Performance monitoring method and system of power quality control device suitable for plateau environment

Abstract

The invention discloses a performance monitoring method and system for an electric energy quality management device suitable for a plateau environment, and relates to the technical field of crossing of electrical engineering and computer science, and the method comprises the steps: collecting the operation data of the electric energy quality management device through a high-precision sensor array; through a deep learning fusion method, automatically extracting deep feature vectors of data of each sensor, and fusing the deep feature vectors into a comprehensive monitoring data set; a long-short-term memory network structure is defined through a historical monitoring data set, a preliminary intelligent electric energy monitoring model is constructed based on a softmax activation function and a cross entropy loss function, local features are efficiently recognized by using convolution operation and pooling layer processing preprocessed operation data of an electric energy quality management device, data dimensions are remarkably reduced, and the accuracy of data processing is improved. A more accurate and efficient intelligent electric energy monitoring model is constructed, the accuracy and reliability of performance monitoring of the electric energy quality management device in the plateau environment are improved, and the stability of an electric power system is guaranteed.

Description

Technical Field

[0001] The present invention relates to the cross - technical field of electrical engineering and computer science, and particularly to a performance monitoring method and system for power quality governance devices applicable to plateau environments. Background Art

[0002] With the development of modern power systems, the importance of power quality governance devices has become increasingly prominent. In plateau environments, due to their special geographical and climatic conditions, such as low air pressure, low temperature, etc., these factors have a significant impact on the performance of electrical equipment. To ensure the stable operation of the power system, effective measures must be taken to monitor and improve the performance of power quality governance devices. Traditional power quality monitoring methods mainly rely on basic sensor technologies and simple data analysis means. Although these technologies can meet the requirements in conventional environments to a certain extent, they are inadequate in the complex and changeable plateau environment. With the progress of information technology, especially the emergence of emerging technologies such as big data analysis and machine learning, new solutions and development directions have been provided for power quality monitoring.

[0003] There have been many current studies and technical applications on power quality monitoring, but the existing solutions still have obvious deficiencies. There are also limitations in noise removal and data normalization in the existing technologies. Commonly used filtering algorithms often cannot effectively eliminate the influence of non - linear noise, resulting in deviations in subsequent analysis results. These problems directly affect the accuracy and reliability of the performance monitoring of power quality governance devices. Summary of the Invention

[0004] In view of the above - mentioned existing problems, the present invention is proposed.

[0005] Therefore, the problem of incomplete removal of operation noise of power quality governance devices in the prior art is solved.

[0006] To solve the above - mentioned technical problems, the present invention provides the following technical solutions:

[0007] In a first aspect, the present invention provides a performance monitoring method for a power quality management device suitable for a plateau environment, which includes collecting operating data of the power quality management device through a high-precision sensor array; preliminarily screening the operating data of the power quality management device through a statistical method, and using an adaptive filtering algorithm to remove noise and perform normalization processing; automatically extracting deep feature vectors of each sensor data through a deep learning fusion method, and fusing them into a comprehensive monitoring data set; defining a long short-term memory network structure through a historical monitoring data set, and constructing a preliminary intelligent power monitoring model based on a softmax activation function and a cross entropy loss function; training the preliminary intelligent power monitoring model through the monitoring data set to obtain prediction results, calculate evaluation indicators, and obtain an intelligent power monitoring model; inputting the real-time monitoring data set into the intelligent power monitoring model to obtain performance indicators of the power quality management device and generate a performance report.

[0008] As a preferred solution of the method for monitoring the performance of a power quality control device suitable for a plateau environment described in the present invention, wherein: the operation data of the power quality control device is collected by a high-precision sensor array, and the specific steps are as follows:

[0009] The grid voltage level, current intensity and equipment operating temperature are collected through a high-precision sensor array, connected to the edge data processor, and uploaded to the nearby edge computing node through wireless communication methods to obtain the operating data of the power quality management device.

[0010] As a preferred solution of the method for monitoring the performance of a power quality control device suitable for a plateau environment described in the present invention, the operating data of the power quality control device is preliminarily screened by a statistical method, and an adaptive filtering algorithm is used to remove noise and perform normalization processing. The specific steps are as follows:

[0011] Through statistical methods, the operation data of the power quality management device is plotted into a time series graph, and the abnormal points are checked through the time series graph to preliminarily screen out the obviously erroneous operation data of the power quality management device, and the random noise is removed through the adaptive filtering algorithm;

[0012] The operation data of the power quality management device after noise removal is normalized through minimum and maximum scaling normalization to obtain the preprocessed operation data of the power quality management device.

[0013] As a preferred solution of the performance monitoring method of the power quality management device suitable for plateau environment described in the present invention, the deep feature vectors of each sensor data are automatically extracted through the deep learning fusion method and fused into a comprehensive monitoring data set. The specific steps are as follows:

[0014] The convolution operation is used to identify the pre-processed power quality management device operation data by sliding the convolution kernel on the time series to extract local features;

[0015] The pooling layer reduces the dimension of the pre-processed power quality management device operation data, and obtains a deep feature vector by stacking multiple convolutional layers and pooling layers;

[0016] All deep feature vectors are concatenated column by column through splicing and fusion to obtain a comprehensive monitoring dataset.

[0017] As a preferred solution of the performance monitoring method of the power quality management device suitable for plateau environment described in the present invention, wherein: the long short-term memory network structure is defined by the historical monitoring data set, and a preliminary intelligent power monitoring model is constructed based on the softmax activation function and the cross entropy loss function. The specific steps are as follows:

[0018] Based on the historical monitoring data set, the input layer is defined, the time step and the number of features are set, multiple LSTM layers are added, and the output layer is defined through the softmax activation function and the cross entropy loss function to obtain a preliminary intelligent power monitoring model.

[0019] As a preferred solution of the performance monitoring method of the power quality management device suitable for plateau environment described in the present invention, wherein: the preliminary intelligent power monitoring model is trained by monitoring the data set to obtain the prediction result, calculate the evaluation index, and obtain the intelligent power monitoring model. The specific steps are as follows:

[0020] The monitoring data set is divided into a training set and a test set, the preliminary intelligent power monitoring model is trained using the training set, and the weight of the preliminary intelligent power monitoring model is gradually adjusted;

[0021] The test set is input into the preliminary intelligent power monitoring model to obtain the prediction results, the prediction results are compared with the actual results, and the evaluation indicators are calculated to obtain the intelligent power monitoring model.

[0022] As a preferred solution of the performance monitoring method of the power quality management device suitable for plateau environment described in the present invention, wherein: the real-time monitoring data set is input into the intelligent power monitoring model to obtain the performance index of the power quality management device and generate a performance report. The specific steps are as follows:

[0023] The real-time monitoring data set is input into the intelligent power monitoring model to obtain the performance indicators of the power quality management device, evaluate the operating status of the power quality management device, and draw a line graph. The changing trends of various performance indicators are observed through the graph, multi-dimensional variables are obtained, and a performance report is generated.

[0024] In a second aspect, the present invention provides a method for monitoring the performance of a power quality control device suitable for a plateau environment, comprising:

[0025] Data acquisition module, data preprocessing module, feature fusion module, monitoring model construction module, model training module, real-time monitoring module; data acquisition module, used to collect power quality management device operation data through a high-precision sensor array; data preprocessing module, used to preliminarily screen the power quality management device operation data through statistical methods, and use adaptive filtering algorithm to remove noise and perform normalization processing; feature fusion module, used to automatically extract deep feature vectors of each sensor data through deep learning fusion method, and fuse them into a comprehensive monitoring data set; monitoring model construction module, used to define the long short-term memory network structure through historical monitoring data sets, and build a preliminary intelligent power monitoring model based on softmax activation function and cross entropy loss function; model training module, used to train the preliminary intelligent power monitoring model through the monitoring data set, obtain prediction results, calculate evaluation indicators, and obtain the intelligent power monitoring model; real-time monitoring module, used to input the real-time monitoring data set into the intelligent power monitoring model, obtain the performance indicators of the power quality management device, and generate a performance report.

[0026] In a third aspect, the present invention provides a computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: when the computer program is executed by the processor, it implements any step of the performance monitoring method of a power quality management device suitable for a plateau environment as described in the first aspect of the present invention.

[0027] In a fourth aspect, the present invention provides a computer-readable storage medium having a computer program stored thereon, wherein: when the computer program is executed by a processor, it implements any step of the method for monitoring the performance of a power quality management device suitable for a plateau environment as described in the first aspect of the present invention.

[0028] The beneficial effects of the present invention are as follows: by using convolution operations and pooling layers to process the preprocessed power quality management device operation data, the present invention can efficiently identify local features and significantly reduce data dimensions, build a more accurate and efficient intelligent power monitoring model, and improve the accuracy and reliability of performance monitoring of power quality management devices in plateau environments, thereby ensuring the stability and safety of the power system. BRIEF DESCRIPTION OF THE DRAWINGS

[0029] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other accompanying drawings can be obtained based on these accompanying drawings without paying creative work.

[0030] Figure 1 This is a flow chart of the performance monitoring method of the power quality management device suitable for plateau environment in Example 1.

[0031] Figure 2 This is a schematic diagram of the performance monitoring system of the power quality management device suitable for plateau environment in Example 1.

[0032] Figure 3 Flow chart of processing performance data of power quality management device in Example 1

[0033] Figure 4 This is a flow chart of training the intelligent power monitoring model in Example 1. DETAILED DESCRIPTION

[0034] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the specific implementation methods of the present invention are described in detail below in conjunction with the accompanying drawings.

[0035] In the following description, many specific details are set forth to facilitate a full understanding of the present invention, but the present invention may also be implemented in other ways different from those described herein, and those skilled in the art may make similar generalizations without violating the connotation of the present invention. Therefore, the present invention is not limited to the specific embodiments disclosed below.

[0036] Secondly, the term "one embodiment" or "embodiment" as used herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The term "in one embodiment" that appears in different places in this specification does not necessarily refer to the same embodiment, nor does it refer to a separate or selective embodiment that is mutually exclusive with other embodiments.

[0037] Example 1, reference Figures 1 to 4 , which is the first embodiment of the present invention, and provides a method for monitoring the performance of a power quality control device suitable for a plateau environment, comprising the following steps:

[0038] S1. Collect the operation data of the power quality management device through a high-precision sensor array.

[0039] The grid voltage level, current intensity and equipment operating temperature are collected through a high-precision sensor array, connected to the edge data processor, and uploaded to the nearby edge computing node through wireless communication methods to obtain the operating data of the power quality management device.

[0040] It should be noted that the voltage sensor, current sensor, and temperature sensor are installed first. These sensors are respectively used to accurately measure the grid voltage level, monitor the current intensity, and monitor the operating temperature of the device, ensuring comprehensive coverage of the key parameters for power quality governance. According to the specific layout requirements of the device, the sensors are carefully placed in appropriate positions, such as near the main circuit board and key heat-generating components. All sensors are connected through standardized interfaces (such as analog signal output or digital I / O) and are interconnected using a bus system (such as I2C, SPI, or CAN bus) to form a modular sensor array.

[0041] The sensor data is first transmitted to the edge data processor for preliminary processing of the original sensor data, including operations such as filtering, format conversion, and simple anomaly detection, to reduce the amount of sensor data uploaded to the cloud and improve the real-time response speed. Subsequently, the preprocessed sensor data is uploaded to a nearby edge computing node through wireless communication technologies (such as Wi-Fi, Zigbee, or LoRaWAN). The choice of which wireless communication method depends on the requirements of the specific application scenario, such as factors like the sensor data transmission rate, coverage range, and power consumption. Finally, these sensor data are further analyzed and processed on the edge computing node for long-term storage and in-depth analysis. The sensors collect data once per second and transmit it to the edge computing node through a high-bandwidth communication interface.

[0042] S2. Through statistical methods, the operation data of the power quality governance device is preliminarily screened, and an adaptive filtering algorithm is used to remove noise and perform normalization processing.

[0043] Through statistical methods, the operation data of the power quality governance device is plotted as a time series graph. By viewing the time series graph for anomaly points, the operation data of the power quality governance device with obvious errors is preliminarily screened out, and random noise is removed through an adaptive filtering algorithm.

[0044] It should be noted that through statistical methods, the operation data of the power quality control device is preliminarily screened. First, the operation data of the power quality control device after preliminary screening is plotted as a time series graph. First, ensure that the data is arranged in chronological order and check for missing or outlier values. Fill in or remove them if necessary. Then, select a suitable plotting tool or software (such as the Matplotlib library in Python, Excel, or other data analysis platforms), set the timestamp as the X-axis, and the corresponding measured value as the Y-axis. Then, import the data set into the plotting tool and use its built-in time series plotting function to generate a preliminary time series graph. Further, legends, titles, and axis labels can be added to enhance the readability of the graph, and the line styles, colors, and markers can be adjusted as needed to distinguish different types of measured data. Finally, carefully review the generated time series graph to identify possible trends, periodicities, and outliers, providing an intuitive basis for subsequent data analysis. Obvious outliers are identified by observing the time series graph, and these outliers may represent obvious error extreme values. The preliminary screening process includes calculating basic statistics (such as mean, standard deviation) and setting thresholds to mark data points outside the normal range, thereby removing obviously incorrect samples. Then, the remaining operation data of the power quality control device is further processed using an adaptive filtering algorithm to remove random noise and smooth the signal, ensuring the authenticity and reliability of the operation data of the power quality control device. The adaptive filtering algorithm can automatically adjust the filter parameters according to the dynamic changes of the input signal, effectively eliminating noise without affecting useful information. The operation data of the power quality control device after removing noise is normalized through min-max scaling normalization to obtain the preprocessed operation data of the power quality control device.

[0045] The operation data of the power quality control device after removing noise is normalized through min-max scaling normalization to obtain the preprocessed operation data of the power quality control device.

[0046] It should be noted that the operation data of the power quality control device after removing noise is normalized through the min-max scaling normalization method. First, the operation data of the power quality control device is mapped to a specific range (usually between [0,1]) to eliminate the influence of dimensionality between different features. Specifically, for each feature, calculate its minimum and maximum values in the operation data set of the power quality control device and use the following formula for conversion. This process ensures that all features are on the same scale, thereby improving the efficiency and stability of subsequent preliminary intelligent power monitoring model training. Finally, the operation data of the power quality control device after normalization not only removes noise interference but also has better numerical attributes, providing high-quality input data for the deep learning model.

[0047] It should be noted that the expression is:

[0048]

[0049] Among them, X is the operation data point of the original power quality governance device, and X min is the minimum value in the operation data of the power quality governance device, and X max is the maximum value in the operation data of the power quality governance device, and X' is the operation data point of the power quality governance device after scaling.

[0050] S3. Through the deep learning fusion method, automatically extract the deep feature vectors of each sensor data and fuse them into a comprehensive monitoring data set.

[0051] Use the convolution operation to identify the preprocessed operation data of the power quality governance device by sliding the convolution kernel on the time series and extract local features;

[0052] It should be noted that the deep learning fusion method is used to automatically extract the deep feature vectors of each sensor data and fuse them into a comprehensive monitoring data set. First, perform a convolution operation on the preprocessed operation data of the power quality governance device. By designing convolution kernels of different sizes and sliding them on the time series, local features can be identified and extracted. For example, for multi-source sensor data such as voltage, current, and temperature, the data stream of each sensor can be regarded as a time series. The convolution operation can effectively capture local patterns and change trends in these time series, such as key information like voltage fluctuations, current mutations, and temperature anomalies. Specifically, in the convolution layer, multiple convolution kernels act simultaneously on the time dimension of the input operation data of the power quality governance device, and each convolution kernel is responsible for detecting specific types of local features. The parameters of the convolution kernel are automatically adjusted during the training process so that the most representative features can be learned adaptively. By stacking multiple convolution layers, higher-level abstract features can be gradually extracted, from simple edges and textures (corresponding to basic waveforms and transient changes in signal processing) to complex structures and patterns (such as periodic fluctuations and long-term trends).

[0053] The pooling layer reduces the dimension of the preprocessed operation data of the power quality governance device. By stacking multiple convolution layers and pooling layers, deep feature vectors are obtained;

[0054] It should be noted that by applying the pooling layer, the spatial dimension of the feature map is reduced by selecting the maximum value in each local area, while the most important feature information is retained. By stacking multiple convolutional layers and pooling layers, deep feature vectors are automatically extracted to further capture complex patterns and high-level abstract information in the operation data of the power quality governance device. In addition, the pooling layer is usually used alternately with the convolutional layer to reduce the dimension of the operation data of the power quality governance device and lower the computational complexity. Next, through methods such as concatenation or weighted averaging, the deep feature vectors extracted from the data of each sensor are fused into a comprehensive monitoring data set. This fusion not only retains the unique features of the data of each sensor, but also integrates multi-source information, providing a more comprehensive perspective to evaluate the overall performance of the power quality governance device. For example, feature vectors such as voltage fluctuations, current mutations, and temperature anomalies are concatenated by columns to form a high-dimensional feature matrix, which contains the key information of all sensor data. Then, the deep feature vectors extracted from the data of each sensor are fused to form a comprehensive monitoring data set, which integrates information from different sources and provides a more comprehensive and richer perspective to evaluate the overall performance of the device. This process not only enhances the ability to understand multi-source heterogeneous data, but also improves the accuracy and robustness of subsequent analysis tasks such as fault diagnosis and performance evaluation, providing strong support for achieving precise power quality monitoring and governance.

[0055] Through concatenation fusion, all deep feature vectors are concatenated by columns to obtain a comprehensive monitoring data set.

[0056] It should be noted that through the concatenation fusion method, the deep feature vectors extracted from the data of each sensor are concatenated by columns to form a comprehensive monitoring data set. Specifically, the data of each sensor generates corresponding deep feature vectors after being processed by the convolutional layer and the pooling layer. These feature vectors contain rich local and global information and reflect the operation status of the power quality governance device in different aspects. Next, these feature vectors are concatenated according to the column dimension, that is, each feature vector is regarded as a column and stacked vertically, thus constructing a high-dimensional feature matrix containing all sensor information. This concatenation method not only retains the unique features of the data of each sensor, but also integrates multi-source information, providing a more comprehensive perspective to evaluate the overall performance of the device. Finally, the obtained comprehensive monitoring data set provides strong support for subsequent tasks such as anomaly detection, fault diagnosis, and performance evaluation.

[0057] S4. Define the long short-term memory network structure through the historical monitoring data set, and construct a preliminary intelligent power monitoring model based on the softmax activation function and the cross-entropy loss function.

[0058] Define the input layer based on the historical monitoring dataset, set the time step and the number of features, add multiple LSTM layers, and define the output layer through the softmax activation function and the cross-entropy loss function to obtain a preliminary intelligent power monitoring model.

[0059] It should be noted that first, it depends on the historical monitoring dataset to define the long short-term memory network (LSTM) structure. This process begins with the design of the input layer, where the time step and the number of features need to be specified. For example, if each sample contains 5 time steps and each time step has 10 features (such as voltage, current, frequency, etc.), the shape of the input layer can be set to (number of samples, 5, 10). Such a setting ensures that the preliminary intelligent power monitoring model can process power quality data with time series characteristics, thereby capturing the dynamic change relationships between different time points. Next, add multiple LSTM layers after the input layer. The LSTM layer is a neural network layer specifically designed to process and predict time series data. They effectively learn complex patterns and dependencies within a long time span through a unique gating mechanism (input gate, forget gate, and output gate). The multi-layer LSTM architecture further enhances the learning ability of the preliminary intelligent power monitoring model, enabling it to not only capture short-term fluctuations but also understand long-term trends. Each layer of LSTM units is connected in a fully connected manner, allowing information to be transmitted between different levels to improve the overall expression ability of the preliminary intelligent power monitoring model. After the LSTM layer, introduce a fully connected layer and use the softmax activation function to convert the output into a probability distribution of each category. The Softmax function is particularly suitable for multi-classification problems because it can convert the output values of each node into probability values, and the sum of all output probabilities is 1. This means that for each input sample, the preliminary intelligent power monitoring model will give a probability estimate of belonging to each category. In addition, the cross-entropy loss function is used as the optimization objective. This loss function can quantify the difference between the predicted value and the true label, especially suitable for classification tasks. By minimizing the cross-entropy loss, the parameters of the preliminary intelligent power monitoring model can be gradually adjusted to make the prediction results as close as possible to the actual labels. During the entire training process, calculate the output and its loss value through forward propagation, and then use the backpropagation algorithm and an optimizer (such as Adam or SGD) to update the weight parameters, gradually reducing the value of the loss function. After multiple iterations, the preliminary intelligent power monitoring model gradually converges and achieves good performance. Finally, the preliminary intelligent power monitoring model constructed based on this structure can not only accurately identify and classify power states but also has good generalization ability and robustness, suitable for complex practical application scenarios.

[0060] S5. Train the preliminary intelligent power monitoring model through the monitoring dataset to obtain prediction results, calculate evaluation metrics, and obtain the intelligent power monitoring model.

[0061] Divide the monitoring data set into a training set and a test set, and use the training set to train the preliminary intelligent power monitoring model, gradually adjusting the weights of the preliminary intelligent power monitoring model;

[0062] It should be noted that the monitoring data set is divided into a training set and a test set by random sampling to ensure consistent and unbiased distribution between the two. Usually, it is divided according to a certain ratio (for example, 80% of the monitoring data is used for training and 20% of the monitoring data is used for testing) to ensure that the preliminary intelligent power monitoring model has sufficient monitoring data for learning and can effectively evaluate its performance. Use the training set to train the preliminary intelligent power monitoring model. First, initialize the weight parameters of the preliminary intelligent power monitoring model, and calculate the output result and the loss value between it and the true label through forward propagation. Then, use the backpropagation algorithm and an optimizer (such as Adam or SGD) to gradually adjust the weights of the preliminary intelligent power monitoring model to minimize the loss function (such as cross-entropy loss), so that the preliminary intelligent power monitoring model can accurately predict the power state. In each iteration process, the weight parameters are continuously updated according to the feedback of the loss value until the preliminary intelligent power monitoring model converges or reaches the predetermined number of training epochs. This process not only ensures that the preliminary intelligent power monitoring model can effectively learn the features and patterns in the training set, but also improves the generalization ability and robustness of the preliminary intelligent power monitoring model through multiple iterations of optimization. Finally, an intelligent power monitoring model with excellent performance is obtained, providing a reliable guarantee for the power state classification in practical applications. This method significantly improves the performance of the intelligent power monitoring model on unseen monitoring data through systematic training and weight adjustment.

[0063] It should also be noted that the test set is input into the preliminary intelligent power monitoring model to obtain the prediction results of each sample, which represent the classification predictions of the preliminary intelligent power monitoring model for the power states at each time point. Then, the prediction results are compared with the true labels in the test set one by one, and a series of evaluation metrics (such as accuracy, precision, recall, and F1-score) are calculated to quantify the performance of the intelligent power monitoring model. Specifically, accuracy measures the proportion of correct predictions among all predictions, precision reflects the proportion of samples predicted as positive classes that are actually positive classes, recall represents the proportion of samples that are actually positive classes and are correctly predicted as positive classes, and the F1-score is the harmonic mean of precision and recall, providing a comprehensive evaluation. Finally, through this systematic evaluation method, the high accuracy and robustness of the intelligent power monitoring model are ensured, enabling it to reliably identify and classify power states in practical applications, thus providing strong support for power management.

[0064] S6. Input the real-time monitoring data set into the intelligent power monitoring model to obtain the performance indicators of the power quality governance device and generate a performance report. The specific steps are as follows.

[0065] Input the real-time monitoring data set into the intelligent power quality monitoring model to obtain the performance indicators of the power quality control device, evaluate the operating status of the power quality control device, draw a line chart, observe the change trends of various performance indicators through the chart, obtain multi-dimensional variables, and generate a performance report.

[0066] It should be noted that inputting the real-time monitoring data set into the trained intelligent power quality monitoring model can obtain various performance indicators of the power quality control device, such as voltage deviation, frequency fluctuation, harmonic content, etc. These indicators reflect the performance of the power quality control device during actual operation. Then, evaluate these performance indicators, and observe the change trends of each indicator through a chart. For example, draw a line chart to show the voltage deviation and frequency fluctuation in different time periods to identify potential problems or optimization opportunities. The analysis of multi-dimensional variables not only includes the trend changes of individual indicators but also covers the mutual relationships between different indicators, providing a basis for comprehensively understanding the device operating status. Finally, generate a detailed performance report based on these analysis results. This report summarizes all key performance indicators and their change trends and provides targeted improvement suggestions. For example, if it is found that the power factor is lower than the standard value in some periods, it is recommended to install or upgrade reactive power compensation equipment; if the current harmonic content is detected to exceed the standard, it is recommended to add filters to purify the power system; for the continuous voltage fluctuation situation, consider optimizing the power grid structure or introducing voltage stabilizing devices to stabilize the voltage level. This method ensures the accuracy and comprehensiveness of the performance evaluation of the power quality control device through systematic data analysis and visualization means of the power quality control device performance, providing strong support for subsequent maintenance and optimization work. This process not only improves the reliability and efficiency of the device operation but also provides clear and intuitive reference information for decision-makers.

[0067] This embodiment also provides a performance monitoring method for a power quality control device applicable to a plateau environment, including: a data acquisition module, a data preprocessing module, a feature fusion module, a monitoring model construction module, a model training module, and a real-time monitoring module;

[0068] The data acquisition module is used to collect the operation data of the power quality control device through a high-precision sensor array;

[0069] The data preprocessing module is used to preliminarily screen the operation data of the power quality control device through statistical methods, and use an adaptive filtering algorithm to remove noise and perform normalization processing;

[0070] The feature fusion module is used to automatically extract the deep feature vectors of each sensor data through a deep learning fusion method and fuse them into a comprehensive monitoring data set;

[0071] The monitoring model building module is used to define the long short-term memory network structure through the historical monitoring data set and build a preliminary intelligent power monitoring model based on the softmax activation function and the cross entropy loss function;

[0072] A model training module is used to train the preliminary intelligent power monitoring model through the monitoring data set, obtain the prediction results, calculate the evaluation index, and obtain the intelligent power monitoring model;

[0073] The real-time monitoring module is used to input the real-time monitoring data set into the intelligent power monitoring model, obtain the performance indicators of the power quality management device, and generate a performance report.

[0074] This embodiment also provides a computer device, which is suitable for the performance monitoring method of the power quality management device in a plateau environment, including: a memory and a processor; the memory is used to store computer executable instructions, and the processor is used to execute computer executable instructions to implement the performance monitoring method of the power quality management device suitable for a plateau environment as proposed in the above embodiment.

[0075] The computer device may be a terminal, and the computer device includes a processor, a memory, a communication interface, a display screen and an input device connected via a system bus. The processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The communication interface of the computer device is used to communicate with an external terminal in a wired or wireless manner, and the wireless manner can be achieved through WIFI, an operator network, NFC (near field communication) or other technologies. The display screen of the computer device may be a liquid crystal display screen or an electronic ink display screen, and the input device of the computer device may be a touch layer covering the display screen, or a key, trackball or touchpad provided on the housing of the computer device, or an external keyboard, touchpad or mouse, etc.

[0076] This embodiment also provides a storage medium, on which a computer program is stored. When the program is executed by a processor, it implements the method for monitoring the performance of a power quality governance device applicable to the plateau environment as proposed in the above embodiment; the storage medium can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (SRAM for short), electrically erasable programmable read-only memory (EEPROM for short), erasable programmable read-only memory (EPROM for short), programmable read-only memory (PROM for short), read-only memory (ROM for short), magnetic memory, flash memory, magnetic disk or optical disc.

[0077] In summary, through the following steps: using convolution operations and pooling layers to process the preprocessed operation data of the power quality governance device, the present invention can efficiently identify local features, significantly reduce the data dimension, construct a more accurate and efficient intelligent power monitoring model, improve the accuracy and reliability of the performance monitoring of the power quality governance device in the plateau environment, and thus ensure the stability and security of the power system.

[0078] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit them. Although the present invention has been described in detail with reference to the preferred embodiments, those of ordinary skill in the art should understand that the technical solutions of the present invention can be modified or equivalently replaced without departing from the spirit and scope of the technical solutions of the present invention, and they should all be covered by the scope of the claims of the present invention.

Claims

1. A method for monitoring the performance of a power quality control device suitable for plateau environments, characterized in that: include, Collect the operation data of power quality management devices through high-precision sensor array; The operating data of the power quality management device is preliminarily screened by statistical methods, and the adaptive filtering algorithm is used to remove noise and perform normalization processing; Through deep learning fusion methods, deep feature vectors of each sensor data are automatically extracted and fused into a comprehensive monitoring data set; The long short-term memory network structure is defined through the historical monitoring data set, and a preliminary intelligent power monitoring model is constructed based on the softmax activation function and the cross entropy loss function; The preliminary intelligent power monitoring model is trained through the monitoring data set, the prediction results are obtained, the evaluation index is calculated, and the intelligent power monitoring model is obtained; The real-time monitoring data set is input into the intelligent power monitoring model to obtain the performance indicators of the power quality management device and generate a performance report.

2. The method for monitoring the performance of a power quality control device suitable for a plateau environment as claimed in claim 1, characterized in that: The operation data of the power quality control device is collected by a high-precision sensor array. The specific steps are as follows: The grid voltage level, current intensity and equipment operating temperature are collected through a high-precision sensor array, connected to the edge data processor, and uploaded to the nearby edge computing node through wireless communication methods to obtain the operating data of the power quality management device.

3. The method for monitoring the performance of a power quality control device suitable for a plateau environment as claimed in claim 2, characterized in that: The statistical method is used to preliminarily screen the operation data of the power quality management device, and the adaptive filtering algorithm is used to remove noise and perform normalization. The specific steps are as follows: Through statistical methods, the operation data of the power quality management device is plotted into a time series graph, and the abnormal points are checked through the time series graph to preliminarily screen out the obviously erroneous operation data of the power quality management device, and the random noise is removed through the adaptive filtering algorithm; The operation data of the power quality management device after noise removal is normalized through minimum and maximum scaling normalization to obtain the preprocessed operation data of the power quality management device.

4. The method for monitoring the performance of a power quality control device suitable for a plateau environment as claimed in claim 3 is characterized in that: The deep learning fusion method is used to automatically extract the deep feature vectors of each sensor data and fuse them into a comprehensive monitoring data set. The specific steps are as follows: The convolution operation is used to identify the pre-processed power quality management device operation data by sliding the convolution kernel on the time series to extract local features; The pooling layer reduces the dimension of the pre-processed power quality management device operation data, and obtains a deep feature vector by stacking multiple convolutional layers and pooling layers; All deep feature vectors are concatenated column by column through splicing and fusion to obtain a comprehensive monitoring dataset.

5. The method for monitoring the performance of a power quality control device suitable for a plateau environment as claimed in claim 4, characterized in that: The long short-term memory network structure is defined through the historical monitoring data set, and a preliminary intelligent power monitoring model is constructed based on the softmax activation function and the cross entropy loss function. The specific steps are as follows: Based on the historical monitoring data set, the input layer is defined, the time step and the number of features are set, multiple LSTM layers are added, and the output layer is defined through the softmax activation function and the cross entropy loss function to obtain a preliminary intelligent power monitoring model.

6. The method for monitoring the performance of a power quality control device suitable for a plateau environment as claimed in claim 5, characterized in that: The specific steps of training the preliminary intelligent power monitoring model through the monitoring data set, obtaining the prediction results, calculating the evaluation index, and obtaining the intelligent power monitoring model are as follows: The monitoring data set is divided into a training set and a test set, the preliminary intelligent power monitoring model is trained using the training set, and the weight of the preliminary intelligent power monitoring model is gradually adjusted; The test set is input into the preliminary intelligent power monitoring model to obtain the prediction results, the prediction results are compared with the actual results, and the evaluation indicators are calculated to obtain the intelligent power monitoring model.

7. The method for monitoring the performance of a power quality control device suitable for a plateau environment as claimed in claim 6, characterized in that: The specific steps of inputting the real-time monitoring data set into the intelligent power monitoring model, obtaining the performance index of the power quality management device, and generating a performance report are as follows: The real-time monitoring data set is input into the intelligent power monitoring model to obtain the performance indicators of the power quality management device, evaluate the operating status of the power quality management device, and draw a line graph. The changing trends of various performance indicators are observed through the graph, multi-dimensional variables are obtained, and a performance report is generated.

8. A power quality control device performance monitoring system suitable for plateau environments, based on the power quality control device performance monitoring method suitable for plateau environments as described in any one of claims 1 to 7, characterized in that: Including data acquisition module, data preprocessing module, feature fusion module, monitoring model building module, model training module, real-time monitoring module; The data acquisition module is used to collect the operation data of the power quality management device through a high-precision sensor array; The data preprocessing module is used to preliminarily screen the operation data of the power quality management device through statistical methods, and use the adaptive filtering algorithm to remove noise and perform normalization processing; The feature fusion module is used to automatically extract the deep feature vectors of each sensor data through deep learning fusion method and fuse them into a comprehensive monitoring data set; The monitoring model building module is used to define the long short-term memory network structure through the historical monitoring data set and build a preliminary intelligent power monitoring model based on the softmax activation function and the cross entropy loss function; A model training module is used to train the preliminary intelligent power monitoring model through the monitoring data set, obtain the prediction results, calculate the evaluation index, and obtain the intelligent power monitoring model; The real-time monitoring module is used to input the real-time monitoring data set into the intelligent power monitoring model, obtain the performance indicators of the power quality management device, and generate a performance report.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the method for monitoring the performance of a power quality management device suitable for a plateau environment as described in any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method for monitoring the performance of a power quality management device suitable for a plateau environment as described in any one of claims 1 to 7 are implemented.

Images