Self-adaptive operation and maintenance quality monitoring and management method

Abstract

The invention provides a self-adaptive operation and maintenance quality monitoring and management method, which comprises the following steps that: monitoring indexes are dynamically adjusted according to real-time data and business requirements so as to ensure the flexibility and adaptability of monitoring, and the monitoring indexes comprise monitoring frequency; the method comprises the following steps: acquiring a quality data set on line through data, wherein the quality data set comprises quality information data of a dimension service object, system service environment data, fault abnormal data and key element index data; analyzing the quality data set, detecting abnormity by adopting a high-dimensional algorithm, and triggering early warning if a detected abnormal value exceeds an abnormal threshold value; performing nonlinear mapping on the current actual operation data and the performance index reference value, and triggering a dynamic adjustment mechanism service to deal with abnormity; and the system calculates a quality problem score according to the operation and maintenance performance, and gives a qualified result according to the quality problem score. According to the invention, by optimizing resource configuration, the operation and maintenance quality is ensured to be matched with service requirements, and finally higher-level service quality and management effect are realized.

Description

Technical Field

[0001] The present invention relates to the field of information system operation and maintenance, and in particular to an adaptive operation and maintenance quality monitoring and management method. Background Art

[0002] In modern operation and maintenance management, with the complexity and diversification of business, traditional operation and maintenance monitoring methods have been unable to meet the requirements of real-time and accuracy. Traditional methods often rely on static indicators and lack flexibility, resulting in an inability to respond to changing business needs in a timely manner. This situation not only affects the efficiency of operation and maintenance, but may also lead to system failures and service interruptions, causing significant economic losses to enterprises.

[0003] In addition, with the development of big data technology, the types and amount of operation and maintenance data have increased dramatically, and how to effectively collect, analyze and use this data has become a challenge. Many existing systems have lags in data collection and anomaly detection, and cannot achieve real-time monitoring and rapid response, resulting in potential problems not being identified and handled in a timely manner, thus affecting the overall operation and maintenance quality.

[0004] To solve the above problems, the present invention proposes an adaptive operation and maintenance quality monitoring and management method, which aims to improve the real-time and effectiveness of operation and maintenance management, optimize resource allocation, and improve overall operation and maintenance efficiency and service quality by dynamically adjusting monitoring indicators, real-time data collection and analysis, and automated anomaly detection mechanisms. Summary of the invention

[0005] The present invention proposes an adaptive operation and maintenance quality monitoring and management method, which solves the shortcomings of traditional operation and maintenance monitoring methods in terms of lack of flexibility and real-time performance. It can dynamically adjust monitoring indicators, collect and analyze operation and maintenance data in real time, and automatically identify abnormal situations, thereby improving the efficiency and accuracy of operation and maintenance management. The present invention optimizes resource allocation to ensure that operation and maintenance quality matches business needs, ultimately achieving a higher level of service quality and management effects.

[0006] An adaptive operation and maintenance quality monitoring and management method comprises the following steps:

[0007] Dynamically adjust monitoring indicators: Dynamically adjust monitoring indicators according to real-time data and business needs to ensure the flexibility and adaptability of monitoring. The monitoring indicators include monitoring frequency;

[0008] Data collection and integration: collect quality data sets online, including quality information data of maintenance service objects, system service environment data, fault anomaly data and key element indicator data;

[0009] Anomaly detection and data warning: Analyze the quality data set, use high-dimensional algorithms to detect anomalies, and trigger an alarm if the detected anomaly value exceeds the anomaly threshold;

[0010] Service standard comparison and dynamic adaptation: nonlinear mapping of current actual operation data and performance index reference values ​​is performed to trigger dynamic adjustment mechanism services to deal with abnormalities;

[0011] Quality analysis and judgment: The system calculates the quality problem score based on the operation and maintenance performance, and gives a qualified result based on the quality problem score.

[0012] Furthermore, when dynamically adjusting monitoring indicators, the comprehensive monitoring indicators are calculated:

[0013]

[0014] Where D(t) is the response time of the system real-time acquisition; B i (t) is the dynamic demand of user request volume; Ri(t) is the resource status, which includes CPU or bandwidth utilization; β is the time decay coefficient; ω i is the frequency; φ i is the phase; M(t) represents the comprehensive monitoring index at time t, which is the weighted result of multiple dynamic factors, integrating the response time, dynamic demand of user request volume, and resource status to reflect the comprehensive performance of the current operation and maintenance system. The dynamic adjustment of the monitoring index is based on an adaptive algorithm. When the comprehensive monitoring index M(t) or the response time, dynamic demand of user request volume, and resource status exceed the preset threshold, indicating that the system load has exceeded the normal range, the monitoring frequency is dynamically adjusted.

[0015] Furthermore, the quality information data includes request response time and throughput, the system service environment data includes CPU, memory, storage space, and bandwidth usage, the fault exception data includes system fault logs and error codes, and the key element indicator data includes service response time.

[0016] Furthermore, the quality data set is defined as:

[0017]

[0018] Among them, Q info，j (t) is the quality information data, S env，j (t) is the system service environment data, F exc，k (t) is the system service environment data, K elem，k (t) is the key factor indicator data; these data are integrated through the model to generate a real-time operation and maintenance data set Q data (t) serves as the input for anomaly detection.

[0019] Furthermore, the following anomaly detection function is used when the high-dimensional algorithm is used to detect anomalies:

[0020]

[0021] Among them, g(Q data (t)) is an anomaly detection algorithm based on high-dimensional statistical analysis, which is used to identify abnormal signals in complex, multi-dimensional data; H(x) is the expected state or threshold model used to describe the data; ∈ is the abnormal threshold.

[0022] Furthermore, when comparing and dynamically adapting the service standards, the adaptive deviation value Aadapt(t) of the system at time t is calculated:

[0023]

[0024] Among them, Aadapt(t) represents the adaptive deviation value of the system at time t, S std，k For operation and maintenance service standards, S curr，k is the current service status. The higher Aadapt(t) is, the more the current system status deviates from the standard value. adapt When (t)>γ, the system will trigger the dynamic adjustment mechanism, where γ is the preset adaptability threshold.

[0025] Furthermore, the dynamic adjustment mechanism includes: increasing computing resources, adjusting task priorities or optimizing load distribution strategies, wherein increasing computing resources includes allocating more CPU or memory, adjusting task priorities includes compressing low-priority task bandwidth, and optimizing load distribution strategies includes reallocating data processing tasks.

[0026] Furthermore, the system calculates the quality problem score based on the operation and maintenance performance, specifically including:

[0027]

[0028] Where Q(t) represents the quality problem score; A i (t) represents the actual monitoring value of a specific monitoring indicator at time t; α i is the weight; T i (t) is the target standard; σ i is the standard deviation;

[0029] When Q(t)>δ, a prompt is given, where δ is the judgment threshold, and a quality problem judgment list is formed. The final result judgment basis is:

[0030]

[0031] The present invention has the following advantages:

[0032] 1. Dynamic adjustment: This method can dynamically adjust operation and maintenance monitoring indicators according to real-time data and business needs to ensure that the operation and maintenance quality matches the actual needs.

[0033] 2. Real-time data collection: Through comprehensive data collection and analysis, the quality information and system status of the operation and maintenance service objects can be obtained in a timely manner, enhancing the real-time nature of decision-making.

[0034] 3. Automated anomaly detection: Establish an automatic anomaly detection mechanism that can quickly identify potential quality issues, improve response speed, and reduce human intervention.

[0035] 4. Customizability and scalability: The system is flexible and customizable, and can adapt to the operation and maintenance scenarios and standards of different enterprises, thus enhancing its applicability.

[0036] 5. Comprehensive quality assessment: Through the quality scoring model, it provides comprehensive quality analysis and judgment to support decision optimization and service improvement. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] Figure 1 It is a flow chart of an adaptive operation and maintenance quality monitoring and management method according to an embodiment of the present invention. DETAILED DESCRIPTION

[0038] In order to make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution 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 part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0039] The embodiment of the present invention provides an adaptive operation and maintenance quality monitoring and management method, which aims to realize intelligent monitoring and optimization of the power information system through real-time data collection, integration and analysis, and includes the following steps:

[0040] (1) Dynamically adjust monitoring indicators: Dynamically adjust monitoring indicators according to real-time data and business needs to ensure the flexibility and adaptability of monitoring. The monitoring indicators include monitoring frequency. The dynamic adjustment is based on an adaptive algorithm, as follows:

[0041]

[0042] Where D(t) is the response time of the system real-time acquisition; B i (t) is the dynamic demand of user request volume; R i (t) is the resource status, which includes CPU or bandwidth usage; β is the time decay coefficient; ωi is the frequency; φ i is the phase; M(t) represents the comprehensive monitoring index at time t, which is the weighted result of multiple dynamic factors, integrating the response time, dynamic demand of user request volume, and resource status to reflect the comprehensive performance of the current operation and maintenance system. The dynamic adjustment of the monitoring index is based on an adaptive algorithm. When the comprehensive monitoring index M(t) or the response time, dynamic demand of user request volume, and resource status exceed the preset threshold, indicating that the system load has exceeded the normal range, the monitoring frequency is dynamically adjusted.

[0043] Example: In an e-commerce festival event, the real-time response time D(t) is 0.8 seconds. Request volume B i (t): 500 times per second. CPU usage R i (t): 85%. Time attenuation coefficient β = 0.5; frequency ω i =2π, phase φ i =π / 4. According to the monitoring model, the system finds that the current response time is about to exceed the target threshold of 1 second due to high request volume and excessive resource usage. The system triggers an adjustment to increase the monitoring frequency.

[0044] (2) Data collection and integration: Through online data collection, quality data sets are collected. Specifically, quality information data of operation and maintenance service objects (such as request response time, throughput, etc.), system service environment data (such as CPU, memory, storage space, bandwidth, etc.), fault exception data (such as system fault logs, error codes) and key factor indicator data (such as service response time) are collected comprehensively. The quality data set is defined as:

[0045]

[0046] Among them, Q data (t): The collected quality data set. Q info，j (t): Quality information data, such as request response time and throughput. env，j (t): System service environment data, such as CPU, memory, and bandwidth usage. exc，k (t): System service environment data, such as system logs and error codes. elem，k (t): Key factor indicator data, such as service response time target.

[0047] Example: In an e-commerce festival, Q info，j (t) = 0.8 seconds (response time), the throughput is 1000 requests per second; S env，j (t) = 85% (CPU usage), memory usage is 60%; F exc，k (r): Crash log found, error code 500; K elem，k(t) = 1 second (target response time). These data are integrated through the model to generate a real-time operation and maintenance data set Q data (t), as the input for anomaly detection.

[0048] (3) Anomaly detection and data warning: Analyze the quality data set and use a high-dimensional algorithm to detect anomalies. If the detected anomaly value exceeds the anomaly threshold, an alarm is triggered. The following anomaly detection function is used when using a high-dimensional algorithm to detect anomalies:

[0049]

[0050] Where A(t): anomaly detection function. g(Q data (t)): Anomaly detection algorithm based on high-dimensional statistical analysis, used to identify abnormal signals in complex, multidimensional data. H(x): Used to describe the expected state or threshold model of data. ∈: Anomaly threshold, used to determine whether further processing is needed.

[0051] Example: When the system detects Q data When the CPU usage reaches 85% and there are 3 crash logs in (t), A(t)=1.2 (greater than the threshold ∈=1) is calculated. Therefore, the system triggers an alert, displays the exception and records it in the list.

[0052] (4) Service standard comparison and dynamic adaptation: Nonlinear mapping of current actual operation data with performance indicator reference values ​​is performed to trigger dynamic adjustment mechanism services to respond to anomalies.

[0053] Specifically, the adaptive deviation value Aadapt(t) of the system at time t is calculated as:

[0054]

[0055] Among them, S std，k : Operation and maintenance service standards (such as CPU usage ≤ 75%). curr，k : Current service status (such as CPU usage 85%). adapt When (t)>γ, the system will trigger the dynamic adjustment mechanism, where γ is the preset adaptability threshold.

[0056] Example: Dynamic adjustment of the system based on the adaptive deviation value A adapt (t) and threshold γ:

[0057] 1. Scenario description: The current system detects that the CPU usage is S curr，k = 85%, while the standard performance index is S std，k =75%.

[0058] 2. Calculate the adaptive deviation value Aadapt(t):

[0059]

[0060] A adapt (t)=0.72+log(11)≈0.72+2.40=3.12

[0061] 3. Threshold determination: The preset adaptive threshold is γ=2.5.

[0062] Compare: A adapt (t)=3.12>γ=2.5

[0063] Because Aadapt(t) exceeds the threshold, the system determines that the current state deviation is too large and dynamic expansion is required.

[0064] 4. Dynamic adjustment decisions: The system triggers the expansion mechanism and starts a new server instance through an automated process; the new server diverts part of the traffic, reducing the CPU usage of the main server, thereby optimizing the overall system performance.

[0065] 5. Adjusted status:

[0066] After the system adjustment, the CPU usage dropped to S curr，k =72%, the adaptive deviation value is recalculated as:

[0067]

[0068] A adapt (t)≈0.79+log(4)≈0.79+1.39=2.18

[0069] At this time A adapt (t)=2.18≤γ=2.5, the system returns to normal state and no further adjustment is required.

[0070] (5) Quality analysis and judgment: The system calculates the quality problem score based on the operation and maintenance performance, and gives a qualified result based on the quality problem score:

[0071]

[0072] Among them, A i (t): represents the actual monitoring value of a specific monitoring indicator at time t; Q(t): quality problem score. i : Weight. T i (t): target standard. σ i : Standard deviation. When Q(t)>δ (where δ is the judgment threshold), a prompt is given and a quality problem judgment list is formed. The final result judgment is based on:

[0073]

[0074] Example: Actual response time A i (t) = 0.8 seconds, target T i (t) = 1 second; weight α1 = 2, standard deviation σ1 = 0.1. Final score Q(t) = 102 (exceeding threshold δ = 100). The system determines it as "unqualified" and generates a list, suggesting optimization of database query performance.

[0075] The following is a specific example to illustrate:

[0076] 1. Dynamically adjust monitoring indicators

[0077] Test data table 1: Monitoring index data

[0078]

[0079] Calculation process:

[0080] (1) Calculate the weighted integral of the response time

[0081]

[0082] M1(t)=∫0 10 D(τ)·e -0.5(10-τ) dτ

[0083] Calculate the integral part first:

[0084] From 0 to 5 seconds: ∫0 5 0.5 e -0.5(10-τ) dτ≈0.0377

[0085] From 5 to 10 seconds: ∫5 10 0.6 e -0.5(10-τ) dτ≈1.1092

[0086] Comprehensive calculation: M1(10)≈1.1469

[0087] (2) Impact of Calculation Request Volume

[0088]

[0089] (3) Comprehensive monitoring indicators

[0090] M(10)=M1(10)+M2(10)≈5.3055

[0091] Set to trigger dynamic adjustment of monitoring indicators when M(t) is greater than 5 and increase the monitoring frequency.

[0092] 2. Data collection and integration

[0093] Test data table 2: Quality data collection

[0094] Data integration model calculation:

[0095]

[0096] Test data table 3: Data integration calculation

[0097] 3. Anomaly detection and data warning

[0098] Experimental data Table 4: Anomaly detection data

[0099] Time (seconds) Data Qdata(t) Smoothing noise H(t) Abnormal threshold ∈ Test result A(t) Perform an action 0 35 50 1 0 No action required 5 72.5 50 1 1.45 No action required 10 471 50 1 9.42 (trigger warning) Send warning information 15 2880 50 1 57.60 (trigger warning) Send warning information 20 5760 50 1 115.20 (trigger warning) Send warning information

[0100] 4. Service standard comparison and dynamic adaptation

[0101] Test data table 5: comparison between service status and standard

[0102] Calculation process:

[0103]

[0104] 5. Quality scoring and result determination

[0105] Test data table 6: Quality score calculation

[0106]

[0107] According to the experimental results, the present invention has the following characteristics and effects:

[0108] (1) Dynamically adjust monitoring indicators: The existing technology mainly relies on fixed monitoring indicators, which cannot quickly adapt to changes in business needs, resulting in the difficulty of operation and maintenance quality to reflect the actual situation in real time. However, the present invention automatically adjusts the operation and maintenance monitoring indicators by dynamically combining real-time data with business needs, ensuring that the monitoring system can respond to changes in a timely manner, thereby greatly improving the adaptability of operation and maintenance quality.

[0109] (2) Comprehensive data collection: Traditional operation and maintenance methods usually focus on a single data source, such as system performance or environmental status, which limits data support and may miss important information. In contrast, the present invention integrates quality information, environmental status and fault data through digital templates to achieve comprehensive data collection. This approach not only improves the diversity and accuracy of data, but also provides a solid foundation for subsequent analysis.

[0110] (3) Automatic anomaly detection mechanism: In the prior art, anomaly detection mostly relies on manual or semi-automatic methods, which usually takes a long time to identify and respond to potential quality issues, causing delays. The present invention introduces an automated anomaly detection mechanism, which analyzes a large amount of collected data in real time, can quickly identify potential quality anomalies, and significantly improves response speed and processing efficiency.

[0111] (4) Flexible quality analysis and judgment: The scoring methods in the prior art are often relatively simple and lack complex analysis models, making it difficult to accurately evaluate the quality of operation and maintenance, and the results are often not detailed enough. The present invention adopts a complex quality scoring model, comprehensively considers multiple parameters for evaluation, and generates a detailed quality problem judgment list, which can more effectively support decision optimization.

[0112] (5) Clarity of final results: The feedback mechanism of traditional operation and maintenance methods is usually not clear enough, relying on verbal reports or simple documents, lacking systematicity and transparency, resulting in low management efficiency. However, the present invention sets clear judgment criteria and adopts a clear result feedback mechanism, making the operation and maintenance results more transparent, significantly improving the efficiency and service level of operation and maintenance management.

[0113] The above is only a specific embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Any changes or substitutions that can be easily thought of by a person skilled in the art within the technical scope disclosed by the present invention should be included in the protection scope of the present invention. Therefore, the protection scope of the present invention should be based on the protection scope of the claims.

Claims

1. An adaptive operation and maintenance quality monitoring and management method, characterized in that: The following steps are involved: Dynamically adjust monitoring indicators: Dynamically adjust monitoring indicators according to real-time data and business needs to ensure the flexibility and adaptability of monitoring. The monitoring indicators include monitoring frequency; Data collection and integration: collect quality data sets online, including quality information data of maintenance service objects, system service environment data, fault anomaly data and key element indicator data; Anomaly detection and data warning: Analyze the quality data set, use high-dimensional algorithms to detect anomalies, and trigger an alarm if the detected anomaly value exceeds the anomaly threshold; Service standard comparison and dynamic adaptation: nonlinear mapping of current actual operation data and performance index reference values ​​is performed to trigger dynamic adjustment mechanism services to deal with abnormalities; Quality analysis and judgment: The system calculates the quality problem score based on the operation and maintenance performance, and gives a qualified result based on the quality problem score.

2. The method according to claim 1, characterized in that When dynamically adjusting monitoring indicators, calculate comprehensive monitoring indicators: Where D(t) is the response time of the system real-time acquisition; B i (t) is the dynamic demand of user request volume; R i (t) is the resource status, which includes CPU or bandwidth usage; β is the time decay coefficient; ω i is the frequency; φ i is the phase; M(t) represents the comprehensive monitoring index at time t, which is the weighted result of multiple dynamic factors, integrating the response time, dynamic demand of user request volume, and resource status to reflect the comprehensive performance of the current operation and maintenance system. The dynamic adjustment of the monitoring index is based on an adaptive algorithm. When the comprehensive monitoring index M(t) or the response time, dynamic demand of user request volume, and resource status exceed the preset threshold, indicating that the system load has exceeded the normal range, the monitoring frequency is dynamically adjusted.

3. The method according to claim 1, characterized in that The quality information data includes request response time and throughput, the system service environment data includes CPU, memory, storage space, and bandwidth usage, the fault exception data includes system fault logs and error codes, and the key element indicator data includes service response time.

4. The method according to claim 3, characterized in that The quality data set is defined as: Among them, Q info，j (t) is the quality information data, S env，j (t) is the system service environment data, F exc，k (t) is the system service environment data, K elem，k (t) is the key factor indicator data; these data are integrated through the model to generate a real-time operation and maintenance data set Q data (t) serves as the input for anomaly detection.

5. The method according to claim 4, characterized in that When the high-dimensional algorithm is used to detect anomalies, the following anomaly detection function is used: Among them, g(Q data (t)) is an anomaly detection algorithm based on high-dimensional statistical analysis, which is used to identify abnormal signals in complex, multi-dimensional data; H(x) is the expected state or threshold model used to describe the data; ∈ is the abnormal threshold.

6. The method according to claim 4, characterized in that When comparing service standards and dynamically adapting, the adaptive deviation value Aadapt(t) of the system at time t is calculated: Among them, Aadapt(t) represents the adaptive deviation value of the system at time t, S std，k For operation and maintenance service standards, S curr，k is the current service status. The higher Aadapt(t) is, the more the current system status deviates from the standard value. adapt When (t)>γ, the system will trigger the dynamic adjustment mechanism, where γ is the preset adaptability threshold.

7. The method according to claim 6, characterized in that The dynamic adjustment mechanism includes: increasing computing resources, adjusting task priorities or optimizing load distribution strategies, wherein increasing computing resources includes allocating more CPU or memory, adjusting task priorities includes compressing low-priority task bandwidth, and optimizing load distribution strategies includes reallocating data processing tasks.

8. The method according to claim 4, characterized in that The system calculates quality problem scores based on operation and maintenance performance, including: Where Q(t) represents the quality problem score; A i (t) represents the actual monitoring value of a specific monitoring indicator at time t; α i is the weight; T i (t) is the target standard; σ i is the standard deviation; When Q(t)>δ, a prompt is given, where δ is the judgment threshold, and a quality problem judgment list is formed. The final result judgment basis is:

Images