Blasting vibration monitoring data processing method and device based on edge computing

Abstract

The present invention discloses a blasting vibration monitoring data processing method and device based on edge computing, which relates to the technical field of blasting vibration monitoring. The blasting point monitoring range of the blasting site is divided into multiple edge monitoring units, and multiple edge data analysis units are respectively deployed at the geometric centers of the multiple edge monitoring units; the multiple vibration sensors in the edge monitoring units are respectively communicatively connected to the edge data analysis unit at the geometric center. The edge data analysis unit of the present invention realizes localized processing of data, reduces the delay of data transmission, and improves the ability of real-time monitoring and rapid response. The localized processing of the edge computing unit enables mutation analysis to be performed in real time. Once a mutation is detected, corresponding measures can be taken immediately to improve the rapid response capability. Through step-by-step feature extraction and comparison, the algorithm flow is clear, which is convenient for optimization and adjustment to adapt to different monitoring environments and needs.

Description

Technical Field

[0001] The present invention relates to the technical field of blasting vibration monitoring, and in particular to a blasting vibration monitoring data processing method and device based on edge computing. Background Art

[0002] Blasting operations are widely used in many engineering fields, including mining, tunnel construction, and road construction. However, the vibrations generated by blasting can have serious impacts on surrounding buildings, infrastructure, and the ecological environment. To effectively control the hazards of blasting vibration, accurate blasting vibration monitoring is crucial. With the continuous development of technology, blasting vibration monitoring technology is also continuously evolving;

[0003] Sensor Technology: Commonly used sensors include piezoelectric accelerometers and strain gauges. Piezoelectric accelerometers detect vibration by measuring the acceleration of a particle, offering high sensitivity and a wide frequency response range. Strain gauges are primarily used to measure the strain of structures caused by blasting. This measurement indirectly assesses the impact of blasting vibration on the structure. For example, in some large bridge demolition projects involving blasting, strain gauges are deployed at key locations to monitor changes in structural strain.

[0004] Data Acquisition System: The data acquisition system is responsible for converting analog signals collected by sensors into digital signals and performing preliminary data storage and transmission. Common data acquisition devices include data acquisition cards and portable data acquisition instruments. These devices have multi-channel acquisition capabilities and can simultaneously collect data from multiple sensors. Some advanced data acquisition instruments also support wireless transmission, facilitating remote data acquisition.

[0005] The main methods for analyzing blasting monitoring data are as follows:

[0006] The Fourier transform converts vibration signals from the time domain into the frequency domain, analyzes the signal's frequency components, and determines parameters such as the dominant frequency. For example, the Fourier transform can be used to analyze the energy distribution of different frequency components in a blasting vibration signal and determine the frequency range that may affect different structures.

[0007] Wavelet transforms can perform multi-resolution signal analysis and have good localization properties in both the time and frequency domains, making them suitable for analyzing non-stationary blasting vibration signals. For example, wavelet transforms can be used to extract mutation characteristics in blasting vibration signals and identify the start and end times of blasting vibrations.

[0008] The peak analysis method evaluates the hazard level of blasting vibration by calculating parameters such as the peak particle velocity of the vibration signal and comparing them with safety standards. In actual projects, according to the relevant provisions of the Blasting Safety Regulations, peak particle velocity is used as a key indicator to determine whether blasting vibration is safe.

[0009] Data transmission and storage pressure: In large-scale blasting monitoring scenarios, numerous sensors generate enormous amounts of data. Transmitting all data to remote servers for centralized processing and storage results in excessive network bandwidth usage, significant data transmission delays, and even possible data loss. For example, during blasting operations in large open-pit coal mines, hundreds of sensors simultaneously collect data, placing significant pressure on data transmission.

[0010] Lack of real-time performance: Due to the data transmission and centralized processing process, there is often a long delay between sensor data acquisition and analysis results, making it difficult to meet the requirements for real-time monitoring of blasting vibration and the implementation of timely control measures. For structures sensitive to vibration response, such as ancient buildings and operating precision equipment, this lack of real-time performance may result in the inability to promptly avoid vibration hazards.

[0011] Limited data analysis accuracy: In complex blasting environments, noise interference is severe, and traditional data analysis methods have limitations in removing noise and accurately extracting signal features. For example, blasting operations in urban environments are subject to numerous noise sources, such as electromagnetic interference, which can affect the accurate analysis of blasting vibration signals and lead to inaccurate assessments of blasting vibration hazards.

[0012] Poor system scalability: When adding new monitoring points or expanding the monitoring range, traditional centralized monitoring systems often require large-scale adjustments to the entire data transmission and processing architecture, which is costly and difficult to implement. This limits the flexible application of blasting vibration monitoring systems in projects of different scales. Summary of the Invention

[0013] In order to solve the above-mentioned technical problems of blasting vibration monitoring, the present invention provides a blasting vibration monitoring data processing method and device based on edge computing. The following technical solutions are adopted:

[0014] The blasting vibration monitoring data processing method based on edge computing includes the following steps:

[0015] Step 1: Divide the blasting point monitoring range of the blasting site into multiple edge monitoring units, and deploy multiple edge data analysis units at the geometric centers of the multiple edge monitoring units;

[0016] Step 2: respectively connect the plurality of vibration sensors in the edge monitoring unit to the edge data analysis unit at the geometric center for communication;

[0017] Step 3: Start blasting, and all vibration sensors monitor vibration data during the duration of the blasting impact;

[0018] Step 4: after the blasting impact duration ends, the multiple edge data analysis units respectively collect time-series-based vibration data packets from the multiple vibration sensors in the edge monitoring unit;

[0019] Step 5: The edge data analysis unit parses the time-series-based vibration data packets and performs edge computing analysis on the vibration data of each vibration sensor to obtain a vibration waveform.

[0020] Step 6: The edge data analysis unit extracts waveform features from the vibration waveform corresponding to the vibration data of each vibration sensor in the same edge monitoring unit, and performs mutation analysis based on the waveform features;

[0021] Step 7: If the difference between a feature value and the corresponding feature mean is greater than the set mutation detection threshold, or the first-order derivative of the feature waveform is greater than the mutation detection threshold, or the second-order derivative of the feature waveform is greater than the mutation detection threshold, then the feature value is determined to be a mutation value;

[0022] In step 8, the edge data analysis unit marks the identified mutation value and its corresponding waveform feature, extracts key feature information, and packages the mutation value and key information data into a mutation value data packet; the edge data analysis unit compresses and downsamples the normal vibration data that has not undergone mutation to form a normal vibration data packet;

[0023] Step 9: All edge data analysis units first transmit the mutation value data packets to the central server via the wireless communication network. After the central server receives the mutation value data packets of all edge data analysis units, it sends signals of the normal vibration data packets to all edge data analysis units respectively. All edge data analysis units then transmit the normal vibration data packets to the central server via the wireless communication network.

[0024] Step 10: The central server performs global vibration pattern recognition and anomaly detection on the received mutation value data packets, and performs statistical analysis on the common vibration data packets to generate trend graphs and distribution graphs of the vibration data.

[0025] By implementing the aforementioned technical solution and deploying edge data analysis units within edge monitoring units, localized data processing is achieved, reducing data transmission delays and improving real-time monitoring and rapid response capabilities. By using edge data analysis units for edge computing at the source of data generation, filtering out important information, extracting key features, and performing mutation analysis, abnormal events in blasting vibration can be more accurately identified, improving data reliability, significantly increasing overall data processing efficiency, and reducing the computational burden on central servers.

[0026] Compressing and downsampling common vibration data reduces storage and transmission requirements, optimizing resource utilization. The central server performs global vibration pattern recognition and anomaly detection, combined with local detail analysis by edge units, enabling a more comprehensive understanding of vibration data and generating more accurate statistical analysis results.

[0027] This method can flexibly divide edge monitoring units and deploy edge data analysis units according to the monitoring scope and needs, and can be easily expanded to larger monitoring areas or more complex monitoring scenarios.

[0028] The edge computing unit can adapt to various harsh environments and transmits data to the central server after the duration of the explosion, so wireless communication is not affected by the explosion.

[0029] Optionally, if the central server analyzes that multiple edge monitoring units exceeding a set number threshold have sudden change values at the same time, an alarm mechanism is triggered.

[0030] Optionally, the central server visualizes the analysis results in the form of charts, maps, etc., and displays the blasting vibration situation to the user in a visual manner.

[0031] By adopting the above technical solution, the detailed steps for the central server to visualize the analysis results are as follows: Ensure that the mutation value data packets and normal vibration data packets received from the edge data analysis unit are complete and accurate. Clean the data to remove invalid or erroneous data points.

[0032] All mutation value data packets and normal vibration data packets uploaded by edge data analysis units are integrated.

[0033] Sort data by timestamp to ensure the accuracy of time series data.

[0034] Perform statistical analysis on the integrated data, including calculating the mean, standard deviation, frequency distribution, etc. of the vibration data.

[0035] Trend Graph: Use a time series graph to show the trend of vibration data, such as a line graph or a line graph.

[0036] X-axis: time; Y-axis: vibration intensity or specific vibration characteristics; Distribution graph: Use a histogram or density graph to display the distribution of vibration data.

[0037] X-axis: vibration intensity or characteristic value;

[0038] Y-axis: frequency or probability density;

[0039] Mark the location of each edge monitoring unit on the map.

[0040] According to the intensity or mutation of vibration data, different colors or icons are used to display the vibration conditions of each monitoring unit on the map.

[0041] An interactive interface allows users to view detailed information about different monitoring units by clicking, zooming, and other operations. Filtering and search functions are provided, allowing users to view data based on time, location, vibration intensity, and other conditions.

[0042] Integrate charts and maps into reports or make them available to users via the web interface.

[0043] Provides a download function that allows users to download charts and raw data.

[0044] Through charts and maps, users can intuitively understand the blasting vibration situation and make decisions quickly. The graphical display of complex data lowers the threshold for users to understand the data.

[0045] Users can explore data in depth through an interactive interface to discover potential problems or patterns.

[0046] The integration of various charts and maps provides a comprehensive perspective of information, helping users gain a deeper understanding.

[0047] Through real-time updated charts and maps, users can monitor blasting vibrations in real time and respond to abnormal events promptly.

[0048] Optionally, physical protection is performed on multiple edge computing units separately. The physical protection is to place the edge data analysis unit in a pre-set monitoring pit and install an explosion-proof cover above the edge data analysis unit.

[0049] By adopting the above technical solution, in order to protect the edge computing unit from physical damage caused by blasting, the edge computing unit does not need to be exposed for monitoring, but only needs to receive the monitoring signal of the vibration sensor during the duration of the blasting. Therefore, a pre-set monitoring pit is used, and an explosion-proof cover is installed above the edge data analysis unit. Physical protection ensures that the edge data analysis unit can continuously and stably collect and transmit data during the blasting process, thereby ensuring the integrity and accuracy of the data.

[0050] Optionally, in step 2, the plurality of vibration sensors in the edge monitoring unit are communicatively connected to the edge data analysis unit at the geometric center via a communication line.

[0051] Optionally, the multiple vibration sensors are wireless vibration sensors. While the communication lines are connected, the multiple vibration sensors are connected to the edge data analysis unit at the geometric center based on the wireless communication network for wireless communication.

[0052] By adopting the above technical solution, vibration monitoring values of multiple vibration sensors are collected in wireless and wired ways at the same time, thereby ensuring the integrity of data collection.

[0053] Optionally, step 6 includes the following sub-steps:

[0054] Step 61: Extract and analyze vibration waveform features, where the vibration waveform features include maximum vibration value, minimum vibration value, mean, variance, kurtosis, and skewness;

[0055] Step 62, respectively calculating the first-order waveform derivative and the second-order waveform derivative;

[0056] Step 63, setting a mutation detection threshold;

[0057] In step 64 , the vibration waveform feature, the first-order waveform derivative, and the second-order waveform derivative are compared with the mutation detection threshold. If any value is greater than the vibration waveform feature, it is determined that a mutation exists in the corresponding vibration waveform.

[0058] By employing this technical solution, we extract multiple features of the vibration waveform (maximum, minimum, mean, variance, kurtosis, and skewness) and combine them with first- and second-order waveform derivatives to more comprehensively analyze vibration data, thereby improving the accuracy of mutation detection. The calculation of first- and second-order waveform derivatives helps identify subtle changes and trends in the vibration waveform, enabling the algorithm to more sensitively detect potential mutation points.

[0059] Setting a mutation detection threshold and comparing it with multiple vibration waveform features can effectively reduce false positives due to random noise or normal fluctuations.

[0060] The localized processing of edge computing units enables real-time mutation analysis. Once a mutation is detected, appropriate measures can be taken immediately, improving rapid response capabilities. Through step-by-step feature extraction and comparison, the algorithm flow is clear and easy to optimize and adjust to suit different monitoring environments and needs.

[0061] After mutation analysis is performed at the edge, information about detected mutations needs to be transmitted to the central server first, reducing the amount of data transmission for key data and avoiding network congestion.

[0062] By comprehensively considering multiple waveform features, the algorithm has strong adaptability to different types of vibration events and improves the robustness of the system.

[0063] Optionally, in step 6, the characteristic value of the vibration waveform is calculated:

[0064] Maximum vibration ; Minimum vibration ; mean ;variance ; Kurtosis ; Skewness ;in is the vibration value at time t, N is the total number of

[0065] The maximum vibration value The difference between the mean value of the maximum vibration value and the minimum vibration value is calculated. Calculate the difference between the mean and the minimum value of the vibration, and convert the kurtosis Calculate the difference with the mean of kurtosis and convert the skewness Calculate the difference with the mean of skewness. If any difference is greater than the mutation detection threshold, it is determined that the corresponding vibration waveform has a mutation. Otherwise, continue to calculate the first-order waveform derivative and the second-order waveform derivative;

[0066] First-order waveform derivative The calculation formula is: ;

[0067] in is the time interval; is the vibration value at time t+1;

[0068] Second-order waveform derivative The calculation formula is: ;

[0069] in is the first-order derivative of the waveform at time t+1.

[0070] A blasting vibration monitoring data processing device based on edge computing is used to implement a blasting vibration monitoring data processing method based on edge computing. The blasting vibration monitoring data processing device includes multiple groups of vibration sensors, multiple edge data analysis units, and a central server. The multiple groups of vibration sensors are evenly arranged at multiple edge monitoring units at the blasting site. The multiple edge data analysis units are respectively installed in monitoring pits pre-set at the physical centers of the multiple edge monitoring units, and explosion-proof covers are installed above the edge data analysis units.

[0071] Multiple vibration sensors in the same edge monitoring unit are respectively connected to the edge data analysis unit of the physical center of the edge monitoring unit by wired and wireless means, and the multiple edge data analysis units are respectively connected to the central server by wireless communication.

[0072] Optionally, the edge data analysis unit includes a wireless communication module, a multi-channel data collector, a memory and a data analysis chip. The multiple vibration sensors in the same edge monitoring unit are respectively wirelessly connected to the wireless communication module, and are respectively communicated with the signal input end of the multi-channel data collector using communication lines. The wireless communication module and the multi-channel data collector are respectively communicated with the memory, the data analysis chip is communicated with the memory, and the memory is wirelessly connected to the central server through the wireless communication module.

[0073] In summary, the present invention provides at least one of the following beneficial technical effects: The edge data analysis unit enables localized data processing, reduces data transmission delays, and improves real-time monitoring and rapid response capabilities. The localized processing of the edge computing unit enables mutation analysis to be performed in real time. Once a mutation is detected, appropriate measures can be taken immediately, improving rapid response capabilities. Through step-by-step feature extraction and comparison, the algorithm flow is clear and easy to optimize and adjust to suit different monitoring environments and needs.

[0074] After mutation analysis is performed at the edge, information about detected mutations needs to be transmitted to the central server first, reducing the amount of data transmission for key data and avoiding network congestion.

[0075] By comprehensively considering multiple waveform features, the algorithm has strong adaptability to different types of vibration events and improves the robustness of the system.

[0076] Compressing and downsampling common vibration data reduces storage and transmission requirements, optimizing resource utilization. The central server performs global vibration pattern recognition and anomaly detection, combined with local detail analysis by edge units, enabling a more comprehensive understanding of vibration data and generating more accurate statistical analysis results.

[0077] Edge monitoring units can be flexibly divided and edge data analysis units deployed according to the monitoring scope and needs, which can be easily expanded to larger monitoring areas or more complex monitoring scenarios. BRIEF DESCRIPTION OF THE DRAWINGS

[0078] Figure 1 1 is a flow chart of a blasting vibration monitoring data processing method based on edge computing according to the present invention;

[0079] Figure 2 It is a schematic diagram of the electrical component connection principle of the blasting vibration monitoring data processing device based on edge computing of the present invention.

[0080] Explanation of the accompanying drawings: 1. Edge data analysis unit; 11. Wireless communication module; 12. Multi-channel data collector; 13. Memory; 14. Data analysis chip; 2. Vibration sensor; 3. Central server. DETAILED DESCRIPTION

[0081] The present invention will be further described in detail below with reference to the accompanying drawings.

[0082] The embodiments of the present invention disclose a method and device for processing blasting vibration monitoring data based on edge computing.

[0083] Reference Figure 1 and Figure 2 ,Example 1, a blasting vibration monitoring data processing method based on edge computing, comprising the following steps:

[0084] Step 1: Divide the blasting point monitoring range of the blasting site into multiple edge monitoring units, and deploy multiple edge data analysis units 1 at the geometric centers of the multiple edge monitoring units;

[0085] Step 2: respectively connect the plurality of vibration sensors 2 in the edge monitoring unit to the edge data analysis unit 1 at the geometric center for communication;

[0086] Step 3: Start blasting, and all vibration sensors 2 monitor vibration data during the duration of the blasting impact;

[0087] Step 4: after the blasting impact duration ends, the multiple edge data analysis units 1 respectively collect time-series-based vibration data packets from the multiple vibration sensors 2 in the edge monitoring unit;

[0088] Step 5: The edge data analysis unit 1 parses the vibration data packet based on the time series, and performs edge computing analysis on the vibration data of each vibration sensor 2 to obtain a vibration waveform;

[0089] Step 6: The edge data analysis unit 1 extracts waveform features from the vibration waveform corresponding to the vibration data of each vibration sensor 2 in the same edge monitoring unit, and performs mutation analysis based on the waveform features;

[0090] Step 7: If the difference between a feature value and the corresponding feature mean is greater than the set mutation detection threshold, or the first-order derivative of the feature waveform is greater than the mutation detection threshold, or the second-order derivative of the feature waveform is greater than the mutation detection threshold, then the feature value is determined to be a mutation value;

[0091] In step 8, the edge data analysis unit 1 marks the identified mutation value and its corresponding waveform feature, extracts key feature information, and packages the mutation value and key information data into a mutation value data packet; the edge data analysis unit 1 compresses and downsamples the normal vibration data that has not undergone mutation to form a normal vibration data packet;

[0092] Step 9: All edge data analysis units 1 first transmit the mutation value data packets to the central server 3 via the wireless communication network. After the central server 3 receives the mutation value data packets of all edge data analysis units 1, it sends signals of the normal vibration data packets to all edge data analysis units 1 respectively. All edge data analysis units 1 then transmit the normal vibration data packets to the central server 3 via the wireless communication network.

[0093] In step 10, the central server 3 performs global vibration pattern recognition and anomaly detection on the received mutation value data packets, and the central server 3 performs statistical analysis on the common vibration data packets to generate trend graphs and distribution graphs of the vibration data.

[0094] By deploying the edge data analysis unit 1 at the edge monitoring unit, localized data processing is achieved, reducing data transmission delays and improving real-time monitoring and rapid response capabilities. Using the edge data analysis unit 1 for edge computing at the data source, filtering out important information, extracting key features, and performing mutation analysis, it can more accurately identify abnormal events in blasting vibrations, improve data reliability, significantly increase overall data processing efficiency, and reduce the computational burden on the central server 3.

[0095] Compressing and downsampling common vibration data reduces storage and transmission requirements, optimizing resource utilization. Central Server 3 performs global vibration pattern recognition and anomaly detection, combined with local detail analysis by edge units, enabling a more comprehensive understanding of vibration data and generating more accurate statistical analysis results.

[0096] The method can flexibly divide edge monitoring units and deploy edge data analysis units 1 according to the monitoring scope and needs, and can be easily expanded to larger monitoring areas or more complex monitoring scenarios.

[0097] The edge computing unit 1 can adapt to various harsh environments and transmits data to the central server 3 after the duration of the explosion impact, so wireless communication is not affected by the explosion.

[0098] In embodiment 2, if the central server 3 analyzes that a plurality of edge monitoring units 1 exceeding a set number threshold have mutation values at the same time, an alarm mechanism is triggered.

[0099] In embodiment 3, the central server 3 visualizes the analysis results in the form of charts, maps, etc., and displays the blasting vibration situation to the user in a visual manner.

[0100] The detailed steps for the central server 3 to visualize the analysis results are as follows: Ensure that the mutation value data packet and normal vibration data packet received from the edge data analysis unit 1 are complete and accurate. Clean the data to remove invalid or erroneous data points.

[0101] All mutation value data packets and normal vibration data packets uploaded by the edge data analysis unit 1 are integrated.

[0102] Sort data by timestamp to ensure the accuracy of time series data.

[0103] Perform statistical analysis on the integrated data, including calculating the mean, standard deviation, frequency distribution, etc. of the vibration data.

[0104] Trend Graph: Use a time series graph to show the trend of vibration data, such as a line graph or a line graph.

[0105] X-axis: time; Y-axis: vibration intensity or specific vibration characteristics; Distribution graph: Use a histogram or density graph to display the distribution of vibration data.

[0106] X-axis: vibration intensity or characteristic value;

[0107] Y-axis: frequency or probability density;

[0108] Mark the location of each edge monitoring unit on the map.

[0109] According to the intensity or mutation of vibration data, different colors or icons are used to display the vibration conditions of each monitoring unit on the map.

[0110] An interactive interface allows users to view detailed information about different monitoring units by clicking, zooming, and other operations. Filtering and search functions are provided, allowing users to view data based on time, location, vibration intensity, and other conditions.

[0111] Integrate charts and maps into reports or make them available to users via the web interface.

[0112] Provides a download function that allows users to download charts and raw data.

[0113] Through charts and maps, users can intuitively understand the blasting vibration situation and make decisions quickly. The graphical display of complex data lowers the threshold for users to understand the data.

[0114] Users can explore data in depth through an interactive interface to discover potential problems or patterns.

[0115] The integration of various charts and maps provides a comprehensive perspective of information, helping users gain a deeper understanding.

[0116] Through real-time updated charts and maps, users can monitor blasting vibrations in real time and respond to abnormal events promptly.

[0117] In Example 4, physical protection is performed on multiple edge computing units 1 respectively. The physical protection is to place the edge data analysis unit 1 in a pre-set monitoring pit and install an explosion-proof cover above the edge data analysis unit 1.

[0118] In order to protect the edge computing unit 1 from physical damage caused by blasting, the edge computing unit 1 does not need to be exposed for monitoring. It only needs to receive the monitoring signal of the vibration sensor 2 during the duration of the blasting. Therefore, a pre-set monitoring pit is used, and an explosion-proof cover is installed above the edge data analysis unit 1. Physical protection ensures that the edge data analysis unit 1 can continuously and stably collect and transmit data during the blasting process, ensuring the integrity and accuracy of the data.

[0119] In Example 5, in step 2, the plurality of vibration sensors 2 in the edge monitoring unit are communicatively connected to the edge data analysis unit 1 at the geometric center via a communication line.

[0120] In Example 6, the multiple vibration sensors 2 are wireless vibration sensors. While the communication lines are connected, the multiple vibration sensors 2 are connected to the edge data analysis unit 1 at the geometric center through wireless communication networks.

[0121] At the same time, wireless and wired methods are used to collect vibration monitoring values of multiple vibration sensors 2 to ensure the integrity of data collection.

[0122] In Example 7, step 6 includes the following sub-steps:

[0123] Step 61: Extract and analyze vibration waveform features, where the vibration waveform features include maximum vibration value, minimum vibration value, mean, variance, kurtosis, and skewness;

[0124] Step 62, respectively calculating the first-order waveform derivative and the second-order waveform derivative;

[0125] Step 63, setting a mutation detection threshold;

[0126] In step 64 , the vibration waveform feature, the first-order waveform derivative, and the second-order waveform derivative are compared with the mutation detection threshold. If any value is greater than the vibration waveform feature, it is determined that a mutation exists in the corresponding vibration waveform.

[0127] By extracting multiple features of the vibration waveform (maximum, minimum, mean, variance, kurtosis, and skewness) and combining them with first- and second-order waveform derivatives, a more comprehensive analysis of vibration data can be achieved, thereby improving the accuracy of sudden change detection. Calculating first- and second-order waveform derivatives helps identify subtle changes and trends in the vibration waveform, enabling the algorithm to more sensitively detect potential sudden change points.

[0128] Setting a mutation detection threshold and comparing it with multiple vibration waveform features can effectively reduce false positives due to random noise or normal fluctuations.

[0129] The localized processing of the edge computing unit 1 enables real-time mutation analysis. Once a mutation is detected, appropriate measures can be taken immediately, improving rapid response capabilities. Through step-by-step feature extraction and comparison, the algorithm flow is clear and easy to optimize and adjust to suit different monitoring environments and needs.

[0130] After mutation analysis is performed at the edge, information about detected mutations needs to be transmitted to the central server first, reducing the amount of data transmission for key data and avoiding network congestion.

[0131] By comprehensively considering multiple waveform features, the algorithm has strong adaptability to different types of vibration events and improves the robustness of the system.

[0132] In Example 8, in step 6, the characteristic value of the vibration waveform is calculated:

[0133] Maximum vibration ; Minimum vibration ; mean ;variance ; Kurtosis ; Skewness ;in is the vibration value at time t, N is the total number of

[0134] The maximum vibration value The difference between the mean value of the maximum vibration value and the minimum vibration value is calculated. Calculate the difference between the mean and the minimum value of the vibration, and convert the kurtosis Calculate the difference with the mean of kurtosis and convert the skewness Calculate the difference with the mean of skewness. If any difference is greater than the mutation detection threshold, it is determined that the corresponding vibration waveform has a mutation. Otherwise, continue to calculate the first-order waveform derivative and the second-order waveform derivative;

[0135] First-order waveform derivative The calculation formula is: ;

[0136] in is the time interval; is the vibration value at time t+1;

[0137] Second-order waveform derivative The calculation formula is: ;

[0138] in is the first-order derivative of the waveform at time t+1.

[0139] Example 9, a blasting vibration monitoring data processing device based on edge computing, is used to implement a blasting vibration monitoring data processing method based on edge computing. The blasting vibration monitoring data processing device includes multiple groups of vibration sensors 2, multiple edge data analysis units 1, and a central server 3. The multiple groups of vibration sensors 2 are evenly arranged at multiple edge monitoring units at the blasting site, and the multiple edge data analysis units 1 are respectively installed in monitoring pits pre-set at the physical centers of the multiple edge monitoring units, and explosion-proof covers are installed above the edge data analysis units 1.

[0140] The multiple vibration sensors 2 in the same edge monitoring unit are respectively connected to the edge data analysis unit 1 of the physical center of the edge monitoring unit by wired and wireless means, and the multiple edge data analysis units 1 are respectively connected to the central server 3 by wireless communication.

[0141] Example 10, the edge data analysis unit 1 includes a wireless communication module 11, a multi-channel data collector 12, a memory 13 and a data analysis chip 14. The multiple vibration sensors 2 in the same edge monitoring unit are respectively connected to the wireless communication module 11 wirelessly, and are respectively connected to the signal input end of the multi-channel data collector 12 using communication lines. The wireless communication module 11 and the multi-channel data collector 12 are respectively connected to the memory 13, and the data analysis chip 14 is connected to the memory 13. The memory 13 is connected to the central server 3 through the wireless communication module 11.

[0142] The above are all preferred embodiments of the present invention, and are not intended to limit the scope of protection of the present invention. Therefore, any equivalent changes made based on the structure, shape, and principle of the present invention should be included in the scope of protection of the present invention.

Claims

1. A blasting vibration monitoring data processing method based on edge computing is characterized by: The following steps are involved: Step 1: Divide the blasting point monitoring range of the blasting site into multiple edge monitoring units, and deploy multiple edge data analysis units (1) at the geometric centers of the multiple edge monitoring units. Step 2, respectively connecting the plurality of vibration sensors (2) in the edge monitoring unit to the edge data analysis unit (1) at the geometric center for communication; Step 3, blasting is started, and all vibration sensors (2) monitor vibration data during the duration of the blasting impact; Step 4, after the blasting impact duration ends, the plurality of edge data analysis units (1) respectively collect time-series-based vibration data packets from the plurality of vibration sensors (2) in the edge monitoring unit; Step 5, the edge data analysis unit (1) parses the vibration data packet based on the time sequence, and performs edge computing analysis on the vibration data of each vibration sensor (2) to obtain a vibration waveform; Step 6, the edge data analysis unit (1) extracts waveform features from the vibration waveform corresponding to the vibration data of each vibration sensor (2) in the same edge monitoring unit, and performs mutation analysis based on the waveform features; Step 7: If the difference between a feature value and the corresponding feature mean is greater than the set mutation detection threshold, or the first-order derivative of the feature waveform is greater than the mutation detection threshold, or the second-order derivative of the feature waveform is greater than the mutation detection threshold, then the feature value is determined to be a mutation value; Step 8, the edge data analysis unit (1) marks the identified mutation value and its corresponding waveform feature, extracts the key feature information, and packages the mutation value and key information data to form a mutation value data packet; the edge data analysis unit (1) performs data compression and downsampling processing on the normal vibration data that has not undergone mutation to form a normal vibration data packet; Step 9: All edge data analysis units (1) first transmit the mutation value data packets to the central server (3) via the wireless communication network. After the central server (3) receives the mutation value data packets from all edge data analysis units (1), it sends signals of the normal vibration data packets to all edge data analysis units (1). All edge data analysis units (1) then transmit the normal vibration data packets to the central server (3) via the wireless communication network. In step 10, the central server (3) performs global vibration pattern recognition and anomaly detection on the received mutation value data packets, and the central server (3) performs statistical analysis on the common vibration data packets to generate trend graphs and distribution graphs of the vibration data.

2. The blasting vibration monitoring data processing method based on edge computing according to claim 1 is characterized in that: If the central server (3) analyzes that multiple edge monitoring units (1) that exceed a set number threshold have mutation values at the same time, an alarm mechanism is triggered.

3. The blasting vibration monitoring data processing method based on edge computing according to claim 2 is characterized in that: The central server (3) visualizes the analysis results in the form of charts and maps, and visually displays the blasting vibration situation to the user.

4. The method for processing blasting vibration monitoring data based on edge computing according to claim 1, characterized in that: Physical protection is performed on each of the plurality of edge computing units (1). The physical protection is to place the edge data analysis unit (1) in a pre-set monitoring pit and to install an explosion-proof cover above the edge data analysis unit (1).

5. The method for processing blasting vibration monitoring data based on edge computing according to claim 4 is characterized in that: In step 2, a plurality of vibration sensors (2) in the edge monitoring unit are communicatively connected to the edge data analysis unit (1) at the geometric center via a communication line.

6. The method for processing blasting vibration monitoring data based on edge computing according to claim 5, characterized in that: The plurality of vibration sensors (2) are wireless vibration sensors. While the communication lines are communicating, the plurality of vibration sensors (2) are wirelessly connected to the edge data analysis unit (1) at the geometric center based on the wireless communication network.

7. The method for processing blasting vibration monitoring data based on edge computing according to claim 1, characterized in that: Step 6 includes the following sub-steps: Step 61: Extract and analyze vibration waveform features, where the vibration waveform features include maximum vibration value, minimum vibration value, mean, variance, kurtosis, and skewness; Step 62, respectively calculating the first-order waveform derivative and the second-order waveform derivative; Step 63, setting a mutation detection threshold; In step 64, the difference between the characteristic value and the corresponding characteristic mean, the first-order waveform derivative, and the second-order waveform derivative are compared with the mutation detection threshold. If any value is greater than the mutation detection threshold, it is determined that the corresponding vibration waveform has a mutation.

8. The method for processing blasting vibration monitoring data based on edge computing according to claim 5, characterized in that: In step 6, the characteristic value of the vibration waveform is calculated: Maximum vibration ; Minimum vibration ; mean ;variance ; Kurtosis ; Skewness ;in is the vibration value at time t, N is the total number of The maximum vibration value The difference between the mean value of the maximum vibration value and the minimum vibration value is calculated. Calculate the difference between the mean and the minimum value of the vibration, and convert the kurtosis Calculate the difference with the mean of kurtosis and convert the skewness Calculate the difference with the mean of skewness. If any difference is greater than the mutation detection threshold, it is determined that the corresponding vibration waveform has a mutation. Otherwise, continue to calculate the first-order waveform derivative and the second-order waveform derivative; First-order waveform derivative The calculation formula is: ; in is the time interval; is the vibration value at time t+1; Second-order waveform derivative The calculation formula is: ; in is the first-order derivative of the waveform at time t+1.

9. A blasting vibration monitoring data processing device based on edge computing, characterized by: A blasting vibration monitoring data processing method based on edge computing for implementing any one of claims 1 to 8, wherein the blasting vibration monitoring data processing device comprises a plurality of groups of vibration sensors (2), a plurality of edge data analysis units (1) and a central server (3), wherein the plurality of groups of vibration sensors (2) are evenly arranged at a plurality of edge monitoring units at a blasting site, the plurality of edge data analysis units (1) are respectively installed in monitoring pits pre-set at the physical centers of the plurality of edge monitoring units, and explosion-proof covers are installed above the edge data analysis units (1); Multiple vibration sensors (2) in the same edge monitoring unit are respectively connected to an edge data analysis unit (1) in a physical center of the edge monitoring unit in a wired and wireless manner, and the multiple edge data analysis units (1) are respectively connected to a central server (3) in wireless communication.

10. The blasting vibration monitoring data processing device based on edge computing according to claim 9, characterized in that: The edge data analysis unit (1) comprises a wireless communication module (11), a multi-channel data collector (12), a memory (13) and a data analysis chip (14); the plurality of vibration sensors (2) in the same edge monitoring unit are respectively wirelessly connected to the wireless communication module (11) and are respectively communicatively connected to the signal input end of the multi-channel data collector (12) using a communication line; the wireless communication module (11) and the multi-channel data collector (12) are respectively communicatively connected to the memory (13); the data analysis chip (14) is communicatively connected to the memory (13); and the memory (13) is wirelessly connected to the central server (3) via the wireless communication module (11).

Images