Intelligent early warning management system and method for information center machine room

Abstract

The invention relates to the technical field of information center machine rooms, in particular to an intelligent early warning management system and method for an information center machine room. The system is composed of a data acquisition module, a data preprocessing module, an intelligent analysis module, an early warning decision module and an interactive response interface, and the data acquisition module collects various environmental parameter data and equipment operation state data in real time through a sensor of a central machine room; the data preprocessing module is used for cleaning, denoising and complementing environment parameter data and equipment operation state data, the intelligent analysis module is used for establishing a multi-parameter correlation analysis model and identifying an abnormal mode and potential risks by using a graph model algorithm, and the early warning decision module is used for automatically generating an early warning signal based on an output result of the intelligent analysis module. A diagnosis report and suggested measures are provided according to the early warning level, and the interactive response interface displays early warning information and real-time data. According to the invention, early signs of data change of the information center machine room can be captured, early warning is carried out in advance, and occurrence of sudden faults is reduced.

Description

Technical Field

[0001] The present invention relates to the technical field of information center computer rooms, and in particular to an intelligent early warning management system and method for an information center computer room. Background Art

[0002] With the continuous development of the big data era, the equipment of the data processing center is also constantly updated and developed. Among them, the information center computer room is the terminal equipment used for massive data transmission. It needs to adapt to the corresponding early warning management system. Although the traditional monitoring system can detect environmental anomalies based on fixed thresholds, its limitations are becoming increasingly prominent. First, in the face of subtle gradual changes or abnormal trends in environmental parameters, such as the temperature slowly rising but not reaching the warning line, the system often reacts slowly and misses the opportunity for early intervention. Secondly, the system lacks deep intelligent analysis capabilities and it is difficult to gain insight into the complex correlations and mutual influences between various parameters, which limits its ability to predict potential faults. Furthermore, once the alarm is triggered, the information provided is mostly a single-dimensional notification of exceeding the standard, lacking multi-dimensional comprehensive diagnosis and suggestions, which requires operation and maintenance personnel to spend a lot of time on fault location and cause analysis, seriously affecting fault response and processing efficiency.

[0003] Therefore, the present invention provides an intelligent early warning management system and method for an information center computer room, which significantly improves the intelligence level of the monitoring system and achieves keen insight into and efficient response to changes in the information center environment. Summary of the invention

[0004] The present invention provides an intelligent early warning management system and method for an information center computer room, which are used to solve the technical problems of slow response and lack of deep intelligent analysis capability of the existing early warning management system.

[0005] In order to achieve the above-mentioned purpose, the technical scheme adopted by the present invention is as follows: an intelligent early warning management system for an information center computer room, consisting of a data acquisition module, a data preprocessing module, an intelligent analysis module, an early warning decision module and an interactive response interface. The data acquisition module collects a variety of environmental parameter data and equipment operation status data in real time through sensors in the central computer room. The data preprocessing module cleans, denoises and completes the environmental parameter data and equipment operation status data. The preprocessed environmental parameter data and equipment operation status data are used by the intelligent analysis module to apply a graphical model algorithm to establish a multi-parameter correlation analysis model to identify abnormal patterns and potential risks. The early warning decision module automatically generates an early warning signal based on the output results of the intelligent analysis module, and provides a diagnostic report and recommended measures according to the early warning level. The interactive response interface displays the early warning information and the real-time collected environmental parameter data and equipment operation data.

[0006] Furthermore, the environmental parameter data includes temperature, humidity, and air quality, and the equipment operation status data includes power status, equipment load, and network traffic.

[0007] Furthermore, the data cleaning adopts the LOF algorithm to identify outliers by calculating the local density deviation of each data point relative to the data points in the neighborhood.

[0008] Furthermore, the data denoising adopts improved wavelet denoising, and wavelet transform is performed on the cleaned data to obtain wavelet coefficients, and the wavelet coefficients are processed with a threshold, and then the processed wavelet coefficients are used to perform inverse wavelet transform to obtain denoised signal data.

[0009] Furthermore, the data completion adopts a multiple interpolation method, and the specific process is: identifying missing data, establishing an interpolation model, generating an interpolation data set, analyzing the interpolation data set, and merging the results.

[0010] Furthermore, the multi-parameter association analysis model establishment process is as follows:

[0011] (3) Graph Construction

[0012] 1) Node construction:

[0013] Each data is regarded as a node in the graph, namely temperature, humidity, air quality, power status, device load, network traffic, and the node contains the current value, historical data, threshold information, and status identification of the data.

[0014] 2) While building:

[0015] The edge represents the association relationship between data. The association relationship is the correlation coefficient based on the data relevance, and the edge weight represents the strength of the association.

[0016] 3) Graph construction:

[0017] Build a complete graph structure based on the relationship between data;

[0018] (4) Algorithm process

[0019] 1) Anomaly detection rules:

[0020] Defining anomaly detection rules for each of the nodes, and selecting detection rules based on thresholds;

[0021] 2) Graph traversal algorithm

[0022] A graph traversal algorithm is used to detect anomalous patterns across parameters and a graph algorithm using depth-first search (DFS) is used to traverse the graph and calculate the anomaly score or risk level of the nodes.

[0023] 3) Abnormal pattern recognition:

[0024] By analyzing the association paths and anomaly scores between nodes, complex anomaly patterns are identified.

[0025] Furthermore, the early warning decision module divides abnormal events into different warning levels, namely urgent, important, and general, according to the risk assessment results, and generates corresponding early warning signals, which include sound alarms, pop-up prompts, and SMS notifications. For each early warning event, the early warning decision module generates a detailed diagnostic report, including the specific values ​​of the abnormal parameters, change trends, causes, and historical similar cases.

[0026] Furthermore, the intelligent analysis module also includes an anomaly detection algorithm that learns normal behavior patterns based on historical data and compares current data in real time to detect anomalies that deviate from the normal pattern.

[0027] A method for an intelligent early warning management system for an information center computer room includes the following steps:

[0028] (4) Data collection: sensors in the central computer room collect data on various environmental parameters and equipment operating status in real time;

[0029] (5) Data preprocessing: cleaning, denoising and completing environmental parameter data and equipment operation status data;

[0030] (6) Intelligent analysis: Use graph model algorithms to establish multi-parameter correlation analysis models to identify abnormal patterns and potential risks;

[0031] (4) Early warning decision-making: Based on the output results of intelligent analysis, early warning signals are automatically generated, and diagnostic reports and recommended measures are provided according to the warning level;

[0032] (5) Interactive response: Display warning information and real-time collected environmental parameter data and equipment operation data.

[0033] Furthermore, interactive responses can also provide historical data query and analysis functions.

[0034] Compared with the prior art, the present invention has the following beneficial effects:

[0035] (1) The present invention uses a graphical model algorithm to establish a multi-parameter correlation analysis model to identify abnormal patterns and potential risks, including abnormal information such as the slow rise in temperature in the central computer room, abnormal humidity fluctuations, and abnormal power load. It can capture early signs of data changes, provide early warnings, and reduce the occurrence of sudden failures.

[0036] (2) The present invention reduces the burden on management personnel and improves overall operation and maintenance efficiency by automating the early warning and diagnosis process. In addition, the machine learning model can be continuously optimized based on the actual operation data of the computer room to adapt to different computer room environments and management requirements. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] Figure 1This is a structural diagram of the intelligent early warning management system of the information center computer room of the present invention. DETAILED DESCRIPTION

[0038] The present invention is described in detail below in conjunction with specific embodiments. The following embodiments will help those skilled in the art to further understand the present invention, but are not intended to limit the present invention in any form. It should be noted that, for those of ordinary skill in the art, several changes and improvements can also be made without departing from the concept of the present invention. These all belong to the protection scope of the present invention.

[0039] like Figure 1 As shown, an intelligent early warning management system for an information center computer room is composed of a data acquisition module, a data preprocessing module, an intelligent analysis module, an early warning decision module and an interactive response interface. The data acquisition module collects a variety of environmental parameter data and equipment operation status data in real time through sensors in the central computer room. The data preprocessing module cleans, denoises and completes the environmental parameter data and equipment operation status data. The preprocessed environmental parameter data and equipment operation status data are used by the intelligent analysis module to apply a graph model algorithm to establish a multi-parameter correlation analysis model to identify abnormal patterns and potential risks. The early warning decision module automatically generates an early warning signal based on the output result of the intelligent analysis module, and provides a diagnostic report and recommended measures according to the early warning level. The interactive response interface displays the early warning information and the real-time collected environmental parameter data and equipment operation data.

[0040] The data acquisition module collects real-time data on various environmental parameters and equipment operating status, including temperature T, humidity W, air quality Q, power status S, equipment load T, and network traffic F, through a sensor network deployed at key locations in the computer room. To ensure the comprehensiveness and accuracy of the data, the early warning management system supports a variety of communication protocols and interface standards, and can seamlessly connect monitoring equipment of different types and manufacturers.

[0041] x(t)=ω T ·T(t)+ω W ·W(t)+ω Q Q(t)+ω s ·S(t)+ω T ·T(t)+ω F F(t),t∈{1,2,3...},

[0042] The collected raw data often contains quality problems such as noise and missing values, which will affect the accuracy of the results if used directly for analysis. Therefore, the data preprocessing module is responsible for cleaning, denoising, completing, and formatting the raw data x(t) to ensure the smooth progress of subsequent analysis. In addition, the module also has data compression and encryption functions to reduce storage and transmission costs while ensuring data security.

[0043] Data cleaning uses the LOF algorithm to identify outliers by calculating the local density deviation of each data point relative to other data points in its neighborhood. If the local density of a data point is significantly lower than other data points in its neighborhood, the point is considered an outlier.

[0044] Specific process:

[0045] (1) Parameter settings:

[0046] Determine the number of neighbors k (i.e. consider the k nearest neighbors around each data point).

[0047] Set the distance measurement method, such as Euclidean distance, Manhattan distance, etc.

[0048] (2) Calculate k-distance:

[0049] For each point p in the dataset, calculate its k-distance d k (p), that is, the distance from point p to its kth nearest neighbor.

[0050] (3) Determine k-distance neighbors:

[0051] For each point p, find all points whose distance to p is less than or equal to d. k (p), these points constitute the k-distance neighbor set N of p. k (p).

[0052] (4) Calculate the reachable distance:

[0053] For each k-distance neighbor o of point p, calculate the reachable distance rd k (p,o), according to the formula rd k (p,o)=max{d k (o),d(p,o)}, where d(p,o) is the actual distance between p and o.

[0054] (5) Calculate the local reachable density:

[0055] For each point p, according to p's k-distance neighbor set N k (p) and reachable distance, calculate the local reachable density lrd k (p), according to the formula

[0056] (6) Calculate the local outlier factor:

[0057] For each point p, according to the local reachability density lrd of p k (p) and its k-distance neighbors’ local reachability density, and calculate the local outlier factor LOF k (p),

[0058] According to the formula

[0059] (7) Abnormal identification:

[0060] According to LOF k The value of (p) identifies outliers. k Points with (p) values ​​significantly greater than 1 are considered outliers because their local density is significantly lower than the local density of their neighbors.

[0061] (8) Output results:

[0062] Output the LOF of each point k values ​​and mark outliers as needed.

[0063] Data denoising uses improved wavelet denoising. For given data x(t), wavelet transform is performed to obtain wavelet coefficients W(a,b), where a is the scale parameter and b is the translation parameter. In practical applications, discrete wavelet transform (DWT) or fast wavelet transform (FWT) algorithms are usually used.

[0064] For each scale j, the wavelet coefficient W j , apply threshold λ for processing:

[0065]

[0066] Use the processed wavelet coefficients Perform inverse wavelet transform to obtain the denoised signal

[0067] The processed data are then completed using a multiple interpolation method. The specific process is as follows:

[0068] (1) Identify missing data:

[0069] Identify missing data X.

[0070] (2) Establishing interpolation model:

[0071] Create an interpolation model for each variable containing missing values, this time using a linear regression model:

[0072] Y mis =β0+β1X1+β2X2+...+β p X p +ε,

[0073] where Ymis is the missing response variable, X1, X2, …, Xp are predictor variables, β0, β1, …, βp are regression coefficients, is the error term.

[0074] (3) Generate interpolation data set:

[0075] For each missing value, multiple (usually several to dozens) imputed values ​​are generated using the corresponding imputation model.

[0076] This is usually done by sampling from the posterior distribution of the imputation model, or by adding random noise to the predicted values ​​to model uncertainty.

[0077] This process is repeated to generate multiple sets of imputed values ​​for each missing value, creating multiple complete imputed data sets.

[0078] (4) Analysis of interpolation data set:

[0079] Perform the desired data analysis (such as statistical tests, regression analysis, classification, etc.) on each imputed data set, saving the results of each analysis for subsequent merging.

[0080] (5) Merger results:

[0081] Use Rubin's rule or other suitable method to merge the results from different imputed datasets.

[0082] For the point estimate θ, calculate the results θ for all imputed data sets j The average value of . That is:

[0083]

[0084] The intelligent analysis module is the core of the system. It uses machine learning algorithms to conduct deep mining and correlation analysis on pre-processed data to identify potential abnormal patterns and risks.

[0085] Different from the traditional single-parameter threshold judgment, the system of the present invention constructs a multi-parameter correlation analysis model. By analyzing the mutual influence and dependency between different parameters, the model can capture complex abnormal patterns that are difficult to detect with a single parameter threshold judgment. For example, when the room temperature gradually rises and the air conditioning cooling efficiency decreases, the system can combine the changing trends of these two parameters and predict the risk of air conditioning system failure in advance.

[0086] Select the graph-based association analysis model to establish a multi-parameter association analysis model. The establishment process is as follows:

[0087] (1) Graph Construction

[0088] Node construction:

[0089] Each data (such as temperature, humidity, air quality, power status, device load, network traffic) is considered a node in the graph. The node contains the current value, historical data, and threshold information of the data.

[0090] Side Build:

[0091] Edges represent the association between parameters. This association is based on the correlation coefficient of data relevance.

[0092] The weight of an edge can represent the strength of the association or the degree of influence.

[0093] Graph construction:

[0094] Use a graph data structure such as an adjacency matrix, adjacency list, or graph database to represent the entire system.

[0095] According to the relationship between parameters, a complete graph structure is constructed.

[0096] (2) Algorithm Design

[0097] Anomaly detection rules:

[0098] Define anomaly detection rules for each node. For example, a temperature node may have a rule that marks an anomaly when the temperature rises above a certain threshold every day for three consecutive days. Use threshold-based detection rules.

[0099] Graph traversal algorithm:

[0100] A graph algorithm called Depth First Search (DFS) is used to traverse the graph and calculate the anomaly score or risk level of the nodes.

[0101] Abnormal pattern recognition:

[0102] By analyzing the association paths and anomaly scores between nodes, complex anomaly patterns can be identified. For example, when temperature and humidity are abnormal at the same time and they are associated with the state changes of the air conditioning system, it can be inferred that the air conditioning system may be faulty.

[0103] Based on the risk assessment results, the early warning decision module divides abnormal events into different warning levels, namely urgent, important, and general, and generates corresponding early warning signals, which include sound alarms, pop-up prompts, and SMS notifications. For each early warning event, the early warning decision module generates a detailed diagnostic report, including the specific values ​​of the abnormal parameters, change trends, causes, and historical similar cases.

[0104] The intelligent analysis module also includes an anomaly detection algorithm that learns normal behavior patterns based on historical data and compares current data in real time to detect anomalies that deviate from normal patterns.

[0105] The interactive response interface is a bridge for managers to interact with the system. This interface ensures that managers can understand the status of the computer room anytime and anywhere and make effective interventions.

[0106] (1) Real-time data display

[0107] The interface displays the environmental parameters and equipment status data in the computer room in real time in the form of a dashboard, allowing managers to understand the overall operating status of the computer room at a glance.

[0108] (2) Alarm information management

[0109] The interface provides centralized management of alarm information, including alarm viewing, confirmation, and processing status tracking. Managers can quickly locate the alarm source, view alarm details, and take measures to solve the problem based on the system's suggestions.

[0110] (3) Historical data analysis

[0111] The interface supports historical data query and analysis functions, allowing managers to review the operation status of the computer room over a period of time and analyze potential patterns and root causes of problems.

[0112] A method for an intelligent early warning management system for an information center computer room includes the following steps:

[0113] (7) Data collection: sensors in the central computer room collect data on various environmental parameters and equipment operating status in real time;

[0114] (8) Data preprocessing: cleaning, denoising and completing environmental parameter data and equipment operation status data;

[0115] (9) Intelligent analysis: Use graph model algorithms to establish multi-parameter correlation analysis models to identify abnormal patterns and potential risks;

[0116] (4) Early warning decision-making: Based on the output results of intelligent analysis, early warning signals are automatically generated, and diagnostic reports and recommended measures are provided according to the warning level;

[0117] (5) Interactive response: Display warning information and real-time collected environmental parameter data and equipment operation data.

[0118] The data acquisition and preprocessing module collects various environmental parameters and equipment operation status data in the computer room in real time, and performs preliminary cleaning and formatting. The intelligent analysis module uses the graph model algorithm to establish a multi-parameter correlation analysis model to identify abnormal patterns and potential risks, including but not limited to slow temperature rise, abnormal humidity fluctuations, abnormal power load, etc. The early warning decision module automatically generates early warning signals based on the analysis results, and provides preliminary diagnostic reports and recommended measures according to the early warning level. The interactive response interface provides managers with intuitive alarm information and operation guidance, and supports rapid response and processing.

[0119] The above describes the specific embodiments of the present invention. It should be understood that the present invention is not limited to the above specific embodiments, and those skilled in the art can make various changes or modifications within the scope of the claims, which does not affect the essence of the present invention. In the absence of conflict, the embodiments of the present application and the features in the embodiments can be combined with each other arbitrarily.

Claims

1. An intelligent early warning management system for an information center computer room, characterized in that: It consists of a data acquisition module, a data preprocessing module, an intelligent analysis module, an early warning decision module and an interactive response interface. The data acquisition module collects a variety of environmental parameter data and equipment operation status data in real time through sensors in the central computer room. The data preprocessing module cleans, denoises and completes the environmental parameter data and equipment operation status data. The intelligent analysis module uses the graph model algorithm to establish a multi-parameter correlation analysis model through the preprocessed environmental parameter data and equipment operation status data to identify abnormal patterns and potential risks. The early warning decision module automatically generates an early warning signal based on the output results of the intelligent analysis module, and provides a diagnostic report and recommended measures according to the early warning level. The interactive response interface displays the early warning information and the real-time collected environmental parameter data and equipment operation data.

2. According to claim 1, the intelligent early warning management system for the information center computer room is characterized in that: The environmental parameter data include temperature, humidity, and air quality; the device operation status data include power status, device load, and network traffic.

3. According to claim 1, the intelligent early warning management system for the information center computer room is characterized in that: The data cleaning adopts the LOF algorithm to identify outliers by calculating the local density deviation of each data point relative to the data points in the neighborhood.

4. The intelligent early warning management method for an information center computer room according to claim 1 is characterized in that: The data denoising adopts improved wavelet denoising, and for the cleaned data, wavelet transform is performed to obtain wavelet coefficients, and the wavelet coefficients are processed with thresholds, and then the processed wavelet coefficients are used to perform inverse wavelet transform to obtain denoised signal data.

5. According to claim 1, the intelligent early warning management system for the information center computer room is characterized in that: The data completion adopts a multiple interpolation method, and the specific process is: identifying missing data, establishing an interpolation model, generating an interpolation data set, analyzing the interpolation data set, and merging the results.

6. The intelligent early warning management system for an information center computer room according to claim 1 is characterized in that: The process of establishing the multi-parameter association analysis model is as follows: (1) Graph Construction 1) Node construction: Each data is regarded as a node in the graph, namely temperature, humidity, air quality, power status, device load, network traffic, and the node contains the current value, historical data, threshold information, and status identification of the data. 2) While building: The edge represents the association relationship between data. The association relationship is the correlation coefficient based on the data relevance, and the edge weight represents the strength of the association. 3) Graph construction: Build a complete graph structure based on the relationship between data; (2) Algorithm Process 1) Anomaly detection rules: Defining anomaly detection rules for each of the nodes, and selecting detection rules based on thresholds; 2) Graph traversal algorithm A graph traversal algorithm is used to detect anomalous patterns across parameters and a graph algorithm using depth-first search (DFS) is used to traverse the graph and calculate the anomaly score or risk level of the nodes. 3) Abnormal pattern recognition: By analyzing the association paths and anomaly scores between nodes, complex anomaly patterns are identified.

7. The intelligent early warning management system for an information center computer room according to claim 1 is characterized in that: Based on the risk assessment results, the early warning decision module divides abnormal events into different warning levels, namely urgent, important, and general, and generates corresponding early warning signals, which include sound alarms, pop-up prompts, and SMS notifications. For each early warning event, the early warning decision module generates a detailed diagnostic report, including the specific values ​​of the abnormal parameters, change trends, causes, and historical similar cases.

8. The intelligent early warning management method for an information center computer room according to claim 1 is characterized in that: The intelligent analysis module also includes an anomaly detection algorithm that learns normal behavior patterns based on historical data and compares current data in real time to detect anomalies that deviate from normal patterns.

9. A method for an intelligent early warning management system for an information center computer room comprises the following steps: (1) Data collection: sensors in the central computer room collect data on various environmental parameters and equipment operating status in real time; (2) Data preprocessing: cleaning, denoising, and completing environmental parameter data and equipment operation status data; (3) Intelligent analysis: Use graph model algorithms to establish multi-parameter correlation analysis models to identify abnormal patterns and potential risks; (4) Early warning decision-making: Based on the output results of intelligent analysis, early warning signals are automatically generated, and diagnostic reports and recommended measures are provided according to the warning level; (5) Interactive response: Display warning information and real-time collected environmental parameter data and equipment operation data.

10. The intelligent early warning management method for an information center computer room according to claim 9, characterized in that: Interactive responses can also provide historical data query and analysis capabilities.

Images