equipped with a galvanometer control system. In the explosion experiment, automatically adjust the irradiation angle and position of the laser head, and control the laser beam to move according to the scanning path.

9. The dynamic detection method for coal mine explosion-proof equipment based on laser Doppler and pressure data according to claim 1, wherein Denoise and calibrate the laser Doppler data and the pressure data before Step 4.

Images