A communication fault sniffing method and system based on the Internet of Things

Abstract

The present application provides a communication fault sniffing method and system based on the Internet of Things. Among them, a status update signal from an Internet of Things device is received; a potential fault source is identified based on an abnormal indicator in the status update signal; a distributed sensor network is used to implement real-time monitoring of the potential fault source, and a fluctuation signal related to the potential fault source is captured; the time series characteristics of the fluctuation signal are analyzed to determine whether there is an abnormal fluctuation that meets the preset fault mode; when it is detected that the abnormal fluctuation meets the preset fault mode, a fault alarm is triggered and the location information of the potential fault source and the specific parameters of the abnormal fluctuation are recorded. Through an in-depth analysis of the time series characteristics of the fluctuation signal, the present application can capture weak early fault signals, which helps to discover and handle faults in advance, thereby effectively improving the subsequent detection accuracy.

Description

Technical Field

[0001] The embodiments of the present application relate to the field of fault detection technology, and in particular to a communication fault sniffing method and system based on the Internet of Things. Background Art

[0002] With the rapid development of IoT technology, a large number of smart devices are widely used in industrial production, smart homes, urban management and public services. These devices are connected through the network to achieve functions such as remote monitoring, data transmission and automatic control. However, IoT devices may encounter various communication failures during operation, such as signal loss and data transmission errors. These problems will not only affect the normal operation of the equipment, but may also cause serious safety accidents. Therefore, effective monitoring and early warning of communication failures of IoT devices has become one of the important requirements to ensure the stable operation of IoT systems.

[0003] At present, the detection methods for IoT device communication failures mainly include rule-based fault detection, statistical analysis-based fault detection, and machine learning-based fault detection. Rule-based fault detection mainly relies on pre-set fault rules to determine whether the device has a fault, while the statistical analysis-based method identifies anomalies by analyzing the statistical characteristics of the device operation data. Machine learning-based fault detection uses historical fault data to train models to automatically identify new faults.

[0004] Although the above methods can detect communication failures of IoT devices to a certain extent, there are still some shortcomings. First, rule-based methods are difficult to deal with complex failure modes, especially when the failure modes are constantly changing, and the update of rules often lags behind actual needs. Secondly, although the statistical analysis-based method can find some obvious anomalies, it has limited ability to identify weak and early fault signals. Finally, although the machine learning-based method has high accuracy, it requires a large amount of labeled data for training, and the generalization ability and real-time performance of the model still need to be improved. In other words, the current conventional communication failure detection methods all have the problem of poor detection accuracy. Summary of the invention

[0005] The embodiments of the present application provide a communication fault sniffing method and system based on the Internet of Things, which are used to solve the problem of poor detection accuracy of communication fault detection methods in the prior art.

[0006] In a first aspect, an embodiment of the present application provides a communication fault sniffing method based on the Internet of Things, including:

[0007] Receive status update signals from IoT devices;

[0008] identifying a potential fault source based on an abnormality indicator in the status update signal;

[0009] Using a distributed sensor network to monitor the potential fault source in real time and capture fluctuation signals related to the potential fault source;

[0010] Analyze the time series characteristics of the fluctuation signal to determine whether there is an abnormal fluctuation that meets the preset fault mode;

[0011] When it is detected that the abnormal fluctuation meets the preset fault mode, a fault alarm is triggered and the location information of the potential fault source and the specific parameters of the abnormal fluctuation are recorded.

[0012] In a second aspect, an embodiment of the present application provides a communication fault sniffing system based on the Internet of Things, including:

[0013] A receiving module, used for receiving a status update signal from an IoT device;

[0014] an identification module, configured to identify a potential fault source based on an abnormality indicator in the status update signal;

[0015] A capture module, used to implement real-time monitoring of the potential fault source using a distributed sensor network, and capture fluctuation signals related to the potential fault source;

[0016] A determination module, used to analyze the time series characteristics of the fluctuation signal and determine whether there is an abnormal fluctuation that meets a preset fault mode;

[0017] The processing module is used to trigger a fault alarm and record the location information of the potential fault source and the specific parameters of the abnormal fluctuation when it is detected that the abnormal fluctuation meets the preset fault mode.

[0018] In a third aspect, an embodiment of the present application provides a computing device, comprising a processing component and a storage component; the storage component stores one or more computer instructions; the one or more computer instructions are used to be called and executed by the processing component to implement a communication fault sniffing method based on the Internet of Things as described in the first aspect above.

[0019] In a fourth aspect, an embodiment of the present application provides a computer storage medium storing a computer program. When the computer program is executed by a computer, a communication fault sniffing method based on the Internet of Things as described in the first aspect is implemented.

[0020] In the embodiment of the present application, by receiving the status update signal from the IoT device, identifying the potential fault source according to the abnormal indicator, using the distributed sensor network to monitor the fluctuation signal of the potential fault source in real time, and by analyzing the time series characteristics of the fluctuation signal to determine whether there is an abnormal fluctuation that meets the preset fault mode. When an abnormal fluctuation is detected, a fault alarm is triggered and relevant information is recorded.

[0021] The main beneficial effects of this method are as follows:

[0022] Improved accuracy and real-time performance of fault detection: By combining the analysis of status update signals and fluctuation signals, this method can more accurately identify potential fault sources and achieve real-time monitoring and rapid response, thereby improving the efficiency and reliability of fault detection.

[0023] Enhanced recognition of complex fault modes: Through in-depth analysis of the time series characteristics of fluctuation signals, this method can capture weak early fault signals, which helps to detect and handle faults in advance and avoid further deterioration of faults.

[0024] Reduced reliance on large amounts of labeled data: By introducing multi-dimensional feature vectors and pre-trained pattern recognition models, this method can achieve higher detection performance with less training data, reducing the workload and cost of data preparation.

[0025] Provides detailed fault diagnosis information: This method can not only trigger a fault alarm, but also record the location information of potential fault sources and specific parameters of abnormal fluctuations, providing an important reference for subsequent fault analysis and maintenance.

[0026] These and other aspects of the present application will become more clearly understood in the description of the following embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0027] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, a brief introduction will be given below to the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0028] Figure 1 A flow chart of a communication fault sniffing method based on the Internet of Things provided in an embodiment of the present application;

[0029] Figure 2 A schematic diagram of the structure of a communication fault sniffing system based on the Internet of Things provided in an embodiment of the present application;

[0030] Figure 3A schematic diagram of the structure of a computing device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0031] In order to enable those skilled in the art to better understand the solution of the present application, the technical solution in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application.

[0032] In some of the processes described in the specification and claims of this application and the above-mentioned figures, multiple operations that appear in a specific order are included, but it should be clearly understood that these operations may not be executed in the order in which they appear in this article or executed in parallel. The serial numbers of the operations, such as 101, 102, etc., are only used to distinguish between different operations, and the serial numbers themselves do not represent any execution order. In addition, these processes may include more or fewer operations, and these operations may be executed in sequence or in parallel. It should be noted that the descriptions of "first", "second", etc. in this article are used to distinguish different messages, devices, modules, etc., do not represent the order of precedence, and do not limit the "first" and "second" to be different types.

[0033] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative work are within the scope of protection of this application.

[0034] Figure 1 A flow chart of a communication fault sniffing method based on the Internet of Things is provided for an embodiment of the present application, such as Figure 1 As shown, the method includes:

[0035] 101. Receive a status update signal from an IoT device;

[0036] This step involves collecting data from various physical or virtual devices connected to the Internet, such as machines, sensors, etc. This data can include key indicators such as the operating status of the equipment, temperature, humidity, pressure, etc.

[0037] In the embodiment of the present application, it is assumed that in an industrial environment, multiple temperature and vibration sensors are installed to monitor the working status of a large motor. These sensors periodically (for example, once a minute) send their readings to a central server, including the current temperature value and vibration level.

[0038] 102. Identify a potential fault source according to an abnormality indicator in the status update signal;

[0039] In this step, by analyzing the received data, the system is able to identify values ​​that are not consistent with the normal operating range. These outliers may be early signs of impending equipment failure.

[0040] In the embodiment of the present application, continuing with the above-mentioned motor as an example, if the system detects that the temperature in a specific area suddenly rises, or the vibration level exceeds a preset safety threshold, then the area may become a potential failure point.

[0041] Optionally, the "identifying potential fault sources based on the abnormal indicators in the status update signal" in step 102 includes: parsing the metadata in the status update signal to extract key indicators reflecting the operating status of the Internet of Things device; comparing the key indicators with a preset normal operating range to identify abnormal indicators that exceed the range; and mapping the abnormal indicators to a predefined equipment failure model to determine the potential fault source corresponding to the abnormal indicators.

[0042] Metadata: Data that describes data and is used to provide detailed information about a dataset, such as creation date, author, format, etc.

[0043] Key indicators: refers to important parameters that can effectively reflect the operating status of the equipment, such as temperature, pressure, current, etc.

[0044] Normal Operating Range: A set of parameter ranges within which a device is considered to operate normally, based on specifications provided by the device manufacturer or historical operating data.

[0045] Equipment failure model: A mathematical model based on historical failure data, used to predict the possible types of equipment failures and their corresponding performance indicators.

[0046] First, separate the part containing information such as device identification and collection time from the received status update signal. Select appropriate parameters as monitoring objects according to the device type and application scenario. For example, for motors, the focus is on temperature and vibration level. Secondly, determine the safety threshold of each key indicator based on the equipment manual or historical data. Compare the key indicator values ​​collected in real time with the preset normal operating range, and those outside the range are considered abnormal. Finally, use historical fault cases to train the machine learning model so that it can infer possible causes of faults based on abnormal indicators. Input the identified abnormal indicators into the fault model to obtain the most likely fault source and its severity assessment.

[0047] In the embodiment of the present application, it is assumed that an industrial motor is being monitored, and the motor is equipped with a temperature sensor and a vibration sensor for real-time monitoring of its working status. The system operates according to the following steps:

[0048] First, every time a status update signal is received, the metadata contained therein, such as device ID, acquisition timestamp, etc., is parsed. Secondly, the readings of the temperature sensor and vibration sensor are extracted. It is known that when the motor is working normally, the temperature should not exceed 80°C and the vibration amplitude should not exceed 0.5mm / s. Suppose a reading shows that the temperature is 90°C and the vibration amplitude is 0.6mm / s, both of which are beyond the normal range. The above abnormal indicators are input into the pre-trained fault model. Based on historical data, the model understands that when the temperature and vibration increase at the same time, it is usually due to damage to the bearing inside the motor. It is finally determined that there may be a problem with the motor's bearing, and the corresponding alarm is triggered to notify the maintenance personnel to check.

[0049] By adopting the above solution, not only the accuracy of anomaly detection is improved and the false alarm rate is reduced, but also the source of the fault can be quickly located, providing a guarantee for taking timely measures to avoid greater losses. In addition, this method also has strong adaptability and scalability, and is suitable for various types of equipment and different application scenarios.

[0050] 103. Using a distributed sensor network to implement real-time monitoring of the potential fault source and capture fluctuation signals related to the potential fault source;

[0051] In this step, once the potential source of the fault has been identified, more sensor resources can be deployed to observe the area in more detail to obtain more information about the nature of the problem.

[0052] In an embodiment of the present application, after an abnormality is found in a part of the motor, the sensor density in the area can be increased, such as adding more temperature and vibration sensors, so as to more accurately capture subtle changes in the area.

[0053] Optionally, in step 103, "using a distributed sensor network to implement real-time monitoring of the potential fault source and capture fluctuation signals related to the potential fault source" includes: collecting physical quantity signals related to the potential fault source in real time through multiple sensor nodes deployed in an Internet of Things environment; processing the physical quantity signals to extract characteristic frequency components of the potential fault source; and continuously monitoring the characteristic frequency components to identify fluctuation signals occurring near the potential fault source.

[0054] Distributed sensor network: A network composed of multiple sensor nodes that are distributed in the monitoring area and can work independently or collaboratively to achieve large-scale, multi-dimensional data collection.

[0055] Physical quantity signal: refers to the measurement results of various physical quantities that can reflect the operating status of the equipment, such as temperature, vibration, sound, etc.

[0056] Characteristic frequency components: In fault signals, certain specific frequency components are often closely related to specific types of faults. These frequencies are called characteristic frequencies.

[0057] Fluctuation signal: refers to the part of the physical quantity signal that changes over time, which is usually related to the working status of the equipment or the development of a fault.

[0058] First, add multiple sensor nodes around the potential fault source to ensure that all key parts of the fault source are covered. These sensor nodes continuously collect physical quantity signals such as temperature and vibration at a high frequency (such as 1,000 times per second) and transmit the data back to the central processing unit. Secondly, remove noise and perform filtering and other operations to ensure signal quality. Use spectrum analysis techniques such as Fourier transform (FFT) to extract characteristic frequency components from the signal. For example, motor bearing failures are often accompanied by vibrations of a specific frequency. Finally, set an alarm threshold to continuously monitor the characteristic frequency components. When the amplitude of the characteristic frequency component exceeds the preset threshold, it is considered that a fluctuation signal appears, indicating that the fault may be intensifying.

[0059] In the embodiment of the present application, it is assumed that in a factory, a large motor is initially judged to have a potential fault of bearing wear. In order to further verify this judgment, the following steps are taken:

[0060] Multiple vibration sensors and temperature sensors are installed at key parts such as the front and rear end covers and bearing seats of the motor. These sensors collect vibration and temperature data at a frequency of 1000 times per second and send the data to the central processing unit through the wireless communication module. The central processing unit preprocesses the received signal to remove background noise. Then the FFT algorithm is used to perform spectrum analysis on the vibration signal to extract the characteristic frequency components. Assuming that the motor speed is 1500rpm, according to experience, bearing faults usually produce significant vibrations at multiples such as 1x (i.e. 25Hz) and 2x (50Hz). The amplitude thresholds of the 25Hz and 50Hz frequency components are set to 0.1mm / s and 0.2mm / s respectively. If the amplitudes of these two frequency components exceed the thresholds at any time, it is considered that a fluctuation signal has occurred, and the system will automatically record the data at that time and trigger an alarm.

[0061] By adopting the above solution, not only can potential fault sources be monitored more comprehensively and accurately, but also the trend of fault development can be discovered in time, providing a scientific basis for taking preventive measures in advance. In addition, the design of the distributed sensor network makes the system more flexible and reliable, and can operate stably in complex and changing industrial environments. This method effectively solves the problems of limited monitoring range and high false alarm rate in traditional methods, and significantly improves the efficiency and safety of equipment maintenance.

[0062] 104. Analyze the time series characteristics of the fluctuation signal to determine whether there is an abnormal fluctuation that meets the preset fault mode;

[0063] In this step, the collected data is analyzed in depth by applying statistical methods or machine learning algorithms to identify whether there are signatures of known failure modes.

[0064] In the embodiment of the present application, a machine learning model is constructed using historical failure cases, which can predict whether a specific type of failure will occur based on the input temperature and vibration data. For example, if the model detects that a specific frequency component related to bearing wear appears in the vibration pattern, it is considered that there may be a risk of such a failure.

[0065] Optionally, the "analyzing the time series characteristics of the fluctuation signal to determine whether there is an abnormal fluctuation that meets the preset fault mode" in step 104 includes: obtaining time series data of the fluctuation signal; extracting short-term trends in the time series data to identify signal change characteristics in the short term; calculating key statistics of the time series data, including mean, variance, skewness and kurtosis, and the key statistics are used to characterize the distribution characteristics of the time series data; constructing a multidimensional feature vector based on the signal change characteristics and key statistics of the time series data; inputting the multidimensional feature vector into a pre-trained pattern recognition model to output the fluctuation signal corresponding to the multidimensional feature vector through the pattern recognition model, wherein the pattern recognition model has learned the fluctuation signal characteristics under multiple fault modes; determining whether the fluctuation signal meets the probability estimate of the preset fault mode and whether it exceeds the predetermined threshold of the probability estimate, and if so, determining that the fluctuation signal has abnormal fluctuations that meet the preset fault mode.

[0066] Time series data: A series of data points arranged in chronological order, reflecting the change of a variable over time.

[0067] Short-term trend: The trend changes shown by time series data within a shorter time window.

[0068] Key statistics: statistical indicators used to describe the distribution characteristics of time series data, including mean, variance, skewness, and kurtosis.

[0069] Multidimensional feature vector: combines multiple features into a vector to represent the comprehensive characteristics of time series data.

[0070] Pattern recognition model: A machine learning model that learns the characteristics of fluctuation signals under different types of fault modes through training and can classify and identify new fluctuation signals.

[0071] Probability estimation: A value output by a pattern recognition model that indicates the likelihood that the input fluctuation signal belongs to a certain fault mode.

[0072] Predetermined threshold: A pre-set critical value used to determine whether the fluctuation signal meets the preset fault mode.

[0073] First, the fluctuation signal is collected from the sensor node in real time and recorded in chronological order to form time series data. Secondly, the short-term trend in the time series data is extracted using techniques such as the moving average method or the sliding window method. The rise, fall, mutation and other characteristics in the short-term trend are analyzed for the subsequent construction of feature vectors.

[0074] Mean: The average value of time series data, indicating the center position of the data.

[0075] ;

[0076] Variance: The degree of dispersion of time series data, indicating the fluctuation of data.

[0077] ;

[0078] Skewness: The asymmetry of time series data, indicating the degree of skewness of the data distribution.

[0079] ;

[0080] Kurtosis: The degree of sharpness of time series data, indicating the peak state of the data distribution.

[0081] ;

[0082] Based on the signal change characteristics and key statistics of the time series data, a multidimensional feature vector is constructed.

[0083] Feature vector construction: Short-term trend characteristics and key statistics are combined into a multi-dimensional feature vector for input into the pattern recognition model.

[0084] ;

[0085] in, Indicates short-term trend characteristics.

[0086] Further use of machine learning models trained with historical fault data, such as support vector machines (SVM), random forests, etc. The model outputs a probability estimate based on the input multi-dimensional feature vector, indicating the possibility that the fluctuation signal belongs to a certain fault mode.

[0087] Probability estimate: The probability value output by a pattern recognition model.

[0088] Predetermined threshold: a pre-set critical value, such as 0.7.

[0089] Fault determination: If the probability estimate is greater than a predetermined threshold, it is determined that the fluctuation signal has abnormal fluctuations that conform to a preset fault mode.

[0090] In the embodiment of the present application, it is assumed that the vibration signal of a large motor is monitored in an industrial environment, and it is preliminarily determined that the motor may have a potential fault of bearing wear. In order to further confirm this judgment, the following steps are taken:

[0091] First, the vibration data is collected from the vibration sensor node of the motor at a frequency of 1000 times per second to form time series data Using the sliding window method, the window size is 100 data points, and the average value in each window is calculated to obtain the short-term trend. . Assume that the short-term trend over a certain period of time is as follows:

[0092] ;

[0093] Calculate key statistics:

[0094] Mean :

[0095] ;

[0096] variance :

[0097] ;

[0098] Skewness:

[0099] ;

[0100] Kurtosis:

[0101] ;

[0102] Construct multidimensional feature vector:

[0103] ;

[0104] Finally, the multidimensional feature vector The input is fed into the pre-trained random forest model, and the model outputs a probability estimate of 0.85. The predetermined threshold is set to 0.7 because Therefore, it is determined that the fluctuation signal has abnormal fluctuations that meet the preset fault mode, that is, the motor bearings may indeed have wear problems.

[0105] By adopting the above scheme, not only can the time series characteristics of the fluctuation signal be analyzed more comprehensively and accurately, but also the background noise and the real fault signal can be effectively distinguished through the multi-dimensional feature vector and pattern recognition model, thus improving the accuracy and reliability of fault diagnosis. This method solves the limitations of the traditional threshold judgment method, can timely detect potential faults in complex and changing industrial environments, provides a scientific basis for equipment maintenance, and significantly improves the safety and stability of equipment operation.

[0106] 105. When it is detected that the abnormal fluctuation meets the preset fault mode, a fault alarm is triggered and the location information of the potential fault source and the specific parameters of the abnormal fluctuation are recorded.

[0107] In this step, once the risk of failure is confirmed, the system will automatically issue an alarm to notify relevant personnel and save all relevant data for subsequent analysis.

[0108] In an embodiment of the present application, when the above-mentioned machine learning model confirms that there is a risk of bearing wear, the system will immediately send an email and text message to the maintenance team, and record the specific location of the fault (for example, "the left bearing of motor A"), the specific parameters of the abnormal fluctuation (such as the vibration intensity at a specific frequency), and the time of occurrence in the database.

[0109] Optionally, extracting the short-term trend from the time series data and identifying the signal change characteristics within a short period include: extracting the short-term trend from the time series data; and calculating a variety of signal change characteristics including high-order differences, autocorrelation functions, volatility and peak frequency based on the short-term trend.

[0110] Short-term trend: refers to the changing trend of time series data within a short period of time (such as a few hours to a few days), which reflects the dynamic changes of the data within this period of time.

[0111] Higher-order differences: By calculating the differences between time series data points to eliminate trend components, higher-order differences can further remove seasonal components in the data and help identify deeper data structures.

[0112] Autocorrelation function: measures the correlation between data at different time points in a time series and is used to identify periodic patterns in the data.

[0113] Volatility: A measure of the degree of change in time series data, often used to assess risk levels.

[0114] Peak frequency: In the frequency domain, the frequency corresponding to the highest amplitude of the time series data after conversion can help understand the main periodic characteristics of the data.

[0115] First, use moving average or other filtering techniques to smooth the original data, remove noise, and retain short-term fluctuation characteristics. Secondly, perform one or more differential operations on the smoothed time series data to further reduce the trend and seasonality of the data and make the data more stable. Then use the extracted short-term trend data to calculate the autocorrelation coefficients of different lags to find the periodic laws that may exist within the data. Further, use statistics such as standard deviation or variance to quantify the degree of fluctuation of short-term trend data and evaluate the instability of the data. Finally, convert the short-term trend data to the frequency domain, and use methods such as Fourier transform to find the frequency point with the largest amplitude in the spectrum diagram to determine the main period of the data.

[0116] In the embodiment of the present application, it is assumed that the time series data of stock prices is being analyzed, and the goal is to predict the price trend in the short term in the future. First, the data is smoothed using the 5-day moving average method to obtain a short-term trend. Next, a quadratic difference is applied to this short-term trend to further remove any remaining trend components. Subsequently, the autocorrelation coefficients within the 1 to 20-day lag period were calculated, and it was found that the autocorrelation coefficients for 10 days were significantly higher than those for other lag periods, indicating that there is a periodic pattern of about 10 days. At the same time, by calculating the volatility, it is known that price fluctuations have increased significantly in the near future, suggesting that there may be high uncertainty in the market. Finally, the frequency domain characteristics are analyzed by Fourier transform, and the peak frequency is found to be 0.1 (i.e., every 10 trading days), which is consistent with the results of the autocorrelation analysis, and further confirms the 10-day periodic characteristics.

[0117] Through the above steps, this solution can not only effectively capture the short-term trends in time series data, but also deeply explore the intrinsic structure of the data, such as periodicity, volatility and trend, thus providing a richer and more accurate information basis for subsequent forecasting modeling. This has important application value for improving the performance of forecasting models, especially in highly dynamic fields such as financial markets.

[0118] The present application takes into account that when processing time series data, existing technologies often rely on simple statistical methods or time domain analysis, which have limitations when processing complex, nonlinear time series data, especially in extracting signal change characteristics within a short period. Traditional time domain analysis methods are difficult to effectively separate high-frequency and low-frequency components in the signal, which leads to the easy omission of important short-term trend information in fault detection and prediction. In addition, existing methods perform poorly when processing noise and non-stationary signals, affecting the accuracy and reliability of the analysis. Therefore, an embodiment of the present invention proposes a time series data processing method based on wavelet transform to extract short-term trends and calculate multiple signal change characteristics, aiming to solve the problems existing in the prior art and improve the accuracy of fault detection and prediction.

[0119] The options are as follows:

[0120] Optionally, the process of extracting the short-term trend in the time series data and identifying the signal change characteristics within a short period includes the following steps:

[0121] Step 1: Select the wavelet basis function and scale parameter. Assume that the Morlet wavelet basis function is used and the scale parameter and translation parameters .

[0122] Morlet wavelet basis function: ;

[0123] in, Represents the Morlet wavelet function, over time The function of change, is the center frequency, is the bandwidth parameter, is an imaginary unit, is the time variable, is the complex exponential part, is the Gaussian envelope, is the normalization factor.

[0124] Step 2: Calculate wavelet coefficients

[0125] For each scale and pan , calculate the wavelet coefficients .

[0126] formula: ;

[0127] in, is the original time series, is the wavelet basis function, is the complex conjugate of the wavelet basis function, is the normalization factor, is the integral symbol.

[0128] Step 3: Discretization calculation

[0129] In practical applications, time series data is discrete, so the above integral formula needs to be discretized.

[0130] Discretization formula: ;

[0131] in, is the summation symbol, which indicates the summation of the entire discrete time signal The sum is calculated, where is the length of the signal, is a discrete sample of the time series, is the sampling interval.

[0132] Step 4: Choose appropriate scale and translation parameters

[0133] Choose the right scale and translation parameters Typically, the scale parameter Select as ,in is an integer, the translation parameter Select as ,in is an integer.

[0134] Scale and translation parameters: ;

[0135] Step 5: Reconstruct the short-term trend

[0136] Reconstruct short-term trends through inverse wavelet transform.

[0137] formula:

[0138] ;

[0139] Discretization formula:

[0140] ;

[0141] The process of calculating the signal change characteristics includes:

[0142] Step 1: Calculate the higher-order differences of the short-term trend

[0143] First-order difference:

[0144] ;

[0145] Second order difference:

[0146] ;

[0147] Third order difference:

[0148] ;

[0149] Step 2: Calculate the autocorrelation function of the short-term trend. The autocorrelation function can reflect the autocorrelation of the time series and help identify the periodic characteristics of the signal.

[0150] formula:

[0151] Autocorrelation function ;

[0152] in, is the mean of the short-term trend: ;

[0153] Step 3: Calculate the volatility of short-term trends. Volatility can reflect the magnitude of changes in short-term trends.

[0154] formula: ;

[0155] in, is the mean of the first differences: ;

[0156] Step 4: Calculate the peak frequency of the short-term trend. The peak frequency can reflect the fluctuation frequency of the short-term trend. First, perform fast Fourier transform (FFT) on the short-term trend, and then find the frequency corresponding to the maximum value in the spectrum.

[0157] formula:

[0158] Fast Fourier Transform: ;

[0159] Peak frequency: ;

[0160] The constructing of a multidimensional feature vector based on the signal change characteristics and key statistics of the time series data includes:

[0161] Calculate key statistics for time series data

[0162] Mean:

[0163] ;

[0164] variance:

[0165] ;

[0166] Skewness:

[0167] ;

[0168] Kurtosis:

[0169] ;

[0170] Combine the signal change features calculated above into a multi-dimensional feature vector:

[0171] ;

[0172] In the above scheme, the wavelet basis function is a mathematical function used for wavelet transform, and common wavelet basis functions include Morlet wavelet, Haar wavelet, etc. The scale parameter is used to control the degree of expansion and contraction of the wavelet function, and different scale parameters correspond to different frequency resolutions.

[0173] Translation parameter: controls the position of the wavelet function on the time axis. Different translation parameters correspond to different time resolutions. The wavelet coefficients are obtained by wavelet transform to represent the intensity values ​​of the signal at different scales and time positions. Inverse wavelet transform is used to convert wavelet coefficients back to time domain signals to reconstruct short-term trends. High-order differences are used to remove trend components and seasonal components by performing multiple difference operations on time series data. The autocorrelation function is used to measure the correlation between data at different time points in the time series and to identify periodic patterns in the data. Volatility is used to measure the degree of change in time series data and is usually used to assess risk levels. Peak frequency is used in the frequency domain to obtain the frequency corresponding to the highest amplitude of the time series data after conversion, which can help understand the main periodic characteristics of the data. Multidimensional feature vectors are used to combine multiple features into a vector for subsequent analysis and modeling.

[0174] In the embodiment of the present application, it is assumed that there is a set of vibration signal data, the sampling frequency is 1000 Hz, the sampling time is 1 second, and a total of 1000 data points are collected. The above method will be used to process the time series data, extract short-term trends and calculate signal change characteristics.

[0175] Select Morlet wavelet basis function, center frequency , bandwidth parameter . Scale parameter ,in . Translation parameters ,in , sampling interval Second.

[0176] For each scale and pan , calculate the wavelet coefficients .

[0177] Calculate the wavelet coefficients using the discretization formula:

[0178] ;

[0179] Select Scale and translation parameters .

[0180] Reconstruct the short-term trend by inverse wavelet transform:

[0181] ;

[0182] First-order difference:

[0183] ;

[0184] Second order difference:

[0185] ;

[0186] Third order difference:

[0187] ;

[0188] Calculate the autocorrelation function of the short-term trend

[0189] Autocorrelation function:

[0190] ;

[0191] Mean:

[0192] ;

[0193] Volatility:

[0194] ;

[0195] Mean of the first differences:

[0196] ;

[0197] Fast Fourier Transform:

[0198] ;

[0199] Peak frequency:

[0200] ;

[0201] Mean:

[0202] ;

[0203] variance:

[0204] ;

[0205] Skewness:

[0206] ;

[0207] Kurtosis:

[0208] ;

[0209] Construct a multidimensional feature vector:

[0210] ;

[0211] Through the above scheme, the embodiment of the present invention can effectively extract short-term trends from complex time series data and calculate multiple signal change characteristics, thereby providing more accurate fault detection and prediction information. And the wavelet transform can effectively separate the high-frequency and low-frequency components in the signal, especially in extracting the signal change characteristics within a short period, thereby improving the accuracy of fault diagnosis. Secondly, through the construction of a multi-dimensional feature vector, multiple features can be comprehensively considered to reduce false alarms caused by a single feature and improve the reliability of the system. The extraction of characteristic frequency components helps to discover potential problems in the early stages of a fault, take preventive measures in time, and avoid further damage to the equipment. And the method is applicable to various types of time series data and has broad application prospects, especially in fault diagnosis and health monitoring of industrial equipment.

[0212] Optionally, it also includes: when it is determined that the fluctuation signal has abnormal fluctuations that conform to a preset fault mode, the fluctuation signal is regarded as an abnormal fluctuation signal, and features are extracted from the abnormal fluctuation signal to form a feature description of the abnormal fluctuation signal; the feature description is matched with historical fault cases in a database to find qualified historical fault cases and corresponding fault solutions; using a pre-established diagnostic model, the similarity between the abnormal fluctuation signal and the historical fault cases is evaluated to determine the possible fault type; relevant parameters under current environmental conditions, including temperature, humidity, and workload, are collected to supplement fault diagnosis information; based on the possible fault type and the relevant parameters, the weight factor of the diagnostic model is adjusted, and a fault diagnosis report is generated by integrating the feature description of the abnormal fluctuation signal, the matching results of historical fault cases, and the influencing factors of the current environmental conditions.

[0213] Fluctuation signal: An instantaneous change in time series data, which may be a normal fluctuation or an abnormal fluctuation.

[0214] Preset failure mode: A specific abnormal mode summarized from historical failure data to identify potential failure signals.

[0215] Feature extraction: Extract key information that can represent the characteristics of the data from the raw data, such as mean, variance, peak, etc.

[0216] Feature description: The result after feature extraction, usually expressed in vector form, is used for subsequent analysis and matching.

[0217] Historical failure cases: past failure events and their detailed failure information records, including failure type, cause, solution, etc.

[0218] Diagnostic model: A model trained based on machine learning or deep learning algorithms to evaluate the similarity between the current fault signal and historical fault cases.

[0219] Similarity: An indicator to measure the similarity between two objects. Commonly used methods include cosine similarity, Euclidean distance, etc.

[0220] Environmental conditions: external factors that affect the operating status of the equipment, such as temperature, humidity, workload, etc.

[0221] Weight factor: A parameter used to adjust the importance of different features in the diagnostic model, which can be adjusted according to actual conditions.

[0222] First, when the fluctuation signal in the monitored time series data meets the preset fault mode, it is marked as an abnormal fluctuation signal, and key features are extracted from it to form a feature description. Secondly, the extracted feature description is compared with the historical fault cases stored in the database to find the closest match and obtain the corresponding fault type and solution. Further, the pre-trained diagnostic model is used to calculate the similarity between the abnormal fluctuation signal and the matched historical fault case to determine the most likely fault type. Then, relevant parameters under the current environmental conditions, such as temperature, humidity, workload, etc., are collected. This information helps to more accurately evaluate the cause of the fault. And according to the collected environmental condition information, the weight factor in the diagnostic model is adjusted to optimize the prediction performance of the model. Finally, the feature description of the abnormal fluctuation signal, the matching results of the historical fault case and the influence of environmental factors are integrated to generate the final fault diagnosis report.

[0223] In the embodiment of the present application, it is assumed that a motor in a factory has abnormal vibration, and the monitoring system detects the fluctuation signal and determines that it meets the preset fault mode. The system automatically performs the following steps:

[0224] First, extract the mean, variance, peak and other features from the abnormal fluctuation signal to form a feature description vector . Secondly, the feature description vector is matched with the historical fault cases in the database, and a case with high similarity is found, which records the damage of the motor bearing. Using the pre-trained diagnostic model, the similarity between the abnormal fluctuation signal and the historical fault case is calculated. , assuming similarity , indicating that the two are very similar. The system further collects the temperature under the current environmental conditions as , humidity is , the workload is According to the environmental conditions, the weight factors in the diagnostic model are adjusted to suit the current specific situation, such as increasing the weight of temperature and humidity and reducing the weight of workload. Finally, after integrating all the information, the system generates a detailed fault diagnosis report, indicating that the motor bearing may be damaged and recommending immediate shutdown for inspection to avoid greater losses.

[0225] Through the above scheme, the embodiment of the present invention can not only quickly and accurately identify abnormal fluctuation signals of equipment, but also provide more accurate fault diagnosis results by combining historical fault cases and current environmental conditions. This helps enterprises take timely measures to reduce downtime and maintenance costs, and improve production efficiency and equipment reliability. In addition, by continuously updating and improving the database, the solution also has the ability of self-learning and optimization, and its diagnostic accuracy will be further improved over time.

[0226] Optionally, the processing of the physical quantity signal to extract the characteristic frequency components of the potential fault source includes: performing frequency domain conversion on the physical quantity signal to convert the time domain signal into a frequency domain signal; identifying and extracting characteristic frequency components related to the potential fault source in the frequency domain signal, the characteristic frequency components being but not limited to fundamental frequency and harmonic components.

[0227] Physical quantity signal: a signal collected from sensors and other devices that reflects the operating status of the equipment, such as vibration signals, temperature signals, etc.

[0228] Frequency domain conversion: the process of converting time domain signals into frequency domain signals. Commonly used conversion methods include Fourier transform (FT), discrete Fourier transform (DFT), fast Fourier transform (FFT), etc.

[0229] Frequency domain signal: After frequency domain conversion, it represents the distribution of the signal at different frequencies.

[0230] Characteristic frequency components: Specific frequency components related to potential fault sources, including the fundamental frequency and its harmonic components.

[0231] Fundamental Frequency: The basic frequency of a signal, usually the main frequency produced by the fault source.

[0232] Harmonic component: frequency that is an integer multiple of the fundamental frequency, which is usually also the characteristic frequency of the fault source.

[0233] First, the time domain signal is converted into a frequency domain signal using Fourier transform (FT) or fast Fourier transform (FFT). Secondly, the characteristic frequency components related to the potential fault source are identified in the frequency domain signal, including the fundamental frequency and harmonic components.

[0234] Suppose you are analyzing the vibration signal of a rotating machine with the goal of diagnosing whether it has a bearing fault. Here are the specific implementation steps:

[0235] Assume that the vibration signal collected from the sensor is , the sampling frequency is , the sampling time is 1 second, and a total of 1000 data points are collected.

[0236] Use Fast Fourier Transform (FFT) to transform the time domain signal Convert to frequency domain signal :

[0237] ;

[0238] Since the FFT result is symmetrical, only the frequency components in the first half, i.e. from 0 to 500 Hz, need to be considered.

[0239] Assume that the rotation frequency of the machine is 50 Hz, that is, the fundamental frequency In the frequency domain, the signal In the example, the fundamental frequency 50 Hz and its harmonic components are identified. By observing the spectrum, it is found that the amplitudes at 50Hz and 100Hz are significantly higher than other frequency components, which indicates that there may be a bearing fault. Calculate the amplitudes of these frequency components, assuming that the amplitude at 50Hz is The amplitude at is 5.

[0240] Through the above scheme, the embodiment of the present invention can effectively extract characteristic frequency components related to potential fault sources from complex time domain signals, thereby providing more accurate fault diagnosis information. And the frequency domain analysis method can effectively separate different frequency components in the signal, especially the characteristic frequency components related to potential fault sources, thereby improving the accuracy of fault diagnosis. Secondly, by identifying and extracting characteristic frequency components, false alarms caused by noise and interference can be reduced, and the reliability of the system can be improved. Finally, the extraction of characteristic frequency components helps to discover potential problems in the early stages of faults, take preventive measures in time, and avoid further damage to the equipment.

[0241] Figure 2 A schematic diagram of the structure of a communication fault sniffing system based on the Internet of Things is provided for the embodiment of the present application, such as Figure 2 As shown, the device comprises:

[0242] A receiving module 21, used for receiving a status update signal from an IoT device;

[0243] An identification module 22, configured to identify a potential fault source according to an abnormality indicator in the status update signal;

[0244] A capture module 23, used to implement real-time monitoring of the potential fault source using a distributed sensor network, and capture fluctuation signals related to the potential fault source;

[0245] A determination module 24 is used to analyze the time series characteristics of the fluctuation signal and determine whether there is an abnormal fluctuation that meets a preset fault mode;

[0246] The processing module 25 is used to trigger a fault alarm and record the location information of the potential fault source and specific parameters of the abnormal fluctuation when it is detected that the abnormal fluctuation meets the preset fault mode.

[0247] Figure 2 The communication fault sniffing system based on the Internet of Things can be executed Figure 1 The implementation principle and technical effect of the communication fault sniffing method based on the Internet of Things described in the illustrated embodiment will not be repeated. The specific manner in which each module and unit performs operations in the communication fault sniffing system based on the Internet of Things in the above embodiment has been described in detail in the embodiment of the method, and will not be elaborated here.

[0248] In one possible design, Figure 2 The communication fault sniffing system based on the Internet of Things of the embodiment shown can be implemented as a computing device, such as Figure 3 As shown, the computing device may include a storage component 31 and a processing component 32;

[0249] The storage component 31 stores one or more computer instructions, wherein the one or more computer instructions are called and executed by the processing component 32 .

[0250] The processing component 32 is used to: receive a status update signal from an IoT device; identify a potential fault source based on an abnormal indicator in the status update signal; use a distributed sensor network to implement real-time monitoring of the potential fault source and capture fluctuation signals related to the potential fault source; analyze the time series characteristics of the fluctuation signal to determine whether there is an abnormal fluctuation that meets a preset fault mode; when it is detected that the abnormal fluctuation meets the preset fault mode, trigger a fault alarm and record the location information of the potential fault source and the specific parameters of the abnormal fluctuation.

[0251] The processing component 32 may include one or more processors to execute computer instructions to complete all or part of the steps in the above method. Of course, the processing component may also be implemented by one or more application-specific integrated circuits (ASICs), digital signal processors (DSPs), digital signal processing devices (DSPDs), programmable logic devices (PLDs), field programmable gate arrays (FPGAs), controllers, microcontrollers, microprocessors or other electronic components to perform the above method.

[0252] The storage component 31 is configured to store various types of data to support operations at the terminal. The storage component can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic memory, flash memory, magnetic disk or optical disk.

[0253] Of course, the computing device may also include other components, such as input / output interfaces, display components, communication components, etc.

[0254] The input / output interface provides an interface between the processing component and the peripheral interface module, which may be an output device, an input device, etc.

[0255] The communication component is configured to facilitate, among other things, wired or wireless communications between the computing device and other devices.

[0256] Among them, the computing device can be a physical device or an elastic computing host provided by a cloud computing platform, etc. In this case, the computing device can refer to a cloud server, and the above-mentioned processing components, storage components, etc. can be basic server resources rented or purchased from the cloud computing platform.

[0257] The present application also provides a computer storage medium storing a computer program, wherein the computer program can achieve the above-mentioned Figure 1 A communication fault sniffing method based on the Internet of Things is shown in the embodiment.

[0258] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0259] The device embodiments described above are merely illustrative, wherein the units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the scheme of this embodiment. Ordinary technicians in this field can understand and implement it without paying creative labor.

[0260] Through the description of the above implementation methods, those skilled in the art can clearly understand that each implementation method can be implemented by means of software plus a necessary general hardware platform, and of course, can also be implemented by hardware. Based on this understanding, the above technical solution is essentially or the part that contributes to the prior art can be embodied in the form of a software product, and the computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, a disk, an optical disk, etc., including a number of instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in each embodiment or some parts of the embodiments.

[0261] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present application, rather than to limit it. Although the present application has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present application.

Claims

1. A communication fault sniffing method based on the Internet of Things, characterized in that: include: Receive status update signals from IoT devices; identifying a potential fault source based on an abnormality indicator in the status update signal; Using a distributed sensor network to monitor the potential fault source in real time and capture fluctuation signals related to the potential fault source; Analyze the time series characteristics of the fluctuation signal to determine whether there is an abnormal fluctuation that meets the preset fault mode; When it is detected that the abnormal fluctuation meets the preset fault mode, a fault alarm is triggered and the location information of the potential fault source and the specific parameters of the abnormal fluctuation are recorded; The analyzing the time series characteristics of the fluctuation signal to determine whether there is an abnormal fluctuation that meets the preset fault mode includes: Acquiring time series data of the fluctuation signal; Extracting short-term trends from the time series data and identifying short-term signal change characteristics; Calculate the key statistics of the time series data, which are mean, variance, skewness and kurtosis, and the key statistics are used to characterize the distribution characteristics of the time series data; Constructing a multidimensional feature vector based on the signal change characteristics and key statistics of the time series data; Inputting the multidimensional feature vector into a pre-trained pattern recognition model to output a fluctuation signal corresponding to the multidimensional feature vector through the pattern recognition model, wherein the pattern recognition model has learned the fluctuation signal characteristics under multiple fault modes; Determine whether the fluctuation signal meets the probability estimation of the preset fault mode and whether it exceeds the predetermined threshold of the probability estimation, and if so, determine that the fluctuation signal has abnormal fluctuations that meet the preset fault mode; The extracting of short-term trends in the time series data and identifying signal change characteristics within a short period include: extracting short-term trends from the time series data; Calculating various signal change characteristics including high-order differences, autocorrelation functions, volatility, and peak frequency based on the short-term trends; The identifying a potential fault source according to the abnormal indicator in the status update signal comprises: Parsing metadata in the status update signal to extract key indicators reflecting the operating status of the IoT device; Comparing the key indicators with a preset normal operating range, and identifying abnormal indicators that exceed the range; According to the abnormal indicator, mapping to a predefined equipment failure model, determining a potential failure source corresponding to the abnormal indicator; The method of using a distributed sensor network to monitor the potential fault source in real time and capture fluctuation signals related to the potential fault source includes: Collecting physical quantity signals related to the potential fault source in real time through multiple sensor nodes deployed in the Internet of Things environment; Processing the physical quantity signal to extract characteristic frequency components of the potential fault source; The characteristic frequency components are continuously monitored to identify fluctuation signals occurring near the potential fault source.

2. The method according to claim 1, characterized in that Also includes: In the case where it is determined that the fluctuation signal has abnormal fluctuations that meet the preset fault mode, the fluctuation signal is regarded as an abnormal fluctuation signal, and features of the abnormal fluctuation signal are extracted to form a feature description of the abnormal fluctuation signal; Match the feature description with historical fault cases in a database to find historical fault cases that meet the conditions and corresponding fault solutions; Using a pre-established diagnostic model, evaluating the similarity between the abnormal fluctuation signal and the historical fault case to determine a possible fault type; Collect relevant parameters under current environmental conditions, such as temperature, humidity, and workload, to supplement fault diagnosis information; Based on the possible fault types and the relevant parameters, the weight factors of the diagnostic model are adjusted, and a fault diagnosis report is generated by integrating the feature description of the abnormal fluctuation signal, the matching results of historical fault cases, and the influencing factors of the current environmental conditions.

3. The method according to claim 1, characterized in that The step of processing the physical quantity signal to extract the characteristic frequency component of the potential fault source includes: Performing frequency domain conversion on the physical quantity signal to convert the time domain signal into a frequency domain signal; Characteristic frequency components related to the potential fault source are identified and extracted from the frequency domain signal, wherein the characteristic frequency components are fundamental frequency and harmonic components.

4. A communication fault sniffing system based on the Internet of Things is used to execute a communication fault sniffing method based on the Internet of Things as described in any one of claims 1 to 3, characterized in that: include: A receiving module, used for receiving a status update signal from an IoT device; an identification module, configured to identify a potential fault source based on an abnormality indicator in the status update signal; A capture module, used to implement real-time monitoring of the potential fault source using a distributed sensor network, and capture fluctuation signals related to the potential fault source; A determination module, used to analyze the time series characteristics of the fluctuation signal and determine whether there is an abnormal fluctuation that meets a preset fault mode; A processing module, configured to trigger a fault alarm and record the location information of the potential fault source and specific parameters of the abnormal fluctuation when detecting that the abnormal fluctuation meets a preset fault mode; The analyzing the time series characteristics of the fluctuation signal to determine whether there is an abnormal fluctuation that meets the preset fault mode includes: Acquiring time series data of the fluctuation signal; Extracting short-term trends from the time series data and identifying short-term signal change characteristics; Calculate the key statistics of the time series data, which are mean, variance, skewness and kurtosis, and the key statistics are used to characterize the distribution characteristics of the time series data; Constructing a multidimensional feature vector based on the signal change characteristics and key statistics of the time series data; Inputting the multidimensional feature vector into a pre-trained pattern recognition model to output a fluctuation signal corresponding to the multidimensional feature vector through the pattern recognition model, wherein the pattern recognition model has learned the fluctuation signal characteristics under multiple fault modes; Determine whether the fluctuation signal meets the probability estimation of the preset fault mode and whether it exceeds the predetermined threshold of the probability estimation, and if so, determine that the fluctuation signal has abnormal fluctuations that meet the preset fault mode; The extracting of short-term trends in the time series data and identifying signal change characteristics within a short period include: extracting short-term trends from the time series data; Calculating various signal change characteristics including high-order differences, autocorrelation functions, volatility, and peak frequency based on the short-term trends; The identifying a potential fault source according to the abnormal indicator in the status update signal comprises: Parsing metadata in the status update signal to extract key indicators reflecting the operating status of the IoT device; Comparing the key indicators with a preset normal operating range, and identifying abnormal indicators that exceed the range; According to the abnormal indicator, mapping to a predefined equipment failure model, determining a potential failure source corresponding to the abnormal indicator; The method of using a distributed sensor network to monitor the potential fault source in real time and capture fluctuation signals related to the potential fault source includes: Collecting physical quantity signals related to the potential fault source in real time through multiple sensor nodes deployed in the Internet of Things environment; Processing the physical quantity signal to extract characteristic frequency components of the potential fault source; The characteristic frequency components are continuously monitored to identify fluctuation signals occurring near the potential fault source.

5. A computing device, characterized in that: It comprises a processing component and a storage component; the storage component stores one or more computer instructions; the one or more computer instructions are used to be called and executed by the processing component to implement a communication fault sniffing method based on the Internet of Things as described in any one of claims 1 to 3.

6. A computer storage medium, characterized in that: A computer program is stored, and when the computer program is executed by a computer, a communication fault sniffing method based on the Internet of Things as described in any one of claims 1 to 3 is implemented.

Images