Electric power communication network operation state prediction method based on artificial intelligence

Abstract

The invention relates to the technical field of electric power communication and artificial intelligence, and discloses an electric power communication network operation state prediction method based on artificial intelligence, the method comprises the steps of data acquisition, index calculation, state prediction and dynamic adjustment which are executed in sequence, and the dynamic adjustment result is output to the state prediction. According to the invention, the operation state of the electric power communication network can be predicted more accurately; the network state is comprehensively and deeply analyzed by calculating the response duration and the accuracy rate; compared with a mode of generating the network management alarm data only depending on the equipment operation data, information of the network management security data is mined; a preset standard and a calculation algorithm are optimized based on historical data and real-time conditions, so that prediction is more suitable for network changes; based on the multi-dimensional data acquisition and intelligent dynamic adjustment mode, the problems of large data volume and poor analysis timeliness in the prior art are solved, and stable operation of the power communication network is effectively guaranteed.

Description

Technical Field

[0001] The present application relates to the field of power communication and artificial intelligence technology, and specifically to a method for predicting the operating status of a power communication network based on artificial intelligence. Background Art

[0002] With the rapid development of intelligent power systems, power communication networks, as an important support for power systems, have a direct impact on the reliable operation of power systems due to the stability of their operating status. In their research on existing power communication network operating status monitoring, the inventors found that most existing methods analyze equipment operating data to generate network management alarm data, thereby enabling monitoring and prediction of operating status. However, this approach has many problems, including the high processing costs caused by the large amount of data, and the significant lag in fault warnings due to poor analysis timeliness, making it difficult to meet the high real-time and accuracy requirements of power communication networks.

[0003] Existing technologies attempt to improve prediction results by leveraging artificial intelligence methods such as deep learning. For example, Chinese patent number CN118821021A discloses a method and system for predicting faults in a power backbone communication network based on deep learning. This invention performs predictions by building a neural network model, but it primarily relies on extracting features from raw data and does not fully consider relevant information during the generation of network management alarm data. In actual power communication networks, the process data and feedback data generated by network management alarm data contain a wealth of information. Ignoring this data will result in predictions that fail to accurately reflect the actual operating conditions of the network and fail to fully consider complex situations such as the network's oversaturation state and potential equipment problems. This reduces the accuracy and reliability of the predictions and makes it difficult to effectively ensure the stable operation of the power communication network.

[0004] In summary, there is an urgent need for a new technical solution for power communication network operation status prediction based on artificial intelligence to solve the above technical problems. Summary of the Invention

[0005] The purpose of this application is to provide an artificial intelligence-based method for predicting the operating status of an electric power communication network to solve the technical problems raised in the above-mentioned background technology.

[0006] To achieve the above objectives, the present application discloses the following technical solutions: a method for predicting the operating status of a power communication network based on artificial intelligence, comprising sequentially performing data acquisition, indicator calculation, status prediction, and dynamic adjustment, wherein the result of the dynamic adjustment is output to the status prediction;

[0007] The data acquisition includes: collecting process data from the process of generating network management alarm data based on equipment operation data in the power communication network, the process data including equipment operation data processing steps, processing algorithms and intermediate calculation results; collecting feedback data after the corresponding network management alarm data is generated, the feedback data including the difference between the actual alarm situation and the expected alarm, and the correlation between the actual fault and the alarm;

[0008] The indicator is calculated as follows: based on the process data and the feedback data, by using timestamp comparison and statistical analysis, the response time for generating network management alarm data based on the equipment operation data is calculated, and the response time is the time interval from the occurrence of an abnormality in the equipment operation data to the generation of the network management alarm data; based on the consistency between the actual alarm situation in the feedback data and the expected alarm, the accuracy of the generated network management alarm data is calculated;

[0009] The state prediction is as follows: setting preset standards for response time and accuracy, comparing and analyzing the calculated response time and accuracy with the preset standards; if both the response time and the accuracy meet the preset standards, it is determined that the power communication network is in a normal operating state; if the response time exceeds the preset standard and the accuracy meets the preset standard, it is determined that the power communication network is in an oversaturated operating state with a high network load and a risk of performance degradation; if the response time meets the preset standard but the accuracy does not meet the preset standard, it is determined that there is a potential problem in the power communication network that affects the accuracy of alarm data generation; if both the response time and the accuracy do not meet the preset standard, it is determined that there is a serious problem in the power communication network that poses a significant threat to the reliability and stability of the network operation state;

[0010] The dynamic adjustment is to use artificial intelligence technology to establish a dynamic adjustment model based on historical process data, feedback data, response time and accuracy. The dynamic adjustment model automatically optimizes the preset standards for response time and accuracy based on the real-time operation of the power communication network, and adjusts the algorithm for calculating response time and accuracy.

[0011] Preferably, in the data acquisition step, the collected process data and feedback data are encrypted and stored.

[0012] Preferably, in the indicator calculation step, the response time T for generating network management alarm data based on the device operation data is calculated as follows:

[0013] The timestamp t1 when the abnormality of the collected equipment operation data occurs and the timestamp t2 when the network management alarm data is generated, the response time T = t2-t1.

[0014] Preferably, in the indicator calculation step, the accuracy rate A of generating the network management alarm data is calculated as follows:

[0015] The number of actual accurate alarms collected is n1, the number of actual faults associated with alarms is n2, and the total number of alarms is n, then the accuracy rate is

[0016] Preferably, in the state prediction step, the preset standard is obtained by regression analysis based on historical operating data and business needs of the power communication network.

[0017] Preferably, when it is determined that the power communication network is in the oversaturated operating state, the traffic load of each node in the network is calculated, and nodes with excessively high traffic occupancy are found based on the traffic load situation for targeted traffic adjustment; wherein the traffic load situation is calculated as follows:

[0018] Get the actual traffic f of node i i , the maximum carrying capacity of node i is F i , then the traffic occupancy rate of node i is Among them, T i is the response time of node i, T i_τ is the preset standard corresponding to the response time of node i.

[0019] Preferably, when it is determined that the electric power communication network has the potential problem, key operating parameters of the device are analyzed, the key operating parameters including the CPU usage and memory usage of the device, and the health index of the device is calculated by comparing the real-time monitoring of the key operating parameters with historical data. The potential problem is determined based on the size of the health index. The calculation of the health index is as follows:

[0020]

[0021] Among them, ω j_1 is the weight of the CPU usage of the preset device j, ω j_2 is the weight of the memory usage of the preset device j, U j_CPU_τ is the upper limit of the normal range of CPU usage of device j, U j_mem_τ is the upper limit of the normal range of memory usage of device j, A j is the accuracy of device j, A j_τ is the preset standard of accuracy of device j, H j is the calculated health index of device j.

[0022] Preferably, in the dynamic adjustment step, the corresponding preset standard is calculated using the response time preset standard calculation formula and the accuracy preset standard calculation formula;

[0023] The preset standard calculation formula for the response time is:

[0024]

[0025] in, is the preset standard for the p-th response time of node i, is the average value of the preset standard of the response time of node i p times, is the absolute value of the difference between the preset standard of the response time of node i for the pth time and the preset standard of the response time of node i for the p-1th time, ΔE i_p|i_p-1 is the deviation between the p-1th prediction result obtained by node i based on the feedback data and the pth actual operating state, The preset standard for the calculated response time of node i (p+1) is:

[0026] The calculation formula for the preset accuracy standard is:

[0027]

[0028] in, is the preset standard for the accuracy of the p-th time of device j, is the average value of the preset standard of the accuracy of device j p times, is the absolute value of the difference between the preset standard of the accuracy of the pth time and the preset standard of the accuracy of the p-1th time of device j, ΔE j_p|j_p-1 is the deviation between the p-1th prediction result obtained by device j based on the feedback data and the pth actual operating state, is the preset standard for the calculated accuracy of device j at the p+1th time.

[0029] Preferably, the data acquisition step further includes performing feature extraction and dimensionality reduction processing on the process data and the feedback data.

[0030] Preferably, when it is determined that the serious problem exists in the power communication network, a corresponding prediction report is output based on the corresponding data acquisition, the indicator calculation and the state prediction.

[0031] Beneficial effects: The artificial intelligence-based power communication network operation status prediction method of the present application realizes a more accurate prediction of the power communication network operation status; by calculating the response time and accuracy, the network status is comprehensively and deeply analyzed; compared with the method of relying solely on equipment operation data to generate network management alarm data, it realizes the mining of network management security data information; based on historical data and real-time conditions, the preset standards and calculation algorithms are optimized to make the prediction more adaptable to network changes; based on the multi-dimensional data collection and intelligent dynamic adjustment method, the problems of large data volume and poor analysis timeliness in the existing technology are solved, and the stable operation of the power communication network is effectively guaranteed. BRIEF DESCRIPTION OF THE DRAWINGS

[0032] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are only some embodiments of the present application. For those skilled in the art, other drawings can be obtained based on these drawings without paying any creative work.

[0033] Figure 1 A flowchart of an artificial intelligence-based method for predicting the operating status of a power communication network provided in an embodiment of the present application. DETAILED DESCRIPTION

[0034] The following is a clear and complete description of the technical solutions in the embodiments of this application. Obviously, the embodiments described are only part of the embodiments of this application, not all of them. Based on the embodiments of this application, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of this application.

[0035] In this document, the term "comprising" is intended to encompass non-exclusive inclusion, such that a process, method, article, or apparatus that includes a list of elements includes not only those elements but also other elements not explicitly listed, or elements inherent to such process, method, article, or apparatus. In the absence of further limitations, an element defined by the phrase "comprising..." does not preclude the presence of additional identical elements in the process, method, article, or apparatus that includes the element.

[0036] The first aspect of this embodiment discloses Figure 1 The method for predicting the operating status of a power communication network based on artificial intelligence includes sequentially performing data acquisition, indicator calculation, status prediction, and dynamic adjustment, wherein the result of the dynamic adjustment is output to the status prediction;

[0037] Data acquisition involves collecting process data from the power communication network during the process of generating network management alarm data based on equipment operation data. The process data includes the equipment operation data processing steps, processing algorithms, and intermediate calculation results; and collecting feedback data after the corresponding network management alarm data is generated. The feedback data includes the difference between the actual alarm situation and the expected alarm, and the correlation between the actual fault and the alarm.

[0038] The indicator is calculated as follows: Based on process data and feedback data, using timestamp comparison and statistical analysis, the response time for generating network management alarm data based on device operation data is calculated. The response time is the interval from the occurrence of an abnormality in device operation data to the generation of network management alarm data. The accuracy of the generated network management alarm data is calculated based on the consistency between the actual alarm situation in the feedback data and the expected alarm situation.

[0039] The state prediction is as follows: preset standards for response time and accuracy are set, and the calculated response time and accuracy are compared and analyzed with the preset standards; if both the response time and the accuracy meet the preset standards, it is determined that the power communication network is in a normal operating state; if the response time exceeds the preset standard and the accuracy meets the preset standard, it is determined that the power communication network is in an oversaturated operating state with excessive network load and the risk of performance degradation; if the response time meets the preset standard but the accuracy does not meet the preset standard, it is determined that there are potential problems in the power communication network that affect the accuracy of alarm data generation; if both the response time and the accuracy do not meet the preset standard, it is determined that there are serious problems in the power communication network that pose a significant threat to the reliability and stability of the network operation state;

[0040] Dynamic adjustment is: using artificial intelligence technology to establish a dynamic adjustment model based on historical process data, feedback data, response time and accuracy. This dynamic adjustment model automatically optimizes the preset standards for response time and accuracy based on the real-time operation of the power communication network, and adjusts the algorithm for calculating response time and accuracy.

[0041] Based on the above, this embodiment uses the process data and feedback data collected in the power communication network to generate network management alarm data based on equipment operation data, thereby realizing a more accurate prediction of the operation status of the power communication network; by obtaining the equipment operation data processing steps, algorithms and intermediate results, as well as data such as the actual and expected alarm differences, alarm and fault associations, the response time and accuracy are calculated, and the network status can be comprehensively and deeply analyzed; compared with the method of relying solely on equipment operation data to generate network management alarm data, it is possible to mine information from network management security data; combining existing artificial intelligence technology to establish a dynamic adjustment model, optimizing preset standards and calculation algorithms based on historical data and real-time conditions, so that predictions are more adaptable to network changes; based on this multi-dimensional data collection and intelligent dynamic adjustment method, the problems of large data volume and poor analysis timeliness of existing technologies are solved, and it can more accurately judge whether the network is operating normally, oversaturated, has potential problems or serious problems, and effectively ensure the stable operation of the power communication network.

[0042] Specifically, in the data acquisition step, the collected process data and feedback data are encrypted and stored.

[0043] Through the above, this embodiment achieves the protection of data security and privacy. In the data acquisition stage, any existing encryption technology is used to encrypt data including key information for equipment operation data processing and alarm feedback information, which can effectively prevent the data from being stolen or tampered with during storage and transmission. In actual power communication networks, these data involve key information for network operation. Once leaked, it may lead to increased network security risks and affect the stable operation of the power system. In a specific example, by regularly replacing the encryption key, the confidentiality of the data is further enhanced to ensure that the security and privacy of the data are continuously protected, thereby building a solid security line for subsequent analysis and prediction work based on these data and improving the reliability and practicality of the entire prediction method.

[0044] Specifically, in the indicator calculation step, the response time T for generating network management alarm data based on device operation data is calculated as follows:

[0045] The timestamp t1 when the abnormality of the collected equipment operation data occurs and the timestamp t2 when the network management alarm data is generated, the response time T = t2-t1.

[0046] Based on the above, this embodiment uses the timestamps of the abnormality of the collected equipment operation data and the generation of the network management alarm data to calculate the response time, thereby achieving accurate acquisition of the response time. By clearly recording these two key time points, the time interval from the abnormality of the equipment operation data to the generation of the alarm data can be accurately obtained. Compared with the fuzzy estimation of the response time, the calculation method based on timestamp comparison in this embodiment is more scientific and accurate. In the actual operation of the power communication network, accurate response time is an important basis for judging the real-time status of the network. If the response time is too long, even if the alarm accuracy is high, it may mean that the network has problems such as oversaturation. Accurate response time calculation helps to timely discover potential network risks and provide reliable data support for the subsequent accurate judgment of the network status, so that operation and maintenance personnel can take more targeted measures to ensure the efficient operation of the power communication network.

[0047] Specifically, in the indicator calculation step, the accuracy rate A of generating network management alarm data is calculated as follows:

[0048] The number of actual accurate alarms collected is n1, the number of actual faults associated with alarms is n2, and the total number of alarms is n, then the accuracy rate is

[0049] Based on the above, this embodiment uses the number of actual accurate alarms, the number of actual fault-alarm associations, and the total number of alarms to calculate the accuracy rate, thereby achieving a quantitative assessment of the accuracy of the alarm data. By counting these data, it is possible to intuitively understand the degree of fit between the network management alarm data and the actual situation. In the power communication network, the accuracy of the alarm data is crucial, but in actual applications, it is often only judged as accurate or inaccurate. This extensive classification method is not conducive to the use of network management security data. Based on this, this embodiment also includes the number of times that actual faults are associated with alarms in the calculation of the accuracy rate, thereby achieving a reasonable quantification of the accuracy rate, and then taking corresponding measures to improve the quality of alarms and ensure the reliability of network operation status monitoring.

[0050] Specifically, in the state prediction step, the preset standard is obtained by regression analysis based on the historical operation data and business needs of the power communication network.

[0051] Through the above, this embodiment realizes the scientific setting of preset standards. Through the regression analysis of existing technologies, it is possible to explore the patterns and trends in historical data, and combine them with the actual needs of the business to determine the preset standards for response time and accuracy that are preliminarily in line with the characteristics of network operation. The operating status of the power communication network is complex and changeable, and the requirements for network performance are different in different periods and business scenarios. For example, during peak and off-peak periods of electricity consumption, the requirements for network response time and alarm accuracy may be different. Scientifically setting preliminary preset standards can more accurately reflect the actual needs of the network, provide a reliable benchmark for accurate judgment of network status, and provide accurate initial values for subsequent dynamic adjustments of the business, thereby improving the adaptability and effectiveness of the prediction method.

[0052] Specifically, when it is determined that the power communication network is in an oversaturated operating state, the traffic load of each node in the network is calculated, and nodes with excessively high traffic occupancy rates are found based on the traffic load ranking for targeted traffic adjustment; wherein, the traffic load is calculated as follows:

[0053] Get the actual traffic f of node i i , the maximum carrying capacity of node i is F i , then the traffic occupancy rate of node i is Among them, T i is the response time of node i, T i_τ is the preset standard corresponding to the response time of node i.

[0054] In a simple example, consider node A in a power communication network. During peak hours, the actual traffic flow at node A is 80 Mbps, while the node's maximum traffic capacity is 100 Mbps. Under normal circumstances, the default response time from when an abnormality in device operating data occurs to when network management alarm data is generated is 5 seconds. However, during this monitoring period, the corresponding response time for node A was 8 seconds. Using the above formula, the traffic utilization ratio for node A is calculated as 2.4. This indicates a high traffic utilization ratio for this node. Combined with the response time exceeding the default standard, this indicates a high network load and possible congestion risk. Exceeding the default response time indicates a delay in processing network management alarm data, likely due to excessive network traffic. Therefore, incorporating response time into the traffic utilization ratio calculation allows for a more comprehensive assessment of node load, providing a more accurate basis for subsequent adjustments to network traffic. For example, operations and maintenance personnel can use this information to adjust data transmission paths, redirecting some data traffic to other nodes with lower loads, or limiting traffic to non-critical services at the node to alleviate network pressure and ensure stable operation of the power communication network.

[0055] Through the above, this embodiment achieves targeted optimization of the oversaturated network state. When it is determined that the network is in an oversaturated operating state, the actual traffic and maximum carrying traffic of the node are obtained, and the traffic occupancy rate is optimized based on the deviation of the response time. The nodes with excessive traffic are sorted based on this, so as to quickly locate the network bottleneck and take traffic adjustment measures in time. In the power communication network, only considering the traffic occupancy rate of the actual traffic and the maximum carrying traffic is an existing conventional analysis method, but it is limited to the device itself. In order to fully consider the impact of the topological structure of the device on the response, further, considering the calculation of the topological structure in the existing technology greatly increases the calculation. Therefore, this embodiment reasonably and concisely quantifies the impact of the topological structure on the response based on the deviation of the response time, providing a data basis for the reasonable allocation of network resources (such as adjusting the data transmission path, limiting some non-critical business traffic, etc.), effectively alleviating network pressure, improving network performance, and ensuring that the power communication network can still operate stably under high load conditions.

[0056] Specifically, when a potential problem is determined to exist in the power communication network, the key operating parameters of the device are analyzed. The key operating parameters include the device's CPU usage and memory usage. By comparing the real-time monitoring of the key operating parameters with historical data, the device's health index is calculated. The potential problem is determined based on the size of the health index. The health index is calculated as follows:

[0057]

[0058] Among them, ω j_1is the weight of the CPU usage of the preset device j, ω j_2 is the weight of the memory usage of the preset device j, U j_CPU_τ is the upper limit of the normal range of CPU usage of device j, U j_mem_τ is the upper limit of the normal range of memory usage of device j, A j is the accuracy of device j, A j_τ is the preset standard of accuracy of device j, H j is the calculated health index of device j.

[0059] In a simple example, consider a power communication device B. The preset weight for device CPU usage is 0.6, and the preset weight for device memory usage is 0.4. The upper limit of the normal range for device CPU usage is 80%, the upper limit of the normal range for memory usage is 90%, and the preset accuracy standard is 95%. At a certain moment, device B's CPU usage is monitored to be 60%, and its memory usage is 70%. The accuracy of generating network management alarm data is 90%. According to the formula, the health index of device B is approximately -0.0119. This low health index indicates potential device issues. The calculation process shows that although the CPU and memory usage have not reached the upper limits, the difference between the upper limits and the difference between the accuracy and the preset standards jointly affect the health index. This helps operation and maintenance personnel evaluate the equipment status based on a variety of factors. If the health index remains low, operation and maintenance personnel can further check the CPU and memory usage of the equipment, and analyze the reasons for the low accuracy of the alarm data, such as whether there is a device hardware failure affecting data processing, or whether the alarm data generation algorithm needs to be optimized, etc., so as to discover and solve equipment problems in advance and ensure the stable operation of the power communication network.

[0060] Through the above, this embodiment achieves in-depth diagnosis of potential problems in the power communication network. By calculating the health index, the operating status of the equipment can be quantitatively evaluated. In this embodiment, based on the deviation of the accuracy rate, efficient use of feedback data is achieved. Compared with the method of relying on only a single parameter or subjectively judging equipment problems, the method of comprehensively calculating the health index based on multiple parameters is more comprehensive and accurate. In the power communication network, the stable operation of the equipment is the basis for ensuring the normal operation of the network. When the alarm accuracy rate does not meet the standard, by analyzing the equipment health index, we can gain an in-depth understanding of whether the equipment has potential fault hazards. For example, if the health index is low, it may mean that there is a problem with the CPU or memory of the device, affecting the accuracy of the alarm data generated. This helps operation and maintenance personnel to discover and solve equipment problems in advance, avoid the problem from worsening and affecting the overall operation of the network, and improve the reliability and stability of the network.

[0061] Specifically, in the dynamic adjustment step, the corresponding preset standard is calculated using the response time preset standard calculation formula and the accuracy preset standard calculation formula;

[0062] The preset standard calculation formula for response time is:

[0063]

[0064] in, is the preset standard for the p-th response time of node i, is the average value of the preset standard of the response time of node i p times, is the absolute value of the difference between the preset standard of the response time of node i for the pth time and the preset standard of the response time of node i for the p-1th time, ΔE i_p|i_p-1 is the deviation between the p-1th prediction result obtained by node i based on feedback data and the pth actual operating state, The preset standard for the calculated response time of node i (p+1) is:

[0065] In a simple example, consider node C in a power communication network. Initially, the first three response times are set to 4 seconds, 5 seconds, and 6 seconds, respectively. The average of these three responses is 5 seconds. During the fourth monitoring phase, the third response time is set to 6 seconds. Assume that the deviation between the fourth prediction based on feedback data and the actual operating state is -0.5 seconds (indicating that the prediction is 0.5 seconds faster than the actual operating state). To calculate the fourth response time preset standard according to the formula: first, the absolute value of the difference between the third and second preset standards is calculated as 1. The fourth response time preset standard is 6 + (6 / 5) * (-0.5 / 1) = 5.4 seconds. This calculation shows that the fourth response time preset standard is dynamically adjusted, taking into account the average of historical preset standards, changes in adjacent preset standards, and the deviation between the current prediction and the actual operating state. If new interference factors emerge in the network, causing a general increase in response time, this formula can be used to promptly adjust the preset standard, making subsequent judgments on network status more accurate. For example, when the network load suddenly increases, the response time may become longer, and the deviation between the predicted result and the actual operating status will also change accordingly. Using this formula, the preset standards can be adjusted according to these changes, thereby more accurately determining whether the network is in a normal operating state.

[0066] The calculation formula for the preset accuracy standard is:

[0067]

[0068] in, is the preset standard for the accuracy of the p-th time of device j, is the average value of the preset standard of the accuracy of device j p times, is the absolute value of the difference between the preset standard of the accuracy of the pth time and the preset standard of the accuracy of the p-1th time of device j, ΔE j_p|j_p-1 is the deviation between the p-1th prediction result obtained by device j based on feedback data and the pth actual operating state, is the preset standard for the calculated accuracy of device j at the p+1th time.

[0069] In a simple example, consider a power communication device D. The preset accuracy standards for the first three times are 90%, 92%, and 91%, respectively. The average of these three times is 91%. During the fourth monitoring, the third preset accuracy standard is 91%. Assume that the deviation between the fourth prediction result based on feedback data and the actual operating status is 0.03 (indicating that the predicted accuracy is 3% higher than the actual accuracy). To calculate the fourth preset accuracy standard, first calculate the absolute value of the 1% difference between the third and second preset standards. The fourth preset accuracy standard is then 91% + (91% / 91%) * (0.03 / 1%) * (1 / 100) = 91.3%. This calculation process demonstrates the dynamic adjustment of the preset accuracy standard based on the average of the device's historical preset accuracy standards, the difference between adjacent preset standards, and the deviation between the predicted result and the actual operating status. In actual power communication networks, the accuracy of alarm data may change due to factors such as equipment aging and changes in the network environment. This formula allows for timely adjustment of preset standards based on these changes, ensuring that network status assessments are more accurate. For example, when equipment aging causes significant fluctuations in alarm data accuracy, adjusting the preset standards using this formula can more accurately assess the operating status of network devices and determine whether the network is operating normally.

[0070] As described above, this embodiment utilizes a preset standard calculation formula for response time and a preset standard calculation formula for accuracy to achieve dynamic optimization of the preset standards. By introducing factors such as the historical preset standard average value of the node or device, the absolute value of the adjacent preset standard deviation values, and the deviation between the predicted result and the actual operating status into the formula, the preset standards can be continuously adjusted according to the real-time network conditions, thereby better adapting to changes in the power communication network. The power communication network is affected by various factors, such as equipment aging and business volume fluctuations. Fixed preset standards are difficult to accurately reflect the actual state of the network. Dynamically optimized preset standards can be adjusted in time with network changes, making the prediction results more accurate. For example, when new interference factors appear in the network, causing a general increase in response time, the preset standards can be adjusted accordingly to ensure that the judgment of the network status is consistent with the actual situation, thereby improving the accuracy and adaptability of the prediction method.

[0071] Specifically, the data acquisition step also includes feature extraction and dimensionality reduction processing of process data and feedback data.

[0072] Through the above, this embodiment achieves improved data processing efficiency and prediction accuracy. By utilizing existing data processing technologies, such as principal component analysis, through feature extraction, it is possible to screen out key information valuable for network status prediction from a large amount of data and reduce the interference of redundant data; dimensionality reduction processing maps high-dimensional data to low-dimensional space, reducing computational complexity, thereby speeding up data processing and improving model training efficiency. In the power communication network, the amount of data is huge and complex. If the raw data is processed directly, it will not only increase the consumption of computing resources, but also may affect the accuracy of prediction due to data noise. After feature extraction and dimensionality reduction processing, the key features of the data can be highlighted, making subsequent indicator calculations and status predictions more efficient and accurate, and improving the performance of the entire prediction method.

[0073] Specifically, when it is determined that there is a serious problem in the power communication network, a corresponding prediction report is output based on corresponding data acquisition, indicator calculation and status prediction.

[0074] Through the above, this embodiment provides comprehensive decision support for operation and maintenance personnel. By utilizing existing report generation technology, a forecast report is generated based on the results of data acquisition, indicator calculation and status prediction. The report includes information such as possible causes of failures, affected equipment and business scope, and recommended repair measures, so that operation and maintenance personnel can quickly understand the overall picture of network problems and make more scientific decisions. When serious problems occur in the power communication network, time is running out and operation and maintenance personnel need to take effective measures quickly to restore the normal operation of the network. The forecast report can help them quickly locate the root cause of the problem, assess the scope of the problem, formulate and implement repair plans in a timely manner, reduce the impact of network failures on the operation of the power system, improve the efficiency and quality of network fault handling, and ensure the reliable operation of the power communication network.

[0075] In summary, the artificial intelligence-based power communication network operation status prediction method of this embodiment realizes a more accurate prediction of the power communication network operation status; by calculating the response time and accuracy, the network status is comprehensively and deeply analyzed; compared with the method of relying solely on equipment operation data to generate network management alarm data, it realizes the mining of network management security data information; based on historical data and real-time conditions, the preset standards and calculation algorithms are optimized to make the prediction more adaptable to network changes; based on this multi-dimensional data collection and intelligent dynamic adjustment method, the problems of large data volume and poor analysis timeliness of the existing technology are solved, and the stable operation of the power communication network is effectively guaranteed.

[0076] In the embodiments provided herein, it should be understood that the embodiments described herein can be implemented in hardware, software, firmware, middleware, code, or any appropriate combination thereof. For hardware implementation, the processor can be implemented in one or more of the following units: an application-specific integrated circuit (ASIC), a digital signal processor (DSP), a digital signal processing device (DSPD), a programmable logic device (PLD), a field programmable gate array (FPGA), a processor, a controller, a microcontroller, a microprocessor, other electronic units designed to implement the functions described herein, or a combination thereof. For software implementation, part or all of the processes of the embodiments can be completed by instructing the relevant hardware through a computer program. When implemented, the above program can be stored in a computer-readable storage medium or transmitted as one or more instructions or codes on a computer-readable storage medium. Computer-readable storage media include computer storage media and communication media, wherein the communication media include any medium that facilitates the transmission of a computer program from one place to another. The storage medium can be any available medium that a computer can access. The computer-readable storage medium can include, but is not limited to, RAM, ROM, EEPROM, CD-ROM or other optical disk storage, magnetic disk storage media or other magnetic storage devices, or any other medium that can be used to carry or store the desired program code in the form of instructions or data structures and can be accessed by a computer.

[0077] Finally, it should be noted that the above is only a preferred embodiment of the present application and is not intended to limit the present application. Although the present application has been described in detail with reference to the aforementioned embodiments, those skilled in the art can still modify the technical solutions described in the aforementioned embodiments or make equivalent replacements for some of the technical features therein. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principles of the present application should be included in the scope of protection of the present application.

Claims

1. A method for predicting the operating status of a power communication network based on artificial intelligence, characterized in that: The method comprises sequentially performing data acquisition, indicator calculation, state prediction and dynamic adjustment, wherein the result of the dynamic adjustment is output to the state prediction; The data acquisition includes: collecting process data from the process of generating network management alarm data based on equipment operation data in the power communication network, the process data including equipment operation data processing steps, processing algorithms and intermediate calculation results; collecting feedback data after the corresponding network management alarm data is generated, the feedback data including the difference between the actual alarm situation and the expected alarm, and the correlation between the actual fault and the alarm; The indicator is calculated as follows: based on the process data and the feedback data, by using timestamp comparison and statistical analysis, the response time for generating network management alarm data based on the equipment operation data is calculated, and the response time is the time interval from the occurrence of an abnormality in the equipment operation data to the generation of the network management alarm data; based on the consistency between the actual alarm situation in the feedback data and the expected alarm, the accuracy of the generated network management alarm data is calculated; The state prediction is as follows: setting preset standards for response time and accuracy, comparing and analyzing the calculated response time and accuracy with the preset standards; if both the response time and the accuracy meet the preset standards, it is determined that the power communication network is in a normal operating state; if the response time exceeds the preset standard and the accuracy meets the preset standard, it is determined that the power communication network is in an oversaturated operating state with a high network load and a risk of performance degradation; if the response time meets the preset standard but the accuracy does not meet the preset standard, it is determined that there is a potential problem in the power communication network that affects the accuracy of alarm data generation; if both the response time and the accuracy do not meet the preset standard, it is determined that there is a serious problem in the power communication network that poses a significant threat to the reliability and stability of the network operation state; The dynamic adjustment is to use artificial intelligence technology to establish a dynamic adjustment model based on historical process data, feedback data, response time and accuracy. The dynamic adjustment model automatically optimizes the preset standards for response time and accuracy based on the real-time operation of the power communication network, and adjusts the algorithm for calculating response time and accuracy.

2. The method for predicting the operating status of a power communication network based on artificial intelligence according to claim 1, characterized in that: In the data acquisition step, the collected process data and feedback data are encrypted and stored.

3. The method for predicting the operating status of a power communication network based on artificial intelligence according to claim 1, characterized in that: In the indicator calculation step, the response time T for generating network management alarm data based on device operation data is calculated as follows: The timestamp t1 when the abnormality of the collected equipment operation data occurs and the timestamp t2 when the network management alarm data is generated, the response time T = t2-t1.

4. The method for predicting the operating status of a power communication network based on artificial intelligence according to claim 1, characterized in that: In the indicator calculation step, the accuracy rate A of generating network management alarm data is calculated as follows: The number of actual accurate alarms collected is n1, the number of actual faults associated with alarms is n2, and the total number of alarms is n, then the accuracy rate is 5. The method for predicting the operating status of a power communication network based on artificial intelligence according to claim 1, characterized in that: In the state prediction step, the preset standard is obtained by regression analysis based on historical operation data and business requirements of the power communication network.

6. The method for predicting the operating status of a power communication network based on artificial intelligence according to claim 1, characterized in that: When it is determined that the power communication network is in the oversaturated operating state, the traffic load of each node in the network is calculated, and nodes with excessively high traffic occupancy rates are found based on the traffic load ranking for targeted traffic adjustment; wherein the traffic load is calculated as follows: Get the actual traffic f of node i i , the maximum carrying capacity of node i is F i , then the traffic occupancy rate of node i is Among them, T i is the response time of node i, T i_τ is the preset standard corresponding to the response time of node i.

7. The method for predicting the operating status of a power communication network based on artificial intelligence according to claim 1, characterized in that: When it is determined that the power communication network has the potential problem, key operating parameters of the device are analyzed. The key operating parameters include the CPU usage and memory usage of the device. The health index of the device is calculated by comparing the real-time monitoring of the key operating parameters with historical data. The potential problem is determined based on the size of the health index. The health index is calculated as follows: Among them, ω j_1 is the weight of the CPU usage of the preset device j, ω j_2 is the weight of the memory usage of the preset device j, U j_CPU_τ is the upper limit of the normal range of CPU usage of device j, U j_mem_τ is the upper limit of the normal range of memory usage of device j, A j is the accuracy of device j, A j_τ is the preset standard of accuracy of device j, H j is the calculated health index of device j.

8. The method for predicting the operating status of a power communication network based on artificial intelligence according to claim 1, characterized in that: In the dynamic adjustment step, the corresponding preset standard is calculated using the response time preset standard calculation formula and the accuracy preset standard calculation formula; The preset standard calculation formula for the response time is: in, is the preset standard for the p-th response time of node i, is the average value of the preset standard of the response time of node i p times, is the absolute value of the difference between the preset standard of the response time of node i for the pth time and the preset standard of the response time of node i for the p-1th time, ΔE i_p|i_p-1 is the deviation between the p-1th prediction result obtained by node i based on the feedback data and the pth actual operating state, The preset standard for the calculated response time of node i (p+1) is: The calculation formula for the preset accuracy standard is: in, is the preset standard for the accuracy of the p-th time of device j, is the average value of the preset standard of the accuracy of device j p times, is the absolute value of the difference between the preset standard of the accuracy of the pth time and the preset standard of the accuracy of the p-1th time of device j, ΔE j_p|j_p-1 is the deviation between the p-1th prediction result obtained by device j based on the feedback data and the pth actual operating state, is the preset standard for the calculated accuracy of device j at the p+1th time.

9. The method for predicting the operating status of a power communication network based on artificial intelligence according to claim 1, characterized in that: The data acquisition step also includes performing feature extraction and dimensionality reduction processing on the process data and the feedback data.

10. The method for predicting the operating status of a power communication network based on artificial intelligence according to claim 1, characterized in that: When it is determined that the serious problem exists in the power communication network, a corresponding prediction report is output based on the corresponding data acquisition, the indicator calculation and the state prediction.

Images