Method, system and device for enhancing message processing stability of Internet of Things and medium

Abstract

The invention discloses a method, a system, equipment and a medium for enhancing message processing stability of the Internet of Things, and the method comprises the steps: a data source sends data to a monitoring module, and the monitoring module recognizes and filters the data; sending the abnormal data to a compensation module for processing; abnormal data can also be sent to the alarm module for strategy alarm; the normally circulated data is directly sent to the service module to be processed; if the business system encounters an error in the processing process, the error data and related abnormal information are transferred to the data mining module; the data mining module distributes the data to mining sub-modules of all professions according to data characteristics and error types; according to the method, advanced data mining algorithms such as TF-IDF are introduced and used for mining error keywords from error data and generating rules capable of predicting future errors. Compared with traditional statistical analysis or simple rule matching, the targeted data mining method has higher accuracy and intelligent level.

Description

Technical Field

[0001] The present invention relates to the technical field of information processing of the Internet of Things, and in particular to a method, system, device and storage medium for enhancing the stability of message processing of the Internet of Things. Background Art

[0002] In complex scenarios where the Internet of Things (IoT) platform is connected to many terminals, the stability of data reporting and sending faces multiple challenges. The IoT platform needs to access various types of terminal devices, and these devices have significant differences in performance. Some devices may have powerful processing capabilities and stable communication modules, and can efficiently process and transmit data; while other devices may be limited by factors such as cost and power consumption, and have relatively weak performance, which is prone to data processing delays, communication interruptions and other problems. In actual applications, the network environment is often unstable and may be affected by a variety of factors, such as signal interference, network congestion, device movement, etc. These factors may cause errors, failures or failures in the data transmission process, thereby affecting the overall performance of the IoT system.

[0003] Glossary: Dead letter storage: When encountering data that cannot be processed, it is generally sent to the dead letter module for storage and recording to facilitate subsequent problem troubleshooting.

[0004] TF-IDF: A statistical method for evaluating the importance of a term to a document set or a document in a corpus. Summary of the invention

[0005] The present invention provides a method, system, device and storage medium for enhancing the stability of Internet of Things message processing, which can at least solve one of the technical problems in the background technology.

[0006] To achieve the above object, the present invention adopts the following technical solutions: A method for enhancing the stability of IoT message processing comprises the following steps: The data source sends the data to the monitoring module, which identifies and filters the data; Send abnormal data to the compensation module for processing; Abnormal data will also be sent to the alarm module for strategic alarms; Normally flowing data is sent directly to the business module for processing; If the business system encounters an error during processing, the error data and related exception information will be forwarded to the data mining module; The data mining module intelligently distributes the data to various specialized mining submodules based on data characteristics and error types; The prediction rules generated by the data mining module, the penalty mechanism for the equipment, and the data placed in the dead letter storage module can be manually reviewed in the audit module to prevent misjudgment. If it is a misjudgment, the rules and data can be adjusted in the audit module.

[0007] Furthermore, the alarm module analyzes and evaluates the anomaly, and sends an alarm notification to relevant personnel based on preset thresholds or business rules combined with the alarm strategy.

[0008] Furthermore, the data source sends the data to the monitoring module, which identifies and filters the data according to the following steps, including: The monitoring module sets data invalidation rules, which read the message reporting time, compare it with the current time, and mark the messages that exceed the preset time threshold as invalid; The monitoring module sets a data missing rule, which counts the message reporting status of each terminal and marks the message that fails to be reported on time as missing; The monitoring module sets error keywords, data sources, and data target prediction rules, and predicts, identifies, and filters the data through the rules provided by the data mining module.

[0009] Further, the abnormal data is sent to the compensation module for processing, which specifically includes: According to the terminal identification information in the abnormal data, the terminal is instructed to re-report the missing, erroneous, and invalid data, thus realizing active real-time compensation of data; Transferring erroneous and invalid data to the dead letter storage module for archiving; The data that failed to be processed by the business system is sent to the delayed retry penalty module to punish the devices that failed to send data multiple times.

[0010] Furthermore, the data that failed to be processed by the business system is sent to the delayed retry penalty module, and the device that failed to send multiple times is punished, including: The failed messages are recorded in the error queue, and the dispatcher counts the number of failures of each message. Different life values ​​are assigned to the messages according to the number of failures. The higher the life value, the more failures there are. The dispatching program distributes these messages to queues with different delay levels according to their life values. The higher the life value, the longer the delay, and punishes device messages that fail multiple times; The delay program is responsible for monitoring the messages in these queues, and sending the messages that have reached the delay requirement to the retry queue for retry. If the retry still fails, the message is put into the error queue again.

[0011] Furthermore, the data mining module intelligently distributes the data to various professional mining sub-modules according to data characteristics and error types, including: a) Error keyword data mining module: collects error data and normal data, uses TF-IDF keyword mining algorithm to clean and segment the data, calculates the TF-IDF value of each word, and selects the top-ranked words as error keywords; generates prediction rules for data messages, such as this rule: when the keyword "null" appears in the data text, it is judged as abnormal data; b) Data source mining module, which collects error data and mines data source information, classifies data sources according to abnormal information, clusters the frequency of abnormal data in data sources using K-means clustering algorithm, divides data sources into high-quality data sources, low-quality data sources, and abnormal data sources, and generates prediction rules for corresponding data sources; c) Data target mining module, which collects data targets and times that are prone to errors when issuing instructions, and uses the Apriori algorithm to perform association rule mining on the collected data. The algorithm will output a series of association rules, identify the error mode of each device, and generate rules for predicting error time in the future using the fitted ARIMA model.

[0012] On the other hand, the present invention also discloses a system for enhancing the stability of Internet of Things message processing, which is used to implement the above-mentioned method for enhancing the stability of Internet of Things message processing, including a monitoring module, an error keyword data mining module, a data compensation module, an audit module, and an alarm module; The data source sends data to the monitoring module, which filters the data according to the rules; normal data is sent to the business processing module, and abnormal data is sent to the compensation module; The business system sends the failed data in the processing process to the data compensation module, and the error data and abnormal information to the data mining module; The data compensation module performs active real-time compensation based on the cause of the abnormality; Abnormal data is sent to the alarm module, and combined with the alarm strategy, an alarm notification is sent to relevant personnel; The data mining module analyzes the data to generate predictions and identification rules; The audit module manually reviews the prediction rules generated by the data mining module, the penalty mechanism for the device, and the data stored in the dead letter storage module.

[0013] Furthermore, the system also includes a delayed retry penalty module, which sends data that fails to be processed by the business system to the delayed retry penalty module and punishes devices that fail to send data multiple times.

[0014] On the other hand, the present invention further discloses a computer-readable storage medium storing a computer program, wherein when the computer program is executed by a processor, the processor executes the steps of the above method.

[0015] On the other hand, the present invention further discloses a computer device, including a memory and a processor, wherein the memory stores a computer program, and when the computer program is executed by the processor, the processor executes the steps of the above method.

[0016] It can be seen from the above technical solutions that the method for enhancing the stability of IoT message processing of the present invention has the following beneficial effects: There is a lack of an automatic troubleshooting, compensation, and penalty processing mechanism for processing failures caused by the uncertainty of the Internet of Things to improve the stability and processing efficiency of the message processing system.

[0017] When a business system handles errors, it usually has a retry mechanism. However, the present invention provides a universal and complete dynamic adjustment strategy (such as a penalty mechanism) based on the number of device failures.

[0018] Data Mining Technology In IoT message processing systems, there are relatively few systems specifically designed to mine erroneous data and generate prediction rules.

[0019] Active real-time data compensation based on terminal identification.

[0020] Deep data mining and intelligent prediction take preventive measures to reduce the generation of erroneous data and improve data processing efficiency and accuracy.

[0021] Comprehensive data filtering and real-time compensation mechanisms ensure the integrity and timeliness of data, avoiding business interruptions or decision-making errors caused by data errors or omissions.

[0022] Flexible penalty mechanism and dynamic retry strategy improve the reliability and stability of data transmission.

[0023] The smart strategy of this method is auditable.

[0024] The entire system operates without intrusion into the business system.

[0025] Specifically, compared with the prior art, the main creativity of the present invention lies in: The present invention introduces advanced data mining algorithms such as TF-IDF, which are specifically used to mine error keywords from error data and generate rules that can predict future errors. This targeted data mining method has higher accuracy and intelligence level than traditional statistical analysis or simple rule matching.

[0026] The present invention introduces a penalty mechanism based on the number of device failures and retry queues with different delay levels into the business system. This mechanism dynamically adjusts the retry strategy according to the performance of the device.

[0027] The present invention closely combines anomaly mining, prediction, data filtering and real-time compensation mechanism, and forms a closed-loop data processing flow with the original Internet of Things data processing module.

[0028] Misjudgments can be reversed through audits. BRIEF DESCRIPTION OF THE DRAWINGS

[0029] Figure 1 is a system block diagram of the present invention; Figure 2 Schematic diagram of the delayed retry penalty module of the present invention. DETAILED DESCRIPTION

[0030] In order to make the purpose, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments.

[0031] Due to the performance differences of various terminals and the instability of the environment, data may encounter errors, failures, and invalidations during transmission, which poses a challenge to the stability and processing efficiency of the system. In order to effectively solve these problems and improve system performance, an embodiment of the present invention proposes a system for enhancing the stability of processing IoT messages, as follows: The embodiments of the present invention can be used in various Internet of Things data transmission systems, such as smart meter systems, smart lighting systems, and smart sensor system lights.

[0032] like Figure 1 As shown, the system includes a monitoring module, an error keyword data mining module, a data compensation module, an audit module, a delay retry penalty module, and an alarm module; The data source sends data to the monitoring module, which filters the data according to the rules. Normal data is sent to the business processing module, and abnormal data is sent to the compensation module.

[0033] The business system sends failed data during processing to the data compensation module, and sends error data and exception information to the data mining module.

[0034] The compensation module performs active real-time compensation based on the cause of the abnormality.

[0035] Abnormal data is sent to the alarm module, and combined with the alarm strategy, alarm notifications are sent to relevant personnel.

[0036] The data mining module analyzes the data to generate predictions and identification rules.

[0037] The audit module manually reviews the prediction rules generated by the data mining module, the penalty mechanism for the device, and the data stored in the dead letter storage module.

[0038] Accordingly, the method steps for enhancing the stability of IoT message processing disclosed in the embodiment of the present invention are as follows: 1) The data source sends data to the monitoring module, which identifies and filters the data according to the following steps.

[0039] The monitoring module can set data invalidation rules. The rule reads the message reporting time, compares it with the current time, and marks messages that exceed the preset time threshold as invalid.

[0040] The monitoring module can set data missing rules. The rules will count the message reports of each terminal, and mark the messages that fail to report on time as missing.

[0041] The monitoring module can set error keywords, data sources, and data target prediction rules. The data is predicted and identified and filtered through the rules provided by the data mining module.

[0042] 2) These abnormal data will be sent to the compensation module for processing a) Based on the terminal identification information in the abnormal data, the terminal is instructed to re-report the missing, erroneous, or invalid data, thereby realizing active real-time compensation of the data.

[0043] b) Transfer erroneous and invalid data to the dead letter storage module for archiving.

[0044] c) Send the data that failed to be processed by the business system to the delayed retry penalty module, and punish the device that failed to send data multiple times. Figure 2 As shown, the steps are as follows: 1. The failed messages are recorded in the error queue, and the dispatcher counts the number of failures of each message. According to the number of failures, different life values ​​are assigned to the messages (initial 1, up to 5), and the higher the life value, the more failures.

[0045] 2. The dispatching program distributes these messages to queues with different delay levels according to their life values. The higher the life value, the longer the delay. For example, messages with a life value of 1 are sent to a queue with a delay of 3 seconds. Messages with a life value of 3 are sent to a queue with a delay of 5 minutes. Device messages that fail multiple times are penalized.

[0046] 3. The delay program is responsible for monitoring the messages in these queues, and sends the messages that have reached the delay requirement to the retry queue for retry. If the retry still fails, the message will be put into the error queue again.

[0047] 3) Abnormal data will also be sent to the alarm module for strategic alarm. The alarm module will analyze and evaluate the abnormality, and send alarm notifications to relevant personnel based on preset thresholds or business rules combined with alarm strategies.

[0048] 4) For data that flows normally, they will be sent directly to the business module for processing. If the business system encounters an error during processing, the error data and related exception information will be transferred to the data mining module. This module intelligently distributes this data to various professional mining sub-modules based on data characteristics and error types for the following processing.

[0049] a) Error keyword data mining module: collects error data and normal data, uses TF-IDF keyword mining algorithm to clean and segment the data, calculates the TF-IDF value of each word, and selects the top-ranked words as error keywords. Generates prediction rules for data messages, such as this rule: when the keyword "null" appears in the data text, it is judged as abnormal data.

[0050] b) Data source mining module, collects error data and mines data source information, classifies data sources according to abnormal information, uses K-means clustering algorithm to cluster the frequency of abnormal data in data sources, divides data sources into high-quality data sources, low-quality data sources, and abnormal data sources, and generates prediction rules for corresponding data sources. For example, the data source numbered 1 is marked as an abnormal data source, and its data is likely to be abnormal data, and its data is sent to a low-priority thread for identification.

[0051] c) Data target mining module, collects data targets and times that are prone to errors when issuing instructions, and uses the Apriori algorithm to mine association rules on the collected data. The algorithm will output a series of association rules, such as "Device A fails between 3 and 4 pm, and the confidence that device B also fails in this time period is 80%." Identify the error mode of each device, such as periodic errors, trend errors, etc. Generate rules for predicting error time in the future using the fitted ARIMA model. For example: Data target A is predicted to fail at 3 o'clock, and the issued data will be sent to a low-priority thread.

[0052] 5) The prediction rules generated by the data mining module, the penalty mechanism for the device, and the data put into the dead letter storage module can be manually reviewed in the audit module to prevent misjudgment. If it is a misjudgment, the rules and data can be adjusted in the audit module.

[0053] The following takes the IoT data transmission system of a smart meter as an example; under normal circumstances, the smart meter will report its own energy consumption data every hour.

[0054] 1) Some electric meters are installed in the meter box in the corner, and the network signal is not good. The data reported at the hour fails to reach the IoT system. When counting the meter readings, missing data may occur. The embodiment of the present invention uses the following method for processing.

[0055] a) The monitoring system will record the reporting of statistical data according to the preset data missing business rules. For example, the rule can be configured as follows: if a certain data is not reported within 20 minutes, it will be marked as missing data.

[0056] b) The monitoring system will send the meter terminal number and its corresponding time information of the meter that failed to report data on time to the data compensation module. The module will then issue instructions to prompt the relevant equipment to actively report its status information to complete the missing data.

[0057] 2) Due to network congestion, the data originally planned to be reported by the device at 0 o'clock was delayed for half an hour before it was successfully reported. For the business system, since the information of the device has greatly exceeded the scheduled receiving time limit, it is considered invalid. This patent uses the following method for processing.

[0058] a) The data source mining module will classify and mine the timeout data sources and normal data sources provided by the business system, and classify these data sources into low-quality data sources, high-quality data sources, etc.

[0059] b) The monitoring system will predict the data based on the data source classification rules provided by the data source mining module. If it is a low-quality data source, it is very likely that timeout data will be reported, which will be sent to a high-priority thread for judgment and processing.

[0060] c) After setting the timeout rule in the monitoring system, the system will compare the data receiving time with the device reporting time. For data exceeding the preset threshold, the monitoring system will not send it directly to the business system, but will transfer it to the data compensation module. The data compensation module will complete the invalid data by issuing instructions. At the same time, the timed data will also be sent to the dead letter storage module for future viewing and analysis.

[0061] 3) When the meter reports data, due to differences in version models, data format errors may occur occasionally. Specifically, some meters should transmit integer data, but they transmit floating point data by mistake, which makes it impossible for the business system to perform correct data format conversion during processing. This patent uses the following method for processing.

[0062] a) The error keyword module uses TF-IDF to deeply compare the differences between normal data and abnormal data, and identifies features containing floating point data as the source of error data. After the rule is confirmed by manual audit, it is sent to the monitoring platform to enter and apply this rule.

[0063] b) The monitoring system predicts the data according to the error keyword rules provided. If floating point data appears, it is erroneous data and the message is filtered. At the same time, it is sent to the data compensation module to let the device re-report the data.

[0064] 4) The IoT platform will send some instructions to the meter. Due to network environment and other issues, the instructions may fail to be sent. Regularly retrying to send instructions puts too much pressure on the platform. This patent uses the following method to handle this problem.

[0065] a) If data often fails to be sent to the meter at 1 pm, it will be sent to the data target mining module to collect the error-prone meter device identification and sending time, and mine the error-prone data target rules. Then, the next time it is sent to the meter at this time, it will be sent using a low-priority thread to reduce system pressure.

[0066] b) At the same time, the message of the failed download of the meter is sent to the error queue, and the dispatch program counts the number of failures of this message and assigns a life value of 1. The dispatch program sends this message to the queue with a delay time of 3 seconds. After blocking for 3 seconds, the delay program sends this message to the retry queue for retrying.

[0067] c) If the data delivery still fails, the system will put the message back into the error queue, and the dispatcher will assign a life value of 2 to the message, and send it to the queue with a 1-minute delay. After the message is blocked for 1 minute, the system will send it to the retry queue again and try to deliver it again.

[0068] d) If the data is sent successfully, the processing flow ends; if it fails, the above steps are continued to send the message to a queue with a higher delay level for blocking and retrying. This process is intended to implement "punitive" delay processing for messages that have failed to be sent multiple times, thereby reducing the system burden and achieving the purpose of optimizing system pressure.

[0069] 5) For the floating-point data keyword rules mined by the error keyword mining module, the low-quality data source rules mined by the data source mining module, the error-prone data target rules mined by the data target mining module, the penalty mechanism for the meter equipment that failed to send, and the data judged as dead letters, etc., they can all be manually reviewed in the audit module to prevent misjudgment. If it is a misjudgment, the rules and data can be adjusted in the audit module.

[0070] In summary, the embodiments of the present invention lack an automatic troubleshooting, compensation, and penalty processing mechanism for processing failures caused by uncertainty in the Internet of Things to improve the stability and processing efficiency of the message processing system. When processing errors, business systems usually have a retry mechanism. However, the embodiments of the present invention provide a general and complete dynamic adjustment strategy (such as a penalty mechanism) based on the number of device failures. Data mining technology is rare in the Internet of Things message processing system, and systems specifically used to mine and generate prediction rules from erroneous data. Active real-time data compensation based on terminal identification.

[0071] On the other hand, the present invention further discloses a computer-readable storage medium storing a computer program, wherein when the computer program is executed by a processor, the processor executes the steps of the above method.

[0072] On the other hand, the present invention further discloses a computer device, including a memory and a processor, wherein the memory stores a computer program, and when the computer program is executed by the processor, the processor executes the steps of the above method.

[0073] In another embodiment provided by the present application, a computer program product including instructions is also provided, which, when executed on a computer, enables the computer to execute any of the methods for enhancing the stability of IoT message processing in the above embodiments.

[0074] It is understandable that the system, device and storage medium provided in the embodiments of the present invention correspond to the method provided in the embodiments of the present invention, and the explanation, examples and beneficial effects of the relevant contents can refer to the corresponding parts in the above methods.

[0075] In the above embodiments, it can be implemented in whole or in part by software, hardware, firmware or any combination thereof. When implemented using software, it can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, the process or function described in the embodiment of the present application is generated in whole or in part. The computer may be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions may be stored in a computer-readable storage medium, or transmitted from one computer-readable storage medium to another computer-readable storage medium. For example, the computer instructions may be transmitted from a website site, a computer, a server or a data center by wired (e.g., coaxial cable, optical fiber, digital subscriber line (DSL)) or wireless (e.g., infrared, wireless, microwave, etc.) mode to another website site, computer, server or data center. The computer-readable storage medium may be any available medium that a computer can access or a data storage device such as a server or a data center that includes one or more available media integrated. The available medium may be a magnetic medium (e.g., a floppy disk, a hard disk, a tape), an optical medium (e.g., a DVD), or a semiconductor medium (e.g., a solid-state drive Solid State Disk (SSD)), etc.

[0076] It should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, the elements defined by the sentence "comprise a ..." do not exclude the existence of other identical elements in the process, method, article or device including the elements.

[0077] Each embodiment in this specification is described in a related manner, and the same or similar parts between the embodiments can be referred to each other, and each embodiment focuses on the differences from other embodiments. In particular, for the system embodiment, since it is basically similar to the method embodiment, the description is relatively simple, and the relevant parts can be referred to the partial description of the method embodiment.

[0078] The above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit the same. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that the technical solutions described in the aforementioned embodiments may still be modified, or some of the technical features may be replaced by equivalents. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for enhancing the stability of IoT message processing, characterized in that: The following steps are included: The data source sends the data to the monitoring module, which identifies and filters the data; Send abnormal data to the compensation module for processing; Abnormal data will also be sent to the alarm module for strategic alarms; Normally flowing data is sent directly to the business module for processing; If the business system encounters an error during processing, the error data and related exception information will be forwarded to the data mining module; The data mining module intelligently distributes the data to various specialized mining submodules based on data characteristics and error types; The prediction rules generated by the data mining module, the penalty mechanism for the equipment, and the data placed in the dead letter storage module can be manually reviewed in the audit module to prevent misjudgment. If it is a misjudgment, the rules and data can be adjusted in the audit module.

2. The method for enhancing the stability of Internet of Things message processing according to claim 1, characterized in that: It also includes an alarm module that analyzes and evaluates anomalies, and sends alarm notifications to relevant personnel based on preset thresholds or business rules combined with alarm strategies.

3. The method for enhancing the stability of Internet of Things message processing according to claim 1, characterized in that: The data source sends the data to the monitoring module, which identifies and filters the data according to the following steps, including: The monitoring module sets data invalidation rules, which read the message reporting time, compare it with the current time, and mark the messages that exceed the preset time threshold as invalid; The monitoring module sets a data missing rule, which counts the message reporting status of each terminal and marks the message that fails to be reported on time as missing; The monitoring module sets error keywords, data sources, and data target prediction rules, and predicts, identifies, and filters the data through the rules provided by the data mining module.

4. The method for enhancing the stability of Internet of Things message processing according to claim 1, characterized in that: Send abnormal data to the compensation module for processing, including: According to the terminal identification information in the abnormal data, the terminal is instructed to re-report the missing, erroneous, and invalid data, thus realizing active real-time compensation of data; Transferring erroneous and invalid data to the dead letter storage module for archiving; The data that failed to be processed by the business system is sent to the delayed retry penalty module to punish the devices that failed to send data multiple times.

5. The method for enhancing the stability of Internet of Things message processing according to claim 4, characterized in that: The data that failed to be processed by the business system is sent to the delayed retry penalty module, and the device that failed to send multiple times is punished, including: The failed messages are recorded in the error queue, and the dispatcher counts the number of failures for each message; Different life values ​​are assigned to messages according to the number of failures. A higher life value means more failures. The dispatching program distributes these messages to queues with different delay levels according to their life values. The higher the life value, the longer the delay, and punishes device messages that fail multiple times; The delay program is responsible for monitoring the messages in these queues, and sending the messages that have reached the delay requirement to the retry queue for retry. If the retry still fails, the message is put into the error queue again.

6. The method for enhancing the stability of Internet of Things message processing according to claim 1, characterized in that: The data mining module intelligently distributes the data to various professional mining sub-modules according to data characteristics and error types, including: a) Error keyword data mining module: collects error data and normal data, uses TF-IDF keyword mining algorithm to clean and segment the data, calculates the TF-IDF value of each word, and selects the top-ranked words as error keywords; Generate prediction rules for data packets; b) Data source mining module, which collects error data and mines data source information, classifies data sources according to abnormal information, clusters the frequency of abnormal data in data sources using K-means clustering algorithm, divides data sources into high-quality data sources, low-quality data sources, and abnormal data sources, and generates prediction rules for corresponding data sources; c) Data target mining module, which collects data targets and times that are prone to errors when issuing instructions, and uses the Apriori algorithm to perform association rule mining on the collected data. The algorithm will output a series of association rules, identify the error mode of each device, and generate rules for predicting error time in the future using the fitted ARIMA model.

7. A system for enhancing the stability of Internet of Things message processing, used to implement the method for enhancing the stability of Internet of Things message processing as described in any one of claims 1 to 5, characterized in that: It includes monitoring module, error keyword data mining module, data compensation module, audit module and alarm module; The data source sends data to the monitoring module, which filters the data according to the rules; normal data is sent to the business processing module, and abnormal data is sent to the compensation module; The business system sends the failed data in the processing process to the data compensation module, and the error data and abnormal information to the data mining module; The data compensation module performs active real-time compensation based on the cause of the abnormality; Abnormal data is sent to the alarm module, and combined with the alarm strategy, an alarm notification is sent to relevant personnel; The data mining module analyzes the data to generate predictions and identification rules; The audit module manually reviews the prediction rules generated by the data mining module, the penalty mechanism for the device, and the data stored in the dead letter storage module.

8. The system for enhancing the stability of IoT message processing according to claim 6, characterized in that: It also includes a delayed retry penalty module, which sends data that fails to be processed by the business system to the delayed retry penalty module and punishes devices that fail to send data multiple times.

9. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the processor is caused to perform the steps of the method according to any one of claims 1 to 6.

10. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that: When the computer program is executed by the processor, the processor is caused to perform the steps of the method according to any one of claims 1 to 6.

Images