Important load standby power manager

Abstract

The invention discloses an important load standby power manager, which relates to the technical field of battery load management, and is characterized in that a communication equipment monitoring area is set, communication equipment state information data is acquired in real time by installing Internet of Things equipment, and the communication equipment state information data acquired in real time is processed in a data processing mode; processing the state information data of the communication equipment to obtain the processed state information data of the communication equipment, and analyzing through a data analysis mode based on the processed state information data of the communication equipment to obtain analyzed state information data of the communication equipment; and meanwhile, a load standby power regulation and control model is constructed based on the analyzed communication equipment state information data, finally, when it is detected that the communication equipment state information data collected in real time is abnormal, standby power supply equipment is arranged to supply power, and real-time regulation and control are performed based on the load standby power regulation and control model, so that the power supply stability of load equipment is improved.

Description

Technical Field

[0001] This invention relates to the field of battery load management technology, and more specifically to a backup power manager for critical loads. Background Technology

[0002] The loads within a communication base station are critical equipment. In the event of a mains power outage or other faults, these critical loads require uninterrupted power supply to prevent widespread communication disruptions. The traditional solution involves connecting more energy storage batteries to the communication base station's power conversion equipment ("communication switching power supply") to extend the power supply time for the load equipment. However, this solution has several drawbacks, such as: 1. All loads within the base station (critical or uncritical equipment) consume the energy storage batteries after a mains power outage, resulting in insufficient backup power for critical equipment and premature shutdown; 2. Due to the complex structure of the communication switching power supply and its relatively high probability of failure due to lightning strikes or other external factors, power outages for critical load equipment can also occur. Another solution is to provide a backup power system with its own energy storage battery, specifically for backing up power for critical loads. However, such systems are bulky and expensive. The circuit that supplies power to critical loads has a secondary DC-DC converter. If the secondary DC-DC converter fails, it will also cause a power outage for critical loads. Moreover, maintenance requires disassembling the entire bulky device, which means that all load lines need to be disconnected, inevitably leading to a long period of equipment interruption.

[0003] The prior art, such as the invention patent application with publication number CN118282002A, discloses a load power supply control circuit and an AC charging and discharging device. The method includes: a load power supply control circuit comprising a charging and discharging circuit, an electricity metering circuit, and a main control circuit; a charging and discharging circuit connected to the power grid and the charging and discharging load, performing charging and discharging processing on the load and outputting charging and discharging information; an electricity metering circuit connected to the power grid and the load, measuring first electricity consumption information of the power grid and second electricity consumption information of the load; and a main control circuit connected to the power grid and the load, and connected to the charging and discharging circuit and the electricity metering circuit, controlling the power grid or the charging and discharging circuit to supply power to the load, and / or controlling the charging and discharging circuit to supply power to the load, based on the first electricity consumption information, the second electricity consumption information, and the charging and discharging information.

[0004] As can be seen from the above solutions, the traditional solution for backup power supply of communication base stations is to connect more energy storage batteries to the power conversion equipment "communication switching power supply" of the communication base station in order to extend the power supply time of the load equipment. However, this solution has some shortcomings. Summary of the Invention

[0005] The purpose of this invention is to provide a critical load backup power manager that solves the problems existing in the background art.

[0006] To solve the above-mentioned technical problems, the present invention adopts the following technical solution: The present invention provides a critical load backup power manager, specifically including the following steps: S1. Set up a monitoring area for communication equipment and collect real-time status information data of communication equipment by installing IoT devices; S2. Process the real-time collected communication device status information data through data processing methods to obtain the processed communication device status information data. S3. Based on the processed communication device status information data, analyze it using data analysis methods to obtain the analyzed communication device status information data, including the following steps: S31. Obtain feature information from the processed communication device status information data through data feature extraction; S32. Based on the acquired feature information, the decision tree analysis method is used to analyze the data to obtain the analyzed communication device status information data. S4. Construct a load backup power control model based on the analyzed communication equipment status information data; S5. When an anomaly is detected in the real-time data acquisition of communication equipment status information, a backup power supply is arranged to provide power, and real-time control is performed based on the load backup power control model.

[0007] Preferably, the step of setting a monitoring area for the communication equipment and collecting real-time status information data of the communication equipment by installing IoT devices includes the following steps: S11. Set up a monitoring area for communication equipment, and evenly set up multiple detection points within the monitoring area. At the same time, collect real-time status information data of communication equipment based on the set detection points. The data set for real-time collection of communication device status information at the detection point is set as follows: , Indicates the first The status information data of communication equipment collected at each detection point This indicates the total number of detection points set; S12. Set the data collection interval for the detection points, and summarize the collected communication device status information data based on the set data collection interval; Set the communication device status information data set at different data collection intervals as follows: , This represents the communication device status information data collected within the i-th data collection interval. Indicates the number of data collection intervals; The collected communication device status information data is summarized to obtain communication device status information data. .

[0008] Preferably, the step of processing the real-time collected communication device status information data to obtain the processed communication device status information data includes the following steps: S21. Filter the collected communication device status information data to obtain filtered communication device status information data. S22. Process the filtered communication device status information data to obtain processed communication device status information data.

[0009] Preferably, the step of filtering the collected communication device status information data to obtain filtered communication device status information data includes the following steps: The communication equipment status parameter data collected under normal communication equipment conditions is set to the communication equipment status information data when the circuit is normal; The status parameter data collected in real time when the communication equipment is in an abnormal state is set to the status information data of the communication equipment when the circuit is abnormal; Set a valid data threshold, and based on the set valid data threshold, remove communication device status information data that exceeds the set threshold from the collected communication device status information data to obtain filtered communication device status information data.

[0010] Preferably, processing the filtered communication device status information data to obtain processed communication device status information data includes the following steps: The standardized processing formula for communication equipment status information data is as follows: ; Where Z represents the data before standardization. This represents the data after standardization. This represents the maximum value in the filtered communication device status information data. This represents the minimum value in the filtered communication device status information data.

[0011] Preferably, the step of obtaining feature information from the processed communication device status information data through data feature extraction includes the following steps: S311. Extract features from the processed communication device status information data using a convolutional neural network; A convolutional neural network is defined as including: an input layer, convolutional layers, pooling layers, and fully connected layers; S312. Input the processed communication device status information data into the input layer of the neural network; S313. After receiving the processed communication device status information data, the input layer transmits the received communication device status information data to the convolutional layer. The convolutional layer extracts local features from the processed communication device status information data through convolution calculation. S314. After the convolutional layer extracts local features from the processed communication device status information data, the pooling layer processes the extracted local features and downsamples the features to reduce data dimensionality. S315. The processed communication device status information data is subjected to feature extraction through continuous convolution and pooling until the feature extraction converges, the convolution stops, and the extracted features are summarized and input into the fully connected layer. S316. The fully connected layer integrates the extracted features and outputs the final feature extraction result. S317. Concatenate the feature extraction results from each group of processed communication device status information data to form a feature vector and save it. The feature vector is defined as the feature information in the processed communication device status information data.

[0012] Preferably, the step of analyzing the acquired feature information using decision tree analysis to obtain the analyzed communication device status information data includes the following steps: S321. For the feature information in the processed communication device status information data, Q groups of feature information are randomly selected with replacement, and a decision tree is constructed based on the selected Q groups of feature information. The selected Q groups of feature information are set as samples at the root node of the decision tree, and each sample represents a set of feature information in the processed communication device status information data. S322. Set that there are K indicators for each sample. When classification is required at each node of the decision tree, randomly select one indicator from these K indicators, and set the range threshold of the selected indicator according to the maximum and minimum values ​​of each indicator, and use the set range threshold as the classification indicator for that node. Each category metric can only be selected once, and each category will only generate two nodes; S323. Classify each node in the decision tree according to step S322 until the sample can no longer be classified, and construct a decision tree based on the classified nodes. The classification results of each decision tree are summarized, identical results are merged, and the classification index is used as the cluster center of the current classification result. Each cluster center represents the index range of a processed communication device status information data. The processed communication equipment status information data indexes are summarized and categorized to obtain the analyzed communication equipment status information data.

[0013] Preferably, the process of constructing a load backup power control model based on the analyzed communication equipment status information data includes the following steps: The expected value of the communication equipment is determined based on the range of the analyzed communication equipment status information data indicators. The expected value of the communication equipment is set as the average value of the range of the analyzed communication equipment status information data indicators. Based on the minimum difference between the output value of the backup power supply equipment and the expected value of the communication equipment, a control objective function is established, and the formula of the control objective function is as follows: ; in, This represents the minimum difference between the output value of the backup power supply equipment and the expected value of the communication equipment. This indicates the fitness of the backup power supply's output value with respect to time t. This represents the fitness of the communication device's expected value with respect to time t, where t represents time. When the backup power supply equipment is working, the minimum difference between the output value of the backup power supply equipment and the expected value of the communication equipment is calculated over time based on the established control objective function, and a set of differences is established based on the calculation results. By using the maximum membership function to fuzzify the output value of the backup power supply equipment, the output of the backup power supply equipment becomes specifically customized. The customized formula for the output of the backup power supply equipment is as follows: ; in, This represents the minimum difference between the output value of the backup power supply equipment and the expected value of the communication equipment at time t-1. The membership decision function is represented by the function name. The output formula of the backup power supply equipment is set as the load backup power control model.

[0014] Preferably, the step of arranging backup power supply equipment to provide power when an anomaly is detected in the real-time collected communication equipment status information data, and performing real-time regulation based on the load backup power control model, includes the following steps: S51. Synchronize the backup power supply equipment with the communication equipment using a clock synchronization method; After the communication equipment is turned on, it will send a request for synchronization data to the backup power supply equipment, and indicate the synchronization data signal number; After receiving a request for synchronization data from the communication equipment, the backup power supply equipment will send a feedback data as a response, indicating the feedback data sequence number; After receiving the feedback data, the communication equipment will send a confirmation synchronization data to the backup power supply equipment. At this time, the clock synchronization between the backup power supply equipment and the communication equipment is completed. S52. After clock synchronization is completed, when an abnormality is detected in the communication equipment, real-time control is performed based on the load backup power control model. Collect communication equipment status data within the time interval preceding an anomaly, and input the collected communication equipment status data as the expected value of the communication equipment into the load backup power control model; After receiving the expected value from the communication equipment, the load backup power control model calculates the corresponding output value of the backup power supply equipment and supplies power to the communication equipment based on the calculated backup power supply equipment.

[0015] This invention also discloses an important load backup power manager comprising: a data acquisition module, a data processing module, a data analysis module, a load regulation module, and a real-time regulation module; The data acquisition module is used to collect real-time status information data of communication equipment; The data processing module is used to process the real-time collected communication device status information data to obtain the processed communication device status information data. The data analysis module is used to analyze the processed communication device status information data through data analysis methods to obtain the analyzed communication device status information data. The load control module is used to construct a load backup power control model based on the analyzed communication equipment status information data. The real-time control module is used to adjust the output of the backup power supply equipment in real time according to the constructed load backup power control model.

[0016] The beneficial effects of this invention are as follows: (1) This invention sets up a monitoring area for communication equipment and collects real-time status information data of communication equipment by installing Internet of Things (IoT) devices. At the same time, it processes the real-time collected status information data of communication equipment through data processing to obtain processed status information data of communication equipment. Based on the processed status information data of communication equipment, it analyzes the data to obtain analyzed status information data of communication equipment. At the same time, it constructs a load backup power control model based on the analyzed status information data of communication equipment. Finally, when an anomaly is detected in the real-time collected status information data of communication equipment, a backup power supply device is arranged to supply power, and real-time control is performed based on the load backup power control model, thereby improving the stability of power supply to the load equipment.

[0017] (2) The present invention improves the reliability of data processing by setting a data validity threshold and filtering the collected communication device status information data based on the set data validity threshold, and by processing the filtered communication device status information data through data standardization.

[0018] (3) The present invention obtains feature information in the processed communication device status information data through data feature extraction. At the same time, based on the obtained feature information, it performs analysis through decision tree analysis to determine the index range of each category of communication device status information data, thereby improving the accuracy of data analysis.

[0019] (4) The present invention determines the expected value of the communication equipment based on the range of the analyzed communication equipment status information data indicators, and constructs an objective function based on the output value of the backup power supply equipment and the expected value of the communication equipment; at the same time, it improves the real-time performance of the backup power supply equipment output by customizing the output of the backup power supply equipment through the maximum membership function. Attached Figure Description

[0020] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0021] Figure 1 This is a schematic diagram of the operation method of the critical load backup power manager of the present invention. Detailed Implementation

[0022] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0023] In a specific embodiment of the present invention, Reference Figure 1 As shown, the present invention provides a critical load backup power manager, comprising: S1. Set up a monitoring area for communication equipment and collect real-time status information data of communication equipment by installing IoT devices; S2. Process the real-time collected communication device status information data through data processing methods to obtain the processed communication device status information data. S3. Based on the processed communication device status information data, analyze it using data analysis methods to obtain the analyzed communication device status information data, including the following steps: S31. Obtain feature information from the processed communication device status information data through data feature extraction; S32. Based on the acquired feature information, the decision tree analysis method is used to analyze the data to obtain the analyzed communication device status information data. S4. Construct a load backup power control model based on the analyzed communication equipment status information data; S5. When an anomaly is detected in the real-time data acquisition of communication equipment status information, arrange for backup power supply equipment to provide power, and perform real-time control based on the load backup power control model. Furthermore, referring to Figure 1 As shown, setting up a monitoring area for communication equipment and collecting real-time status information data of the communication equipment by installing IoT devices includes the following steps: S11. Set up a monitoring area for communication equipment, and evenly set up multiple detection points within the monitoring area. At the same time, collect real-time status information data of communication equipment based on the set detection points. The data set for real-time collection of communication device status information at the detection point is set as follows: , Indicates the first The status information data of communication equipment collected at each detection point This indicates the total number of detection points set; S12. Set the data collection interval for the detection points, and summarize the collected communication device status information data based on the set data collection interval; Set the communication device status information data set at different data collection intervals as follows: , This represents the communication device status information data collected within the i-th data collection interval. Indicates the number of data collection intervals; Furthermore, the collected communication device status information data is summarized to obtain communication device status information data. As shown below: ; in, Indicates the time within the j-th data collection interval. The status information data of communication equipment collected at each detection point; Furthermore, referring to Figure 1 As shown, the real-time collected communication device status information data is processed using data processing methods to obtain the processed communication device status information data, including the following steps: S21. Filter the collected communication device status information data to obtain filtered communication device status information data. The communication equipment status parameter data collected under normal communication equipment conditions is set to the communication equipment status information data when the circuit is normal; The status parameter data collected in real time when the communication equipment is in an abnormal state is set to the status information data of the communication equipment when the circuit is abnormal; Furthermore, a valid data threshold is set, and based on the set valid data threshold, communication device status information data exceeding the set threshold is removed from the collected communication device status information data to obtain filtered communication device status information data; S22. Process the filtered communication device status information data to obtain processed communication device status information data. The standardized processing formula for communication equipment status information data is as follows: ; Where Z represents the data before standardization. This represents the data after standardization. This represents the maximum value in the filtered communication device status information data. This represents the minimum value in the filtered communication device status information data. Furthermore, referring to Figure 1 As shown, obtaining feature information from processed communication device status information data through data feature extraction includes the following steps: S311. Extract features from the processed communication device status information data using a convolutional neural network; A convolutional neural network is defined as including: an input layer, convolutional layers, pooling layers, and fully connected layers; S312. Input the processed communication device status information data into the input layer of the neural network; S313. After receiving the processed communication device status information data, the input layer transmits the received communication device status information data to the convolutional layer. The convolutional layer extracts local features from the processed communication device status information data through convolution calculation. The formula for calculating convolution is as follows:

[0024] in, This represents the input communication device status information data. represents the weights of the corresponding convolution kernel, and b represents the bias value. Indicates output features; S314. After the convolutional layer extracts local features from the processed communication device status information data, the pooling layer processes the extracted local features and downsamples the features to reduce data dimensionality. S315. The processed communication device status information data is subjected to feature extraction through continuous convolution and pooling until the feature extraction converges, the convolution stops, and the extracted features are summarized and input into the fully connected layer. S316. The fully connected layer integrates the extracted features and outputs the final feature extraction result. S317. Concatenate the feature extraction results from each group of processed communication device status information data to form a feature vector and save it. The feature vector is defined as the feature information in the processed communication device status information data; Furthermore, referring to Figure 1 As shown, based on the acquired feature information, the analysis is performed using decision tree analysis to obtain the analyzed communication device status information data, including the following steps: S321. For the feature information in the processed communication device status information data, Q groups of feature information are randomly selected with replacement, and a decision tree is constructed based on the selected Q groups of feature information. The selected Q groups of feature information are set as samples at the root node of the decision tree, and each sample represents a set of feature information in the processed communication device status information data. S322. Set that there are K indicators for each sample. When classification is required at each node of the decision tree, randomly select one indicator from these K indicators, and set the range threshold of the selected indicator according to the maximum and minimum values ​​of each indicator, and use the set range threshold as the classification indicator for that node. Each category metric can only be selected once, and each category will only generate two nodes; S323. Classify each node in the decision tree according to step S322 until the sample can no longer be classified, and construct a decision tree based on the classified nodes. The classification results of each decision tree are summarized, identical results are merged, and the classification index is used as the cluster center of the current classification result. Each cluster center represents the index range of a processed communication device status information data. The range of indicators for the processed communication equipment status information data after classification is summarized to obtain the analyzed communication equipment status information data. Furthermore, referring to Figure 1 As shown, the process of constructing a load backup power control model based on the analyzed communication equipment status information data includes the following steps: The expected value of the communication equipment is determined based on the range of the analyzed communication equipment status information data indicators. The expected value of the communication equipment is set as the average value of the range of the analyzed communication equipment status information data indicators. Furthermore, based on the minimum difference between the output value of the backup power supply equipment and the expected value of the communication equipment, a control objective function is established, and the formula for the control objective function is as follows: ; in, This represents the minimum difference between the output value of the backup power supply equipment and the expected value of the communication equipment. This indicates the fitness of the backup power supply's output value with respect to time t. This represents the fitness of the communication device's expected value with respect to time t, where t represents time. Furthermore, when the backup power supply equipment is working, the minimum difference between the output value of the backup power supply equipment and the expected value of the communication equipment is calculated over time based on the established control objective function, and a set of differences is established based on the calculation results. Furthermore, by using the maximum membership function, the output value of the backup power supply equipment is fuzzified, making the output of the backup power supply equipment specific and customized. The formula for customizing the output of the backup power supply equipment is as follows: ; in, This represents the minimum difference between the output value of the backup power supply equipment and the expected value of the communication equipment at time t-1. The membership decision function is represented by the function name. The output formula of the backup power supply equipment is set as the load backup power control model; Furthermore, referring to Figure 1 As shown, when an anomaly is detected in the real-time data acquisition of communication equipment status information, a backup power supply is arranged to provide power, and real-time control is performed based on the load backup power control model, including the following steps: S51. Synchronize the backup power supply equipment with the communication equipment using a clock synchronization method; After the communication equipment is turned on, it will send a request for synchronization data to the backup power supply equipment, and indicate the synchronization data signal number; After receiving a request for synchronization data from the communication equipment, the backup power supply equipment will send a feedback data as a response, indicating the feedback data sequence number; Furthermore, after receiving the feedback data, the communication equipment will send a confirmation synchronization data to the backup power supply equipment. At this time, the clock synchronization between the backup power supply equipment and the communication equipment is completed. S52. After clock synchronization is completed, when an abnormality is detected in the communication equipment, real-time control is performed based on the load backup power control model. Collect communication equipment status data within the time interval preceding an anomaly, and input the collected communication equipment status data as the expected value of the communication equipment into the load backup power control model; After receiving the expected value of the communication equipment, the load backup power control model calculates the corresponding output value of the backup power supply equipment and supplies power to the communication equipment based on the calculated backup power supply equipment. In one specific embodiment, the critical load backup power manager includes: a data acquisition module, a data processing module, a data analysis module, a load regulation module, and a real-time regulation module; The data acquisition module is used to collect real-time status information data of communication equipment; The data processing module is used to process the real-time collected communication device status information data to obtain the processed communication device status information data. The data analysis module is used to analyze the processed communication device status information data through data analysis methods to obtain the analyzed communication device status information data. The load control module is used to construct a load backup power control model based on the analyzed communication equipment status information data. The real-time control module is used to adjust the output of the backup power supply equipment in real time according to the constructed load backup power control model.

[0025] It should be noted that, The above content is merely an example and illustration of the concept of the present invention. Those skilled in the art can make various modifications or additions to the specific embodiments described, or use similar methods to replace them, as long as they do not deviate from the concept of the invention or exceed the scope defined by the present invention, and all such modifications and additions should fall within the protection scope of the present invention.

Claims

1. An important load backup power manager, characterized in that, The method comprises the following steps: S1, setting a communication equipment monitoring area, and collecting communication equipment state information data in real time through the installation of Internet of Things equipment; S2, processing the real-time collected communication equipment state information data through a data processing method to obtain processed communication equipment state information data; S3, based on the processed communication equipment state information data, analyzing through a data analysis method to obtain analyzed communication equipment state information data, comprising the following steps: S31, obtaining feature information in the processed communication equipment state information data through a data feature extraction method; S32, based on the obtained feature information, analyzing through a decision tree analysis method to obtain analyzed communication equipment state information data; S4, constructing a load standby power control model based on the analyzed communication equipment state information data; S5, when detecting that the real-time collected communication equipment state information data is abnormal, arranging standby power supply equipment to supply power, and based on the load standby power control model, real-time control.

2. The critical load backup power manager of claim 1, wherein, The setting of the communication equipment monitoring area and the real-time collection of the communication equipment state information data through the installation of the Internet of Things equipment comprises the following steps: S11, setting a communication equipment monitoring area, and uniformly setting multiple detection points in the communication equipment monitoring area, and collecting communication equipment state information data in real time based on the set detection points; The set of communication equipment state information data collected in real time at the set detection point is , The communication equipment state information data collected at the first detection point is represented by The communication equipment state information data collected at the first detection point is represented by The total number of the set detection points is represented by S12, setting a detection point data collection interval, and based on the set data collection interval, collecting the communication equipment state information data; Set the communication equipment state information data set under different collection data intervals as , denotes the communication equipment state information data collected in the i-th collection data interval, denotes the number of collection data intervals; The communication device state information data collected is summarized to obtain communication device state information data .

3. The critical load backup power manager of claim 1, wherein, The processing of the real-time collected communication equipment state information data through the data processing method to obtain the processed communication equipment state information data comprises the following steps: S21, filtering the collected communication equipment state information data to obtain filtered communication equipment state information data; S22, processing the filtered communication equipment state information data to obtain processed communication equipment state information data.

4. The critical load backup power manager of claim 3, wherein, The filtering of the collected communication equipment state information data to obtain filtered communication equipment state information data comprises the following steps: Setting the communication equipment state parameter data collected in the normal state of the communication equipment as the communication equipment state information data when the circuit is normal; Setting the state parameter data collected in real time in the abnormal state of the communication equipment as the communication equipment state information data when the circuit is abnormal; Setting a data effective threshold, and based on the set data effective threshold, removing the communication equipment state information data exceeding the set threshold from the collected communication equipment state information data to obtain the filtered communication equipment state information data.

5. The critical load backup power manager of claim 3, wherein, The processing of the filtered communication equipment state information data to obtain the processed communication equipment state information data comprises the following steps: The standardization processing formula of the communication equipment state information data is as follows: ； wherein Z represents data before normalization processing, represents data after normalization processing; represents a maximum value in the filtered communication device status information data; represents a minimum value in the filtered communication device status information data.

6. The critical load backup power manager of claim 1, wherein, The obtaining of the feature information in the processed communication equipment state information data through the data feature extraction method comprises the following steps: S311, performing feature extraction on the processed communication equipment state information data through a convolutional neural network; The convolutional neural network comprises an input layer, a convolutional layer, a pooling layer, and a full connection layer; S312, input the processed communication device state information data to the input layer of the neural network; S313, after receiving the processed communication device state information data, the input layer transmits the received communication device state information data to the convolution layer, and the convolution layer extracts the local features in the processed communication device state information data through convolution calculation; S314, after the convolution layer extracts the local features in the processed communication device state information data, the extracted local features are processed through the pooling layer, and the features are down-sampled through the pooling layer to reduce the data dimension; S315, the processed communication device state information data is continuously convolved and pooled to extract features until the extracted features converge, and the convolution stops. The extracted features are input to the fully connected layer; S316, the fully connected layer integrates the extracted features and outputs the final feature extraction result; S317, the feature extraction results in each group of processed communication device state information data are connected to form a feature vector, which is saved; The feature vector is the feature information in the processed communication device state information data.

7. The critical load backup power manager of claim 1, wherein, Based on the obtained feature information, the decision tree analysis method is used for analysis to obtain the analyzed communication device state information data, including the following steps: S321, randomly select Q groups of feature information from the feature information in the processed communication device state information data with replacement, and construct a decision tree based on the selected Q groups of feature information. Set the selected Q groups of feature information as the samples at the root node of the constructed decision tree, and set each sample to represent a group of feature information in the processed communication device state information data; S322, set K indicators for each sample. When each node of the decision tree needs to be classified, randomly select one indicator from the K indicators, set the range threshold of the selected one indicator according to the maximum value and the minimum value in each indicator, and set the set range threshold as the classification indicator of the node; Each classification indicator can only be selected once, and each classification will only produce two nodes; S323, classify each node in the decision tree according to step S322 until the samples cannot be classified any more, and construct a decision tree based on the classified nodes; Summarize the classification results of each decision tree, merge the same results, and set the classification indicators as the clustering centers of the current classification results. Each clustering center represents an indicator range of the processed communication device state information data; Summarize the processed communication device state information data indicator range after classification to obtain the analyzed communication device state information data.

8. The critical load backup power manager of claim 1, wherein, The load standby power regulation model is constructed based on the analyzed communication device state information data, including the following steps: Determine the expected value of the communication device based on the analyzed communication device state information data indicator range. Set the expected value of the communication device as the average value of the analyzed communication device state information data indicator range; Based on the minimum difference between the standby power supply device output value and the communication device expected value as the control target, establish a control target function, and the control target function formula is as follows: ； wherein represents the minimum difference between the backup power supply device output value and the communication device desired value, represents the fitness of the backup power supply device output value with respect to time t, represents the fitness of the communication device desired value with respect to time t, t represents time; When the standby power supply device works, the minimum difference between the output value of the standby power supply device and the expected value of the communication device is calculated based on the established control target function, and the time-varying quantity of the difference is calculated based on the calculation result, and a difference set is established based on the calculation result; The output value of the standby power supply device is fuzzed by the maximum membership function, so that the output of the standby power supply device becomes a specific customization, and the output customization formula of the standby power supply device is as follows: ； wherein, represents the minimum difference between the output value of the backup power supply device at time t-1 and the expected value of the communication device, represents the membership decision function; The output customization formula of the standby power supply device is set as the load standby regulation and control model.

9. The critical load backup power manager of claim 1, wherein, When it is detected that the real-time collected communication device state information data is abnormal, the standby power supply device is arranged to supply power, and real-time regulation and control is performed based on the load standby regulation and control model, which includes the following steps: S51, the standby power supply device and the communication device are clock-synchronized by a clock synchronization method; After the communication device is turned on, it will send a request synchronization data to the standby power supply device, and mark the synchronization data signal serial number; After receiving the request synchronization data from the communication device, the standby power supply device will send a feedback data as a response, and mark the feedback data serial number; After receiving the feedback data, the communication device will send a confirmation synchronization data to the standby power supply device, and at this time the standby power supply device and the communication device are clock-synchronized; S52, after the clock synchronization is completed, when it is detected that the communication device is abnormal, real-time regulation and control is performed based on the load standby regulation and control model; The communication device state data within a time interval before the communication device is detected to be abnormal is collected, and the collected communication device state data is input as the expected value of the communication device into the load standby regulation and control model; After receiving the expected value of the communication device, the load standby regulation and control model calculates the corresponding output value of the standby power supply device, and supplies power to the communication device based on the calculated standby power supply device.

10. The critical load backup power manager of claims 1-9, wherein, It includes: Data acquisition module, data processing module, data analysis module, load regulation and control module, and real-time regulation and control module; The data acquisition module is used for real-time acquisition of communication device state information data; The data processing module is used for processing the real-time collected communication device state information data to obtain processed communication device state information data; The data analysis module is used for analyzing the processed communication device state information data by a data analysis method to obtain analyzed communication device state information data; The load regulation and control module is used for constructing a load standby regulation and control model according to the analyzed communication device state information data; The real-time regulation and control module is used for real-time regulation and control of the output of the standby power supply device based on the constructed load standby regulation and control model.

Images