Machine room operation and maintenance management system based on big data

Abstract

The invention discloses a machine room operation and maintenance management system based on big data, and relates to the technical field of operation and maintenance management, and the system comprises a plurality of data collection modules which are respectively matched with a set of machine room equipment, and each data collection module comprises a transmission unit and a comparison unit; the transmission unit collects real-time data of the corresponding machine room; a comparison model is arranged in the comparison unit, after the real-time data are received, corresponding comparison model data are collected, and the comparison model data and the real-time data are synchronously transmitted; the analysis module is used for receiving the real-time data and the comparison model data and cleaning the real-time data and the comparison model data; and the relationship network module is used for generating a relationship network for each machine room device according to the relationship between the workflow and the region. Real-time data of the data acquisition module and comparison model data are synchronously compared, the analysis module can quickly and accurately detect the abnormal state of the equipment, the hysteresis of traditional manual monitoring is reduced, and meanwhile, the states of different equipment can be analyzed and monitored in a multi-dimensional mode by means of the process relation module and the regional relation module.

Description

Technical Field

[0001] The present invention relates to the technical field of operation and maintenance management, and in particular to a computer room operation and maintenance management system based on big data. Background Art

[0002] With the development of data-driven technology, the operation and maintenance management of computer room equipment has gradually shifted towards intelligence and automation. Traditional computer room operation and maintenance management methods usually rely on manual inspections and static threshold judgments, which makes it difficult to detect equipment failures in a timely and accurate manner. With the development of the Internet of Things and big data technologies, real-time monitoring and intelligent analysis methods have provided new solutions for equipment operation and maintenance. How to achieve comprehensive monitoring of equipment status through real-time data collection and intelligent analysis has become an important research direction of current operation and maintenance management systems.

[0003] After searching, a Chinese patent (publication number: CN117575559A) discloses an equipment operation and maintenance management system, which includes: an operation and maintenance personnel database, which is used to record and update the operation and maintenance personnel information; a trigger module, which is used to parse the user-side information to obtain semantic information and match the operation and maintenance commands corresponding to the semantic information, and trigger new work tasks according to the operation and maintenance commands; a work order generation module, which is used to generate corresponding work orders according to the work tasks; an allocation module, which is used to allocate the work orders to appropriate operation and maintenance personnel according to the work tasks and the skill information and work arrangement information of the operation and maintenance personnel in the operation and maintenance personnel database; a recording module, which is used to record the execution process of the work tasks; a confirmation module, which is used to confirm the completion of the work tasks; an evaluation module, which is used to evaluate the completion effect of the work tasks and generate an operation and maintenance score; and an end module, which is used to end the work tasks.

[0004] In the prior art, if manual inspections and regular maintenance are used, it is easy to cause fault detection delays, making it difficult to detect equipment abnormalities in a timely manner, affecting the stability and security of the system. In addition, equipment status monitoring is mostly based on static threshold judgments and cannot flexibly respond to dynamic environmental changes, which can easily cause false alarms or missed alarms, thereby reducing the accuracy of fault detection. Therefore, the present invention proposes a computer room operation and maintenance management system based on big data. Summary of the invention

[0005] The purpose of the present invention is to provide a computer room operation and maintenance management system based on big data to solve the problems mentioned in the above background technology.

[0006] The present invention can be implemented through the following technical solutions: a computer room operation and maintenance management system based on big data, including a data acquisition module, an analysis module, a relationship network module and an operation and maintenance platform;

[0007] The plurality of data acquisition modules are respectively matched with a group of equipment in the computer room, which includes:

[0008] The transmission unit is used to collect the real-time data of the corresponding computer room;

[0009] A comparison unit has a built-in comparison model, and after receiving real-time data, retrieves corresponding comparison model data and transmits it synchronously with the real-time data;

[0010] The analysis module is used to receive real-time data and comparison model data, and clean the real-time data after receiving it;

[0011] The relationship network module generates a relationship network for each computer room device according to the corresponding relationship, including a workflow relationship unit and a regional relationship unit;

[0012] The process relationship unit locates the working order and data flow between each node of each computer room equipment according to the corresponding work process, and generates a directed graph of each computer room equipment according to the working order node;

[0013] The regional relationship unit divides the computer room into different areas based on the physical layout of the computer room and the actual location of the equipment in the computer room, and allocates each computer room equipment to a corresponding location node according to the location, and the regional relationship unit generates a regional relationship model representing the spatial distribution based on the location of each computer room equipment;

[0014] The process relationship unit processes the real-time data and the comparison model data in each work process in the directed graph to obtain a process data set;

[0015] The regional relationship unit processes the real-time data and the comparison model data in each region to obtain a regional data set;

[0016] The analysis module receives the process data set and the regional data set, and aggregates and performs multi-dimensional analysis on the process data set and the regional data set to obtain analysis results;

[0017] The operation and maintenance platform includes a prediction unit and a decision support unit;

[0018] The prediction unit predicts the working data and data change trend of each equipment in the computer room based on the analysis result to obtain the prediction result;

[0019] The decision support unit provides decision suggestions to operation and maintenance personnel based on the prediction results and the recommendation algorithm based on big data, so as to facilitate the dynamic adjustment of maintenance strategies and improve management efficiency.

[0020] A further technical improvement of the present invention is that the analysis module cleans the real-time data including deduplication of duplicate data, data type conversion and format correction, standardization, outlier detection and processing, missing data processing and verification of cleaned data.

[0021] A further technical improvement of the present invention is that when processing missing data, the missing data is obtained by calculating the difference between the historical data of the missing data and the corresponding historical data and the comparison model data based on the corresponding data in the comparison model data. The formula used is:

[0022]

[0023] Among them, X m For missing data at time point T m The value of , which is the value used for missing data;

[0024] To compare the model data at time point T m The value of is used as the benchmark value to fill in the missing data;

[0025] X i is the historical data at time point T i The value of

[0026] α is the decay factor, which is used to control the rate of time decay, that is, the greater the time from missing data, the smaller the weight;

[0027] T m is the time point at which the data is missing, T i The time points of historical data used to fill in missing data;

[0028] |T i -T m | is the time point T i With T m The time difference between them, that is, the distance between the time point of historical data and the time point of missing data;

[0029] X i is the historical data at time point T i The value of

[0030] is the weighting coefficient, ensuring that the missing data time point T m The more recent the historical data, the greater the impact on the filling results.

[0031] A further technical improvement of the present invention is that: in a directed graph, the input data of the initial electromechanical device in each workflow is used as a matching point to match the corresponding working model, and the electromechanical device corresponding to each node in the workflow is matched with the corresponding part in the working model;

[0032] After matching, the comparison model of the comparison unit in each data acquisition module is established, and the comparison model data of the comparison model is used for subsequent synchronous transmission with the corresponding real-time data.

[0033] A further technical improvement of the present invention is that the analysis method of the process data set by the analysis module comprises the following steps:

[0034] A1. Take each node of the workflow as the axis and the comparison model data corresponding to each node as the base point to establish a comparison curve;

[0035] A2. Compare the real-time data matching values ​​corresponding to each node with the base point corresponding to the curve, and use the formula: Obtain the total difference value Dt of the comparison model data of each computer room equipment and the corresponding working model in the corresponding workflow;

[0036] Among them, X i is the real-time data of the i-th node, M i is the comparison model data of the ith node, i.e., the corresponding base point on the comparison curve;

[0037] N is the total number of nodes in the workflow;

[0038] A3, comparing the total difference value Dt with a preset difference threshold value T;

[0039] a1. If the total difference value Dt is not greater than the difference threshold T, the fault degree of each equipment room in the workflow is marked as level one;

[0040] a2. If the total difference value Dt is greater than the difference threshold T, the failure degree of each equipment room in the workflow is marked as level 2.

[0041] The further technical improvement of the present invention is: on the basis of a2, identifying the distribution state of each real-time data in the comparison curve;

[0042] b1. If the difference between individual real-time data and the corresponding comparison model data is in an irregular discrete state, the equipment in the computer room corresponding to the irregular discrete real-time data will be marked;

[0043] b2. If the differences between the multiple real-time data and the corresponding comparison model data are in a continuous state, each real-time data with a continuous difference is identified to obtain an identification data segment;

[0044] The marking is based on the equipment in the computer room corresponding to the starting point of the identified data segment.

[0045] A further technical improvement of the present invention is that the analysis method of the analysis module for the regional data set includes:

[0046] Y1, difference detection;

[0047] The preset comparison threshold G is used to determine whether the real-time data of the nodes in the corresponding area exceeds the reasonable range of the comparison model data;

[0048] The real-time data of each node in the corresponding area is compared with the comparison threshold G. If it exceeds the comparison threshold G, it is marked as an abnormal node, and each abnormal node is recorded as a set S;

[0049] Y2, state pattern recognition;

[0050] Perform the following operations on each abnormal node in the set S:

[0051] u1. Retrieve similar records of the abnormal node from the big data storage, and use a similarity matching algorithm to compare the real-time data of the abnormal node with similar data in the big data storage, and determine whether the abnormal state of the abnormal node exists in the same data in the big data storage;

[0052] u2. When the same data exists, the abnormal state of the abnormal node is matched with the result of storing the data in the big data;

[0053] Y3. Generate analysis results;

[0054] For each abnormal node, a corresponding analysis report is generated, and the fault level of the equipment in the corresponding node is marked as level 2. At the same time, the analysis reports of all abnormal nodes in the region are summarized to generate a regional analysis report;

[0055] The regional analysis report includes the number of abnormal nodes in the region and their specific locations, the status classification of each node, and summarizes the overall health status of the region;

[0056] If the analysis reports of all abnormal nodes in the area have the same type of abnormal status, the area is marked as an abnormal area.

[0057] Compared with the prior art, the present invention has the following beneficial effects:

[0058] The present invention synchronously compares the real-time data of the data acquisition module with the comparison model data, and the analysis module can quickly and accurately detect the abnormal state of the equipment, reducing the lag of traditional manual monitoring. At the same time, the process relationship and regional relationship modules can be used to perform multi-dimensional analysis and monitoring of the status of different equipment, adapt to complex operation and maintenance environments, and meet the needs of associated monitoring between different equipment.

[0059] Moreover, through the accumulation of historical data and continuous optimization of comparison models, it is not only possible to detect equipment failures, but also to predict trends, provide decision-making support for operation and maintenance personnel, and take preventive maintenance measures in advance. BRIEF DESCRIPTION OF THE DRAWINGS

[0060] In order to facilitate understanding by those skilled in the art, the present invention is further described below with reference to the accompanying drawings.

[0061] Figure 1 It is a system block diagram of the present invention. DETAILED DESCRIPTION

[0062] In order to further explain the technical means and effects adopted by the present invention to achieve the predetermined invention purpose, the specific implementation methods, structures, features and effects of the present invention are described in detail below in conjunction with the accompanying drawings and preferred embodiments.

[0063] See also Figure 1 As shown, the present invention provides a computer room operation and maintenance management system based on big data, including a data acquisition module, an analysis module, a relationship network module and an operation and maintenance platform;

[0064] Multiple data acquisition modules are matched with a group of equipment in the computer room, including:

[0065] The transmission unit is used to collect the real-time data of the corresponding computer room;

[0066] A comparison unit has a built-in comparison model, and after receiving real-time data, retrieves corresponding comparison model data and transmits it synchronously with the real-time data;

[0067] The analysis module is used to receive real-time data and compare model data, and clean the real-time data after receiving it;

[0068] The analysis module cleans real-time data, including deduplication, data type conversion and format correction, standardization, outlier detection and processing, missing data processing, and post-cleaning data verification;

[0069] When processing missing data, the missing data is obtained by calculating the difference between the historical data of the missing data and the corresponding historical data and the comparison model data based on the corresponding data in the comparison model data. The formula used is:

[0070]

[0071] Among them, X m For missing data at time point T m The value of , which is the value used for missing data;

[0072] To compare the model data at time point T m The value of is used as the benchmark value to fill in the missing data;

[0073] X i is the historical data at time point T i The value of

[0074] α is the decay factor, which is used to control the rate of time decay, that is, the greater the time from missing data, the smaller the weight;

[0075] T m is the time point at which the data is missing, T i The time points of historical data used to fill in missing data;

[0076] |T i -T m | is the time point T i With T m The time difference between them, that is, the distance between the time point of historical data and the time point of missing data;

[0077] X i is the historical data at time point T i The value of

[0078] is the weighting coefficient, ensuring that the missing data time point T m The more recent the historical data, the greater the impact on the filling results;

[0079] The relationship network module generates a relationship network for each computer room device according to the corresponding relationship, including process relationship units and regional relationship units;

[0080] The process relationship unit locates the working order and data flow between each node according to the corresponding work process of each computer room equipment, and generates a directed graph according to the working order node of each computer room equipment;

[0081] In the directed graph, the input data of the initial electromechanical equipment in each workflow is used as the matching point to match the corresponding working model, and the electromechanical equipment corresponding to each node in the workflow is matched with the corresponding part in the working model;

[0082] After matching, the comparison model of the comparison unit in each data acquisition module is established, and the comparison model data of the comparison model is used for subsequent synchronous transmission with the corresponding real-time data;

[0083] The regional relationship unit divides the computer room into different areas based on the physical layout of the computer room and the actual location of the equipment in the computer room, and allocates each computer room equipment to a corresponding location node according to the location, and the regional relationship unit generates a regional relationship model representing the spatial distribution based on the location of each computer room equipment;

[0084] The process relationship unit processes the real-time data and the comparison model data in each workflow in the directed graph to obtain a process data set;

[0085] The regional relationship unit processes the real-time data and the comparison model data in each region to obtain a regional data set;

[0086] The analysis module receives the process data set and the regional data set, aggregates and performs multi-dimensional analysis on the process data set and the regional data set to obtain analysis results;

[0087] The analysis module analyzes the process data set by the following steps:

[0088] A1. Take each node of the workflow as the axis and the comparison model data corresponding to each node as the base point to establish a comparison curve;

[0089] A2. Compare the real-time data matching values ​​corresponding to each node with the base point corresponding to the curve, and use the formula: Obtain the total difference value Dt of the comparison model data of each computer room equipment and the corresponding working model in the corresponding workflow;

[0090] Among them, X i is the real-time data of the i-th node, M i is the comparison model data of the ith node, i.e., the corresponding base point on the comparison curve;

[0091] N is the total number of nodes in the workflow;

[0092] A3, comparing the total difference value Dt with a preset difference threshold value T;

[0093] a1. If the total difference value Dt is not greater than the difference threshold T, the fault degree of each equipment room in the workflow is marked as level one;

[0094] a2. If the total difference value Dt is greater than the difference threshold T, the failure degree of each equipment room in the workflow is marked as level 2.

[0095] Based on a2, identify the distribution status of each real-time data in the comparison curve;

[0096] b1. If the difference between individual real-time data and the corresponding comparison model data is in an irregular discrete state, the equipment in the computer room corresponding to the irregular discrete real-time data will be marked;

[0097] b2. If the differences between the multiple real-time data and the corresponding comparison model data are in a continuous state, each real-time data with a continuous difference is identified to obtain an identification data segment;

[0098] Marking is based on identifying the equipment in the computer room corresponding to the starting point of the data segment;

[0099] In this embodiment, the marking method based on identifying data segments includes:

[0100] z1, preset fluctuation threshold Y;

[0101] z2. Calculate the difference value V between each node and the fluctuation threshold Y;

[0102] z3. Detect continuous difference segments;

[0103] Traverse all nodes of the comparison curve to find a continuous node set, so that the difference value V of each node in the node set is greater than the fluctuation threshold Y, that is, convert the corresponding node set into an identification data segment;

[0104] z4. In the identification data segment, the first node is selected as the starting point, and the node is assumed to be a continuous difference segment from node i to node j. Then the first node i is taken as the starting point of the identification data segment, and the equipment in the computer room corresponding to node i and its first s nodes is marked;

[0105] The analysis module analyzes the process data set by the following steps:

[0106] A1. Take each node of the workflow as the axis and the comparison model data corresponding to each node as the base point to establish a comparison curve;

[0107] A2. Compare the real-time data matching values ​​corresponding to each node with the base point corresponding to the curve, and use the formula: Obtain the total difference value Dt of the comparison model data of each computer room equipment and the corresponding working model in the corresponding workflow;

[0108] Among them, X i is the real-time data of the i-th node, M i is the comparison model data of the ith node, i.e., the corresponding base point on the comparison curve;

[0109] N is the total number of nodes in the workflow;

[0110] A3, comparing the total difference value Dt with a preset difference threshold value T;

[0111] a1. If the total difference value Dt is not greater than the difference threshold T, the fault degree of each equipment room in the workflow is marked as level one;

[0112] a2. If the total difference value Dt is greater than the difference threshold T, the failure degree of each equipment room in the workflow is marked as level 2.

[0113] When the corresponding equipment in the computer room is marked as level 2, it means that the corresponding equipment in the computer room has a fault and needs to be processed;

[0114] Based on a2, identify the distribution status of each real-time data in the comparison curve;

[0115] b1. If the difference between individual real-time data and the corresponding comparison model data is in an irregular discrete state, the equipment in the computer room corresponding to the irregular discrete real-time data will be marked;

[0116] b2. If the differences between the multiple real-time data and the corresponding comparison model data are in a continuous state, each real-time data with a continuous difference is identified to obtain an identification data segment;

[0117] Marking is based on identifying the equipment in the computer room corresponding to the starting point of the data segment;

[0118] In this embodiment, the marking method based on identifying data segments includes:

[0119] z1, preset fluctuation threshold Y;

[0120] z2. Calculate the difference value V between each node and the fluctuation threshold Y;

[0121] z3. Detect continuous difference segments;

[0122] Traverse all nodes of the comparison curve to find a continuous node set, so that the difference value V of each node in the node set is greater than the fluctuation threshold Y, that is, convert the corresponding node set into an identification data segment;

[0123] z4. In the identification data segment, the first node is selected as the starting point, and the node is assumed to be a continuous difference segment from node i to node j. Then the first node i is taken as the starting point of the identification data segment, and the equipment in the computer room corresponding to node i and its first s nodes is marked;

[0124] The analysis module analyzes the regional data set, including:

[0125] Y1, difference detection;

[0126] The preset comparison threshold G is used to determine whether the real-time data of the nodes in the corresponding area exceeds the reasonable range of the comparison model data;

[0127] The real-time data of each node in the corresponding area is compared with the comparison threshold G. If it exceeds the comparison threshold G, it is marked as an abnormal node, and each abnormal node is recorded as a set S;

[0128] Y2, state pattern recognition;

[0129] Perform the following operations on each abnormal node in the set S:

[0130] u1. Retrieve similar records of the abnormal node from the big data storage, and use a similarity matching algorithm to compare the real-time data of the abnormal node with similar data in the big data storage, and determine whether the abnormal state of the abnormal node exists in the same data in the big data storage;

[0131] u2. When the same data exists, the abnormal state of the abnormal node is matched with the result of storing the data in the big data;

[0132] Y3. Generate analysis results;

[0133] For each abnormal node, a corresponding analysis report is generated, and the fault level of the equipment in the corresponding node is marked as level 2. At the same time, the analysis reports of all abnormal nodes in the region are summarized to generate a regional analysis report;

[0134] The regional analysis report includes the number of abnormal nodes in the region and their specific locations, the status classification of each node, and the summary of the overall health status of the region, and provides regional-level operation and maintenance suggestions, including "immediate maintenance recommended" and "continuous monitoring recommended";

[0135] If the analysis reports of all abnormal nodes in the area have the same type of abnormal status, the area is marked as an abnormal area;

[0136] In the analysis results of the analysis module on the process data set and the regional data set, if all the equipment in a certain computer room is marked as level 2, the detection priority of the equipment in the computer room is increased, and the system recommends that the equipment in the computer room be detected first;

[0137] In this embodiment, a certain area includes 5 nodes, whose real-time data P = {15, 18, 30, 12, 22}, the comparison model data L = {10, 18, 25, 14, 20}, and the comparison threshold G = 5;

[0138] Calculate the difference value B between each node and the comparison threshold G:

[0139] Node 1: B1=|15-10|=5;

[0140] Node 2: B2 = |18-18 | = 0;

[0141] Node 3: B3=|30-25|=5;

[0142] Node 4: B4=|12-14|=2;

[0143] Node 5: B5=|22-20|=2;

[0144] Among them, the difference value B between node 1 and node 3 is equal to 5, which is marked as an abnormal node;

[0145] Query the historical status data of nodes 1 and 3, and compare the current status with the historical records through the pattern matching algorithm. It is found that the situation of node 1 is similar to the "overload" mode in the history, and the situation of node 3 is similar to the "temperature fluctuation" mode;

[0146] Generate analysis report, node 1: marked as "overload" mode, status classified as "potential failure".

[0147] Node 3: marked as "temperature fluctuation" mode, and the status is classified as "normal fluctuation";

[0148] Summarize the number of abnormal nodes in the area and their health status, generate an overall analysis report and provide decision support, and recommend that the operation and maintenance personnel conduct further inspections on Node 1;

[0149] The operation and maintenance platform includes a prediction unit and a decision support unit;

[0150] Based on the analysis results, the prediction unit predicts the working data and data change trends of each equipment room to obtain the prediction results;

[0151] Based on the prediction results and recommendation algorithms based on big data, the decision support unit provides decision-making advice to operation and maintenance personnel to facilitate dynamic adjustment of maintenance strategies and improve management efficiency;

[0152] When using the recommendation algorithm, corresponding decision opinions are generated based on the processing plan of the corresponding historical data in the big data, thereby assisting operation and maintenance personnel to process quickly.

[0153] The above description is only a preferred embodiment of the present invention and does not limit the present invention in any form. Although the present invention has been disclosed as a preferred embodiment as above, it is not used to limit the present invention. Any technical personnel in this field can make some changes or modify the technical contents disclosed above into equivalent embodiments without departing from the scope of the technical solution of the present invention. However, any simple modification, equivalent change and modification made to the above embodiments according to the technical essence of the present invention without departing from the content of the technical solution of the present invention still fall within the scope of the technical solution of the present invention.

Claims

1. A computer room operation and maintenance management system based on big data, characterized in that: include: A data acquisition module, wherein a plurality of the data acquisition modules are respectively matched with a group of equipment in the computer room, and each of the modules comprises a transmission unit and a comparison unit; The transmission unit collects the real-time data of the corresponding computer room; The comparison unit has a built-in comparison model. After receiving real-time data, it retrieves the corresponding comparison model data and transmits it synchronously with the real-time data. The analysis module receives real-time data and comparison model data and cleans them; The relationship network module generates a relationship network for each equipment room according to the relationship between work processes and regions, including process relationship units and regional relationship units; The process relationship unit locates the working order and data flow between each node of each computer room equipment according to the corresponding work process, and generates a directed graph according to the working order node of each computer room equipment. At the same time, the process relationship unit processes the real-time data and comparison model data of each computer room equipment to obtain a process data set; The regional relationship unit divides the computer room into different areas based on the physical layout of the computer room and the actual location of the equipment in the computer room, and allocates each computer room equipment to a corresponding location node according to the location to generate a regional relationship model. At the same time, the regional relationship unit processes the real-time data and comparison model data of the computer room equipment in each area to obtain a regional data set; The analysis module receives the process data set and the regional data set, and analyzes the process data set and the regional data set to obtain analysis results; Operation and maintenance platform, including prediction unit and decision support unit; The prediction unit predicts the working data and data change trend of each equipment in the computer room based on the analysis result to obtain the prediction result; The decision support unit provides decision suggestions to the operation and maintenance personnel based on the prediction results and the recommendation algorithm based on big data.

2. According to the big data-based computer room operation and maintenance management system of claim 1, it is characterized in that: The analysis module cleans real-time data by removing duplicate data, converting data types and correcting formats, standardizing, detecting and processing outliers, processing missing data, and verifying cleaned data.

3. The computer room operation and maintenance management system based on big data according to claim 2 is characterized in that: When processing missing data, it is obtained by calculating the difference between the historical data of the missing data and the corresponding historical data and the comparison model data based on the corresponding data in the comparison model data.

4. The computer room operation and maintenance management system based on big data according to claim 3 is characterized in that: The formula used to handle missing data is: Among them, X m For missing data at time point T m The value of To compare the model data at time point T m The value of is used as the benchmark value to fill in the missing data; X i is the historical data at time point T i The value of α is the decay factor, which is used to control the rate of time decay; T m is the time point at which the data is missing, T i The time points of historical data used to fill in missing data; |T i -T m | is the time point T i With T m The time difference between X i is the historical data at time point T i The value of is the weighting coefficient.

5. The computer room operation and maintenance management system based on big data according to claim 1 is characterized in that: In the directed graph, the input data of the initial electromechanical equipment in each workflow is used as the matching point to match the corresponding working model, and the electromechanical equipment corresponding to each node in the workflow is matched with the corresponding part in the working model; After matching, the comparison model of the comparison unit in each data acquisition module is established, and the comparison model data of the comparison model is used for subsequent synchronous transmission with the corresponding real-time data.

6. The computer room operation and maintenance management system based on big data according to claim 5 is characterized in that: The analysis module's method for analyzing the process data set includes the following steps: A1. Take each node of the workflow as the axis and the comparison model data corresponding to each node as the base point to establish a comparison curve; A2. Compare the real-time data matching values ​​corresponding to each node with the base point corresponding to the curve, and use the formula: Obtain the total difference value Dt of the comparison model data of each computer room equipment and the corresponding working model in the corresponding workflow; Among them, X i is the real-time data of the i-th node, M i is the comparison model data of the ith node, i.e., the corresponding base point on the comparison curve; N is the total number of nodes in the workflow; A3, comparing the total difference value Dt with a preset difference threshold value T; a1. If the total difference value Dt is not greater than the difference threshold T, the fault degree of each equipment room in the workflow is marked as level one; a2. If the total difference value Dt is greater than the difference threshold T, the failure degree of each equipment room in the workflow is marked as level 2.

7. The computer room operation and maintenance management system based on big data according to claim 6 is characterized in that: Based on a2, identify the distribution status of each real-time data in the comparison curve; b1. If the difference between individual real-time data and the corresponding comparison model data is in an irregular discrete state, the equipment in the computer room corresponding to the irregular discrete real-time data will be marked; b2. If the differences between the multiple real-time data and the corresponding comparison model data are in a continuous state, each real-time data with a continuous difference is identified to obtain an identification data segment; The marking is based on the equipment in the computer room corresponding to the starting point of the identified data segment.

8. The computer room operation and maintenance management system based on big data according to claim 1 is characterized in that: The analysis module analyzes the regional data set, including: Y1, difference detection; The preset comparison threshold G is used to determine whether the real-time data of the nodes in the corresponding area exceeds the reasonable range of the comparison model data; The real-time data of each node in the corresponding area is compared with the comparison threshold G. If it exceeds the comparison threshold G, it is marked as an abnormal node, and each abnormal node is recorded as a set S; Y2, state pattern recognition; Perform the following operations on each abnormal node in the set S: u1. Retrieve similar records of the abnormal node from the big data storage, and use a similarity matching algorithm to compare the real-time data of the abnormal node with similar data in the big data storage, and determine whether the abnormal state of the abnormal node exists in the same data in the big data storage; u2. When the same data exists, the abnormal state of the abnormal node is matched with the result of storing the data in the big data; Y3. Generate analysis results; For each abnormal node, a corresponding analysis report is generated, and the fault level of the equipment in the corresponding node is marked as level 2. At the same time, the analysis reports of all abnormal nodes in the region are summarized to generate a regional analysis report; If the analysis reports of all abnormal nodes in the area have the same type of abnormal status, the area is marked as an abnormal area.

Images