Dynamic environment monitoring data real-time processing method and system based on edge computing

Abstract

The invention discloses a dynamic environment monitoring data real-time processing method and system based on edge computing, and relates to the technical field of data processing. The method comprises the following steps: respectively deploying an acquisition terminal to obtain different acquisition data, obtaining a plurality of monitoring point locations to deploy a monitoring terminal, obtaining abnormal data sets and corresponding abnormal data features, constructing a first evaluation model according to the abnormal data sets of different fault types and the abnormal data features thereof, obtaining processing parameters of different abnormal data sets, and establishing a second evaluation model according to the processing parameters. Constructing a second evaluation model by combining the processing time consumption of the monitoring terminal on different abnormal data sets under different operation parameters, acquiring an optimal monitoring terminal in real time by utilizing the second evaluation model, and judging whether the abnormal data sets have fault events or not by utilizing the first evaluation model; data processing of the edge end of the power equipment can be realized, and the data processing efficiency can be improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of data processing, and particularly relates to a dynamic environment monitoring data real-time processing method and system based on edge computing. BACKGROUND

[0002] Dynamic environment monitoring, full name of which is dynamic and environmental monitoring, is a system for real-time monitoring of power equipment and environmental parameters of key facilities, which aims to ensure the safe and stable operation of power equipment. In the prior art, the data of power equipment often needs to be sent to the cloud for processing, which inevitably leads to high delay and low efficiency. Therefore, using edge computing technology to process the data of power equipment in real time has become a new technology direction. In the prior art, the deployment and use of edge terminals are often fixed and unchanged, and cannot be flexibly adjusted according to the state of edge terminals and data to be processed, so that the efficiency of data processing is still high. In view of the deficiencies of the prior art, the present application provides a dynamic environment monitoring data real-time processing method and system based on edge computing. SUMMARY

[0003] The present application aims to provide a dynamic environment monitoring data real-time processing method and system based on edge computing.

[0004] The purpose of the present application can be achieved by the following technical solution: a dynamic environment monitoring data real-time processing system based on edge computing, comprising the following modules: A device monitoring module is used to deploy a collection terminal for each power equipment to obtain different collection data, and to obtain a plurality of monitoring points in an application scenario to deploy a monitoring terminal. A fault evaluation module is used to obtain an abnormal data set according to a preset filtering range combined with different collection data, and to obtain corresponding abnormal data features, and to construct a first evaluation model according to the abnormal data set and the abnormal data features of different fault types. A time-consuming evaluation module is used to obtain processing parameters of different abnormal data sets, and to construct a second evaluation model combined with the processing time of different abnormal data sets by the monitoring terminal under different operating parameters. A monitoring feedback module is used to obtain the optimal monitoring terminal of different abnormal data sets in real time by using the second evaluation model, and to determine whether there is a fault event in the abnormal data set by using the first evaluation model in the optimal monitoring terminal and to feed back.

[0005] Further, the process of deploying a collection terminal for each power equipment to obtain different collection data, and obtaining a plurality of monitoring points in an application scenario to deploy a monitoring terminal comprises: Each power equipment in the same application scenario is respectively deployed with a collection terminal, different collection units are arranged in the collection terminal, and corresponding collection data are obtained in real time through each collection unit, including voltage, current, temperature, humidity, pressure and amplitude; The deployment positions of each collection terminal in the same application scenario are obtained, a GIS distribution map of all collection terminals is constructed according to the obtained deployment positions by using GIS technology, an initial monitoring point is obtained in the GIS distribution map, and a probability value of each deployment position being selected as a next monitoring point is obtained; The next monitoring point is constructed at the deployment position corresponding to the maximum probability value, and the step is repeated until n monitoring points are constructed, and a monitoring terminal is respectively deployed at each monitoring point.

[0006] Further, the process of obtaining an abnormal data set according to the preset filtering range and different collection data and obtaining corresponding abnormal data features includes: Corresponding filtering ranges are respectively set for different collection data, a judgment unit is respectively arranged in each collection terminal, and whether each collection data of each collection terminal is within its filtering range is judged; When any collection data of a single collection terminal is not within its filtering range, the current time is marked as an abnormal time, and the values of the collection data obtained by each collection terminal in the same application scenario at the abnormal time are all included in a first abnormal data set; All values of the collection data of the collection terminal between the current time and the last abnormal time are included in a second abnormal data set, and the data features of the first abnormal data set and the second abnormal data set are respectively taken as the first abnormal data features and the second abnormal data features, and the abnormal data set and the abnormal data features of each collection data of each collection terminal are obtained.

[0007] Further, the process of constructing a first evaluation model according to the abnormal data set and the abnormal data features of different fault types includes: The historical fault records of different fault types are obtained, the historical fault records contain the abnormal data set and the abnormal data features at the corresponding abnormal time, a first evaluation set is generated according to the abnormal data set, the abnormal data features and the corresponding fault type contained in different historical fault records, and the first evaluation set is divided into a first training set and a first test set; A first convolutional neural network is constructed, the abnormal data set and the abnormal data features in the first training set are taken as the input data of the first convolutional neural network, and whether a fault event exists and the fault type are taken as the output data of the first convolutional neural network; The first convolutional neural network is trained using the first training set to obtain an initial first convolutional neural network. The initial first convolutional neural network is validated using the first test set, and the initial first convolutional neural network that is less than or equal to a preset first test error threshold is output as the first evaluation model.

[0008] Furthermore, the process of acquiring processing parameters for different abnormal data sets and constructing a second evaluation model based on the processing time of different abnormal data sets under different operating parameters by the monitoring terminal includes: The first evaluation model is deployed in each monitoring terminal. A single abnormal data set and its abnormal data characteristics are input into the first evaluation model in each monitoring terminal, and the evaluation results are output respectively. The time interval from data input to result output of each first evaluation model is used as its processing time. The processing parameters of the single abnormal data set and the operating parameters of each monitoring terminal are obtained. The processing parameters refer to the data volume and data packet size of the single abnormal data set, and the operating parameters refer to the real-time status parameters of the server preset in each monitoring terminal. The processing time and corresponding operating parameters of abnormal data sets with different processing parameters are obtained on each monitoring terminal. A second evaluation set is generated based on the processing time and operating parameters of abnormal data sets with different processing parameters on each monitoring terminal, and it is divided into a second training set and a second test set. Construct a second convolutional neural network, using different processing parameters and running parameters from the second training set as input data for the second convolutional neural network, and using the corresponding processing time from the second training set as output data for the second convolutional neural network; The second convolutional neural network is trained using the second training set to obtain an initial second convolutional neural network. The initial second convolutional neural network is validated using the second test set, and the output of the initial second convolutional neural network that is less than or equal to the preset second test error threshold is used as the second evaluation model.

[0009] Furthermore, the process of using the second evaluation model to obtain the optimal monitoring terminal for different sets of abnormal data in real time includes: A second evaluation model is deployed in each monitoring terminal. When a single collection terminal generates a single abnormal data set, the abnormal data characteristics and processing parameters corresponding to the abnormal data set are obtained, and the processing parameters are sent to the monitoring terminal closest to the collection terminal. The operating parameters of each monitoring terminal are synchronized in real time. In the monitoring terminal closest to the acquisition terminal, the processing parameters of the abnormal data set and the operating parameters of each monitoring terminal are respectively input into the second evaluation model to obtain the processing time of each monitoring terminal for the abnormal data set. The monitoring terminal with the shortest processing time is selected as the optimal monitoring terminal for the abnormal data set.

[0010] Furthermore, the process of using the first evaluation model within the optimal monitoring terminal to determine whether there are fault events in its abnormal data set and then reporting them includes: The single abnormal data set and its abnormal data characteristics of a single acquisition terminal are sent to its optimal monitoring terminal. In the optimal monitoring terminal, the received abnormal data set and abnormal data characteristics are input into the first evaluation model to determine whether there is a fault event in the abnormal data set. If there is, the fault type is obtained and fed back to the relevant personnel. If there is no fault, no other operation is performed.

[0011] A real-time processing method for environmental monitoring data based on edge computing includes the following steps: Step S1: Deploy acquisition terminals for each power device to obtain different acquisition data, and obtain multiple monitoring points in the application scenario to deploy monitoring terminals; Step S2: Obtain an abnormal data set based on the preset filtering range and different collected data, and obtain the corresponding abnormal data features. Construct a first evaluation model based on the abnormal data sets and abnormal data features of different fault types. Step S3: Obtain the processing parameters for different abnormal data sets, and construct a second evaluation model by combining the processing time of different abnormal data sets under different operating parameters of the monitoring terminal; Step S4: Use the second evaluation model to obtain the optimal monitoring terminal for different abnormal data sets in real time, and use the first evaluation model in the optimal monitoring terminal to determine whether there are fault events in its abnormal data set and provide feedback.

[0012] Compared with the prior art, the beneficial effects of the present invention are: This invention enables data processing at the edge of power equipment by setting up corresponding acquisition terminals and several monitoring terminals for each power equipment. This solves the technical problems of high latency and low efficiency in cloud processing. By setting a filtering range, data within the normal range can be filtered out, thereby effectively reducing the amount of data processing and improving the efficiency of data processing. By acquiring two types of abnormal data sets and their abnormal data characteristics from the same acquisition terminal, it is possible to comprehensively reflect the data changes of the power equipment during a fault based on the data characteristics of the same power equipment at different times, as well as the data characteristics of different power equipment at the same time. A first evaluation model is then constructed, which is beneficial for forming an effective fault judgment mechanism for the power equipment. A second evaluation model is constructed based on the processing time of the monitoring terminal under different processing and operating parameters. This model can allocate the optimal monitoring terminal to different abnormal data sets in real time for data processing, thereby improving the efficiency of data processing. Attached Figure Description

[0013] Figure 1 This is a schematic diagram of the modules of the present invention. Detailed Implementation

[0014] like Figure 1 As shown, a real-time environmental monitoring data processing system based on edge computing includes the following modules: The equipment monitoring module is used to deploy acquisition terminals for each power equipment to obtain different acquisition data, and to obtain multiple monitoring points in the application scenario to deploy monitoring terminals. The fault assessment module is used to obtain abnormal data sets based on a preset filtering range and different collected data, and to obtain the corresponding abnormal data characteristics. Based on the abnormal data sets and abnormal data characteristics of different fault types, a first assessment model is constructed. The time consumption evaluation module is used to obtain the processing parameters of different abnormal data sets, and to build a second evaluation model by combining the processing time of different abnormal data sets under different operating parameters of the monitoring terminal. The monitoring feedback module is used to obtain the optimal monitoring terminal for different abnormal data sets in real time using the second evaluation model, and to use the first evaluation model in the optimal monitoring terminal to determine whether there are fault events in its abnormal data set and provide feedback.

[0015] It should be further explained that, in the specific implementation process, data acquisition terminals are deployed separately for each power unit to obtain different data. The process of acquiring multiple monitoring points and deploying monitoring terminals in the application scenario includes: For each power device in the same application scenario, a corresponding data acquisition terminal is deployed. The data acquisition terminal is equipped with different data acquisition units, including voltage acquisition unit, current acquisition unit, temperature acquisition unit, humidity acquisition unit, pressure acquisition unit, and amplitude acquisition unit. The corresponding data, including voltage, current, temperature, humidity, pressure, and amplitude, are acquired in real time through each data acquisition unit. The deployment locations of each data acquisition terminal in the same application scenario are obtained to reflect the specific coordinates of each data acquisition terminal in the application scenario. Based on the obtained deployment locations, a GIS distribution map of all data acquisition terminals is constructed using GIS technology. In the GIS distribution map, the deployment location of any data acquisition terminal is taken as the initial monitoring point, and the shortest distance d between each other deployment location and the existing monitoring point is obtained. i Where i = 1, 2, ..., k, k is the number of each deployment location other than the current monitoring point in this application scenario, and the probability value P of each other deployment location being selected as the next monitoring point is obtained. i ; ;

[0016] Build the next monitoring point at the deployment location corresponding to the highest probability value, and repeat this step until n monitoring points are built. Upload the n built monitoring points to the GIS distribution map for synchronization, and deploy monitoring terminals for data processing at each monitoring point.

[0017] It should be further explained that, in the specific implementation process, the process of obtaining an abnormal data set based on a preset filtering range and different collected data, and obtaining the corresponding abnormal data characteristics, includes: A corresponding filtering range is set for different collected data. The filtering range is a fixed numerical range. A judgment unit is set in each collection terminal to determine whether each collected data of each collection terminal is within (or equal to) its corresponding filtering range. Taking any acquisition terminal as an example, when each acquisition data is within its filtering range, no other operation is performed on it. When a certain acquisition data is not within its filtering range, the current time is marked as an abnormal time, and the values ​​of the acquisition data obtained by each acquisition terminal in the same application scenario under the abnormal time are all included in the first abnormal data set. All values ​​of the data collected by the acquisition terminal between the current time and the previous abnormal time are included in the second abnormal data set. The first abnormal data set and the second abnormal data set are both abnormal data sets of the acquisition terminal. The data characteristics of the first abnormal data set and the second abnormal data set are respectively used as their first abnormal data characteristics and second abnormal data characteristics. Both are abnormal data characteristics of the corresponding acquisition terminal. Since the first abnormal data set consists of multiple values ​​at the same time, its data characteristics include mean and variance. Since the second abnormal data set consists of multiple values ​​at different times, its data characteristics include mean, variance, and trend. This method is used to obtain the abnormal data sets and abnormal data characteristics of each data collection terminal.

[0018] It should be further explained that, in the specific implementation process, the process of constructing the first evaluation model based on the abnormal data sets and their abnormal data characteristics for different fault types includes: Acquire historical fault records for different fault types. The historical fault records refer to the relevant records of fault events that have occurred in the past of the power equipment, which include the abnormal data set and its abnormal data characteristics at the corresponding abnormal time. The first evaluation set is generated based on the abnormal data sets, abnormal data characteristics and corresponding fault types contained in different historical fault records, and is divided into the first training set and the first test set. Construct a first convolutional neural network, using the abnormal data set and abnormal data features in the first training set as input data for the first convolutional neural network, and using the existence of fault events and their fault types as output data for the first convolutional neural network; The first convolutional neural network is trained using the first training set to obtain an initial first convolutional neural network. The initial first convolutional neural network is validated using the first test set, and the initial first convolutional neural network with an error less than or equal to a preset first test error threshold is output as the corresponding first evaluation model.

[0019] It should be further explained that, in the specific implementation process, the process of obtaining processing parameters for different abnormal data sets and constructing a second evaluation model based on the processing time of different abnormal data sets under different operating parameters by the monitoring terminal includes: The constructed first evaluation model is deployed in each monitoring terminal. A single abnormal data set and its abnormal data characteristics are input into the first evaluation model in each monitoring terminal, and the corresponding evaluation results are output. The time interval from data input to result output of each first evaluation model is taken as its processing time. The processing parameters of the single abnormal data set and the operating parameters of each monitoring terminal during the evaluation process of its first evaluation model are obtained. The processing parameters refer to the data volume and data packet size of the single abnormal data set, and the operating parameters refer to the real-time status parameters of the server preset in each monitoring terminal, including CPU utilization, number and frequency of CPU cores, memory utilization, memory read and write speed, hard disk utilization, hard disk read and write speed, network bandwidth utilization, and network latency. The processing time and corresponding operating parameters of abnormal data sets with different processing parameters are obtained on each monitoring terminal. A second evaluation set is generated based on the processing time and operating parameters of abnormal data sets with different processing parameters on each monitoring terminal, and it is divided into a second training set and a second test set. Construct a second convolutional neural network, using different processing parameters and running parameters from the second training set as input data for the second convolutional neural network, and using the corresponding processing time from the second training set as output data for the second convolutional neural network; The second convolutional neural network is trained using the second training set to obtain an initial second convolutional neural network. The initial second convolutional neural network is then validated using the second test set, and the output of the initial second convolutional neural network that is less than or equal to the preset second test error threshold is used as the corresponding second evaluation model.

[0020] It should be further explained that, in the specific implementation process, the process of using the second evaluation model to obtain the optimal monitoring terminal for different abnormal data sets in real time includes: The constructed second evaluation model is deployed in each monitoring terminal. Taking a single acquisition terminal as an example, when the acquisition terminal generates an abnormal data set, the abnormal data characteristics and processing parameters corresponding to the abnormal data set are obtained, and the processing parameters are sent to the monitoring terminal closest to the acquisition terminal. The operating parameters of each monitoring terminal are synchronized in real time. In the monitoring terminal closest to the acquisition terminal, the processing parameters of the abnormal data set and the operating parameters of each monitoring terminal are respectively input into the second evaluation model to obtain the processing time of each monitoring terminal for the abnormal data set. The monitoring terminal with the shortest processing time is selected as the optimal monitoring terminal for the abnormal data set.

[0021] It should be further explained that, in the specific implementation process, the process of using the first evaluation model within the optimal monitoring terminal to determine whether there are fault events in its abnormal data set and to report them includes: The abnormal data set and its abnormal data characteristics from a single acquisition terminal are sent to the corresponding optimal monitoring terminal. In the optimal monitoring terminal, the received abnormal data set and abnormal data characteristics are input into the first evaluation model to determine whether there is a fault event in the abnormal data set. If there is, the fault type is obtained and fed back to the relevant personnel to prompt them to repair the power equipment in a timely manner. If there is no fault, no other operation is performed.

[0022] The embodiments of the present invention also include a real-time processing method for environmental monitoring data based on edge computing, comprising the following steps: Step S1: Deploy acquisition terminals for each power device to obtain different acquisition data, and obtain multiple monitoring points in the application scenario to deploy monitoring terminals; Step S2: Obtain an abnormal data set based on the preset filtering range and different collected data, and obtain the corresponding abnormal data features. Construct a first evaluation model based on the abnormal data sets and abnormal data features of different fault types. Step S3: Obtain the processing parameters for different abnormal data sets, and construct a second evaluation model by combining the processing time of different abnormal data sets under different operating parameters of the monitoring terminal; Step S4: Use the second evaluation model to obtain the optimal monitoring terminal for different abnormal data sets in real time, and use the first evaluation model in the optimal monitoring terminal to determine whether there are fault events in its abnormal data set and provide feedback.

[0023] The above embodiments are only used to illustrate the technical methods of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical methods of the present invention without departing from the spirit and scope of the technical methods of the present invention.

Claims

1. A real-time processing system for environmental monitoring data based on edge computing, characterized in that, Includes the following modules: The equipment monitoring module is used to deploy acquisition terminals for each power equipment to obtain different acquisition data, and to obtain multiple monitoring points in the application scenario to deploy monitoring terminals. The fault assessment module is used to obtain abnormal data sets based on a preset filtering range and different collected data, and to obtain the corresponding abnormal data characteristics. Based on the abnormal data sets and abnormal data characteristics of different fault types, a first assessment model is constructed. The time consumption evaluation module is used to obtain the processing parameters of different abnormal data sets, and to build a second evaluation model by combining the processing time of different abnormal data sets under different operating parameters of the monitoring terminal. The monitoring feedback module is used to obtain the optimal monitoring terminal for different abnormal data sets in real time using the second evaluation model, and to use the first evaluation model in the optimal monitoring terminal to determine whether there are fault events in its abnormal data set and provide feedback.

2. The real-time processing system for environmental monitoring data based on edge computing according to claim 1, characterized in that, The process of acquiring and collecting data and deploying monitoring terminals includes: For each power device in the same application scenario, a data acquisition terminal is deployed. Each data acquisition terminal is equipped with different data acquisition units. The corresponding data is acquired in real time through each data acquisition unit, including voltage, current, temperature, humidity, pressure, and amplitude. The deployment locations of each data acquisition terminal in the same application scenario are obtained. GIS technology is used to construct a GIS distribution map of all data acquisition terminals based on the obtained deployment locations. The initial monitoring point is obtained from the GIS distribution map, and the probability value of each other deployment location being selected as the next monitoring point is obtained. Build the next monitoring point at the deployment location corresponding to the highest probability value, and repeat this step until n monitoring points are built, and deploy monitoring terminals at each monitoring point.

3. The real-time processing system for environmental monitoring data based on edge computing according to claim 2, characterized in that, The process of obtaining a set of anomalous data and identifying its characteristics includes: Set corresponding filtering ranges for different types of collected data, set judgment units in each collection terminal, and determine whether each type of collected data from each collection terminal is within its filtering range. When any data collected by a single acquisition terminal is outside its filtering range, the current time is marked as an abnormal time, and the values ​​of the data collected by each acquisition terminal in the same application scenario at the abnormal time are all included in the first abnormal data set. All values ​​of the data collected by the acquisition terminal between the current time and the previous abnormal time are included in the second abnormal data set. The data features of the first abnormal data set and the second abnormal data set are respectively used as the first abnormal data feature and the second abnormal data feature. The abnormal data set and the abnormal data feature of each acquisition terminal are obtained respectively.

4. The real-time processing system for environmental monitoring data based on edge computing according to claim 3, characterized in that, The process of constructing the first evaluation model includes: Obtain historical fault records for different fault types. The historical fault records contain abnormal data sets and their abnormal data characteristics at the corresponding abnormal time. Generate a first evaluation set based on the abnormal data sets, abnormal data characteristics, and corresponding fault types contained in different historical fault records, and divide it into a first training set and a first test set. Construct a first convolutional neural network, using the abnormal data set and abnormal data features in the first training set as input data for the first convolutional neural network, and using the existence of fault events and their fault types as output data for the first convolutional neural network; The first convolutional neural network is trained using the first training set to obtain an initial first convolutional neural network. The initial first convolutional neural network is validated using the first test set, and the initial first convolutional neural network that is less than or equal to a preset first test error threshold is output as the first evaluation model.

5. The real-time processing system for environmental monitoring data based on edge computing according to claim 4, characterized in that, The process of constructing the second evaluation model includes: The first evaluation model is deployed in each monitoring terminal. A single abnormal data set and its abnormal data characteristics are input into the first evaluation model in each monitoring terminal, and the evaluation results are output respectively. The time interval from data input to result output of each first evaluation model is used as its processing time. The processing parameters of the single abnormal data set and the operating parameters of each monitoring terminal are obtained. The processing parameters refer to the data volume and data packet size of the single abnormal data set, and the operating parameters refer to the real-time status parameters of the server preset in each monitoring terminal. The processing time and corresponding operating parameters of abnormal data sets with different processing parameters are obtained on each monitoring terminal. A second evaluation set is generated based on the processing time and operating parameters of abnormal data sets with different processing parameters on each monitoring terminal, and it is divided into a second training set and a second test set. Construct a second convolutional neural network, using different processing parameters and running parameters from the second training set as input data for the second convolutional neural network, and using the corresponding processing time from the second training set as output data for the second convolutional neural network; The second convolutional neural network is trained using the second training set to obtain an initial second convolutional neural network. The initial second convolutional neural network is validated using the second test set, and the output of the initial second convolutional neural network that is less than or equal to the preset second test error threshold is used as the second evaluation model.

6. The real-time processing system for environmental monitoring data based on edge computing according to claim 5, characterized in that, The process of obtaining the optimal monitoring terminal for different sets of abnormal data includes: A second evaluation model is deployed in each monitoring terminal. When a single collection terminal generates a single abnormal data set, the abnormal data characteristics and processing parameters corresponding to the abnormal data set are obtained, and the processing parameters are sent to the monitoring terminal closest to the collection terminal. The operating parameters of each monitoring terminal are synchronized in real time. In the monitoring terminal closest to the acquisition terminal, the processing parameters of the abnormal data set and the operating parameters of each monitoring terminal are respectively input into the second evaluation model to obtain the processing time of each monitoring terminal for the abnormal data set. The monitoring terminal with the shortest processing time is selected as the optimal monitoring terminal for the abnormal data set.

7. The real-time processing system for environmental monitoring data based on edge computing according to claim 6, characterized in that, The process of determining whether a fault event exists in an abnormal data set includes: The single abnormal data set and its abnormal data characteristics of a single acquisition terminal are sent to its optimal monitoring terminal. In the optimal monitoring terminal, the received abnormal data set and abnormal data characteristics are input into the first evaluation model to determine whether there is a fault event in the abnormal data set. If there is, the fault type is obtained and fed back to the relevant personnel. If there is no fault, no other operation is performed.

8. A real-time processing method for environmental monitoring data based on edge computing, implemented based on the real-time processing system for environmental monitoring data according to any one of claims 1-7, characterized in that, The method includes: Step S1: Deploy acquisition terminals for each power device to obtain different acquisition data, and obtain multiple monitoring points in the application scenario to deploy monitoring terminals; Step S2: Obtain an abnormal data set based on the preset filtering range and different collected data, and obtain the corresponding abnormal data features. Construct a first evaluation model based on the abnormal data sets and abnormal data features of different fault types. Step S3: Obtain the processing parameters for different abnormal data sets, and construct a second evaluation model by combining the processing time of different abnormal data sets under different operating parameters of the monitoring terminal; Step S4: Use the second evaluation model to obtain the optimal monitoring terminal for different abnormal data sets in real time, and use the first evaluation model in the optimal monitoring terminal to determine whether there are fault events in its abnormal data set and provide feedback.

Images