Machine room operation and maintenance risk identification monitoring method

Abstract

The invention relates to the technical field of machine room operation and maintenance management, in particular to a machine room operation and maintenance operation risk identification monitoring method. According to the invention, through generation and distribution of the electronic work ticket and combination of OCR and two-dimensional code scanning technologies, precise authentication of the identity of the machine room operation and maintenance personnel is realized. According to the method, only authorized personnel can perform operation, each operation can be traced, and the compliance and safety of equipment operation are guaranteed. Through operation instruction monitoring and authority management, the risk caused by unauthorized operation or wrong operation is further reduced; according to the invention, by monitoring the voltage, temperature, humidity and other environmental data of the machine room equipment in real time and combining the long and short-term memory network model to analyze and predict the data in real time, potential risks can be evaluated, and abnormal conditions can be found in time. The mechanism is helpful to predict and identify possible equipment faults or environmental problems in advance, so that emergency response measures are taken, and risk expansion and loss are prevented.

Description

Technical Field

[0001] The present invention relates to the technical field of computer room operation and maintenance management, and specifically to a method for identifying and monitoring risks in computer room operation and maintenance operations. Background Art

[0002] With the rapid development of information technology and the sharp increase in data volume, computer rooms, as the core facilities for data storage and processing, have received increasing attention. The security and stability of the equipment in the computer room are crucial for ensuring the continuity of network communication, data storage, and computing services. Traditional computer room operation and maintenance management mainly rely on manual inspections and manual records. Although certain security can be ensured, there are many limitations, such as human operation errors, irregular task arrangements, and untimely emergency responses to sudden events.

[0003] In recent years, with the application of intelligent technologies, many computer room operation and maintenance management tasks have begun to introduce automated and intelligent means. For example, using sensors and monitoring devices to monitor the computer room environment in real time, applying data analysis technologies to identify potential risks, and ensuring the compliance of personnel operations through identity recognition technologies. Nevertheless, there are still some problems in the current technologies, especially in aspects such as personnel identity verification, operation permission management, risk warning and response, lacking effective technical means for comprehensive integration and automated management.

[0004] Moreover, most current computer room operation and maintenance management rely on manual identity verification, or only verify personnel identities through simple access cards, passwords, etc. These methods are easily affected by human factors and are at risk of being misused, lost, or operated incorrectly. In addition, many systems do not combine permission management with the actual details of task operations, resulting in some personnel performing high-risk operations beyond the authorized scope, increasing the security risks of the system.

[0005] In view of the above problems, it is necessary to propose a method for identifying and monitoring risks in computer room operation and maintenance operations. Summary of the Invention

[0006] The purpose of the present invention is to solve the problems existing in the background art and propose a method for identifying and monitoring risks in computer room operation and maintenance operations.

[0007] The purpose of the present invention can be achieved through the following technical solutions:

[0008] A method for identifying and monitoring risks in computer room operation and maintenance operations includes the following steps:

[0009] Step 1: Generation of work tickets;

[0010] Access the operation and maintenance management work order database of the computer room, obtain all the operation and maintenance task arrangement data therein, and extract features from the obtained task arrangement data based on natural language processing technology. Through word segmentation, named entity recognition, and relation extraction, obtain data fields such as the computer room number, equipment number, start time of operation and maintenance operations, end time of operation and maintenance operations, operation task type, as well as the name, personnel identity, and personnel number of the operation and maintenance personnel corresponding to each operation and maintenance task arrangement.

[0011] Among them, the personnel identities include: on-site operators, equipment engineers, safety officers, system administrators, network engineers, auditors, external suppliers, and emergency response personnel.

[0012] Among them, the operation task types include: any one or a combination of multiple items such as equipment failure repair, system maintenance, software update, hardware replacement, safety inspection, data backup, network debugging, and power supply inspection.

[0013] As a preferred embodiment of the present invention, generate a unique corresponding electronic work ticket for each individual operation and maintenance personnel. The electronic work ticket includes a title part, a verification QR code part, and a text part. The title part therein includes the words "electronic work ticket" and the unique number of the electronic work ticket; the text part therein is obtained by automatically filling all the feature data about the operation and maintenance personnel extracted into a preset template in the form of data fields; the QR code part therein is obtained by inputting all the feature data about the operation and maintenance personnel extracted into a QR code generator for conversion and encoding;

[0014] Distribute all the electronic work tickets to the mobile communication devices of each operation and maintenance personnel according to their corresponding operation and maintenance personnel numbers.

[0015] Make a completely identical backup work ticket for each electronic work ticket, and encrypt and store all the backup work tickets. Encrypt the feature data fields in the text part of all the backup work tickets through the SHA-256 hash encryption algorithm, and encrypt the QR code images in all the electronic work tickets through the AES symmetric encryption algorithm. And save all the encrypted backup work tickets to the database.

[0016] Step 2: Personnel identity recognition;

[0017] Authenticate the identity of the personnel entering the computer room through a personnel identity recognition program based on the recognition of the electronic work ticket.

[0018] Collect real-time images of the computer room through a camera, and perform motion tracking and image recognition analysis. Whenever it is detected that a person enters the computer room, output a preset voice asking the entering person to show the electronic work ticket to the camera.

[0019] Subsequently, if the camera recognizes the electronic work ticket image, based on the OCR optical character recognition technology, the electronic work ticket information, including the personnel number, personnel name, personnel identity, operation time, operation task, and equipment number involved, is obtained from the two-dimensional code in the electronic work ticket through the two-dimensional code reader embedded in the camera.

[0020] If the camera fails to recognize the electronic work ticket image, it is determined that the identity authentication fails, and an illegal personnel warning signal is output.

[0021] After obtaining the electronic work ticket information of the entering personnel through the camera, the work ticket information is compared with the backup work ticket in the database, and its authenticity and timeliness are authenticated successively. The specific process is as follows:

[0022] Authenticity authentication: The extracted electronic work ticket information is converted into structured data, and the personnel number, personnel name, and personnel identity in it are extracted for subsequent comparison processing. Connect to the database through the API interface and obtain the encrypted backup work ticket with the same unique number as the electronic work ticket from it. Decrypt the two-dimensional code image in the encrypted backup work ticket through the AES symmetric encryption algorithm, and compare the personnel number, personnel name, and personnel identity obtained after decryption with the personnel number, personnel name, and personnel identity in the electronic work ticket extracted by the camera.

[0023] If the comparison result is consistent, it is determined that the authenticity authentication is passed. After passing the authenticity authentication, the timeliness authentication of the electronic work ticket is carried out.

[0024] Timeliness authentication: When it is determined that the identity authentication is successful, the operation time is checked. Compare the operation time in the two-dimensional code of the electronic work ticket with the task arrangement time obtained after decryption in the database. If the comparison result is consistent, it is further determined whether the current time is within the operation time that conforms to the plan. If so, it is determined that the timeliness authentication is passed.

[0025] If the authenticity authentication and timeliness authentication are passed, a legal personnel signal is output; otherwise, an illegal personnel signal is output.

[0026] If a legal personnel signal is generated, the access, use, and operation permissions of all devices corresponding to the equipment number involved in the electronic work ticket that triggers the legal personnel signal are enabled;

[0027] If an illegal personnel warning signal is generated, the access, use, and operation permissions of all devices in the computer room are closed.

[0028] Step 3: Operation instruction monitoring;

[0029] After the personnel pass the authentication of the personnel identity recognition program and enter the computer room, their real-time operations are automatically recorded.

[0030] Monitor the operation instructions issued by this person, including device startup, shutdown, configuration parameters, parameter modification, operation mode switching, enabling or disabling network interfaces, data backup, data synchronization, data recovery, clearing device data caches and historical data, firmware update, and software version upgrade.

[0031] Record the time when each operation instruction is issued, the device numbers involved, and the current status of the device, and generate an operation instruction log.

[0032] As a preferred embodiment of the present invention, obtain the person's identity and operation task type in the electronic work ticket, and open the corresponding device operation permissions, including device startup, shutdown, configuration parameters, parameter modification, operation mode switching, enabling or disabling network interfaces, data backup, data synchronization, data recovery, clearing device data caches and historical data, firmware update, and software version upgrade.

[0033] As a preferred embodiment of the present invention, analyze each operation instruction in combination with the person's identity and operation task type in the electronic work ticket.

[0034] Automatically determine the authorization scope of operation instructions for each person's identity through a rule engine and intelligent algorithms.

[0035] Limit the authorization of operation instructions containing the operation task type of device restart to specific person identities: emergency response personnel;

[0036] Limit the authorization of operation instructions containing the operation task type of hardware replacement to specific person identities: device engineers;

[0037] Limit the authorization of operation instructions containing the operation task type of software update to specific person identities: network engineers;

[0038] Limit the authorization of operation instructions containing the operation task type of data backup to specific person identities: auditors.

[0039] If the type of an operation instruction exceeds the authorization scope of the person's identity and operation task type, automatically issue an alarm to this person through a voice output device and suspend the execution of this operation. For high-risk operation instructions, such as sensitive operations involving device restart, firmware update, etc., the operator is required to confirm twice to ensure the rationality and necessity of the operation.

[0040] Synchronize the generated operation instruction log to the central management platform for operation and maintenance management personnel to view and review. If the management personnel find any abnormal operations or unauthorized instructions, the management platform will generate a warning in a timely manner and notify the relevant personnel to take emergency measures.

[0041] Step 4: Real-time monitoring and risk identification;

[0042] During the operation and maintenance tasks, continuously monitor the voltage fluctuations and environmental data fluctuations of each device in the computer room, collect real-time data and compare it with historical data, and analyze the dynamic changes of potential risks. If an abnormal situation is monitored, immediate response measures shall be taken, including automatically triggering the emergency response program, pausing the operation tasks, and notifying all relevant responsible persons.

[0043] Combined with environmental sensor data: temperature C(t), humidity H(t), voltage ui(t) of each device, and voltage U(t) of the computer room bus, etc., conduct a comprehensive risk assessment. Where i is the device sequence number, i = 1, 2,..., n; where t is the data acquisition time;

[0044] At preset time intervals, integrate all the environmental sensor data collected into an environmental data multi-dimensional vector X(t) = {C(t), H(t), u1(t), u2(t),..., un(t), U(t)}.

[0045] Perform real-time analysis on the environmental data multi-dimensional vector through the long short-term memory network LSTM model to evaluate potential risks.

[0046] The long short-term memory network LSTM model controls the flow of information through forget gates, input gates, and output gates.

[0047] Among them, the forget gate determines which information in the memory cell at the previous moment is forgotten;

[0048] Among them, the input gate determines which information in the current input will be written into the memory cell;

[0049] Among them, the memory cell combines the outputs of the forget gate and the input gate to update the memory cell;

[0050] Among them, the output gate calculates the hidden state at the current moment based on the memory cell and the input, and calculates the final output h(t);

[0051] As a preferred method of the present invention, predict the risk state at the future moment t + 1 through the final output h(t) of the long short-term memory network LSTM model, and predict the limit values of voltage fluctuations and environmental data fluctuations at the future moment through the regression model to evaluate potential risks.

[0052] Through the preset formula R(t) = α×||h(t) - h threshold || + β×max(0, |U(t) - U threshold|) Calculate the risk assessment value R(t) at the current moment t, where α and β are preset weight factors used to balance the reference weights for the output value of the long short-term memory network (LSTM) model and the actual computer room bus voltage. Both α and β are decimals in the range of 0 to 1, and α + β = 1. Here, h threshold is the maximum threshold of the final output value of the long short-term memory network (LSTM) model, where U threshold is the maximum threshold of the computer room bus voltage.

[0053] As a preferred embodiment of the present invention, the data fields including the computer room number, equipment number, personnel identity, and operation task type contained in all electronic work tickets and their corresponding mapping relationships are written into the Neo4j graph database to establish a comparison graph.

[0054] The comparison graph contains a node set V and a relationship set R. The node set V = {v1, v2, v3,..., vp} contains several nodes, and each node represents the data fields of the specific computer room number, equipment number, personnel identity, and operation task type. Here, p is the total number of nodes, that is, the total number of data fields of the computer room number, equipment number, personnel identity, and operation task type. Among them, v1, v2, v3,..., vp each correspond to a specific data field of the computer room number, equipment number, personnel identity, or operation task type; the relationship set R = {r(v1, v2), r(v2, v3), r(v3, v4),..., r(vp-1, vp)} contains the mapping symbols between each data field; the mapping symbol represents the mapping relationship between any two data fields among the computer room number, equipment number, personnel identity, or operation task type. Let the value of each mapping symbol be equal to the number of times the corresponding two data fields have appeared in the same electronic work ticket; if the two data fields corresponding to the mapping symbol have never appeared in any electronic work ticket, then let the value of the mapping symbol be 0;

[0055] As a preferred embodiment of the present invention, the risk situation is calculated by combining nodes and mapping symbols to evaluate the risk situation included in the equipment number, personnel identity, and operation task type in each electronic work ticket.

[0056] Through a preset formula Calculate the influence relationship coefficients between the computer room number, equipment number, personnel identity, and operation task type in each electronic work ticket, including the influence relationship coefficient R relation 1 between the computer room number and the equipment number, the influence relationship coefficient R relation 2 between the equipment number and the personnel identity, the influence relationship coefficient R relation 3 between the personnel identity and the operation task type, the influence relationship coefficient R relation4. Influence relationship coefficient R between computer room number and operation task type relation 5 and influence relationship coefficient R between equipment number and operation task type relation 6. By analyzing the influence relationship coefficient R relation 1, R relation 2,..., R relation 6, the overall influence relationship coefficient R is obtained by summation relation . Wherein, both vi1 and vi2 are node numbers in the node set V

[0057] Step Five: Alarm and Response

[0058] When a security hazard or operation risk is detected, the system should issue an alarm in a timely manner and start an emergency response procedure

[0059] Based on the predicted risk assessment value R(t), risk judgment is carried out. If it is detected that the risk assessment value of a computer room exceeds the set threshold within consecutive T preset time intervals, the overall influence relationship coefficient R corresponding to the computer room number of this computer room is further detected relation . If the overall influence relationship coefficient R relation is less than the preset threshold, it is determined that there is a risk of equipment damage caused by unconventional maintenance operations in this computer room, and the following measures are taken

[0060] Automatically suspend all operation tasks in this computer room to prevent the risk from intensifying

[0061] Cancel all equipment operation permissions for this computer room

[0062] Send an alarm notification to all system administrators

[0063] Compared with the prior art, the beneficial effects of the present invention are

[0064] 1. The present invention realizes the accurate authentication of the identities of computer room operation and maintenance personnel through the generation and distribution of electronic work tickets, combined with OCR and two-dimensional code scanning technologies. This method ensures that only authorized personnel can perform operations, and each operation is traceable, guaranteeing the compliance and security of equipment operations. Through operation instruction monitoring and permission management, the risk brought by unauthorized operations or incorrect operations is further reduced

[0065] 2. The present invention can evaluate potential risks and detect abnormal situations in a timely manner by real-time monitoring of environmental data such as voltage, temperature, and humidity of computer room equipment and combining real-time analysis and prediction of data using a long short-term memory network model. This mechanism helps to predict and identify possible equipment failures or environmental problems in advance, so as to take emergency response measures to prevent the expansion and loss of risks

[0066] 3. When the present invention identifies an anomaly or a potential safety hazard, it will automatically trigger an emergency response procedure, including pausing the operation task, canceling the device operation permission, and sending an alarm notification. This automated response mechanism reduces the time delay of manual intervention and improves the efficiency of risk response. Through comprehensive risk assessment and management, the safety and stability of the computer room are ensured. BRIEF DESCRIPTION OF THE DRAWINGS

[0067] For the convenience of those skilled in the art to understand, the present invention will be further described below in conjunction with the accompanying drawings:

[0068] Figure 1 is the flowchart of the method of the present invention;

[0069] Figure 2 is the schematic diagram of the electronic work ticket proposed in the embodiment of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0070] The technical solutions of the present invention will be clearly and completely described below in conjunction with the embodiments. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all of the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the scope of protection of the present invention.

[0071] Please refer to Figure 1 shown, a method for identifying and monitoring risks in computer room operation and maintenance work includes the following steps:

[0072] Step 1: Work ticket generation;

[0073] Access the computer room operation and maintenance work order database, obtain all the computer room operation and maintenance task arrangement data therein, and perform feature extraction on the obtained task arrangement data based on natural language processing technology. Through word segmentation, named entity recognition, and relationship extraction, data fields corresponding to the computer room number, device number, start time of the operation and maintenance operation, end time of the operation and maintenance operation, operation task type, as well as the name, identity, and number of the operation and maintenance personnel are obtained.

[0074] Among them, the personnel identities include: on-site operators, equipment engineers, safety officers, system administrators, network engineers, auditors, external suppliers, and emergency response personnel.

[0075] Among them, the operation task types include: any one or a combination of multiple items such as equipment fault repair, system maintenance, software update, hardware replacement, safety inspection, data backup, network debugging, and power supply inspection.

[0076] Please refer to Figure 2As shown, a unique corresponding electronic work ticket is generated for each individual operation and maintenance personnel. The electronic work ticket includes a title part, a verification QR code part, and a text part. The title part of it includes the words "electronic work ticket" and the unique number of the electronic work ticket; the text part is obtained by automatically filling all the feature data about the operation and maintenance personnel extracted into a preset template in the form of data fields; the QR code part is obtained by inputting all the feature data about the operation and maintenance personnel extracted into a QR code generator for conversion and encoding;

[0077] All the electronic work tickets are distributed to the mobile communication devices of each operation and maintenance personnel according to their corresponding operation and maintenance personnel numbers.

[0078] A completely identical backup work ticket is copied for each electronic work ticket, and all the backup work tickets are encrypted and stored. The feature data fields in the text part of all the backup work tickets are encrypted through the SHA-256 hash encryption algorithm, and the QR code images in all the electronic work tickets are encrypted through the AES symmetric encryption algorithm. And all the encrypted backup work tickets are saved to the database.

[0079] Step Two: Personnel identity recognition;

[0080] The personnel entering the computer room are authenticated through a personnel identity recognition program based on the recognition of the electronic work ticket.

[0081] The real-time image of the computer room is collected through a camera, and motion tracking and image recognition analysis are carried out. Whenever it is detected that someone enters the computer room, a preset voice is output to require the entering person to show the electronic work ticket to the camera.

[0082] Subsequently, if the camera recognizes the electronic work ticket image, based on the OCR optical character recognition technology, the electronic work ticket information, including the personnel number, personnel name, personnel identity, operation time, operation task, and equipment number involved, is obtained from the QR code in the electronic work ticket through the QR code reader embedded in the camera.

[0083] If the camera fails to recognize the electronic work ticket image, it is determined that the identity authentication fails, and an alarm signal for illegal personnel is output.

[0084] After obtaining the electronic work ticket information of the entering person through the camera, the work ticket information is compared with the backup work tickets in the database to authenticate the authenticity and timeliness of the electronic work ticket of the entering person successively. The specific process is as follows:

[0085] Authenticity authentication: Convert the extracted electronic work ticket information into structured data, and extract the personnel number, personnel name, and personnel identity therein for subsequent comparison processing. Connect to the database through the API interface, and obtain the encrypted backup work ticket with the same unique number as the electronic work ticket from it. Decrypt the QR code image in the encrypted backup work ticket through the AES symmetric encryption algorithm, and compare the personnel number, personnel name, and personnel identity obtained after decryption with the personnel number, personnel name, and personnel identity in the electronic work ticket extracted through the camera.

[0086] If the comparison result is consistent, it is determined that the electronic work ticket of the entering personnel passes the authenticity authentication. After passing the authenticity authentication, perform timeliness authentication on the electronic work ticket.

[0087] Timeliness authentication: After determining that the identity authentication is successful, check the operation time, compare the operation time in the QR code of the electronic work ticket with the task arrangement time obtained after decryption in the database. If the comparison result is consistent, further determine whether the current time is within the operation time that conforms to the plan. If so, it is determined that the electronic work ticket of the entering personnel passes the timeliness authentication.

[0088] If the electronic work ticket of the entering personnel passes the authenticity authentication and timeliness authentication, output the legal personnel signal of the entering personnel, otherwise output the illegal personnel signal of the entering personnel.

[0089] If a legal personnel signal is generated, enable the access, use, and operation permissions of all devices corresponding to the device numbers involved in the electronic work ticket of the entering personnel who triggered the legal personnel signal;

[0090] If an illegal personnel warning signal is generated, disable the access, use, and operation permissions of the entering personnel who triggered the illegal personnel warning signal to all devices in the computer room.

[0091] Step 3: Operation instruction monitoring;

[0092] After the personnel pass the authentication of the personnel identity recognition program and enter the computer room, automatically record the real-time operations of the personnel.

[0093] Monitor the operation instructions issued by the personnel, including device startup, shutdown, configuration parameter setting, parameter modification, operation mode switching, network interface enabling or disabling, data backup, data synchronization, data recovery, device data cache and historical data clearing, firmware update, and software version upgrade.

[0094] Record the issuance time of each operation instruction, the device number involved, and the current status of the device, and generate an operation instruction log.

[0095] Further, obtain the personnel identity and operation task type in the electronic work ticket of this person, and open the corresponding device operation permissions, including device startup, shutdown, configuration parameter setting, parameter modification, operation mode switching, enabling or disabling network interfaces, data backup, data synchronization, data recovery, clearing device data caches and historical data, firmware update, and software version upgrade.

[0096] Further, analyze each operation instruction in combination with the personnel identity and operation task type in the electronic work ticket of this person.

[0097] Automatically judge the authorization scope of operation instructions for each personnel identity through a rule engine and intelligent algorithms.

[0098] Limit the authorization of operation instructions containing the operation task type of device restart to specific personnel identities: emergency response personnel;

[0099] Limit the authorization of operation instructions containing the operation task type of hardware replacement to specific personnel identities: device engineers;

[0100] Limit the authorization of operation instructions containing the operation task type of software update to specific personnel identities: network engineers;

[0101] Limit the authorization of operation instructions containing the operation task type of data backup to specific personnel identities: auditors.

[0102] If the type of an operation instruction exceeds the authorization scope of this person's personnel identity and operation task type, automatically issue an alarm to this person through a voice output device and suspend the execution of this operation. For high-risk operation instructions, such as sensitive operations involving device restart, firmware update, etc., the operator is required to confirm again to ensure the rationality and necessity of the operation.

[0103] Synchronize the generated operation instruction log to the central management platform for operation and maintenance management personnel to view and review. If the management personnel discover any abnormal operations or unauthorized instructions, the management platform will promptly generate a warning and notify the relevant personnel to take emergency measures.

[0104] Step Four: Real-time Monitoring and Risk Identification;

[0105] Continuously monitor the voltage fluctuations and environmental data fluctuations of each device in the computer room during the operation and maintenance tasks, collect real-time data and compare it with historical data, and analyze the potential dynamic changes of risks. If an abnormal situation is monitored, immediately take response measures. This includes automatically triggering the emergency response program, suspending the operation task, and notifying all relevant responsible persons.

[0106] Combined with environmental sensor data: temperature C(t), humidity H(t), the voltage ui(t) of each device, and the voltage U(t) of the computer room bus, etc., a comprehensive risk assessment is carried out. Where i is the device sequence number, i = 1, 2,..., n; where t is the data acquisition time;

[0107] At every preset time interval, all the collected environmental sensor data are integrated into an environmental data multi-dimensional vector X(t) = {C(t), H(t), u1(t), u2(t),..., un(t), U(t)}.

[0108] The long short-term memory network LSTM model is used to perform real-time analysis on the environmental data multi-dimensional vector to evaluate potential risks.

[0109] The long short-term memory network LSTM model controls the flow of information through forget gates, input gates, and output gates.

[0110] The forget gate determines which information in the memory unit at the previous moment is forgotten. The mathematical formula is:

[0111] f(t) = σ{W f [h(t - 1), X(t)] + b f};

[0112] Where σ is the sigmoid activation function, and the formula is: Where f(t) is the output of the forget gate, representing the proportion of information forgotten, ranging from 0 to 1; where W f is the weight matrix of the forget gate; b f is the bias term of the forget gate; where h(t - 1) is the hidden state at the previous moment, and the specific value is calculated by the output gate; where X(t) is the environmental data multi-dimensional vector input at the current moment; where [h(t - 1), X(t)] is the combination of X(t) input at the current moment t and the hidden state information h(t - 1) at the previous moment t - 1.

[0113] It should be noted that the role of the forget gate is to determine which information in the memory unit C(t - 1) at the previous moment is forgotten and which is retained.

[0114] The input gate determines which information in the current input will be written into the memory unit. The mathematical formula is:

[0115] i(t) = σ{W i [h(t - 1), X(t)] + b i};

[0116] Where i(t) is the output of the input gate, representing the influence of the current input on the memory unit, ranging from 0 to 1; where σ is the sigmoid activation function; where W iis the weight matrix of the input gate; b i is the bias term of the input gate;

[0117] The memory cell combines the outputs of the forget gate and the input gate to update the memory cell. The mathematical formula is:

[0118] C(t) = f(t) × C(t - 1) + i(t) × tanh{W c [h(t - 1), X(t)] + b c};

[0119] where C(t) is the memory cell at the current time; where C(t - 1) is the memory cell at the previous time; where f(t) × C(t - 1) is the retained part of the memory cell at the previous time; where i(t) × tanh{W c [h(t - 1), X(t)] + b c} is the updated part of the input at the current time; where tanh is the hyperbolic tangent function, and the output value is between -1 and 1. The formula is where W c is the weight matrix of the memory cell; where b c is the bias term for updating the memory cell.

[0120] The output gate calculates the hidden state at the current time based on the memory cell and the input, and outputs the final result. The mathematical formula is:

[0121] where W o is the weight matrix of the output gate; where b o is the bias term of the output gate. Where h(t) is the hidden state at the current time and is also the final output of the long short-term memory network LSTM model; where o(t) is the output of the output gate, which determines which information in the memory cell C(t) is output.

[0122] Furthermore, the risk state at the future time t + 1 is predicted through the final output h(t) of the long short-term memory network LSTM model, and the limit values of voltage fluctuations and environmental data fluctuations at the future time are predicted through a regression model to evaluate potential risks.

[0123] The risk assessment value R(t) at the current time t is calculated through the preset formula R(t) = α × ||h(t) - h threshold || + β × max(0, |U(t) - U threshold |), where α and β are preset weight factors used to balance the reference weights for the output value of the long short-term memory network LSTM model and the actual computer room bus voltage. Both α and β are decimals in the range of 0 to 1 and α + β = 1. Where h threshold is the maximum threshold of the final output value of the long short-term memory network LSTM model, where Uthreshold is the maximum threshold of the computer room bus voltage.

[0124] Furthermore, write the data fields of the computer room number, equipment number, personnel identity, and operation task type included in all electronic work tickets and their corresponding mapping relationships into the Neo4j graph database to establish a comparison graph.

[0125] The comparison graph contains a node set V and a relationship set R. The node set V = {v1, v2, v3,..., vp} contains several nodes, and each node represents the data fields of the specific computer room number, equipment number, personnel identity, and operation task type. Here, p is the total number of nodes, that is, the total number of data fields of the computer room number, equipment number, personnel identity, and operation task type. Among them, v1, v2, v3,..., vp each correspond to a specific data field of the computer room number, equipment number, personnel identity, or operation task type; the relationship set R = {r(v1, v2), r(v2, v3), r(v3, v4),..., r(vp - 1, vp)} contains the mapping symbols between each data field; the mapping symbol represents the mapping relationship between any two data fields in the computer room number, equipment number, personnel identity, or operation task type. Let the value of each mapping symbol be equal to the number of times the corresponding two data fields have appeared in the same electronic work ticket; if the two data fields corresponding to the mapping symbol have never appeared in any electronic work ticket, then let the value of this mapping symbol be 0;

[0126] Furthermore, combine nodes and mapping symbols to calculate the risk situation and evaluate the risk situation included in the equipment number, personnel identity, and operation task type in each electronic work ticket.

[0127] Through a preset formula calculate the influence relationship coefficients between the computer room number, equipment number, personnel identity, and operation task type in each electronic work ticket, including the influence relationship coefficient R relation 1 between the equipment number and the personnel identity, and the influence relationship coefficient R relation 2 between the personnel identity and the operation task type, and the influence relationship coefficient R relation 3 between the computer room number and the personnel identity, and the influence relationship coefficient R relation 4 between the computer room number and the operation task type, and the influence relationship coefficient R relation 5 and the influence relationship coefficient R between the equipment number and the operation task type relation 6. By summing the influence relationship coefficients R relation 1, R relation 2,..., R relation 6, obtain the overall influence relationship coefficient R relationAmong them, both vi1 and vi2 are node numbers in the node set V;

[0128] Step Five: Alarm and Response;

[0129] When potential safety hazards or operation risks are detected, the system should promptly issue an alarm and initiate an emergency response procedure.

[0130] Based on the predicted risk assessment value R(t), risk judgment is carried out. If it is detected that the risk assessment value of a machine room exceeds the set threshold within consecutive T preset time intervals, the overall impact relationship coefficient R corresponding to the machine room number of this machine room is further detected relation , if the overall impact relationship coefficient R relation is less than the preset threshold, it is determined that there is a risk of equipment damage caused by unconventional maintenance operations in this machine room, and the following measures are taken:

[0131] Automatically suspend all operation tasks in this machine room to prevent the risk from intensifying;

[0132] Revoke all equipment operation permissions for this machine room;

[0133] Send an alarm notification to all system administrators.

[0134] It should be understood that the terms "including" and "comprising" used in the specification and claims of this disclosure indicate the presence of the described features, wholes, steps, operations, elements, and / or components, but do not exclude the presence or addition of one or more other features, wholes, steps, operations, elements, components, and / or their combinations.

[0135] It should also be understood that the terms used in this disclosure specification are only for the purpose of describing specific embodiments and are not intended to limit this disclosure. As used in this disclosure specification and claims, unless the context clearly indicates otherwise, the singular forms "a", "an", and "the" are intended to include the plural forms. It should be further understood that the term "and / or" used in this disclosure specification and claims refers to any combination and all possible combinations of one or more of the associated listed items, and includes these combinations;

[0136] The preferred embodiments of the present invention disclosed above are only used to help explain the present invention. The preferred embodiments do not elaborate on all details and do not limit the present invention to only the specific implementation manners. Obviously, many modifications and variations can be made according to the content of this specification. These embodiments are selected and specifically described in this specification to better explain the principles and practical applications of the present invention, so that those skilled in the relevant technical fields can well understand and utilize the present invention. The present invention is only limited by the claims and their full scope and equivalents.

Claims

1. A method for identifying and monitoring risks in computer room operation and maintenance operations, characterized in that: The following steps are involved: Step 1: Generate work ticket; Access the computer room operation and maintenance management work order database, obtain all computer room operation and maintenance task arrangement data, generate an electronic work ticket for each operation and maintenance personnel, and distribute it to the mobile devices of the corresponding operation and maintenance personnel, and generate an encrypted backup work ticket for storage; the electronic work ticket includes a title part, a verification QR code part and a text part; Step 2: Personnel identification; The camera collects real-time images of the computer room and uses OCR optical character recognition technology to identify the electronic work ticket information of personnel entering the computer room, and perform personnel identity authentication and timeliness authentication; Control device access rights based on authentication results; if authentication fails, an illegal personnel alarm is triggered; Step 3: monitor operation instructions; After a person enters the computer room and passes identity authentication, the person's operation instruction information is recorded in real time; operation instruction monitoring is performed, an operation instruction log is generated, and compared with the person's authorized tasks. If unauthorized operations occur, an alarm is triggered; Step 4: Real-time monitoring and risk identification; During the operation and maintenance task, the voltage fluctuations and environmental data fluctuations of each device in the computer room are continuously monitored, real-time data is collected and compared with historical data, and the multi-dimensional vectors of environmental data are analyzed in real time through the long short-term memory network LSTM model and the comparison map. The output results are respectively obtained: the operation output of the LSTM model and the operation output result of the comparison map: the overall impact relationship coefficient; Step 5: Alarm and response; Combine the long short-term memory network LSTM model and the calculation output results of the comparison map to detect safety hazards; When a safety hazard risk is detected, an alarm is issued in a timely manner and the emergency response procedure is initiated.

2. A method for identifying and monitoring risks in computer room operation and maintenance operations according to claim 1, characterized in that: The task arrangement data includes: Data fields for the computer room number, equipment number, operation and maintenance operation start time, operation and maintenance operation end time, operation task type, and the name, identity, and number of the operation and maintenance personnel corresponding to each operation and maintenance task arrangement; Among them, personnel identities include: field operators, equipment engineers, security officers, system administrators, network engineers, auditors, external suppliers, and emergency responders; Among them, the operation task types include: any one or a combination of equipment fault repair, system maintenance, software update, hardware replacement, security inspection, data backup, network debugging, and power supply inspection.

3. A method for identifying and monitoring risks in computer room operation and maintenance operations according to claim 1, characterized in that: The specific process of generating an encrypted backup work ticket for storage is as follows: Copy an identical backup work ticket for each electronic work ticket, and encrypt and store all backup work tickets; encrypt the characteristic data fields in the text body of all backup work tickets using the SHA-256 hash encryption algorithm, and encrypt the QR code images in all electronic work tickets using the AES symmetric encryption algorithm; and save all encrypted backup work tickets in the database.

4. A method for identifying and monitoring risks in computer room operation and maintenance operations according to claim 1, characterized in that: The specific process of personnel identity authentication is as follows: The camera collects real-time images of the computer room, and performs motion tracking and image recognition analysis. Whenever a person is detected entering the computer room, a preset voice message is output to require the person to show the electronic work ticket to the camera. Subsequently, if the camera recognizes the electronic work ticket image, the electronic work ticket information, including the personnel number, personnel name, personnel identity, operation time, operation task and the number of the equipment involved, is obtained from the QR code in the electronic work ticket through the QR code reader embedded in the camera based on the OCR optical character recognition technology; If the camera cannot recognize the electronic work ticket image, it will determine that the identity authentication has failed and output an illegal personnel alarm signal; After obtaining the electronic work ticket information of the entrant through the camera, the work ticket information is compared with the backup work ticket in the database to authenticate the authenticity and timeliness of the electronic work ticket of the entrant.

5. A method for identifying and monitoring risks in computer room operation and maintenance operations according to claim 4, characterized in that: The specific process of authenticity authentication and timeliness authentication is as follows: Authenticity authentication: Convert the extracted electronic work ticket information into structured data, and extract the personnel number, personnel name and personnel identity for subsequent comparison processing; connect to the database through the API interface, and obtain the encrypted backup work ticket with the same unique number as the electronic work ticket; decrypt the QR code image in the encrypted backup work ticket through the AES symmetric encryption algorithm, and compare the decrypted personnel number, personnel name and personnel identity with the personnel number, personnel name and personnel identity in the electronic work ticket extracted through the camera; If the comparison result is consistent, it is determined that the electronic work ticket of the entrant has passed the authenticity authentication; after passing the authenticity authentication, the electronic work ticket is authenticated for timeliness; Timeliness authentication: When the identity authentication is successful, the operation time is checked, and the operation time in the QR code of the electronic work ticket is compared with the task schedule time obtained by decryption in the database. If the comparison result is consistent, it is further determined whether the current time is within the planned operation and maintenance operation time; If yes, it is determined that the validity certification has been passed; If the electronic work ticket of the entrant passes the authenticity authentication and timeliness authentication, then the legal personnel signal of the entrant is output, otherwise the illegal personnel signal of the entrant is output; If a legal personnel signal is generated, the access, use and operation rights of all devices corresponding to the device numbers involved in the electronic work ticket of the person who entered and triggered the legal personnel signal are enabled; If an illegal person alarm signal is generated, the access, use and operation rights of the person who triggered the illegal person alarm signal to all equipment in the computer room will be closed.

6. A method for identifying and monitoring risks in computer room operation and maintenance operations according to claim 1, characterized in that: The specific process of monitoring operation instructions is as follows: Obtain the personnel identity and operation task type in the electronic work ticket of the personnel, and open the corresponding equipment operation permissions, including equipment startup, shutdown, parameter configuration, parameter modification, operation mode switching, network interface enable or disable, data backup, data synchronization, data recovery, clearing device data cache and historical data, firmware update and software version upgrade; Analyze each operation instruction based on the personnel identity and operation task type in the electronic work ticket of the personnel; Automatically determine the authorization scope of each personnel's operation instructions through rule engines and intelligent algorithms; Restrict authorization of operational instructions that include equipment restart operational task types to specific personnel identities: emergency response personnel; Restrict authorization of operation instructions containing hardware replacement operation task types to specific personnel identities: equipment engineers; Restrict authorization of operational instructions that include software update operational task types to specific personnel identities: network engineers; Restrict authorization of operation instructions containing data backup operation task types to specific personnel identities: auditors; If the type of operation instruction exceeds the authorized scope of the personnel's personnel identity and operation task type, an alarm will be automatically issued to the personnel through the voice output device, and the execution of the operation will be suspended; for high-risk operation instructions, the operator will be required to make a second confirmation; Synchronize the generated operation instruction logs to the central management platform for operation and maintenance managers to view and review; If the manager finds any abnormal operation or unauthorized instructions, the management platform will generate a warning in time and notify relevant personnel to take emergency measures.

7. A method for identifying and monitoring risks in computer room operation and maintenance operations according to claim 1, characterized in that: The specific process of collecting real-time data is as follows: Collect environmental sensor data: temperature C(t), humidity H(t), voltage ui(t) of each device and voltage U(t) of the computer room bus for comprehensive risk assessment; i is the device sequence number, i=1, 2, ..., n; t is the data collection time; At every preset time interval, all collected environmental sensor data are integrated into an environmental data multidimensional vector X(t)={C(t), H(t), u1(t), u2(t), ..., un(t), U(t)}; The multi-dimensional vector of environmental data is analyzed in real time through a long short-term memory network (LSTM) model including a forget gate, an input gate, and an output gate.

8. A method for identifying and monitoring risks in computer room operation and maintenance operations according to claim 1, characterized in that: The comparison diagram is specifically: The comparison graph includes a node set V and a relationship set R, wherein the node set V = {v1, v2, v3, ..., vp} includes a number of nodes, each node represents a specific data field of a computer room number, a device number, a personnel identity and an operation task type, wherein p is the total number of nodes, that is, the total number of data fields of the computer room number, the device number, the personnel identity and the operation task type, wherein v1, v2, v3, ..., vp each corresponds to a specific data field of a computer room number, a device number, a personnel identity or an operation task type; wherein the relationship set R = {r(v1, v2), r(v2, v3), r(v3, v4), ..., r(vp-1, vp)} includes mapping symbols between various data fields; wherein the mapping symbols represent the mapping relationship between any two data fields in the computer room number, the device number, the personnel identity or the operation task type; let the value of each mapping symbol be equal to the number of times the corresponding two data fields have appeared in the same electronic work ticket; If the two data fields corresponding to the mapping symbol have never appeared in any electronic work ticket, the value of the mapping symbol is set to 0.

9. A method for identifying and monitoring risks in computer room operation and maintenance operations according to claim 1, characterized in that: The specific process of real-time analysis of multi-dimensional vectors of environmental data through the long short-term memory network LSTM model and comparison map is as follows: The multi-dimensional vector of environmental data is analyzed in real time through the long short-term memory network LSTM model including the forget gate, input gate and output gate to obtain the final output h(t); The final output h(t) of the long short-term memory network (LSTM) model is used to predict the risk status at the future time t+1. The limit values ​​of voltage fluctuation and environmental data fluctuation at the future time are predicted through the regression model to evaluate the potential risks. By the preset formula R(t) = α×||h(t)-h threshold ||+β×max(0,|U(t)-U threshold |) Calculate the risk assessment value R(t) at the current time t, where α and β are preset weight factors used to balance the reference weights for the output value of the long short-term memory network LSTM model and the actual room bus voltage, where α and β are both decimals ranging from 0 to 1 and α+β=1; where h threshold is the maximum threshold of the final output value of the long short-term memory network LSTM model, where U threshold is the maximum threshold of the bus voltage in the computer room; Calculate risk situation by combining nodes and mappers to evaluate the risk situation contained in the equipment number, personnel identity and operation task type in each electronic work ticket; By preset formula Calculate the influence relationship coefficients between the room number, equipment number, personnel identity and operation task type in each electronic work ticket, including the influence relationship coefficients R between the room number and equipment number relation 1. The influence coefficient R between equipment number and personnel identity relation 2. The influence coefficient R between personnel identity and operation task type relation 3. The influence coefficient R between the room number and the personnel identity relation 4. The influence coefficient R between the room number and the operation task type relation 5 and the influence relationship coefficient R between the equipment number and the operation task type relation 6. Through the influence relationship coefficient R relation 1. R relation 2, ..., R relation The sum of 6 gives the overall influence relationship coefficient R relation ; Among them, vi1 and vi2 are both node numbers in the node set V; Risk judgment is made based on the predicted risk assessment value and the overall impact relationship coefficient.

10. A method for identifying and monitoring risks in computer room operation and maintenance operations according to claim 9, characterized in that: The specific process of risk judgment based on the predicted risk assessment value and the overall impact relationship coefficient is as follows: Risk judgment is performed based on the predicted risk assessment value R(t). If it is detected that the risk assessment value of a computer room exceeds the set threshold value within T consecutive preset time intervals, the overall impact relationship coefficient R corresponding to the computer room number of the computer room is further detected. relation , if the overall influence relationship coefficient R relation If the value is less than the preset threshold, it is determined that there is a risk of equipment damage caused by non-routine maintenance operations in the equipment room, and the following measures are taken: Automatically suspend all operating tasks in the computer room to prevent the risk from increasing; Cancel all equipment operation permissions of the computer room; Send alert notifications to all system administrators.

Images