Method for monitoring power device on line through Internet of Things

Abstract

A method for monitoring a power device on line through the Internet of Things belongs to the technical field of electrical digital data processing, and comprises the following steps of: performing more accurate processing on data of the power device by applying a formula and an algorithm of a parameter cleaning method of a neural network, and improving the monitoring accuracy. Under different working conditions, the monitoring strategy can be adaptively adjusted, and the monitoring reliability is guaranteed. Compared with other existing monitoring methods, the method provided by the invention has higher flexibility and accuracy, can better meet the requirement of the power system for monitoring the power device, and provides powerful guarantee for safe and stable operation of the power system. The defects that in the prior art, a fixed threshold value monitoring method is prone to misjudgment, a traditional monitoring method is easy to process data, potential information behind the data cannot be fully mined, and complex and changeable power system monitoring requirements are difficult to meet in the aspects of parameter processing and adaptability are effectively overcome.

Description

Technical Field

[0001] The present invention belongs to the technical field of electrical digital data processing, and in particular relates to a method for online monitoring of power devices by the Internet of Things. Background Art

[0002] In the power system sector, real-time and accurate monitoring of power equipment is crucial. With the development of the Internet of Things (IoT), online monitoring of power equipment using the IoT has become a leading trend. Conventional power equipment monitoring methods primarily rely on fixed thresholds, comparing monitored parameters against preset thresholds to determine whether the equipment is functioning normally. For example, for voltage monitoring, a fixed voltage range is set, and voltage outside this range is considered abnormal. However, this approach has significant limitations. In actual operation, power equipment operating conditions are complex and variable. Factors such as varying loads, ambient temperatures, and operating time can affect the normal parameter ranges of power equipment. For example, during high temperatures in summer, the temperature of power equipment rises, and its normal current and voltage parameters also change accordingly. Fixed-threshold monitoring methods are prone to misjudgments. Furthermore, traditional monitoring methods are relatively simple in data processing and fail to fully exploit the potential information behind the data. Although some machine learning-based monitoring methods have emerged in recent years, these methods still lack parameter processing and adaptability, making them difficult to meet the complex and ever-changing needs of power system monitoring. Summary of the Invention

[0003] In order to solve the defects in the existing technology, the present invention proposes an Internet of Things online monitoring power device and method, which effectively avoids the defects of the existing technology's fixed threshold monitoring method that is prone to misjudgment, the traditional monitoring method is relatively simple in processing data, cannot fully explore the potential information behind the data, and is difficult to meet the complex and changeable power system monitoring needs in terms of parameter processing and adaptability.

[0004] The present invention utilizes the following technical solutions.

[0005] A method for online monitoring of an electric power device using the Internet of Things, comprising:

[0006] Step 1: Collect the operating data of power devices in real time through IoT sensors;

[0007] Step 2: Use the neural network parameter cleaning algorithm to clean the collected operation data;

[0008] Step 3: Deeply process the cleaned data based on the neural network algorithm to mine potential information and analyze the operating status and trends of power equipment.

[0009] Furthermore, in step 1, the operation data of the power device is collected in real time by deploying IoT sensors at key locations of the power device. The IoT sensors include voltage sensors, current sensors, temperature sensors, and power sensors. The voltage sensor is used to measure the voltage value of the power device; the current sensor is used to measure the current value; the temperature sensor is used to monitor the temperature change of the power device; and the power sensor calculates the power value of the power device by simultaneously measuring the voltage and current.

[0010] Furthermore, in step 1, the IoT sensor transmits the collected data to the data processing center via wired or wireless means.

[0011] Furthermore, step 2 specifically includes:

[0012] Step 2-1: Data preprocessing;

[0013] Step 2-2: Noise detection and removal;

[0014] Step 2-3: Outlier detection and processing;

[0015] Step 2-4: Duplicate data detection and deletion;

[0016] Step 2-5: Data reconstruction.

[0017] Furthermore, in step 2-1, first, the collected raw data is standardized, and the standardization formula is:

[0018]

[0019] Among them, x is the original data, μ is the mean of the data, σ is the standard deviation of the data, and x' is the standardized data.

[0020] Furthermore, in step 2-2, the specific formula for noise detection and removal is:

[0021]

[0022] Among them, W(a,b) is the wavelet coefficient, a is the scale parameter, b is the translation parameter, and ψ is the wavelet basis function.

[0023] Furthermore, in step 2-3, an outlier detection algorithm based on isolation forest is used. Isolation forest constructs multiple isolated trees by randomly selecting features and split points. The specific formula is:

[0024]

[0025] Among them, h(x) is the average path length of sample x, c(T i) is the path length of sample x in the i-th isolated tree, n is the number of isolated trees, and by setting the outlier threshold, samples with h(x) greater than the threshold are regarded as outliers and processed.

[0026] Furthermore, in steps 2-4, first, a hash calculation is performed on each piece of data to generate a unique hash value. Then, by comparing the hash values, duplicate data is identified and deleted. The specific formula is:

[0027] H(x)=hash(x)

[0028] Where H(x) is the hash value of data x, and hash(x) is the hash function.

[0029] Furthermore, in steps 2-5, after removing noise, outliers, and duplicate data, the data is reconstructed to restore its original format. The reconstruction formula is:

[0030] x=x′·σ+μ

[0031] Among them, x' is the standardized data, σ is the standard deviation of the data, μ is the mean of the data, and x is the reconstructed data.

[0032] Furthermore, step 3 specifically includes:

[0033] Step 3-1: Feature extraction;

[0034] Step 3-2: Timing modeling;

[0035] Step 3-3: State prediction;

[0036] Step 3-4: Trend analysis;

[0037] Step 3-5: Output the results.

[0038] Furthermore, in step 3-1, first, feature extraction is performed on the cleaned data to capture key information about the operating status of the power device. A feature extraction method based on a convolutional neural network is used. The specific formula is:

[0039] F=CNN(X)

[0040] Among them, X is the cleaned data, CNN is the convolutional neural network, and F is the extracted features.

[0041] Furthermore, in step 3-2, a time series modeling method based on long short-term memory network is adopted, and the specific formula is:

[0042] S=LSTM(F)

[0043] Among them, F is the extracted feature, LSTM is the long short-term memory network, and S is the state after time series modeling.

[0044] Furthermore, in step 3-3, a state prediction method based on the attention mechanism is adopted. The specific formula is:

[0045] P=Attention(S)

[0046] Among them, S is the state after time series modeling, Attention is the attention mechanism, and P is the predicted operating state of the power device.

[0047] Furthermore, in step 3-4, a trend analysis method based on an autoregressive integrated moving average model is used, and the specific formula is:

[0048] T=ARIMA(P)

[0049] Where P is the predicted operating status of the power device, ARIMA is the autoregressive integrated moving average model, and T is the trend analysis result.

[0050] Furthermore, in steps 3-5, the trend analysis results are output for subsequent decision-making. The specific formula is:

[0051] R=Output(T)

[0052] Among them, T is the trend analysis result, Output is the output function, and R is the final result.

[0053] An Internet of Things online monitoring power device, comprising:

[0054] A data acquisition module, which is used to collect operating data of power devices in real time through IoT sensors;

[0055] Parameter cleaning module, which is used to clean the collected operation data using the neural network parameter cleaning algorithm;

[0056] The data processing module is used to perform in-depth processing on the cleaned data based on the neural network algorithm, mine potential information, and analyze the operating status and trends of power equipment.

[0057] The beneficial effects of the present invention are that, compared with the prior art, the technical effects of the present invention include:

[0058] The invention aims to solve the shortcomings of existing monitoring methods for power devices in terms of accuracy and adaptability. Specifically, the traditional fixed threshold monitoring method cannot adapt to the complex and changeable working conditions of power devices, which can easily lead to misjudgment and affect the safe and stable operation of the power system. However, some existing machine learning monitoring methods are not sophisticated enough in parameter processing and cannot make full use of data information. The present invention uses the formulas and algorithms of the parameter cleaning method of the neural network to more accurately process the data of the power device and improve the accuracy of monitoring. Under different working conditions, the monitoring strategy can be adaptively adjusted to ensure the reliability of monitoring. Compared with other existing monitoring methods, the method of the present invention has higher flexibility and accuracy, can better meet the needs of the power system for power device monitoring, and provide strong guarantees for the safe and stable operation of the power system. BRIEF DESCRIPTION OF THE DRAWINGS

[0059] Figure 1 This is a flow chart of using the neural network parameter cleaning algorithm to clean the collected operating data as described in the present invention;

[0060] Figure 2 This is a flowchart of the present invention for deeply processing cleaned data based on a neural network algorithm, mining potential information, and analyzing the operating status and trends of power devices;

[0061] Figure 3 It is a partial structural diagram of the Internet of Things online monitoring power device in the present invention. DETAILED DESCRIPTION

[0062] To make the purpose, technical solutions and advantages of the present invention more clear, the following will be combined with the drawings in the embodiments of the present invention to clearly and completely express the technical solutions of the present invention. The embodiments expressed in this application are only some embodiments of the present invention, not all embodiments. According to the spirit of the present invention, other embodiments obtained by those skilled in the art without making creative work shall fall within the scope of protection of the present invention.

[0063] like Figure 1 As shown, the method of online monitoring of power devices by the Internet of Things according to the present invention includes:

[0064] Step 1: Use IoT sensors to collect real-time operating data such as voltage, current, temperature, and power of power devices;

[0065] In a preferred but non-limiting embodiment of the present invention, in step 1, the operating data of the power device is collected in real time by deploying Internet of Things sensors at key locations of the power device; the Internet of Things sensors include voltage sensors, current sensors, temperature sensors, and power sensors. The voltage sensor is used to measure the voltage value of the power device, and is usually implemented using the resistance voltage divider principle or the Hall effect principle; the current sensor is used to measure the current value, and common ones include current sensors based on the Hall effect and current sensors based on the Rogowski coil; the temperature sensor is used to monitor the temperature changes of the power device, and is usually a thermocouple, thermistor, or infrared temperature sensor; the power sensor calculates the power value of the power device by simultaneously measuring the voltage and current.

[0066] In a preferred but non-limiting embodiment of the present invention, in step 1, the IoT sensor transmits the collected data to a data processing center via wired or wireless means. Wired transmission methods include RS-485 and CAN bus, which offer advantages such as long transmission distance and strong anti-interference capabilities. Wireless transmission methods include ZigBee, LoRa, and NB-IoT, which offer advantages such as flexible deployment and low cost. Data compression and encryption technologies are employed during data transmission to improve transmission efficiency and security.

[0067] In specific implementation, taking transformer monitoring at a substation as an example, voltage sensors, current sensors, and temperature sensors are installed on the high-voltage side, low-voltage side, and oil temperature measurement point of the transformer. The voltage sensor uses the resistor divider principle to reduce the high-voltage side voltage to a safe range for measurement. The current sensor uses the Rogowski coil principle to infer the current value by measuring the induced electromotive force in the coil. The temperature sensor uses a thermistor to convert temperature changes into resistance changes for measurement. These sensors transmit the collected data to the substation's data acquisition unit via the RS-485 bus. The data acquisition unit performs preliminary processing on the received data, such as filtering and noise removal, and then transmits the data to the power company's data center via a fiber optic network, providing basic data support for subsequent parameter cleaning and data processing.

[0068] Step 2: Use the neural network parameter cleaning algorithm to clean the collected operating data to remove noise, outliers and duplicate data to improve data quality;

[0069] In a preferred but non-limiting embodiment of the present invention, step 2 specifically comprises:

[0070] The neural network parameter cleaning algorithm is used to clean the collected data to remove noise, outliers and duplicate data, thereby improving data quality.

[0071] The parameter cleaning module is one of the core modules in this invention, responsible for cleaning the collected power device operation data, removing noise, outliers and duplicate data to improve data quality. The specific implementation steps are as follows:

[0072] Step 2-1: Data preprocessing;

[0073] In a preferred but non-limiting embodiment of the present invention, in step 2-1, first, the collected raw data is standardized, and the standardization formula is:

[0074]

[0075] Here, x is the original data, μ is the mean of the data, σ is the standard deviation of the data, and x' is the normalized data. The purpose of this step is to unify data of different dimensions to the same scale for ease of subsequent processing.

[0076] Step 2-2: Noise detection and removal;

[0077] In a preferred but non-limiting embodiment of the present invention, in step 2-2, a noise detection method based on wavelet transform is used. Wavelet transform can decompose the signal into sub-bands of different frequencies and identify the noise components by analyzing the energy distribution of each sub-band. The specific formula for noise detection and removal is:

[0078]

[0079] Where W(a,b) is the wavelet coefficient, a is the scale parameter, b is the translation parameter, and ψ is the wavelet basis function. By setting a threshold, the wavelet coefficients below the threshold are considered as noise and removed.

[0080] Step 2-3: Outlier detection and processing;

[0081] In a preferred but non-limiting embodiment of the present invention, in steps 2-3, an outlier detection algorithm based on an isolation forest is used. The isolation forest constructs multiple isolated trees by randomly selecting features and split points. Outliers are usually isolated within a relatively small number of steps due to their sparse feature values. The specific formula is:

[0082]

[0083] Among them, h(x) is the average path length of sample x, c(T i ) is the path length of sample x in the i-th isolated tree, n is the number of isolated trees, and by setting the outlier threshold, samples with h(x) greater than the threshold are regarded as outliers and processed.

[0084] Step 2-4: Duplicate data detection and deletion;

[0085] In a preferred but non-limiting embodiment of the present invention, in steps 2-4, a duplicate data detection method based on a hash algorithm is used. First, a hash calculation is performed on each piece of data to generate a unique hash value. Then, by comparing the hash values, duplicate data is identified and deleted. The specific formula is:

[0086] H(x)=hash(x)

[0087] Among them, H(x) is the hash value of data x, hash(x) is the hash function, and by comparing H(x), duplicate data can be quickly identified.

[0088] Step 2-5: Data reconstruction.

[0089] In a preferred but non-limiting embodiment of the present invention, in steps 2-5, after removing noise, outliers and duplicate data, the data is reconstructed to restore its original format. The reconstruction formula is:

[0090] x=x'·σ+μ

[0091] Among them, x' is the standardized data, σ is the standard deviation of the data, μ is the mean of the data, and x is the reconstructed data.

[0092] For step 2, the example is as follows:

[0093] Taking transformer monitoring data from a substation as an example, assume the collected voltage data is [220, 221, 222, 223, 224, 225, 226, 227, 228, 229], with a mean of 224.5 and a standard deviation of 3.03. First, the data is normalized to obtain the standardized data [-1.48, -1.15, -0.82, -0.49, -0.16, 0.16, 0.49, 0.82, 1.15, 1.48]. Then, a wavelet transform is used to detect noise, identifying the first and tenth data points as noise and removing them. Next, an isolation forest algorithm is used to detect outliers, identifying the fifth data point as an outlier and removing it. Finally, a hash algorithm is used to detect duplicate data, identifying the third and seventh data points as duplicates and removing them. Finally, the data is reconstructed to obtain cleaned data [221,222,223,224,226,227,228]. Through this series of steps, the quality of the data is effectively improved, providing a reliable foundation for subsequent data processing and analysis.

[0094] Step 3: Deeply process the cleaned data based on a neural network algorithm to mine potential information and analyze the operating status and trends of power equipment;

[0095] In a preferred but non-limiting embodiment of the present invention, step 3 specifically comprises:

[0096] Responsible for in-depth processing of the cleaned power device operation data, mining the potential information behind the data, and analyzing the operating status and trends of the power device. The specific implementation steps are as follows:

[0097] Step 3-1: Feature extraction;

[0098] In a preferred but non-limiting embodiment of the present invention, in step 3-1, first, feature extraction is performed on the cleaned data to capture key information about the operating status of the power device. A feature extraction method based on a convolutional neural network (CNN) is used, and the specific formula is:

[0099] F=CNN(X)

[0100] Where X is the cleaned data, CNN is the convolutional neural network, and F is the extracted features. Convolutional neural networks can effectively extract local and global features from data through multi-layer convolution and pooling operations.

[0101] Step 3-2: Timing modeling;

[0102] In a preferred but non-limiting embodiment of the present invention, in step 3-2, the extracted features are subjected to time series modeling to capture the dynamic changes in the operating state of the power device. A time series modeling method based on a long short-term memory network (LSTM) is used, and the specific formula is:

[0103] S=LSTM(F)

[0104] Here, F represents the extracted features, LSTM represents the long short-term memory network, and S represents the state after time series modeling. Long short-term memory networks, through memory cells and gating mechanisms, can effectively process long-term dependencies in time series data.

[0105] Step 3-3: State prediction;

[0106] In a preferred but non-limiting embodiment of the present invention, in step 3-3, the operating state of the power device is predicted based on the results of the time series modeling. A state prediction method based on the attention mechanism is used. The specific formula is:

[0107] P=Attention(S)

[0108] Here, S represents the state after time series modeling, Attention is the attention mechanism, and P is the predicted operating state of the power device. The attention mechanism effectively captures important information in time series data through weighted summation, improving prediction accuracy.

[0109] Step 3-4: Trend analysis;

[0110] In a preferred but non-limiting embodiment of the present invention, in step 3-4, a trend analysis is performed on the predicted operating state of the power device to identify potential faults and anomalies. A trend analysis method based on the Autoregressive Integrated Moving Average (ARIMA) model is used, and the specific formula is:

[0111] T=ARIMA(P)

[0112] Where P is the predicted operating status of the power plant, ARIMA is the autoregressive integrated moving average model, and T is the trend analysis result. The autoregressive integrated moving average model effectively captures trends and cyclical changes in the data through a combination of autoregression, differencing, and moving average.

[0113] Step 3-5: Output the results.

[0114] In a preferred but non-limiting embodiment of the present invention, in steps 3-5, the trend analysis results are output for subsequent decision-making. The specific formula is:

[0115] R=Output(T)

[0116] Where T is the trend analysis result, Output is the output function, and R is the final result. The output function converts the trend analysis results into an easy-to-understand format, such as fault warnings and operating status reports.

[0117] For step 3, the example is as follows:

[0118] Taking transformer monitoring data from a substation as an example, assume the cleaned voltage data is [221, 222, 223, 224, 226, 227, 228]. First, a convolutional neural network is used for feature extraction to obtain feature F. Then, a long short-term memory network is used for time series modeling to obtain the state S. Next, an attention mechanism is used for state prediction to obtain the predicted power device operating state P. Then, an autoregressive integrated moving average model is used for trend analysis to obtain the trend analysis result T. Finally, the trend analysis result is output to obtain the final result R. Through this series of steps, the potential information behind the data is effectively mined, the operating status and trends of the power devices are analyzed, and a strong guarantee is provided for the safe and stable operation of the power system.

[0119] An Internet of Things online monitoring power device, comprising:

[0120] Data acquisition module, which is used to collect voltage, current, temperature, power and other operating data of power devices in real time through IoT sensors;

[0121] The parameter cleaning module is used to clean the collected operating data using a neural network parameter cleaning algorithm to remove noise, outliers, and duplicate data, thereby improving data quality.

[0122] The data processing module is used to perform in-depth processing on the cleaned data based on the neural network algorithm, mine potential information, and analyze the operating status and trends of power equipment.

[0123] The beneficial effects of the present invention are that, compared with the prior art, the technical effects of the present invention include:

[0124] The invention aims to solve the shortcomings of existing monitoring methods for power devices in terms of accuracy and adaptability. Specifically, the traditional fixed threshold monitoring method cannot adapt to the complex and changeable working conditions of power devices, which can easily lead to misjudgment and affect the safe and stable operation of the power system. However, some existing machine learning monitoring methods are not sophisticated enough in parameter processing and cannot make full use of data information. The present invention uses the formulas and algorithms of the parameter cleaning method of the neural network to more accurately process the data of the power device and improve the accuracy of monitoring. Under different working conditions, the monitoring strategy can be adaptively adjusted to ensure the reliability of monitoring. Compared with other existing monitoring methods, the method of the present invention has higher flexibility and accuracy, can better meet the needs of the power system for power device monitoring, and provide strong guarantees for the safe and stable operation of the power system.

[0125] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art should understand that the specific implementation methods of the present invention can still be modified or replaced with equivalents, and any modifications or equivalent replacements that do not deviate from the spirit and scope of the present invention should be covered within the scope of protection of the claims of the present invention.

Claims

1. A method for online monitoring of power devices using the Internet of Things, characterized in that: include: Step 1: Collect the operating data of power devices in real time through IoT sensors; Step 2: Use the neural network parameter cleaning algorithm to clean the collected operation data; Step 3: Deeply process the cleaned data based on the neural network algorithm to mine potential information and analyze the operating status and trends of power equipment.

2. The method for online monitoring of power devices by the Internet of Things according to claim 1, characterized in that: In step 1, IoT sensors deployed at key locations of the power device collect real-time operating data of the power device. The IoT sensors include voltage sensors, current sensors, temperature sensors, and power sensors. The voltage sensor is used to measure the voltage value of the power device. The current sensor is used to measure the current value; The temperature sensor is used to monitor the temperature changes of the power device; the power sensor calculates the power value of the power device by measuring the voltage and current simultaneously; In step 1, the IoT sensor transmits the collected data to the data processing center via wired or wireless means.

3. The method for online monitoring of power devices by the Internet of Things according to claim 2, characterized in that: Step 2 specifically includes: Step 2-1: Data preprocessing; Step 2-2: Noise detection and removal; Step 2-3: Outlier detection and processing; Step 2-4: Duplicate data detection and deletion; Step 2-5: Data reconstruction.

4. The method for online monitoring of power devices by the Internet of Things according to claim 3, characterized in that: In step 2-1, first, the collected raw data is standardized. The standardization formula is: Among them, x is the original data, μ is the mean of the data, σ is the standard deviation of the data, and x' is the standardized data; In step 2-2, the specific formula for noise detection and removal is: Among them, W(a,b) is the wavelet coefficient, a is the scale parameter, b is the translation parameter, and ψ is the wavelet basis function.

5. The method for online monitoring of power devices by the Internet of Things according to claim 4, characterized in that: In steps 2-3, an outlier detection algorithm based on isolation forest is used. Isolation forest constructs multiple isolated trees by randomly selecting features and split points. The specific formula is: Among them, h(x) is the average path length of sample x, c(T i ) is the path length of sample x in the i-th isolated tree, n is the number of isolated trees, and by setting the outlier threshold, samples with h(x) greater than the threshold are regarded as outliers and processed; In steps 2-4, first, a hash calculation is performed on each piece of data to generate a unique hash value. Then, by comparing the hash values, duplicate data is identified and deleted. The specific formula is: H(x)=hash(x) Where H(x) is the hash value of data x, and hash(x) is the hash function.

6. The method for online monitoring of power devices by the Internet of Things according to claim 5, characterized in that: In steps 2-5, after removing noise, outliers, and duplicate data, the data is reconstructed to restore its original format. The reconstruction formula is: x=x′·σ+μ Among them, x' is the standardized data, σ is the standard deviation of the data, μ is the mean of the data, and x is the reconstructed data.

7. The method for online monitoring of power devices by the Internet of Things according to claim 6, characterized in that: Step 3 specifically includes: Step 3-1: Feature extraction; Step 3-2: Timing modeling; Step 3-3: State prediction; Step 3-4: Trend analysis; Step 3-5: Output the results.

8. The method for online monitoring of power devices by the Internet of Things according to claim 7, characterized in that: In step 3-1, first, feature extraction is performed on the cleaned data to capture key information about the operating status of the power device. A feature extraction method based on a convolutional neural network is used. The specific formula is: F=CNN(X) Where X is the cleaned data, CNN is the convolutional neural network, and F is the extracted features; In step 3-2, a time series modeling method based on long short-term memory network is adopted. The specific formula is: S=LSTM(F) Among them, F is the extracted feature, LSTM is the long short-term memory network, and S is the state after time series modeling.

9. The method for online monitoring of power devices by the Internet of Things according to claim 8, characterized in that: In step 3-3, a state prediction method based on the attention mechanism is adopted. The specific formula is: P=Attention(S) Among them, S is the state after time series modeling, Attention is the attention mechanism, and P is the predicted operating state of the power device; In steps 3-4, a trend analysis method based on the autoregressive integrated moving average model is used. The specific formula is: T=ARIMA(P) Where P is the predicted operating status of the power device, ARIMA is the autoregressive integrated moving average model, and T is the trend analysis result; In steps 3-5, the trend analysis results are output for subsequent decision-making. The specific formula is: R=Output(T) Among them, T is the trend analysis result, Output is the output function, and R is the final result.

10. An Internet of Things online monitoring power device, characterized in that: include: A data acquisition module, which is used to collect operating data of power devices in real time through IoT sensors; Parameter cleaning module, which is used to clean the collected operation data using the neural network parameter cleaning algorithm; The data processing module is used to perform in-depth processing on the cleaned data based on the neural network algorithm, mine potential information, and analyze the operating status and trends of power equipment.

Images