Dynamic scheduling method for IT equipment maintenance tasks integrated with fault prediction

Abstract

The present invention discloses a method for dynamically scheduling IT equipment maintenance tasks that integrates fault prediction, and relates to the technical field of maintenance task scheduling. The method includes the following steps: constructing a spatiotemporal feature matrix of the equipment status and a historical fault frequency matrix of the IT equipment, and determining a frequency matrix of each fault type of the IT equipment based on the working condition correction factors stored in a database; determining the amount of fault repair tasks based on the number of IT equipment and the frequency matrix of each fault type; determining the expected completion status of the fault repair tasks based on the current working status of the on-site maintenance personnel, and adjusting the working mode and personnel of the on-site maintenance personnel based on the expected completion status of the fault repair tasks. The system and method can intelligently assess the amount of faults and optimize the distribution of manual workload, thereby improving fault repair efficiency, reducing downtime, and enhancing user experience.

Description

Technical Field

[0001] The present invention relates to the technical field of maintenance task scheduling, and in particular to a method for dynamic scheduling of IT equipment maintenance tasks integrating fault prediction. Background Art

[0002] With the widespread adoption of smart terminals across various industries, the increasing diversity of their types and functions has significantly increased the frequency of hardware and software failures. In hospitals, a specialized field that serves the public and is crucial to life and health, a malfunction in a smart terminal can directly disrupt doctors' consultations, diagnoses, and interventions, negatively impacting patients' treatment progress. In government offices, smart terminals are crucial tools for administrative processing and information transmission. Failures can lead to delays and poor information flow, severely impacting the efficiency and quality of public services. In business operations, smart terminals play an indispensable role in production management, marketing, and customer service. Failures can disrupt business processes, reduce operational efficiency, and increase costs. Traditional fault repair methods rely primarily on manual inspections and repairs, which are not only inefficient but also difficult to accurately assess the relationship between fault volume and manual workload. This leads to uneven resource allocation, slow response times, and overall operational efficiency. Therefore, there is an urgent need for a system and method that can intelligently assess fault volume and optimize manual workload distribution to improve fault repair efficiency, reduce downtime, and enhance user experience. Summary of the Invention

[0003] To achieve the above objectives, the present invention is implemented through the following technical solutions: a dynamic scheduling method for IT equipment maintenance tasks integrating fault prediction, comprising the following steps: constructing a device status spatiotemporal feature matrix and a historical fault frequency matrix of IT equipment, and determining the frequency matrix of each fault type of IT equipment in combination with the working condition correction factors stored in the database; determining the amount of fault maintenance tasks based on the number of IT equipment and the frequency matrix of each fault type; determining the expected completion status of the fault maintenance task based on the current working status of the on-site maintenance personnel, and adjusting the working mode and personnel of the on-site maintenance personnel based on the expected completion status of the fault maintenance task.

[0004] Furthermore, a spatiotemporal feature matrix of the device status of IT equipment is constructed, including the following steps: cleaning the device ID-timestamp association of historical fault data, extracting the fault component code and severity level in the maintenance log through regular expressions, and generating a structured fault feature vector; calibrating the device operation log to UTC time, using a dynamic time warping algorithm to align load curves of different sampling frequencies, calculating the hourly load change rate through the first-order derivative, and generating an operation status time series matrix; constructing a joint interpolation space for the temperature and humidity sensor and vibration spectrometer data, generating minute-level synchronous data streams through cubic spline interpolation, calculating the temperature-vibration covariance coefficient of each device per hour, and forming an environmental impact tensor; and obtaining the device status spatiotemporal feature matrix by performing a spatiotemporal tensor product operation on the structured fault feature vector, the operation status time series matrix, and the environmental impact tensor.

[0005] Constructing a historical failure frequency matrix includes the following steps: obtaining historical equipment failure data from a database, grouping them by equipment model, calculating the average annual occurrence rate of each failure type, and constructing a historical failure frequency matrix.

[0006] Furthermore, the frequency matrix of each fault type of IT equipment is determined, which includes the following steps: obtaining a trained multi-task prediction model stored in the database, the multi-task prediction model uses a dual-channel gated recurrent unit network, and the main channel input is the device state spatiotemporal feature matrix , the auxiliary channel input is the historical fault frequency matrix and operating condition correction factors ;

[0007] Use modified Poisson regression to output the occurrence frequency of each fault type of each IT device in the future time window;

[0008] ;

[0009] ;

[0010] ;

[0011] in, is the gating signal, is the gating weight matrix, is the Sigmoid function, is the result of fusion feature calculation, is the element-wise product, is the bias term for the cth fault type, is the sensitivity weight of the cth fault type of the i-th IT equipment to the m-th dimension fusion feature, M is the total number of dimensions of the fusion feature, is the occurrence frequency of the cth fault type of the i-th IT equipment in the future time window;

[0012] The predicted occurrence frequency is dynamically adjusted based on real-time data streams to construct a frequency matrix for each fault type of IT equipment:

[0013] ,in, To dynamically adjust the predicted occurrence frequency, it represents the frequency of the cth fault type occurring in the future time window of the i-th IT equipment. For the forgetting factor, is the sliding window length, is the actual number of failures that occurred within the sliding window length.

[0014] Furthermore, the fault repair task volume is determined based on the number of IT devices and the frequency matrix of each fault type. The process is as follows: ,in, is the fault repair task volume, is the number of type i IT equipment.

[0015] Furthermore, the expected completion status of the fault repair task is determined based on the current work status of the on-site maintenance personnel, including the following steps: obtaining the maintenance work efficiency of the current on-site maintenance personnel for each fault type of each IT device, and calculating the average fault repair efficiency of each on-site maintenance personnel; obtaining the current on-site time of each on-site maintenance personnel, and calculating the manual maintenance workload: ,in, For manual maintenance workload, is the average efficiency of the fault maintenance of the a-th station maintenance personnel, is the station time of the ath station maintenance personnel, A is the number of current station maintenance personnel; the manual maintenance workload and the amount of fault repair tasks Perform the subtraction to obtain the difference; if the difference is less than 0, the expected completion status of the fault repair task is incomplete; if the difference is not less than 0 and not greater than the set threshold, the expected completion status of the fault repair task is complete; if the difference is greater than the set threshold, the expected completion status of the fault repair task is excessive on-site personnel.

[0016] Furthermore, the work mode and personnel of the on-site maintenance personnel are adjusted based on the expected completion status of the fault repair task, including the following steps: if the expected completion status of the fault repair task is that it cannot be completed, the maintenance personnel are transferred internally; if the expected completion status of the fault repair task is that it can be completed, the current status of the on-site maintenance personnel is maintained; if the expected completion status of the fault repair task is that there are too many on-site personnel, the maintenance personnel are transferred externally or the work mode of the maintenance personnel is changed.

[0017] Furthermore, the internal adjustment of maintenance personnel includes the following steps: determining the best matching personnel list from the list of dispatchable personnel based on the fitness calculation model; screening the personnel in the best matching personnel list based on the constraint conditions to determine the best combination of internal adjustment personnel.

[0018] Furthermore, based on the fitness calculation model, a list of the most matching personnel is determined from the list of schedulable personnel, including the following steps: determining a work feature matrix of each schedulable personnel in the list of schedulable personnel, including a set of historical maintenance fault types, a set of maintenance completion times for each fault type, a set of maintenance completion time for each fault type, and an average work efficiency; inputting the work feature matrix of each schedulable personnel into the fitness calculation model to obtain the maintenance fitness of each schedulable personnel; arranging the maintenance fitness in descending order, and obtaining the top R schedulable personnel as the list of the most matching personnel.

[0019] Furthermore, the fitness calculation model is expressed as:

[0020] ;

[0021] in, To maintain compatibility, is the average work efficiency, Maintain a collection of fault types for history, is the set of maintenance completion times for each fault type, The maintenance completion time set for each fault type, Maintain a set of fault types for the required history, is the required maintenance completion times for each fault type, The required maintenance completion time for each fault type is a set. is the cosine similarity function.

[0022] Furthermore, based on the constraints, the personnel in the best matching personnel list are screened to determine the best internal transfer personnel combination, including the following steps: randomly combining the original station personnel with the personnel in the best matching personnel list to obtain a maintenance personnel combination, where the maintenance personnel combination includes all the original station personnel and at least one person in the best matching personnel list; and evaluating each maintenance personnel combination based on the constraints to determine the best maintenance personnel combination as the best internal transfer personnel combination, wherein the constraints are:

[0023] ;

[0024] in, is the fault repair task volume, is the average work efficiency of the pth maintenance personnel in the maintenance personnel combination, is the on-site time of the pth maintenance personnel in the maintenance personnel combination, is the redundancy factor, is the operation and maintenance cost function.

[0025] The present invention has the following beneficial effects:

[0026] This dynamic scheduling method for IT equipment maintenance tasks, integrated with fault prediction, constructs a spatiotemporal feature matrix of equipment status and a historical fault frequency matrix, and combines this with operating condition correction factors to determine the occurrence frequency matrix of each fault type. This method accurately determines the amount of fault maintenance tasks to be completed. Furthermore, it determines the expected completion status of fault maintenance tasks based on the work status of on-site maintenance personnel, and rationally adjusts personnel work patterns and personnel accordingly. This method not only improves the accuracy of fault prediction, enables intelligent assessment and dynamic scheduling of maintenance tasks, optimizes the allocation of maintenance resources, improves maintenance efficiency, and reduces equipment downtime, but also enhances overall work efficiency and user experience. It is particularly suitable for critical areas where smart terminals are widely used and have high requirements for fault maintenance. It can effectively address the problems of low maintenance efficiency and uneven resource allocation caused by frequent failures of smart terminals in key areas (such as hospitals, governments, and enterprises).

[0027] Of course, any product implementing the present invention does not necessarily need to achieve all of the advantages described above at the same time. BRIEF DESCRIPTION OF THE DRAWINGS

[0028] Figure 1 This is a flow chart of the method for dynamically scheduling IT equipment maintenance tasks integrating fault prediction according to the present invention.

[0029] Figure 2 This is a schematic diagram of the operating and maintenance cost composition of the dynamic scheduling method for IT equipment maintenance tasks integrated with fault prediction in the present invention.

[0030] Figure 3 This is a schematic diagram of the operation and maintenance cost measurement model of the dynamic scheduling method for IT equipment maintenance tasks integrated with fault prediction in the present invention. DETAILED DESCRIPTION

[0031] See also Figure 1 , an embodiment of the present invention provides a technical solution: a dynamic scheduling method for IT equipment maintenance tasks integrating fault prediction, comprising the following steps: constructing a device status spatiotemporal feature matrix and a historical fault frequency matrix of IT equipment, and combining the working condition correction factors stored in the database to determine the frequency matrix of each fault type of IT equipment.

[0032] Constructing a spatiotemporal feature matrix for IT equipment status involves the following steps: Cleaning historical fault data by correlating device IDs with timestamps, extracting fault component codes and severity levels from maintenance logs using regular expressions, and generating a structured fault feature vector consisting of [device ID, timestamp, fault type code, and repair action hash value]. This process normalizes the disorganized historical fault data, extracting key information and forming a structured fault feature vector. This improves data availability and accuracy, providing a reliable foundation for subsequent fault prediction and analysis. Correlating device IDs with timestamps allows for better tracking and analysis of each device's fault history, helping to identify patterns and trends.

[0033] The device operation logs are calibrated to UTC time, and a dynamic time warping algorithm is used to align load curves with different sampling frequencies. The hourly load change rate is calculated using the first-order derivative to generate an operation status time series matrix containing [device ID, timestamp, operating time, number of starts and stops, and load change gradient]. This allows for temporal unification and regularization of device operation logs, aligning data with different sampling frequencies and more accurately reflecting changes in the device's operating status. By calculating the load change rate, load fluctuations during device operation can be captured, which is crucial for predicting equipment failures, as abnormal load changes are often a precursor to failure.

[0034] A joint interpolation space is constructed for data from temperature and humidity sensors (5 minutes / time) and vibration spectrometers (1 hour / time). Cubic spline interpolation is used to generate minute-by-minute synchronized data streams. The hourly temperature-vibration covariance coefficient is calculated for each device, forming an environmental impact tensor consisting of [device ID, timestamp, mean temperature, RMS vibration, covariance coefficient]. This tensor comprehensively considers the impact of environmental factors (such as temperature and vibration) on device operation. By constructing a joint interpolation space and generating synchronized data streams, the relationship between environmental factors such as temperature, humidity, and vibration and device failures can be more accurately analyzed. Abnormal changes in temperature and vibration can lead to device performance degradation or failure, so this step helps to more comprehensively assess the impact of the device's operating environment on its reliability.

[0035] The structured fault feature vector, the operating state time series matrix, and the environmental impact tensor are combined through a spatiotemporal tensor product operation to obtain the device state spatiotemporal feature matrix. This effectively integrates different types of feature data (fault characteristics, operating state, and environmental impact) to form a comprehensive device state spatiotemporal feature matrix. This integration provides a more comprehensive description of the device state, providing richer information for subsequent fault prediction models and improving prediction accuracy and reliability. The spatiotemporal tensor product operation captures the temporal and spatial variation patterns of the device state, which is critical for accurately predicting the occurrence of faults.

[0036] Constructing a historical failure frequency matrix includes the following steps: obtaining historical equipment failure data from a database, grouping them by equipment model, calculating the average annual occurrence rate of each failure type, and constructing a historical failure frequency matrix.

[0037] Determine the frequency matrix of each fault type of IT equipment, including the following steps: obtain the trained multi-task prediction model stored in the database, the multi-task prediction model uses a dual-channel gated recurrent unit network, and the main channel input is the device state spatiotemporal feature matrix , the auxiliary channel input is the historical fault frequency matrix and operating condition correction factors Leveraging pre-trained multi-task prediction models, we can simultaneously handle multiple related tasks (such as predicting different fault types), improving model efficiency and performance. Our dual-channel gated recurrent unit network effectively processes time series data, capturing trends in device status over time while factoring in historical fault frequencies and operating condition correction factors, resulting in more accurate and comprehensive predictions.

[0038] Use modified Poisson regression to output the occurrence frequency of each fault type of each IT device in the future time window;

[0039] ;

[0040] ;

[0041] ;

[0042] in, is the gating signal, is the gating weight matrix, is the Sigmoid function, is the result of fusion feature calculation, is the element-wise product, is the bias term for the cth fault type, is the sensitivity weight of the cth fault type of the i-th IT equipment to the m-th dimension fusion feature, M is the total number of dimensions of the fusion feature, is the frequency of the cth fault type for the i-th IT device within a future time window. Modified Poisson regression can process count data (such as the number of faults) and predict future fault frequencies. This approach allows for early understanding of potential faults for each IT device, providing a basis for scheduling subsequent maintenance tasks, enabling preventive maintenance and reducing equipment downtime.

[0043] The predicted occurrence frequency is dynamically adjusted based on real-time data streams to construct a frequency matrix for each fault type of IT equipment:

[0044] ,in, To dynamically adjust the predicted occurrence frequency, it represents the frequency of the cth fault type occurring in the future time window of the i-th IT equipment. For the forgetting factor, is the sliding window length, The actual number of faults that occurred within the sliding window length. The predicted frequency is dynamically adjusted based on the latest real-time data, making the prediction more accurate. By introducing a forgetting factor and the sliding window length, the influence of historical and new data can be balanced to a certain extent, improving the accuracy and timeliness of the prediction. The constructed frequency matrix provides the basis for determining the subsequent fault repair workload.

[0045] Determine the amount of fault repair tasks based on the number of IT equipment and the frequency matrix of each fault type.

[0046] ,in, is the fault repair task volume, is the number of IT devices of type i. By comprehensively considering the number of devices and the frequency of failures, the workload for fault repairs is accurately determined, providing a scientific basis for subsequent maintenance task scheduling and resource allocation. This helps address the problems of uneven resource allocation and low maintenance efficiency caused by the inability to accurately assess the matching between fault volume and manual workload in traditional maintenance methods, thereby improving overall maintenance efficiency and resource utilization.

[0047] The expected completion status of the fault repair task is determined based on the current working status of the on-site maintenance personnel, and the working mode and personnel of the on-site maintenance personnel are adjusted based on the expected completion status of the fault repair task.

[0048] Obtain the maintenance efficiency of current on-site maintenance personnel for each IT device fault type and calculate the average fault maintenance efficiency of each on-site maintenance personnel. Quantify and evaluate the work efficiency of each on-site maintenance personnel to provide data support for subsequent task allocation and personnel scheduling. By understanding the maintenance efficiency of different maintenance personnel for different equipment and fault types, maintenance tasks can be more rationally arranged, improving overall maintenance efficiency.

[0049] Get the current on-site maintenance time of each on-site maintenance personnel and calculate the manual maintenance workload: ,in, For manual maintenance workload, is the average efficiency of the fault maintenance of the a-th station maintenance personnel, The on-site maintenance personnel's on-site time is calculated based on their on-site time, and A is the number of on-site maintenance personnel currently on-site. This allows for a more accurate assessment of the overall work capacity of the current maintenance personnel. By combining on-site time with maintenance efficiency, we can determine the maximum workload for each maintenance personnel within a given timeframe, providing a basis for subsequent task allocation.

[0050] Manual maintenance workload and the amount of fault repair tasks The difference is calculated and the difference is calculated. If the difference is less than 0, the expected completion status of the fault repair task is unavailable. If the difference is not less than 0 and is not greater than the set threshold, the expected completion status of the fault repair task is complete. If the difference is greater than the set threshold, the expected completion status of the fault repair task is excessive on-site staffing. This intuitively compares the gap between maintenance personnel's work capacity and the actual maintenance task load to determine whether the current maintenance personnel are capable of completing the task. By calculating the difference, it is clear whether the task can be completed on time and whether there is a shortage or surplus of staff.

[0051] If the expected completion status of the fault repair task is unachievable, the maintenance personnel will be transferred internally. If the expected completion status of the fault repair task is complete, the current on-site maintenance personnel will be maintained. If the expected completion status of the fault repair task is excessive on-site personnel, the maintenance personnel will be transferred externally or the maintenance personnel's work mode will be changed (to a semi-resident or call-based service mode). Specifically, if the difference is less than 0, it indicates that the maintenance personnel's work capacity is insufficient and cannot complete the task. If the difference is not less than 0 and not greater than the set threshold, it indicates that the task can be completed but may require some effort. If the difference is greater than the set threshold, it indicates that the maintenance personnel's work capacity is excessive and there may be staff redundancy.

[0052] Internal transfers of maintenance personnel involve the following steps: Using a fitness calculation model, the best-matched personnel are identified from the list of available personnel; and based on constraints, the personnel in the best-matched list are screened to determine the optimal internal transfer personnel combination. The fitness calculation model comprehensively considers factors such as the maintenance personnel's historical maintenance failure types, number of completed maintenance tasks, maintenance completion time, and average work efficiency. This ensures that the selected personnel have the best skills and experience for the current task, improving the quality and efficiency of maintenance work.

[0053] Based on the most suitable personnel, we further consider practical constraints (such as the volume of repair tasks, average maintenance personnel efficiency, on-site hours, and operating and maintenance costs) to determine the optimal internal transfer personnel combination. This approach not only ensures the smooth completion of maintenance tasks, but also ensures the rationality and cost-effectiveness of the personnel combination, avoiding waste of human resources and unnecessary costs.

[0054] Based on the fitness calculation model, a list of the most matching personnel is determined from the list of schedulable personnel, including the following steps: determining the work characteristic matrix of each schedulable personnel in the list of schedulable personnel, including a set of historical maintenance fault types, a set of maintenance completion times for each fault type, a set of maintenance completion time for each fault type, and an average work efficiency; inputting the work characteristic matrix of each schedulable personnel into the fitness calculation model to obtain the maintenance fitness of each schedulable personnel; arranging the maintenance fitness in descending order, obtaining the top R schedulable personnel and recording them as the list of the most matching personnel.

[0055] The fitness calculation model is expressed as:

[0056] ;

[0057] in, To maintain compatibility, is the average work efficiency, Maintain a collection of fault types for history, is the set of maintenance completion times for each fault type, The maintenance completion time set for each fault type, Maintain a set of fault types for the required history, is the required maintenance completion times for each fault type, The required maintenance completion time for each fault type is a set. is the cosine similarity function. This data provides a detailed understanding of each dispatchable personnel's maintenance experience, number of completions, completion time, and average work efficiency for different fault types. This provides a rich data foundation for subsequent fitness calculations and helps to more accurately assess the degree of match between each dispatchable personnel and the current maintenance task. The cosine similarity function can effectively measure the similarity between a maintenance personnel's historical work characteristics and the current maintenance task requirements. Cosine similarity is not affected by the length of the feature vector and can better reflect the directional relationship between features, thereby more accurately assessing the match between maintenance personnel and tasks. This helps to screen the most suitable personnel for the current task from among the many dispatchable personnel, improving the efficiency and quality of maintenance work.

[0058] Based on the constraints, the personnel in the best matching personnel list are screened to determine the best internal transfer personnel combination, including the following steps: randomly combining the original station personnel with the personnel in the best matching personnel list to obtain a maintenance personnel combination, where the maintenance personnel combination includes all the original station personnel and at least one person in the best matching personnel list; and evaluating each maintenance personnel combination based on the constraints to determine the best maintenance personnel combination as the best internal transfer personnel combination, where the constraints are:

[0059] ;

[0060] in, is the fault repair task volume, is the average work efficiency of the pth maintenance personnel in the maintenance personnel combination, is the on-site time of the pth maintenance personnel in the maintenance personnel combination, is the redundancy factor, The O&M cost function is used. The generated maintenance personnel combinations can be comprehensively evaluated based on practical constraints (such as the volume of repair tasks, average maintenance personnel efficiency, on-site hours, and O&M costs). This method identifies the optimal maintenance personnel combination that meets task requirements while minimizing O&M costs. This not only ensures the smooth completion of maintenance tasks but also optimizes human resource utilization, avoiding redundant personnel and cost waste.

[0061] It should be noted that the operation and maintenance cost function The measurement method is as follows Figure 2 and Figure 3 As shown: Operation and maintenance costs include direct labor costs, direct non-labor costs, indirect labor costs and indirect non-labor costs in operation and maintenance services.

[0062] Direct labor costs include salaries, bonuses, benefits, and other human resource expenses for personnel providing operation and maintenance services. Operation and maintenance personnel include those directly involved in the service process, such as service managers, engineers, and quality assurance personnel. For personnel who do not work full-time on the operation and maintenance services, their human resource expenses are calculated based on the proportion of their workload to their total workload.

[0063] Direct non-labor costs include:

[0064] a) Office expenses, i.e. administrative office expenses incurred by the operation and maintenance service provider in providing the operation and maintenance services, such as office supplies, communications, mailing, printing, meetings, etc.;

[0065] b) Travel expenses, i.e. travel expenses incurred by the operation and maintenance service provider for providing the operation and maintenance services, such as transportation, accommodation, travel allowance, etc.;

[0066] c) Training fees, i.e. the fees incurred by the operation and maintenance service provider for special training arranged for the provision of this operation and maintenance service;

[0067] d) Business expenses, i.e., expenses incurred by the operation and maintenance service provider for auxiliary activities required to complete the operation and maintenance service, such as review fees, acceptance fees, etc.;

[0068] e) Procurement costs, i.e., costs incurred by the operation and maintenance service provider for purchasing specialized assets or services for the operation and maintenance services, such as spare parts, operation and maintenance tools, technical collaboration fees, patent fees, and venues;

[0069] f) Rental Fees: These are the rental fees incurred by the operation and maintenance service provider for the operation and maintenance services, such as office space rental fees (due to the limited office conditions of the operation and maintenance service demander, office space cannot be provided for operation and maintenance service personnel, but timely service is required), equipment rental fees (when handling professional issues, temporary rental of professional equipment is often required; disaster recovery drills are regularly provided as required by the demander, etc.);

[0070] g) Others, i.e., expenses not listed in a) to f) but which are indeed the expenses incurred by the operation and maintenance service provider for this service.

[0071] Indirect labor costs refer to the human resource expenses of non-project team personnel used by the operation and maintenance service provider to meet the overall operation and maintenance needs, including the salary, bonuses, benefits, etc. of the operation and maintenance department head, project management office (PMO) personnel, organizational-level quality assurance personnel, etc.

[0072] Indirect non-labor costs refer to the non-labor costs incurred by the operation and maintenance service provider that are not incurred for a specific project but contribute to the overall operation and maintenance activities. These include office space rent, utilities, property management, daily office expenses, and the rental, maintenance, and depreciation of various office equipment.

[0073] The specific breakdown is basic environment operation and maintenance costs, hardware operation and maintenance costs, software operation and maintenance costs, etc.

[0074] Basic environment operation and maintenance costs :

[0075] Basic environment operation and maintenance services provide the maintenance of infrastructure such as power, air conditioning, fire protection, and security systems necessary to ensure the normal operation of information systems. These services include routine inspections and status monitoring, response support, troubleshooting, and performance optimization of computer room power, fire protection, and security systems. Basic environment operation and maintenance services cover power supply and distribution systems, generator systems, precision air conditioning systems, fresh air systems, lightning protection and grounding systems, fire protection systems, video surveillance systems, and access control systems. Basic environment operation and maintenance cost measurement measures the cost of providing basic environment operation and maintenance services.

[0076] ;

[0077] Direct labor costs for basic environment operation and maintenance services, For the workload of type b basic environment operation and maintenance services, The unit price for the b-type basic environment operation and maintenance service is Price adjustment factor for basic environment operation and maintenance services, Direct non-human cost for basic environment operation and maintenance services, Indirect labor costs for basic environment operation and maintenance services, is the indirect non-human cost of basic environment operation and maintenance services, and B is the number of types of basic environment operation and maintenance services.

[0078] The basic environment operation and maintenance service workload is the sum of the number of type b devices in the basic environment, the unit workload of type b devices, and the product of the workload adjustment factor of type b devices.

[0079] Unit workload is the number of man-days required for the operation and maintenance of a certain type of hardware equipment within a certain period. The workload of operation and maintenance services is divided into the following four categories:

[0080] Routine operation: workload required for a single routine operation service of a single device × number of times;

[0081] Response support: the workload required for a single response support service for a single device × the number of times;

[0082] Optimization and improvement: the workload required for a single optimization and improvement service of a single device × the number of times;

[0083] Survey and evaluation: The workload required for a single survey and evaluation service for a single device × the number of times.

[0084] When calculating the overall workload, add up the total workload of the above four types of operation and maintenance service content, and then tailor the operation and maintenance service content accordingly according to the service catalog.

[0085] Hardware operation and maintenance costs :

[0086] Hardware operation and maintenance services refer to routine inspections and status monitoring, response support, troubleshooting, performance optimization, and other services for hardware equipment (networks, hosts, storage, desktop devices, and other related equipment). The objects of hardware operation and maintenance include: networks and network equipment, host equipment (personal computer servers, minicomputers, and mainframes, etc.), storage devices, desktops and peripheral devices (fixed computing terminals, mobile computing terminals, peripheral input and output devices, peripheral storage devices, and peripheral communication devices), and other hardware. Hardware operation and maintenance cost measurement is the calculation of the cost of providing hardware operation and maintenance services.

[0087] ;

[0088] Direct labor costs for hardware operation and maintenance, For the dth type of hardware operation and maintenance service workload, is the unit price of the dth type of hardware operation and maintenance service, is the price adjustment factor for hardware operation and maintenance services. Direct non-human costs for hardware operation and maintenance services, Indirect labor costs for hardware operation and maintenance services, is the indirect non-labor cost of hardware operation and maintenance services, and D is the number of types of hardware operation and maintenance services;

[0089] The workload of hardware operation and maintenance services is the sum of the number of type d hardware devices, the unit workload of type d hardware devices, and the product of the workload adjustment factor of type d hardware devices.

[0090] Software operation and maintenance costs :

[0091] Software operations and maintenance refers to the comprehensive services provided to the information system operating environment and business systems used by the customer, using information technology means and methods, based on the service level requirements specified by the customer. Software operations and maintenance services include functional modification and improvement of software (including basic software, support software, and application software), performance tuning, routine inspections, status monitoring, and response support. Software operations and maintenance cost measurement involves calculating the cost of providing these services.

[0092] ;

[0093] For the e-th type of software operation and maintenance workload, is the average labor cost rate for the operation and maintenance of the e-th type of software, is the direct non-human cost of software operation and maintenance, and E is the number of software operation and maintenance types.

[0094] If the unit price of operation and maintenance function points has been determined, the software operation and maintenance cost can be calculated based on the unit price of operation and maintenance function points using the following formula: Software operation and maintenance service cost = (software scale × operation and maintenance function point unit price) × operation and maintenance level requirement adjustment factor × operation and maintenance capability adjustment factor × operation and maintenance system and business characteristics adjustment factor + direct non-human cost.

[0095] Software operation and maintenance service workload = (software scale × productivity) × operation and maintenance level requirement adjustment factor × operation and maintenance capability adjustment factor × operation and maintenance system and business characteristics adjustment factor.

[0096] An electronic device includes: a processor; and a memory, wherein computer program instructions are stored in the memory, and when the computer program instructions are executed by the processor, the processor executes the above-mentioned method for dynamically scheduling IT equipment maintenance tasks integrating fault prediction.

[0097] A computer-readable storage medium is used to store a program, which, when executed by a processor, implements the above-mentioned method for dynamically scheduling IT equipment maintenance tasks integrating fault prediction.

[0098] Those skilled in the art will appreciate that embodiments of the present invention may be provided as methods, systems, or computer program products. Thus, the present invention may take the form of an entirely hardware embodiment, an entirely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to magnetic disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0099] The present invention is described with reference to flowcharts and / or block diagrams of systems, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowcharts and / or block diagrams, as well as combinations of processes and / or blocks in the flowcharts and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowcharts and / or block diagrams. Figure 1 a process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0100] These computer program instructions may also be stored in a computer readable memory that can direct a computer or other programmable data processing device to work in a specific manner, so that the instructions stored in the computer readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 a process or multiple processes and / or boxes Figure 1 The function specified in one or more boxes.

[0101] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operational steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing the instructions executed on the computer or other programmable device for implementing the process. Figure 1 a process or multiple processes and / or boxes Figure 1 A step that specifies a function in one or more boxes.

[0102] Although the preferred embodiments of the present invention have been described, those skilled in the art may make additional changes and modifications to these embodiments once they have learned the basic creative concept. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments and all changes and modifications that fall within the scope of the present invention.

[0103] Obviously, those skilled in the art may make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if such changes and modifications fall within the scope of the claims and their equivalents, the present invention is intended to include such changes and modifications.

Claims

1. A dynamic scheduling method for IT equipment maintenance tasks integrating fault prediction, characterized by: The following steps are involved: Construct the spatiotemporal characteristic matrix of the equipment status and the historical fault frequency matrix of the IT equipment, and combine it with the working condition correction factors stored in the database to determine the frequency matrix of each fault type of the IT equipment; Historical fault data is cleaned by device ID-timestamp association, and regular expressions are used to extract fault component codes and severity levels from maintenance logs to generate structured fault feature vectors. The device operation log is calibrated to UTC time, and a dynamic time warping algorithm is used to align load curves with different sampling frequencies. The hourly load change rate is calculated through the first-order derivative to generate an operation status time series matrix. A joint interpolation space is constructed for the data from temperature and humidity sensors and vibration spectrometers. Minute-level synchronized data streams are generated through cubic spline interpolation. The temperature-vibration covariance coefficient for each device per hour is calculated to form an environmental impact tensor. The structured fault feature vector, the operating state time series matrix and the environmental impact tensor are used to obtain the equipment state time-space feature matrix through time-space tensor product operation; Obtain historical equipment failure data from the database, group them by equipment model, calculate the average annual occurrence rate of each failure type, and construct a historical failure frequency matrix; Get the trained multi-task prediction model stored in the database. The multi-task prediction model uses a dual-channel gated recurrent unit network, and the main channel input is the device state spatiotemporal feature matrix , the auxiliary channel input is the historical fault frequency matrix and operating condition correction factors ; Use modified Poisson regression to output the occurrence frequency of each fault type of each IT device in the future time window; ； ； ； in, is the gating signal, is the gating weight matrix, is the Sigmoid function, is the result of fusion feature calculation, is the element-wise product, is the bias term for the cth fault type, is the sensitivity weight of the cth fault type of the i-th IT equipment to the m-th dimension fusion feature, M is the total number of dimensions of the fusion feature, is the occurrence frequency of the cth fault type of the i-th IT equipment in the future time window; The predicted occurrence frequency is dynamically adjusted based on real-time data streams to construct a frequency matrix for each fault type of IT equipment: ,in, To dynamically adjust the predicted occurrence frequency, it represents the frequency of the cth fault type occurring in the future time window of the i-th IT equipment. For the forgetting factor, is the sliding window length, is the actual number of secondary failures that occurred within the sliding window length; Determine the amount of fault repair tasks based on the number of IT equipment and the frequency matrix of each fault type; ,in, is the fault repair task volume, is the number of IT equipment of type i; Determine the expected completion status of the fault repair task based on the current work status of the on-site maintenance personnel, and adjust the work mode and personnel of the on-site maintenance personnel based on the expected completion status of the fault repair task; Obtain the maintenance efficiency of the current on-site maintenance personnel for each fault type of each IT device, and calculate the average fault maintenance efficiency of each on-site maintenance personnel; Get the current on-site maintenance time of each on-site maintenance personnel and calculate the manual maintenance workload: ,in, For manual maintenance workload, is the average efficiency of the fault maintenance of the a-th station maintenance personnel, is the station time of the ath on-site maintenance personnel, and A is the number of on-site maintenance personnel currently on-site; Manual maintenance workload and the amount of fault repair tasks Make a difference and get the difference; If the difference is less than 0, the expected completion status of the fault repair task is incomplete; If the difference is not less than 0 and not greater than the set threshold, the expected completion status of the fault repair task is complete; If the difference is greater than the set threshold, the expected completion status of the fault repair task is excessive on-site personnel.

2. The method for dynamic scheduling of IT equipment maintenance tasks integrating fault prediction according to claim 1 is characterized in that: Adjust the work mode and personnel of on-site maintenance personnel based on the expected completion status of the fault repair task, including the following steps: If the expected completion status of the fault repair task is impossible to complete, the maintenance personnel will be transferred internally; If the expected completion status of the fault repair task is complete, the current status of the on-site maintenance personnel is maintained; If the expected completion status of the fault maintenance task is that there are too many on-site personnel, the maintenance personnel will be transferred or the maintenance personnel's work mode will be changed.

3. The method for dynamic scheduling of IT equipment maintenance tasks integrating fault prediction according to claim 2 is characterized in that: Conducting an internal adjustment of maintenance personnel includes the following steps: Determine the best matching personnel list from the list of schedulable personnel based on the fitness calculation model; Based on the constraints, the personnel in the list of most matching personnel are screened to determine the best combination of internal transfer personnel.

4. The method for dynamic scheduling of IT equipment maintenance tasks integrating fault prediction according to claim 3 is characterized in that: Determining the best matching personnel list from the list of schedulable personnel based on the fitness calculation model includes the following steps: Determine the work characteristic matrix of each dispatchable personnel in the dispatchable personnel list, including the set of historical maintenance fault types, the set of maintenance completion times for each fault type, the set of maintenance completion time for each fault type, and the average work efficiency; Input the work characteristic matrix of each dispatchable personnel into the fitness calculation model to obtain the maintenance fitness of each dispatchable personnel; Arrange the maintenance suitability in descending order, and obtain the top R dispatchable personnel as the best matching personnel list.

5. The method for dynamic scheduling of IT equipment maintenance tasks integrating fault prediction according to claim 4 is characterized in that: The fitness calculation model is expressed as: ； in, To maintain compatibility, is the average work efficiency, Maintain a collection of fault types for history, is the set of maintenance completion times for each fault type, The maintenance completion time set for each fault type, Maintain a set of fault types for the required history, is the required maintenance completion times for each fault type, The required maintenance completion time for each fault type is a set. is the cosine similarity function.

6. The method for dynamic scheduling of IT equipment maintenance tasks integrating fault prediction according to claim 3 is characterized in that: Based on the constraints, the best matching candidates are screened to determine the optimal internal transfer candidate combination, including the following steps: Randomly combine the original on-site personnel with the personnel in the best matching personnel list to obtain a maintenance personnel combination, which includes all the original on-site personnel and at least one person in the best matching personnel list; Based on the constraints, each maintenance personnel combination is evaluated to determine the best maintenance personnel combination as the best internal transfer personnel combination, where the constraints are: ； in, is the fault repair task volume, is the average work efficiency of the pth maintenance personnel in the maintenance personnel combination, is the on-site time of the pth maintenance personnel in the maintenance personnel combination, is the redundancy factor, is the operation and maintenance cost function.

Images