A method and device for intelligent fault early warning in civil aviation systems

Abstract

This invention discloses an intelligent fault early warning method and device for civil aviation systems, belonging to the field of civil aviation information technology. It includes establishing a fault information-related database; acquiring and parsing raw log data to obtain error information data; using a preset machine learning algorithm to analyze the data in the relevant database to obtain computation time configuration and fault probability configuration; analyzing the error information data based on the computation time configuration and fault probability configuration to obtain fault information data; and sending fault information data to trigger a fault alarm according to preset alarm rules. This invention, by real-time parsing of system logs and employing statistical and comparative algorithms to analyze error information, can quickly and accurately identify faults in operational problems, avoiding the tediousness and inefficiency of traditional manual analysis and troubleshooting, and significantly improving the speed of fault identification.

Description

Technical Field

[0001] This invention relates to the field of civil aviation information technology, and in particular to a method and device for intelligent fault early warning in civil aviation systems. Background Technology

[0002] In the daily operation and maintenance of the civil aviation system, fault early warning and rapid response are crucial to ensuring the stable operation of the system. Traditional fault early warning methods mainly rely on the experience and judgment of maintenance personnel and manual analysis, which proves inadequate when faced with complex and ever-changing business problems. In particular, for non-hardware faults such as business logic errors and service access anomalies, due to their diverse manifestations and difficulty in predefinition, it is often necessary to analyze and judge them by comprehensively considering various log information, business indicators, and system status.

[0003] However, with the continuous expansion of the civil aviation system and the increase in business complexity, the amount of log data generated by the system has exploded. Traditional manual analysis methods can no longer meet the needs for efficient and accurate fault early warning. Furthermore, error messages related to business problems often exhibit uncertainty; the same error code may represent different fault degrees in different scenarios, making rule-based early warning methods inapplicable. Therefore, developing an intelligent fault early warning method and device that can automatically and intelligently analyze system logs, quickly and accurately identify faults in business problems, and provide detailed fault information and response suggestions is particularly important. Summary of the Invention

[0004] To address the aforementioned issues, this invention provides an intelligent fault early warning method and device for civil aviation systems. By analyzing system logs in real time and employing statistical and comparative algorithms to analyze error information, it can quickly and accurately identify faults in operational problems, avoiding the tediousness and inefficiency of traditional manual analysis and troubleshooting, and significantly improving the speed of fault identification.

[0005] The above objectives can be achieved through the following approach:

[0006] A method for intelligent fault early warning in civil aviation systems includes: establishing a database related to fault information; acquiring and parsing raw log data to obtain error information data; using a preset machine learning algorithm to analyze the data in the relevant database to obtain computation time configuration and fault probability configuration; analyzing the error information data according to the computation time configuration and the fault probability configuration to obtain fault information data; and sending the fault information data to issue a fault alarm according to preset alarm rules.

[0007] Optionally, obtaining and parsing the raw log data to obtain error information data includes: reading the raw log data line by line; determining whether the raw log data contains error codes or anomalies; if so, extracting the error information data from the raw log data, performing standardization processing, and storing it in a preset error information database; if not, performing standardization processing on the raw log data and storing it in a preset standard information database.

[0008] Optionally, the step of using a preset machine learning algorithm to analyze data in the relevant database to obtain computation time configuration and fault probability configuration includes: acquiring data from the error information database and the standard information database to obtain first sample data; using a preset machine learning algorithm to analyze the impact of different computation time configurations of the first sample data on the error rate in different time periods, calculating the optimal computation time configuration to obtain a preset time; acquiring data from a preset fault information database and a preset error rate database to obtain second sample data; using a preset machine learning algorithm to perform statistical analysis on the second sample data based on the growth of error information data and the status of fault information data for different services, calculating the optimal fault probability configuration to obtain a first threshold, a second threshold, and a third threshold.

[0009] Optionally, the step of analyzing the error information data according to the calculation time configuration and the fault probability configuration to obtain fault information data includes: extracting newly generated unprocessed error information data from the error information database; determining whether the error code of the error information data is generated for the first time according to the error information database; if so, obtaining the standard information data of the service corresponding to the error information data from the standard information database; determining the access volume of the service corresponding to the error information data on the previous day according to the standard information database, and analyzing the determination result to obtain fault information data; and storing the fault information data in the fault information database.

[0010] Optionally, determining the access volume of the service corresponding to the error information data on the previous day based on the standard information database, and analyzing the determination result to obtain fault information data includes: determining whether the service corresponding to the error information data had access volume on the previous day based on the standard information database; if not, the service corresponding to the error information data is a newly launched service and is determined to be normal; if yes, the service corresponding to the error information data is not a newly launched service and is determined to be faulty, thus obtaining fault information data.

[0011] Optionally, determining whether the error code of the error information data is generated for the first time based on the error information database includes: if the error code of the error information data is not generated for the first time, extracting the total number of errors reported by the service corresponding to the error information data within the preset time period from the error information database; obtaining the total number of accesses to the service corresponding to the error information data within the preset time period from the standard information database; obtaining the error information data of the service corresponding to the error information data based on the total number of errors and the total number of accesses; and determining the error rate ER in the error information data of service SY. SY ,have

[0012]

[0013] In the formula, SE SY To serve the total number of errors reported by SY within a preset time, TV SY To handle the total number of accesses to service SY within a preset time period; store the error information data in the error rate database; and analyze the fault information data based on the error rate in the error information data.

[0014] Optionally, the step of analyzing and obtaining fault information data based on the error rate in the error rate information data includes: obtaining the maximum value of the error rate of the service corresponding to the error information data from the error rate database; calculating a fault ratio value using the maximum value of the error rate and the error rate in the error information data; determining the magnitude of the fault ratio value based on the optimal fault probability configuration of the service corresponding to the error information data; and obtaining fault information data based on the determination result.

[0015] Optionally, the step of determining the magnitude of the fault ratio value based on the optimal fault probability configuration of the service corresponding to the error information data, and obtaining the fault information data based on the determination result, includes: determining whether the fault ratio value is greater than the first threshold; if not, determining that the service is normal; if yes, determining whether the fault ratio value is greater than the second threshold; if not, determining that the fault probability is low and obtaining first fault data; if yes, determining whether the fault ratio value is greater than the third threshold; if not, determining that the fault probability is medium and obtaining second fault data; if yes, determining that the fault probability is high and obtaining third fault data; and storing the first fault data, second fault data, and third fault data in the fault information database.

[0016] Optionally, sending the fault information data to trigger a fault alarm according to preset alarm rules includes: obtaining alarm rules from a preset alarm rule database; obtaining current fault information data from the fault information database; generating alarm information and triggering a fault alarm based on the alarm rules and the current fault information data; determining whether a fault exists based on the alarm information; if so, updating the status of the corresponding fault information data in the fault information database to "confirmed" and updating the status of the corresponding error information data in the error rate database to "deleted"; if not, updating the status of the corresponding fault information data in the fault information database to "false alarm".

[0017] Based on the same inventive concept, this invention also provides an intelligent fault early warning device for civil aviation systems. The device includes: a database storage module for establishing a database related to fault information; a log parsing module for acquiring and parsing raw log data to obtain error information data and storing it in the relevant database; a fault rule management module for analyzing data in the relevant database using a preset machine learning algorithm to obtain computation time configuration and fault probability configuration; a fault analysis module for analyzing the error information data based on the computation time configuration and the fault probability configuration to obtain fault information data; and a fault alarm module for sending the fault information data to trigger a fault alarm according to preset alarm rules.

[0018] Compared with the prior art, the present invention has the following advantages:

[0019] 1. This invention analyzes system logs in real time and uses statistical and comparative algorithms to analyze error information, which can quickly and accurately identify faults in business problems, avoiding the tediousness and inefficiency of traditional manual analysis and troubleshooting, and significantly improving the speed of fault identification. At the same time, it judges the probability of faults by the error rate per unit time rather than the number of errors, and dynamically adjusts the threshold by combining historical data, which improves the accuracy of fault judgment and reduces the false alarm rate.

[0020] 2. This invention adopts an automatic threshold algorithm based on error rate, which can dynamically adjust the threshold according to historical data, making the early warning system more adaptable to different service scenarios and time periods, and improving the system's adaptability and flexibility.

[0021] 3. When issuing an early warning, this invention not only provides basic fault information, but also further clarifies the probability of the fault by calculating the multiple relationship between the current error rate and the historical maximum error rate, and classifies it into low, medium and high levels, providing maintenance personnel with more detailed fault information and response basis, which facilitates maintenance personnel to make quick decisions.

[0022] 4. This invention uses statistical machine learning and algorithm analysis to dynamically adjust the optimal computation time configuration and fault probability configuration based on the impact of different computation time configurations on the error rate in different time periods, as well as the growth of error information data and fault information data status of different services, thereby further optimizing the system performance.

[0023] Other features and advantages of the invention will be set forth in the description which follows, and will be apparent in part from the description, or may be learned by practicing the invention. The objects and other advantages of the invention may be realized and obtained by means of the structures pointed out in the description, claims and drawings. Attached Figure Description

[0024] 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 some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0025] Figure 1 This is a flowchart illustrating an intelligent fault early warning method for a civil aviation system according to an embodiment of the present invention.

[0026] Figure 2 This is a schematic diagram of the workflow of the log parsing module in an embodiment of the present invention.

[0027] Figure 3 This is a schematic diagram of the workflow of the fault analysis module in an embodiment of the present invention.

[0028] Figure 4 A schematic diagram of the structure of an intelligent fault early warning device for a civil aviation system according to an embodiment of the present invention. Detailed Implementation

[0029] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, 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, 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.

[0030] Reference Figure 1 One embodiment of the present invention proposes an intelligent fault early warning method for civil aviation systems. By parsing system logs in real time and using statistical and comparative algorithms to analyze error information, it can quickly and accurately identify faults in business problems, avoiding the tediousness and inefficiency of traditional manual analysis and troubleshooting, and significantly improving the speed of fault identification.

[0031] The method described in this embodiment specifically includes:

[0032] Establish a database related to fault information;

[0033] Specifically, the relevant databases include an error information database, a standard information database, a fault information database, an error rate database, a fault rule database, and an alarm rule database.

[0034] Obtain the raw log data and parse it to obtain the error information data;

[0035] Specifically, the raw logs must be recorded according to preset rules and specifications to facilitate log parsing, and can be stored in disk files or databases, etc. The raw logs are parsed to analyze the error codes that appear in the logs, and the error information is stored in the error information database. The raw log data is standardized to remove useless information, and the standardized log data is stored in the standard data storage. The information that needs to be recorded in the raw log data may include: service access time, service name, access result, error code or exception information, other scenario information, and other access details. Some information, such as long processing time and other abnormal information, needs to be judged by the business system and recorded actively.

[0036] Using a pre-defined machine learning algorithm, the data in the relevant database is analyzed to obtain the computation time configuration and failure probability configuration.

[0037] Based on the computation time configuration and the fault probability configuration, the error information data is analyzed to obtain fault information data;

[0038] Specifically, the system retrieves the error information to be analyzed from the error information database; retrieves the fault rules from the fault rule database, and obtains the optimal computing configuration and optimal fault probability configuration corresponding to the service based on the fault rules; analyzes the error information according to the fault rules; the data required for the analysis process mainly relies on the data in the standard information database, the data in the error information database itself, and the data in the fault information database; if the analysis determines that it is a fault, the fault information is stored in the fault information database.

[0039] According to the preset alarm rules, the fault information data is sent to trigger a fault alarm.

[0040] Specifically, fault information is retrieved from the fault information database and communicated to system maintenance personnel via SMS, email, WeChat, etc. After receiving the alarm information, the maintenance personnel manually confirm that a fault does exist, and then update the fault status to "confirmed". Otherwise, if there is no fault, the fault status is updated to "false alarm", thereby providing further information for judgment and improving the accuracy of analysis.

[0041] Specifically, during implementation, fault rules and alarm rules for each service must first be set. If the initial settings are not appropriate, they can be adjusted based on historical data and the effects after implementation. By parsing raw log data in real time and using statistical and comparative algorithms combined with machine learning technology, Ming has achieved automation and intelligence in intelligent fault early warning, significantly improving the speed and accuracy of fault identification, reducing the false alarm rate, and providing maintenance personnel with detailed fault information and response basis. This helps to improve maintenance efficiency, reduce maintenance costs, and enhance the stability and reliability of the civil aviation system.

[0042] Optionally, obtaining and parsing the raw log data to get error information data includes:

[0043] Read raw log data line by line;

[0044] Determine whether the original log data contains error codes or anomalies;

[0045] If so, the error information data in the original log data is extracted, standardized, and stored in a preset error information database;

[0046] Specifically, the error information data in the error information database can include: the service name that caused the error, the error code or exception information, the time of the error, and the server where the error occurred.

[0047] If not, the raw log data will be standardized and stored in a preset standard information database.

[0048] Specifically, the standard information data in the standard information database may include: service name, log time, and the server where the log is located.

[0049] For example, such as Figure 2 As shown, the system logs are parsed in real time, line by line. The system logs are read line by line, and the presence of error information is determined by the error code or exception. If error information is found, it is standardized and recorded in the error information database. The main information recorded may be the service name that caused the error, the error code or exception information, the error time, and the server where the error occurred. If no error information is found, the log is parsed in a standardized manner, and the standardized data is recorded in the standard information database. The main information recorded may be the service name, the log time, and the server where the log was located.

[0050] Optionally, the step of using a preset machine learning algorithm to analyze data in the relevant database to obtain the computation time configuration and failure probability configuration includes:

[0051] Data from the error information database and the standard information database are obtained to obtain the first sample data;

[0052] Using a pre-defined machine learning algorithm, the impact of different computation time configurations on the error rate of the first sample data in different time periods is analyzed, the optimal computation time configuration is calculated, and the pre-defined time is obtained.

[0053] Data from a preset fault information database and a preset error rate database are obtained to obtain second sample data;

[0054] Using a pre-set machine learning algorithm, the second sample data is statistically analyzed based on the growth of error information data and the status of fault information data for different services. The optimal fault probability configuration is calculated to obtain the first threshold, the second threshold, and the third threshold.

[0055] Specifically, fault rules in the fault rule database can include: service name, rule name (e.g., calculation period configuration when an error occurs, fault probability), and rule value (e.g., 6:00-9:00, XX minutes; 9:00-20:00, SS minutes; 20:00-6:00, KK minutes; low fault probability = 2 ≥ HEMR > 1; medium fault probability = 3 ≥ HEMR > 2; high fault probability = HEMR > 3).

[0056] For example, using data from the error information database and the standard information database, a statistical machine learning algorithm is used to analyze the impact of different computation time configuration error rates in different time periods, determine the optimal value of the preset time within the data sample, and update it to the fault rule database; using data from the fault information database and the error rate database, statistical analysis is performed on the fault information data and the error information data, and based on the growth trend of the error rate in the error information data and the actual situation of the fault information data, the optimal fault probability configuration, namely the optimal first threshold, the optimal second threshold, and the optimal third threshold, is determined and updated to the fault rule database.

[0057] Optionally, the step of analyzing the error information data based on the computation time configuration and the failure probability configuration to obtain the failure information data includes:

[0058] Extract newly generated, unprocessed error information data from the error information database;

[0059] Based on the error information database, determine whether the error code of the error information data is generated for the first time;

[0060] If so, then retrieve the standard information data of the service corresponding to the error information data from the standard information database;

[0061] Based on the standard information database, determine the access volume of the service corresponding to the error information data on the previous day, and analyze the determination result to obtain the fault information data;

[0062] The fault information data is stored in the fault information database.

[0063] Specifically, the fault information data in the fault information database may include: service name, error code or exception information, error rate, historical maximum error rate, ratio of the historical maximum error rate, probability of failure, and whether the error is occurring for the first time.

[0064] For example, such as Figure 3 As shown, by analyzing newly generated error information data in the error information database, the error rate is calculated, an alarm is triggered, and the probability of a fault is determined. Newly generated, unprocessed error information data is retrieved from the error information database. This error information data can be: service SY, error code 1001, server 10.201.xx.xx, and time of error occurrence: 2024.6.20 06:15:21. Based on the error information database, it is determined whether error code 1001 for service SY is a first occurrence. If it is, the previous access volume of service SY is analyzed to obtain fault information data, which is then stored in the fault information database. Fault information data can include: service name, error code or exception information, fault probability (low, medium, high), and whether the error is a first occurrence.

[0065] Optionally, the step of determining the service access volume corresponding to the error information data on the previous day based on the standard information database, and analyzing the determination result to obtain the fault information data includes:

[0066] Based on the standard information database, determine whether the service corresponding to the error information data had any access volume the previous day;

[0067] If not, the service corresponding to the error information data is a newly launched service, and the service is determined to be normal;

[0068] If so, the service corresponding to the error information data is not a newly launched service, and the service is determined to be faulty, thus obtaining the fault information data.

[0069] For example, if the error code 1001 of service SY obtained from the error information database is generated for the first time, then check whether service SY had any access volume the previous day from the standard data storage. If there was no access volume the previous day, service SY is a newly launched service; otherwise, it is a non-newly launched service. If it is a non-newly launched service, it is determined to be a fault, and the fault information data is stored in the fault information database. If it is a newly launched service, it is determined to be normal, and the analysis ends.

[0070] Optionally, determining whether the error code of the error information data is generated for the first time based on the error information database includes:

[0071] If the error code in the error information data is not the first time it has been generated, extract the total number of errors reported by the service corresponding to the error information data within the preset time period from the error information database;

[0072] Obtain the total number of accesses to the service corresponding to the error information data within the preset time period from the standard information database;

[0073] Based on the total number of errors and the total number of accesses, the error information data of the service corresponding to the error information data is obtained, and the error rate ER in the error information data of service SY is calculated. SY ,have

[0074]

[0075] In the formula, SE SY To serve the total number of errors reported by SY within a preset time, TV SY The total number of visits to service SY within a preset time period;

[0076] Specifically, the error information data in the error rate database includes the service name, error code or exception information, time, error rate, and status (pending, normal, deleted).

[0077] The error information data is stored in a preset error rate database;

[0078] Based on the error rate in the error information data, fault information data is obtained through analysis.

[0079] For example, the total number of errors SE reported for the service (e.g., SY) and the error code (e.g., 1001) within a preset time period (determined based on machine learning results) is retrieved from the error information database and calculated. SU The total number of visits to the service, such as SY, within a preset time period is obtained and calculated from the standard information database. SY Then according to Get the error rate (ER) for this service. SYThis process generates error information data, which is then stored in the error rate database. The error information data can include: service name, error code or exception information, time, error rate, and status (in this case, pending processing).

[0080] Optionally, the step of analyzing and obtaining fault information data based on the error rate in the error rate information data includes:

[0081] Obtain the maximum value of the error rate of the service corresponding to the error information data from the error rate database;

[0082] The fault ratio is calculated using the maximum error rate and the error rate in the error information data.

[0083] Based on the optimal failure probability configuration of the service corresponding to the error information data, the magnitude of the failure ratio value is determined, and the failure information data is obtained based on the determination result.

[0084] For example, retrieve the historical maximum error rate for this error in the error rate database, such as the maximum error rate P-ER(SY-1001) for error code 1001 in service SY; calculate the current error rate ER for error code 1001 in service SY. SU The ratio of the historical maximum error rate P-ER(SY-1001) to the error rate HMER(SY-1001) = ER SY / P-ER(SY-1001) retrieves fault rules related to the ratio value from the fault rule database. If the fault rule is set to HMER<1, the error is considered normal, and the status of the data in the error rate database is updated to normal; if HMER>1, a fault may exist.

[0085] Optionally, based on the optimal failure probability configuration of the service corresponding to the error information data, the magnitude of the failure ratio value is determined, and the failure information data obtained based on the determination result includes:

[0086] Determine whether the fault ratio value is greater than the first threshold; if not, determine that the service is normal.

[0087] If yes, then determine whether the fault ratio value is greater than the second threshold; if no, then determine that the fault probability is low and obtain the first fault data.

[0088] If yes, then determine whether the fault ratio value is greater than the third threshold; if no, then determine the fault probability as medium and obtain the second fault data.

[0089] If so, the probability of a fault is high, and the third fault data is obtained;

[0090] The first fault data, the second fault data, and the third fault data are stored in the fault information database.

[0091] For example, fault rules are retrieved from the fault rule database, and the probability of a fault (low / medium / high) is determined based on the fault ratio value. When the fault ratio value is less than a first threshold, the status of the error information data in the error rate database is updated to "service normal". For example, if the first threshold is set to 1, the second threshold to 2, and the third threshold to 3, the fault rule is: if 2 ≥ HMER > 1, the probability of a fault is low; if 3 ≥ HMER > 2, the probability of a fault is medium; and if HMER > 3, the probability of a fault is high. Then, the fault information data, namely the first fault data, the second fault data, and the third fault data, are stored in the fault information database. The fault information data may include: service name (SY), error code or exception information (1001), and error rate ER. SY The system includes the historical maximum error rate P-ER (SY-1001), the ratio of the historical maximum error rate HMER (SY-1001), the probability of failure (low / medium / high), and whether the error is the first occurrence. Initial settings can be set with a first threshold of 1, a second threshold of 2, and a third threshold of 3. These settings are then adjusted based on actual alarms and failures, using statistical machine learning results. For example, if the failure ratio HMER (SY-1001) for error code 1001 in service SY is 2, a failure has been confirmed. The corresponding ratio here should be adjusted. Setting HMER (SY-1001) ≥ 2 indicates a high probability of failure. Similarly, the low and medium values ​​can be adjusted proportionally, for example, by a ratio of 2 / 3 downwards.

[0092] Optionally, sending the fault information data to trigger a fault alarm according to preset alarm rules includes:

[0093] Retrieve alarm rules from the preset alarm rule database;

[0094] Specifically, the alarm rules in the alarm rule database can include: fault probability, alarm method, and alarm sender.

[0095] Retrieve current fault information data from the fault information database;

[0096] Based on the alarm rules and current fault information data, generate alarm information and issue a fault alarm;

[0097] For example, fault information data is obtained from the fault information database, and the system maintenance personnel are notified via SMS, email, WeChat notification, etc., according to the alarm rules in the alarm rule database.

[0098] Based on the alarm information, determine whether a fault exists;

[0099] If so, the status of the corresponding fault information data in the fault information database will be updated to "confirmed", and the status of the corresponding error information data in the error rate database will be updated to "deleted".

[0100] If not, the corresponding fault information data status in the fault information database will be updated to false alarm.

[0101] For example, after receiving an alarm message, the system maintenance personnel confirm whether the fault exists. If the fault exists, the status of the fault information data in the fault information database is updated to "confirmed," and the relevant information in the error rate database is updated to "deleted" (to avoid using this as a standard for subsequent fault judgment). If the fault does not exist, the status of the fault information data in the fault information database is updated to "false alarm," and the error rate database is updated according to the actual situation. If the error rate deviates too much from the error rate under normal circumstances, the relevant information in the error rate database can be updated to "deleted" (not used as a basis for subsequent analysis); otherwise, it is updated to "normal."

[0102] Based on the same inventive concept, such as Figure 4 As shown, the present invention also provides an intelligent fault early warning device for civil aviation systems, the device comprising:

[0103] The database storage module is used to establish a database related to fault information;

[0104] The log parsing module is used to obtain raw log data, parse it, obtain error information data, and store it in the relevant database.

[0105] The fault rule management module is used to analyze data in the relevant database using preset machine learning algorithms to obtain computation time configuration and fault probability configuration.

[0106] The fault analysis module is used to analyze the error information data according to the calculation time configuration and the fault probability configuration to obtain fault information data;

[0107] The fault alarm module is used to send the fault information data to trigger a fault alarm according to preset alarm rules.

[0108] It should be noted that the electrical connections between the various units described above do not necessarily represent connections between circuits. Indirect connections are acceptable as long as they achieve the objectives of this invention. The above descriptions are merely exemplary embodiments of this invention and should not be construed as limiting the scope of the invention.

[0109] All equivalent changes and modifications made in accordance with the teachings of this invention are still within the scope of this invention. Those skilled in the art will readily conceive of other embodiments of this invention upon considering the specification and the disclosure of practical truth. This application is intended to cover any variations, uses, or adaptations of this invention that follow the general principles of this invention and include common knowledge or conventional techniques in the art not described herein.

Claims

1. A method for intelligent fault early warning in civil aviation systems, characterized in that, The method includes: Establish a database related to fault information; Obtain the raw log data and parse it to obtain the error information data; Data from the error information database and the standard information database are obtained to obtain the first sample data; Using a pre-defined machine learning algorithm, the impact of different computation time configurations on the error rate of the first sample data in different time periods is analyzed, the optimal computation time configuration is calculated, and the pre-defined time is obtained. Data from a preset fault information database and a preset error rate database are obtained to obtain second sample data; Using a pre-set machine learning algorithm, the second sample data is statistically analyzed based on the growth of error information data and the status of fault information data for different services, and the optimal fault probability configuration is calculated to obtain the first threshold, the second threshold and the third threshold. Extract newly generated, unprocessed error information data from the error information database; Based on the error information database, determine whether the error code of the error information data is generated for the first time; If so, then retrieve the standard information data of the service corresponding to the error information data from the standard information database; Based on the standard information database, determine the access volume of the service corresponding to the error information data on the previous day, and analyze the determination result to obtain the fault information data; The fault information data is stored in the fault information database; According to the preset alarm rules, the fault information data is sent to trigger a fault alarm.

2. The intelligent fault early warning method for civil aviation systems according to claim 1, characterized in that, The process of obtaining and parsing the raw log data to obtain error information data includes: Read raw log data line by line; Determine whether the original log data contains error codes or anomalies; If so, the error information data in the original log data is extracted, standardized, and stored in a preset error information database; If not, the raw log data will be standardized and stored in a preset standard information database.

3. The intelligent fault early warning method for civil aviation systems according to claim 1, characterized in that, The step of determining the service access volume corresponding to the error information data on the previous day based on the standard information database, and analyzing the determination result to obtain the fault information data includes: Based on the standard information database, determine whether the service corresponding to the error information data had any access volume the previous day; If not, the service corresponding to the error information data is a newly launched service, and the service is determined to be normal; If so, the service corresponding to the error information data is not a newly launched service, and the service is determined to be faulty, thus obtaining the fault information data.

4. The intelligent fault early warning method for civil aviation systems according to claim 1, characterized in that, The step of determining whether the error code of the error information data is generated for the first time based on the error information database includes: If the error code in the error information data is not the first time it has been generated, extract the total number of errors reported by the service corresponding to the error information data within the preset time period from the error information database; Obtain the total number of accesses to the service corresponding to the error information data within the preset time period from the standard information database; Based on the total number of errors and the total number of accesses, the error information data corresponding to the service is obtained, and the error rate in the error information data for service SY is calculated. ,have ， In the formula, To serve the total number of errors reported by SY within a preset time, The total number of visits to service SY within a preset time period; The error information data is stored in the error rate database; Based on the error rate in the error information data, fault information data is obtained through analysis.

5. The intelligent fault early warning method for a civil aviation system according to claim 4, characterized in that, The step of analyzing the fault information data based on the error rate in the error rate information data includes: Obtain the maximum value of the error rate of the service corresponding to the error information data from the error rate database; The fault ratio is calculated using the maximum error rate and the error rate in the error information data. Based on the optimal failure probability configuration of the service corresponding to the error information data, the magnitude of the failure ratio value is determined, and the failure information data is obtained based on the determination result.

6. The intelligent fault early warning method for a civil aviation system according to claim 5, characterized in that, The step of determining the fault ratio value based on the optimal fault probability configuration of the service corresponding to the error information data, and obtaining the fault information data based on the determination result, includes: Determine whether the fault ratio value is greater than the first threshold; if not, determine that the service is normal. If yes, then determine whether the fault ratio value is greater than the second threshold; if no, then determine that the fault probability is low and obtain the first fault data. If yes, then determine whether the fault ratio value is greater than the third threshold; if no, then determine the fault probability as medium and obtain the second fault data. If so, the probability of a fault is high, and the third fault data is obtained; The first fault data, the second fault data, and the third fault data are stored in the fault information database.

7. The intelligent fault early warning method for a civil aviation system according to claim 4, characterized in that, The step of sending the fault information data to trigger a fault alarm according to preset alarm rules includes: Retrieve alarm rules from the preset alarm rule database; Retrieve current fault information data from the fault information database; Based on the alarm rules and current fault information data, generate alarm information and issue a fault alarm; Based on the alarm information, determine whether a fault exists; If so, the status of the corresponding fault information data in the fault information database will be updated to "confirmed", and the status of the corresponding error information data in the error rate database will be updated to "deleted". If not, the corresponding fault information data status in the fault information database will be updated to false alarm.

8. A civil aviation system intelligent fault early warning device, characterized in that, The device includes: The database storage module is used to establish a database related to fault information; The log parsing module is used to obtain raw log data, parse it, obtain error information data, and store it in the relevant database. The fault rule management module is used to acquire data from the error information database and the standard information database to obtain first sample data; using a preset machine learning algorithm, it analyzes the impact of different computation time configurations of the first sample data on the error rate in different time periods, calculates the optimal computation time configuration, and obtains the preset time; acquires data from the preset fault information database and the preset error rate database to obtain second sample data; using a preset machine learning algorithm, it performs statistical analysis on the second sample data based on the growth of error information data and the status of fault information data for different services, calculates the optimal fault probability configuration, and obtains the first threshold, the second threshold, and the third threshold. The fault analysis module is used to extract newly generated and unprocessed error information data from the error information database; determine whether the error code of the error information data is generated for the first time according to the error information database; if so, obtain the standard information data of the service corresponding to the error information data from the standard information database; determine the access volume of the service corresponding to the error information data on the previous day according to the standard information database, and analyze the determination result to obtain fault information data; and store the fault information data in the fault information database. The fault alarm module is used to send the fault information data to trigger a fault alarm according to preset alarm rules.

Images