Graded alarm processing and task scheduling method used in multi-user environment

Abstract

The invention discloses a hierarchical alarm processing and task scheduling method used in a multi-user environment, and relates to the technical field of information processing and resource scheduling, and the method comprises the steps: obtaining a technician capability parameter vector, time constraint data and a workload upper limit, and obtaining a candidate distribution scheme; calculating task priority weighting factors of the technicians in different time periods, if time constraint data of the technicians conflicts with task time requirements, marking the time constraint data as an unavailable state, obtaining a first allocation scheme after conflicts are eliminated, obtaining capability parameter vectors and workload allocation proportions of other technicians, and obtaining a second allocation scheme after the conflicts are eliminated; a substitute person list closest to the unavailable person capability parameter vector is generated, and a second distribution scheme is obtained; according to the hierarchical alarm processing and task scheduling method used in the multi-user environment, real-time task execution logs are generated, system state data are updated, intelligent distribution, dynamic adjustment and stability prediction of tasks of technicians are achieved, and the efficiency and rationality of task distribution are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of information processing and resource scheduling, and in particular to a hierarchical alarm processing and task scheduling method for a multi-user environment. Background Art

[0002] Against the backdrop of rapidly developing information technology and intelligent systems, research on hierarchical alarm processing and task scheduling in multi-user environments is crucial. This field is directly related to the efficient operation and resource optimization of complex systems. In particular, in scenarios such as emergency response, industrial control, and intelligent services, proper human resource scheduling can significantly improve task processing efficiency and system reliability.

[0003] Currently, research on human resource scheduling for hierarchical alarm processing has largely focused on automated task allocation, but existing methods still have significant shortcomings in intelligently allocating personnel to multi-user schedules. Many traditional scheduling schemes rely on static rules or simple priority rankings, making them difficult to adapt to the dynamically changing needs of multiple users. These methods often ignore the individual workload limits and ability differences of technicians, resulting in uneven task allocation and inefficiency. Furthermore, existing solutions often lack flexible mechanisms for recommending backup personnel when handling time conflicts, which can easily lead to delays in alarm processing due to unavailability. In particular, accurately matching alarm tasks of varying levels based on task urgency and personnel capabilities remains an unresolved challenge. These limitations highlight the core challenges of human resource scheduling in multi-user environments. First, intelligent staffing requires dynamically integrating multiple constraints, such as time, ability, and task priority, into multi-user schedules, but existing algorithms struggle to achieve efficient global optimization. Second, resolving time conflicts and recommending backup personnel requires the system to rapidly respond to dynamic changes, but currently lack mechanisms that balance real-time performance with accuracy. Finally, the setting of personalized workload upper limits and capability parameters involves a multi-dimensional assessment of technical personnel. Existing methods are often oversimplified and difficult to accurately reflect individual differences. Summary of the Invention

[0004] The purpose of the present invention is to provide a hierarchical alarm processing and task scheduling method for a multi-user environment to solve the problems existing in the prior art.

[0005] To achieve the above objectives, the present invention provides the following technical solution: a hierarchical alarm processing and task scheduling method for a multi-user environment, comprising: Obtain the technical personnel's capability parameter vector, time constraint data and workload upper limit to obtain candidate allocation plans; Calculate the technician's task priority weighting factor in different time periods. If the technician's time constraint data conflicts with the task time requirement, mark it as unavailable, and obtain the first allocation plan after eliminating the conflict; Obtain the capability parameter vectors and workload distribution ratios of other technicians, generate a list of substitute personnel closest to the capability parameter vectors of the unavailable personnel, and obtain a second allocation plan; Combining the task priority weighting factor and time constraint check, the substitute personnel are sorted to obtain the third allocation plan; According to the time constraint verification and capability parameter vector matching requirements of each task, determine the task and personnel mapping table; Obtain the load change trend of technicians after assignment, reallocate tasks and generate an adjusted fifth assignment plan; Predict the stability of task allocation in the future time period, perform batch comparison operations to regenerate the list of substitute personnel, and obtain the final allocation plan; Combining the task priority weighting factor and capability parameter vector, a real-time task execution log is generated to obtain the system status data.

[0006] Preferably, the candidate allocation schemes obtained by obtaining the technician capability parameter vector, time constraint data and workload upper limit include: The technical personnel's ability parameter vectors, time constraint data and workload upper limit are extracted from the schedule database. The multi-user data is parsed using structured query language to obtain the initial personnel data set. The ability parameter vectors in the initial personnel data set are processed by vector normalization. Combined with the task priority list, a personnel and task matching matrix is generated to obtain an intermediate allocation plan that includes time availability and ability score. If the workload of a technician in the intermediate allocation plan exceeds the upper limit, the task allocation is adjusted through a linear programming algorithm to obtain an optimized candidate allocation plan. Based on the optimized candidate allocation plan, the time availability of each technician is obtained, and the time window partitioning method is used to determine the task execution time period. If there is a conflict between the task execution time period and the time constraint data, the task order is adjusted through priority sorting to obtain the allocation plan after the conflict is resolved. By comparing the allocation plan after the conflict is resolved with the ability score, the weighted average method is used to determine the final task allocation result. The final task allocation result is obtained, and an allocation record including technicians, tasks and time periods is generated and stored in the schedule database.

[0007] Preferably, the step of calculating the task priority weighting factors of technicians in different time periods and marking technicians as unavailable if their time constraint data conflicts with the task time requirement, and obtaining a first allocation plan after eliminating the conflict includes: Obtain the initial personnel and task matching matrix, extract technicians, task priorities and time period information, generate preliminary allocation data including time constraints, use dynamic integration algorithm to calculate the task priority weighting factors in each time period for the preliminary allocation data, and obtain a set of weighting factors. If the technician time constraint conflicts with the task time requirement, the corresponding technician will be marked as unavailable according to the conflict marking rule to obtain the allocation data after eliminating the conflict. By eliminating the conflict and combining the allocation data with the weighting factor set, a first allocation plan is generated. For the first allocation plan, obtain the matching degree of each task priority and time period to determine whether it meets the preset allocation balance threshold and obtain the allocation balance evaluation result. According to the allocation balance evaluation result, if the matching degree is lower than the preset threshold, adjust the weighting factor and recalculate the allocation data to obtain the optimized allocation plan. Through the optimized allocation plan, determine the final personnel and task matching relationship and generate the allocation result including time period and task priority.

[0008] Preferably, the step of obtaining the capability parameter vectors and workload distribution ratios of other technicians, generating a list of substitute personnel closest to the capability parameter vectors of the unavailable personnel, and obtaining the second distribution plan comprises: The capability parameter vectors of unavailable technicians and the capability parameter vectors of all technicians and the workload distribution ratio are extracted from the multi-user schedule database to obtain the initial personnel data set. The similarity between the capability parameter vectors of each technician in the initial personnel data set and the capability parameter vectors of unavailable technicians is calculated by cosine distance to obtain a similarity value set, where the cosine distance formula is: cos(θ)=(A·B) / (||A||·||B||), A is the capability parameter vector of unavailable technicians, B is the capability parameter vector of other technicians, ||A|| and ||B|| represent the modulus of vector A and vector B respectively, and the similarity value set is screened according to the preset threshold. If the similarity value is large, the similarity value set is filtered. If the workload distribution ratio is lower than the preset maximum load threshold, the technician will be retained to obtain an optimized substitute list. The optimized substitute list is sorted in descending order according to the similarity value in the optimized substitute list to obtain a sorted substitute sequence. The sorted substitute sequence is combined with the task data corresponding to the unavailable technician to generate a second allocation plan, determine the final substitute allocation result, and update the workload distribution ratio of the technician in the second allocation plan from the multi-user schedule database to obtain an updated schedule data set.

[0009] Preferably, the step of sorting the substitutes by combining the task priority weighting factor and the time constraint check to obtain the third allocation scheme includes: Obtain the list of substitute personnel, task priority weighting factors and time constraints from the second allocation plan, use the weighted sorting algorithm to calculate the personnel priority and obtain the initial priority ranking. Based on the initial priority ranking, obtain the workload data of each substitute personnel and calculate the load ratio. If the load ratio exceeds the upper limit threshold, eliminate the corresponding substitute personnel to obtain the updated priority ranking. From the updated priority ranking, obtain the personnel's time constraints to determine whether the task time requirements are met. If not, adjust the personnel order to obtain the adjusted priority ranking. Based on the adjusted priority ranking, use the allocation algorithm to generate the third allocation plan. Obtain the allocation result from the third allocation plan to determine whether there is a load conflict. If so, recalculate the load ratio and eliminate overloaded personnel to obtain the optimized allocation plan. Based on the optimized allocation plan, generate the final personnel allocation list and determine the third allocation plan.

[0010] Preferably, the time constraint check and capability parameter vector matching requirements for each task and determining the task-personnel mapping table include: By parsing the third allocation plan, obtaining the task time constraints and capability parameter vectors, and determining the basic allocation data, if the task time constraints do not match the capability parameter vectors, adjusting the task priority, and obtaining the verified task sequence, a greedy algorithm is used to allocate technicians based on the verified task sequence and capability parameter vectors, and a preliminary allocation result is determined. Based on the preliminary allocation result, a time schedule is generated, and time conflicts are determined to obtain an optimized schedule. Based on the optimized schedule and the preliminary allocation result, a fourth allocation plan is generated to determine the task allocation details, and the correspondence between tasks and technicians is extracted from the fourth allocation plan to generate a task-personnel mapping table. Based on the task-personnel mapping table, the matching degree between the allocation plan and the capability parameter vector is verified to determine the final allocation result.

[0011] Preferably, the acquiring of the load change trend of the technicians after assignment and reallocating the tasks to generate the adjusted fifth assignment plan comprises: Obtain the technician load change trend from the task mapping table, calculate the deviation between the daily task duration of each technician and the historical average, and obtain the load change trend. According to the load change trend and the preset workload upper limit, calculate the distribution ratio of each technician's current workload and determine the distribution ratio data. If the distribution ratio of a technician exceeds the dynamic threshold, adjust the task distribution through the linear programming algorithm to generate a temporary task distribution plan. Extract the technician load data after task reallocation from the temporary task distribution plan, calculate the load balance index, and obtain the balance evaluation value. Compare the balance evaluation value with the balance of the historical distribution plan. If the current balance is higher than the historical average, use the current plan to generate the fifth distribution plan, extract the task distribution details of each technician from the fifth distribution plan, update the task mapping table, and obtain the optimized distribution result. Based on the optimized distribution result, calculate the load change trend of each technician, verify whether the dynamic threshold constraint is met, and confirm the final distribution plan.

[0012] It can be seen from the above technical solution that the present invention has the following beneficial effects: This hierarchical alarm processing and task scheduling method for a multi-user environment generates an initial personnel-task matching matrix by obtaining the technical personnel's capability parameter vector, time constraints and workload upper limit, combined with task priority. In response to time conflicts, the present invention uses a dynamic integration algorithm to calculate the priority weighting factor and generates a list of substitute personnel through cosine distance calculation. Tasks are assigned using a greedy algorithm and dynamically adjusted according to load change trends. The present invention also uses time series analysis to predict the stability of task assignments and optimizes substitute plans through batch comparison operations. Finally, the present invention generates a real-time task execution log and updates system status data, realizing intelligent assignment, dynamic adjustment and stability prediction of technical personnel tasks, thereby improving the efficiency and rationality of task assignment. BRIEF DESCRIPTION OF THE DRAWINGS

[0013] Figure 1 Flow chart of the method of the present invention. DETAILED DESCRIPTION

[0014] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0015] like Figure 1As shown, the present invention provides a technical solution: a hierarchical alarm processing and task scheduling method for a multi-user environment, comprising obtaining a technician's capability parameter vector, time constraint data and workload upper limit to obtain a candidate allocation scheme; calculating the technician's task priority weighting factor in different time periods, and if the technician's time constraint data conflicts with the task time requirement, marking it as unavailable, and obtaining a first allocation scheme after eliminating the conflict; obtaining the capability parameter vectors and workload distribution ratios of other technicians, generating a list of substitute personnel closest to the capability parameter vector of the unavailable personnel, and obtaining a second allocation scheme; combining the task priority weighting factor and time constraint check to sort the substitute personnel and obtain a third allocation scheme; determining a task and personnel mapping table based on the time constraint check and capability parameter vector matching requirements of each task; obtaining the load change trend of the technician after allocation, reallocating the tasks to generate an adjusted fifth allocation scheme; predicting the stability of task allocation in future time periods, and regenerating the substitute personnel list through batch comparison operations to obtain a final allocation scheme; combining the task priority weighting factor and capability parameter vector to generate a real-time task execution log and obtain system status data.

[0016] This invention achieves efficient matching between tasks and technicians by constructing a multidimensional scheduling model. First, technician capability parameters are recorded in vector form. Each technician's capability parameter consists of several dimensions, such as proficiency in different skill categories, and can range from zero to one. Based on the task requirements, the system also constructs a task parameter vector, representing the task's requirements across each capability dimension. To calculate the compatibility between a technician and a task, the system uses a cosine similarity algorithm. Specifically, the cosine similarity algorithm is used to perform a dot product between the technician and the task's capability parameter vectors, followed by dividing the result by the product of the two vectors' moduli. The closer the similarity value is to one, the higher the compatibility between the technician and the task. The task priority weighting factor takes into account the importance and time urgency of the task. First, each task is assigned a base priority coefficient. Then, a time decay parameter is introduced, which determines the changing trend of the task's priority over time. The further away the current time is from the task's start time, the lower the task's priority weighting value. The system then processes the base priority using an exponential decay function to obtain a weighted priority value for the current time period. To identify time conflicts, the system checks whether the task's time interval overlaps with a technician's available time slot. If there is no overlap, the technician is unable to complete the task within that time slot and is therefore marked as unavailable. When generating a replacement list, the system compares the capability gap between the unavailable technician and the other technicians. This gap is quantified by calculating the Euclidean distance between the two sets of parameter vectors, with smaller distances indicating closer capabilities. The system sorts the two sets of parameters in ascending order and selects the technician with the closest capability as a replacement. Scheduling stability is predicted by setting a sliding time window. The system monitors changes in the number of assigned tasks within this time window and calculates the change ratio of each task's assigned quantity within the time period. If the change ratio of a task exceeds a set threshold, it indicates that the task scheduling is unstable, and the system regenerates the replacement list to improve scheduling robustness. Finally, the system records real-time task execution logs, including fields such as task number, assignee, start time, end time, task status, priority factor, and capability match. This data serves as the core foundation for system status and is used for subsequent visualization analysis and scheduling optimization.

[0017] The hierarchical alarm processing and task scheduling method provided by the present invention has significant advantages in a multi-user environment. By introducing the capability parameter vector and the task priority weighting mechanism, a more accurate match between tasks and technicians is achieved, effectively improving the system's responsiveness to time and skill requirements, ensuring that high-priority tasks can be promptly processed by highly matched personnel; an automatic screening mechanism for substitute personnel is adopted to enhance the system's fault tolerance and scheduling continuity in the event of personnel unavailability; the matching algorithm is based on vector similarity evaluation, avoiding human intervention and improving the objectivity and scientific nature of task allocation; traceable log records are generated throughout the scheduling process, facilitating real-time monitoring of the system's operating status and task traceability management; by dynamically analyzing the load change trend of technicians and the volatility of task allocation, forward-looking adjustment and stability prediction of the task scheduling process are achieved, thereby further optimizing the overall operating efficiency and resource allocation effect of the system, with extremely high practical value and promotion prospects.

[0018] For example, a large manufacturing company has multiple maintenance and technical teams that need to respond to equipment alarms daily. After this invention is implemented on the company's scheduling platform, it automatically recommends the most suitable dispatcher based on the maintenance technician's skill level, shift schedule, and historical task records. If a technician takes a temporary leave or a faulty task times out, the system immediately activates a replacement mechanism to ensure uninterrupted task response. The platform's scheduling accuracy has increased from 75% before implementation to 93%, and the task overtime rate has dropped by approximately 40%, earning high praise from users.

[0019] The technical personnel's capability parameter vector, time constraint data and workload upper limit are obtained to obtain a candidate allocation plan, which includes extracting the technical personnel's capability parameter vector, time constraint data and workload upper limit from a schedule database, parsing multi-user data using structured query language to obtain an initial personnel data set, processing the capability parameter vector in the initial personnel data set by vector normalization, and generating a personnel and task matching matrix in combination with a task priority list to obtain an intermediate allocation plan including time availability and capability score. If the workload of a technician in the intermediate allocation plan exceeds the upper limit, the task allocation is adjusted by a linear programming algorithm to obtain an optimized candidate allocation plan. According to the optimized candidate allocation plan, the time availability of each technician is obtained, and the task execution time period is determined by a time window partitioning method. If there is a conflict between the task execution time period and the time constraint data, the task order is adjusted by priority sorting to obtain an allocation plan after the conflict is resolved. By comparing the allocation plan after the conflict is resolved with the capability score, a weighted average method is used to determine the final task allocation result, and the final task allocation result is obtained. An allocation record including technicians, tasks and time periods is generated and stored in the schedule database.

[0020] This implementation, based on structured scheduling modeling, automatically matches tasks and personnel, and dynamically optimizes scheduling by modeling and scoring technicians' capabilities, time resources, and task requirements. First, the system extracts information on all technicians to be scheduled from the company's internal schedule database. This information includes three key data types: capability parameters, representing the technician's proficiency in multiple skill categories; time constraints, reflecting the technician's available time within the current time period; and workload limits, such as the maximum number of tasks allowed per day or per week or the maximum total working hours. This data is extracted and organized using structured query language to form an initial personnel dataset. After obtaining the raw capability data, the system normalizes each technician's capability parameters to ensure that ratings across different personnel can be compared on the same scale. This normalization method involves subtracting the minimum skill score for each individual from the individual's skill score, dividing the score by the difference between the minimum and maximum skill scores. Ultimately, each capability score is mapped to a value between zero and one. The raw value of each capability score can be obtained from personnel assessment scores, training system records, and historical task completion results. The system then obtains a list of tasks to be scheduled. Each task also defines a list of task competency requirements, indicating the skill dimensions and corresponding skill intensity requirements required to complete the task. The system calculates the skill match between each technician and each task, creating a matching score matrix. This score is calculated by multiplying the person's standardized competency value by the task's skill requirements item by item, then dividing the sum by a factor that factors the combined competency data. The closer the result is to unity, the more closely the technician matches the task. Incorporating time constraints, the system generates an intermediate allocation plan, where each task is assigned to a number of highly compatible and available candidate technicians. However, because each technician's workload has an upper limit, the system must further determine whether the number of tasks or total work hours exceeds this limit. If so, the allocation plan is marked invalid, and the system proceeds to the next task rescheduling process. This rescheduling process optimizes the allocation between personnel and tasks with the goal of maximizing the overall matching score. This process constructs a linear optimization model to select the highest matching score from all combinations of technicians and tasks, ensuring that each task can only be performed by one technician and that the total number of tasks per technician does not exceed an upper limit. This optimization problem can be solved using a linear programming or integer programming solver. After optimization, the system further checks whether the technician's time slot conflicts with the scheduled task time. Each task has a start and end time, and each technician has a unique available time interval. If the task time does not overlap with the technician's available time, the task cannot be performed by that technician.In the event of a conflict, the system will reorder tasks based on priority, prioritizing higher-priority tasks to determine the conflict-resolved assignment. To balance task importance and skill match, the system uses a weighted average strategy to combine these scores. Specifically, the task priority value is combined with the employee skill match score in a certain proportion to produce a final score. Task priority is typically set by the business system and can be scored based on factors such as business urgency and customer impact, with a typical score range of one to five. The weighting is determined by the scheduling policy, indicating whether the system prioritizes the priority of the task itself or the skill match of the employee. Finally, the system selects the optimal person for each task based on the weighted score and generates a detailed task assignment record, including technician number, task number, task execution timeframe, task importance, and match score. All results are stored in the original schedule database for subsequent access and execution monitoring by the scheduling system.

[0021] This implementation method can significantly improve the automation and intelligence level of task allocation. By introducing vector normalization and matching scoring mechanisms, a more accurate assessment of the compatibility between tasks and personnel is achieved, solving the problem of subjective judgment bias in manual allocation. The introduction of linear programming models effectively balances resource utilization and load limits, avoiding scheduling failures caused by overload of individual personnel. The time window division and priority sorting mechanism improves the flexibility of task time matching, making task scheduling more reasonable and maximizing resource utilization. Finally, through the database write mechanism, the entire process of allocation records can be traced and audited, providing data support for subsequent scheduling optimization, performance evaluation and task backtracking, and significantly improving the overall scheduling efficiency and stability of the system.

[0022] For example, at an information technology outsourcing service company, client project teams often need to mobilize multiple technicians in a short period of time to handle alarm tasks for different client systems. After deploying the scheduling system of the present invention, the company extracts the available time, ability level, and historical task volume of hundreds of technicians through the schedule database. The system automatically completes task matching and load balancing to avoid duplicate dispatch or personnel conflicts. Optimal task scheduling is achieved through linear programming and a weighted scoring mechanism. More than 200 tasks can be automatically scheduled daily, the manual intervention rate has dropped by more than 70%, and the scheduling accuracy rate has stabilized at more than 95%, significantly improving customer satisfaction and system processing efficiency.

[0023] The method comprises calculating the task priority weighting factors of technicians in different time periods, marking the technicians as unavailable if their time constraint data conflicts with the task time requirement, and obtaining a first allocation plan after eliminating conflicts. The method comprises obtaining an initial personnel and task matching matrix, extracting technician, task priority, and time period information, generating preliminary allocation data including time constraints, using a dynamic integration algorithm, calculating the task priority weighting factors in each time period for the preliminary allocation data, obtaining a weighting factor set, and if the technician's time constraint conflicts with the task time requirement, marking the corresponding technicians as unavailable according to a conflict marking rule, and obtaining allocation data after eliminating conflicts. The first allocation plan is generated by combining the allocation data after eliminating conflicts with the weighting factor set, obtaining a matching degree between each task priority and time period for the first allocation plan, determining whether a preset allocation balance threshold is met, and obtaining an allocation balance evaluation result. Based on the allocation balance evaluation result, if the matching degree is lower than the preset threshold, adjusting the weighting factor, and recalculating the allocation data to obtain an optimized allocation plan. The optimized allocation plan is used to determine the final personnel and task matching relationship, and generate an allocation result including time period and task priority.

[0024] This implementation aims to address the issue of task allocation and time conflicts in a multi-user environment. By establishing a time constraint and priority weighting model, it achieves an efficient and executable task scheduling mechanism. The core process includes priority factor calculation, time conflict determination, conflict data elimination, match balance determination, and dynamic factor adjustment. First, the system obtains a "personnel-task matching matrix," which consists of three types of information: technician identification (i.e., personnel number or ID), task priority value, and task execution time period. Task priority is a parameter that indicates the importance and urgency of the task, typically ranging from 1 to 5, with 1 being the lowest priority and 5 being the highest. Priority values can be set through business rules, alarm levels, service level agreement (SLA) terms, manual expert evaluation, or directly provided by the task registration module in enterprise information systems (such as CRM and ERP). Task execution time periods are typically expressed as start and end times, for example, "April 11, 2025, 9:00 AM to 11:00 AM." The technician's available time periods are also described as start and end times and are derived from the human resources management system or schedule management module. The system compares the task timeframe with the technician's available timeframe to see if there's any overlap. If there's no overlap, the technician is unavailable to perform the task. A preliminary allocation dataset is constructed, combining the three types of information mentioned above to form the basis for task scheduling. Subsequently, a dynamic integration algorithm is used to aggregate the set of tasks within each timeframe and calculate a task priority weighting factor for that timeframe. This weighting factor represents a composite of the concentration and urgency of tasks within a given timeframe. It is calculated by summing the priority values of all tasks within that timeframe and multiplying it by a value called the "task urgency decay parameter." The task urgency decay parameter is a constant that controls the rate at which task priorities decrease over time. For example, setting this parameter to 0.05 means that the further away from the task start time the current time is, the faster the task's weight decreases. This parameter is typically set by the system administrator during initialization, but can also be determined through modeling of historical task response data, such as by fitting task delay rates within different time windows. The system matches the task timeframe with the technician's available timeframe. If there's no overlap, the technician is marked as "unavailable" for that task according to the conflict marking rules. The marked information is eliminated to form the allocation data after conflict elimination. Next, the system combines the "weighted factor set" with the conflict elimination data set to generate the first round of allocation plan. On this basis, the system further calculates the matching degree, that is, the degree of adaptability of the task to the personnel. The matching degree can be composed of two parts: one is the aforementioned task priority weighting factor, and the other is the matching score between the personnel and the task on the capability vector. The two can be integrated into the final score through the weighted average method. The weighted average ratio parameter is used to set the weight relationship between the priority factor and the capability score. For example, setting it to 0.7:0.3 means that the system pays more attention to the urgency of the task. This ratio can be configured by the administrator and can also be adjusted dynamically according to the nature of the task.The system performs a statistical analysis of the matching degree of all tasks to determine whether a preset allocation balance threshold has been met. This threshold defines the minimum average matching degree required by the system. For example, a setting of 0.75 indicates that an average matching score below 75% is considered poor scheduling performance. This threshold can be set empirically or obtained through training of a scheduling strategy model. If the matching degree falls below the set value, the system triggers a weighting factor adjustment mechanism. This mechanism adjusts task priority factors based on comprehensive indicators such as historical task completion time, complaint records, and timeout frequency, increasing the urgency of high-risk tasks and reducing the scheduling priority of non-critical tasks. The adjusted weighting factors are then re-integrated into the scheduling calculation to produce an optimized scheduling result. Ultimately, the system generates a complete task scheduling dataset that clearly identifies the technician number, task number, task execution time period, and task priority level, and stores it in the system database.

[0025] By introducing a task priority weighting mechanism and a dynamic conflict detection strategy, this invention achieves precise task scheduling and personnel screening in multiple time periods, greatly improving the system's ability to adapt to task urgency and time resources. The technician availability detection mechanism ensures the executability of scheduling, avoiding situations where tasks cannot be implemented due to resource conflicts after assignment; the real-time dynamic adjustment strategy of the weighting factor enables active intervention in task backlogs and resource overloads, ensuring the continuous balance of scheduling and service stability at the system level. The matching evaluation and allocation balance judgment mechanism further enhances the visualization and intelligence of the allocation process, making scheduling decisions more data-based and with more adjustable control logic.

[0026] For example, a national chain of on-site operation and maintenance service companies needs to handle more than a thousand equipment alarm tasks every day, and the distribution of tasks involves multiple time zones and cities. After deploying the method of the present invention, the scheduling system can identify the business priority of the task and the deployability of technical personnel in real time, and can still maintain a task success matching rate of more than 95% during the task outbreak period. The system eliminates technical personnel who do not meet the time conditions through conflict detection, and dynamically adjusts the task priority weights so that tasks can be automatically assigned to the most suitable personnel. The system successfully supported overload operation for 7 consecutive days during peak periods such as the Spring Festival and Double Eleven, and the timely completion rate of tasks increased by 22%, significantly improving customer satisfaction and corporate operating efficiency.

[0027] Obtain the capability parameter vectors and workload distribution ratios of other technicians, generate a list of substitute personnel closest to the capability parameter vector of the unavailable technician, and obtain a second allocation plan, including extracting the capability parameter vector of the unavailable technician and the capability parameter vectors and workload distribution ratios of all technicians from a multi-user schedule database, obtaining an initial personnel data set, and calculating the similarity between the capability parameter vector of each technician in the initial personnel data set and the capability parameter vector of the unavailable technician using cosine distance to obtain a set of similarity values, where the cosine distance formula is: cos(θ)=(A·B) / (||A||·||B||), where A is the capability parameter vector of the unavailable technician, B is the capability parameter vector of the other technicians, and ||A|| and ||B|| represent the moduli of vectors A and B, respectively. The similarity value set is screened according to a preset threshold. If the similarity value is greater than the preset threshold, the corresponding technician is included in the candidate replacement list to obtain a preliminary replacement set. The workload distribution ratio of each technician is obtained from the preliminary replacement set. If the workload distribution ratio is lower than the preset maximum load threshold, the technician is retained to obtain an optimized replacement list. The optimized replacement list is sorted in descending order according to the similarity value to obtain a sorted replacement sequence. The sorted replacement sequence is combined with the task data corresponding to the unavailable technician to generate a second allocation plan and determine the final replacement allocation result. The workload distribution ratio of the technician in the second allocation plan is updated from the multi-user schedule database to obtain an updated schedule dataset.

[0028] This implementation primarily enables rapid screening of replaceable technicians based on ability similarity and current workload levels, and completes replacement task assignments to ensure uninterrupted tasks. First, the system extracts two types of data from the multi-user schedule database: the ability parameter vectors of technicians currently in the "unavailable" state, and the ability parameter vectors of all technicians in the current system and their corresponding workload distribution ratios. The ability parameter vectors reflect the technician's ratings across various skill dimensions. The scoring criteria can be set based on an internal evaluation system or historical task completion performance, and are typically decimals between 0 and 1. Next, the system calculates the similarity between the unavailable technician and the other technicians' ability vectors. This similarity is calculated using the cosine similarity method: First, the scores of the two technicians across all skill dimensions are multiplied and summed to obtain the dot product of the vectors. Each item in the two ability vectors is then squared, summed, and squared to obtain the modulus of the two vectors. Finally, the dot product is divided by the product of the two moduli to obtain a similarity value between 0 and 1. The closer the value is to 1, the closer the two technicians' abilities are. The system then combines the similarity values of all technicians with the unavailable technician into a single similarity set. Next, a similarity threshold parameter is set, for example, 0.8, indicating that only candidates with a similarity of at least 80% are considered suitable replacements. This threshold is set by the system administrator or a policy model, and can also be optimized through learning based on historical replacement success rates. Technicians who meet the similarity criteria are included in a preliminary pool of replacements. The system then extracts each candidate's current workload allocation ratio from this pool. This ratio is the ratio of their current workload to their maximum allowable workload. For example, if a technician has 5 tasks assigned to a maximum of 10, the allocation ratio is 50%. If a technician's workload ratio falls below the preset maximum workload threshold, for example, 80%, indicating they are still capable of taking on new tasks, they will be retained in the optimized replacement pool. The system then sorts the optimized replacements in descending order based on their similarity to the unavailable personnel, forming a ranked replacement sequence. The system then combines the unavailable personnel's task data (such as task priority and time period) to assign tasks to the most suitable replacements, thereby constructing a second allocation plan. This plan ensures that each task is performed by the person with the best matching capabilities and a moderate workload. Finally, the system writes the task allocation in the second allocation plan into the database, updates the workload distribution ratio of each substitute, forms a new schedule data set, and provides data support for subsequent scheduling.

[0029] By introducing a capability similarity screening mechanism and a load ratio control mechanism, this invention enables efficient backup scheduling in the event of personnel unavailability. This not only ensures the continuity and accuracy of task execution, but also minimizes the overload of personnel resources. Using cosine similarity as a basis for capability assessment provides a quantifiable and interpretable basis for backup selection; while the load threshold-based screening logic effectively prevents system bottlenecks caused by the concentration of tasks on a small number of people, thereby improving the load balancing and resource utilization efficiency of the overall scheduling system.

[0030] A power operations and maintenance company has technical response stations in multiple locations, responding to numerous daily alerts. When a technician in one area is temporarily absent for some reason, the dispatch system immediately identifies personnel from other areas with similar expertise and less than 80% of the workload based on their skill profile and current task data. The system then automatically generates a replacement task allocation plan. Emergency replacements, which previously required manual coordination of over 30 minutes, now complete the dispatch within 3 seconds and simultaneously update the database, significantly improving emergency response capabilities and the intelligence of operations and maintenance scheduling.

[0031] Combined with the task priority weighting factor and time constraint verification, the substitute personnel are sorted to obtain a third allocation plan, which includes obtaining a list of substitute personnel, task priority weighting factors and time constraints from the second allocation plan, using a weighted sorting algorithm to calculate the personnel priority, and obtaining an initial priority ranking. Based on the initial priority ranking, the workload data of each substitute personnel is obtained, and the load ratio is calculated. If the load ratio exceeds the upper limit threshold, the corresponding substitute personnel is eliminated to obtain an updated priority ranking. From the updated priority ranking, the time constraints of the personnel are obtained to determine whether the task time requirements are met. If not, the personnel order is adjusted to obtain an adjusted priority ranking. Based on the adjusted priority ranking, the allocation algorithm is used to generate a third allocation plan. The allocation result is obtained from the third allocation plan to determine whether there is a load conflict. If so, the load ratio is recalculated and overloaded personnel are eliminated to obtain an optimized allocation plan. Based on the optimized allocation plan, a final personnel allocation list is generated to determine the third allocation plan.

[0032] The primary goal of this implementation is to optimize task allocation by further considering task priority, workload, and time constraints based on the selected replacements, ensuring efficient and executable resource allocation. First, the system extracts the list of available replacements, along with each task's weighted priority factor and relevant time constraint information from the second allocation plan. A task's weighted priority factor is calculated by multiplying its original priority by its time urgency. Urgency is controlled by a time decay factor, which is set by the scheduling policy and typically ranges from 1% to 10%, with higher values indicating more urgent tasks. The system calculates a ranking score for each replacement, taking into account the task priority factor and the technician's current workload. Specifically, the task priority weighting factor is multiplied by a set task importance weight, and then the technician's current workload ratio is subtracted by a load suppression factor. The weighting ratio can be set to, for example, 7:3 or 6:4, depending on the administrator's task characteristics. The resulting score is used to preliminarily rank all replacements, resulting in an initial priority list. Next, the system obtains the current workload ratio for each candidate in the prioritized list. This ratio is calculated by dividing their current number of assigned tasks by their maximum allowed number of tasks. If a candidate's workload ratio exceeds a system-defined upper threshold (e.g., 80%), they are removed from the prioritized list, forming an updated prioritized list. The system then re-reads the time constraints of each replacement candidate from this updated list, namely their available time periods. The system checks whether these time periods overlap with the task execution timeframe. If not, the candidate's position in the list is adjusted downward, lowering their priority. This process forms an adjusted prioritized list that reflects the overall availability of each candidate. Based on this ranking result, the system applies a task allocation algorithm. This algorithm assigns tasks to the technician with the lowest workload and the best time match, following the priority of the candidate, to generate a third allocation plan. The system then performs a workload conflict check on the allocation results. If a candidate's workload ratio exceeds the limit (e.g., exceeding 100%) after task assignment, the candidate is removed from the list and the workload is redistributed and ranked among the remaining candidates, forming an optimized allocation plan. Finally, the system generates a final personnel allocation list based on the optimization results. The list includes the task number, personnel number, task priority, expected start and end time, and current load ratio, and is written into the system database as the third allocation plan.

[0033] This implementation introduces task priority factors and a dynamic sorting mechanism, combined with time matching and load constraints, to construct a scheduling and sorting logic based on multiple factors. This effectively avoids the scheduling error of "capacity matching only," improving the overall rationality of task execution and the balance of system resource utilization. Furthermore, through a progressive screening and allocation optimization process, automatic backtracking and error correction capabilities are implemented during the scheduling process, making the scheduling solution more stable and practical.

[0034] For example, a smart building equipment operation and maintenance platform received a large number of alarm requests during the holidays, with several frontline technicians marked as unavailable due to vacation. The system constructed a pool of replacement candidates based on historical performance and competency data. It then automatically prioritized the candidates based on task urgency, the individual's current workload, and their shift schedule, eliminating overloaded and conflicting personnel. The final assignment list was automatically sent to personnel terminals, improving overall response efficiency by 32% and reducing holiday overtime by 46%, ensuring stable facility operation and service quality.

[0035] According to the time constraint verification and capability parameter vector matching requirements of each task, the task and personnel mapping table is determined, including parsing the third allocation plan, obtaining the task time constraint and capability parameter vector, determining the basic allocation data, if the task time constraint does not match the capability parameter vector, adjusting the task priority, obtaining the verified task sequence, using a greedy algorithm, according to the verified task sequence and capability parameter vector, assigning technical personnel, determining the preliminary allocation result, generating a time arrangement based on the preliminary allocation result, judging time conflicts, obtaining an optimized timetable, generating a fourth allocation plan based on the optimized timetable and the preliminary allocation result, determining the task allocation details, extracting the correspondence between tasks and technical personnel from the fourth allocation plan, generating a task-personnel mapping table, verifying the matching degree of the allocation plan with the capability parameter vector for the task-personnel mapping table, and judging the final allocation result.

[0036] This implementation aims to build on the third allocation scheme by combining task execution time constraints with personnel capability requirements to generate a final matching mapping table between tasks and personnel, ensuring the feasibility and rationality of task scheduling. First, the system analyzes the third allocation scheme and extracts two core pieces of information from each task: the time constraint and the capability parameter vector requirement. The time constraint is typically expressed as start and end times and is derived from alarm records or scheduling plans. The capability parameter vector represents the personnel capability requirements for the task across several skill dimensions. For example, a task might require "device diagnostic capability ≥ 0.8, communication protocol familiarity ≥ 0.6." These values are modeled based on the system's skill labels for the task type or extracted based on historical successful matching experience. The system then performs a capability time compatibility check on all tasks. Specifically, each technician's capability vector is compared with the task requirements. If a task's capability requirement for a particular skill dimension exceeds the technician's ability score for the corresponding dimension, the system marks the task-person combination as "mismatched." In this case, the system adjusts the task's priority and postpones its scheduling appropriately to allow a more suitable technician to handle it first. Priority adjustments can be weighted downward based on the degree of skill gap, for example, reducing the priority by one level for every 0.1 point difference. After completing the priority adjustment, the system uses a greedy algorithm to assign technicians to the verified task sequence. This algorithm sorts tasks from highest to lowest priority and searches for the first technician who meets the skill requirements and has an available time slot. After completing the initial assignment, the system constructs a preliminary scheduling plan, aligning the start and end times of each task with the technician's schedule. If a particular technician is assigned multiple tasks with overlapping times, the system identifies a time conflict. Conflict detection uses time slot overlap logic: if the time intervals of two tasks overlap, a conflict is identified. To resolve conflicts, the system prioritizes the higher-priority task and attempts to postpone the conflicting task to the next available time slot or assign it to the next most compatible technician, thereby constructing an optimized schedule. This schedule is used to generate a fourth assignment plan, further clarifying the task execution time and performer. The system extracts the task-personnel mapping from the fourth assignment plan to create a task-personnel mapping table. Each task is associated with a personnel number and an execution time slot. To verify the rationality of the assignment, the system performs a capability match check on the mapping table. This is done by calculating the similarity between the assigned personnel's capability vector and the task requirement vector, and then calculating the average match for the overall scheduling. If the match falls below a set threshold (e.g., 75%), the assignment is marked as "low quality," and the system proceeds to the next round of optimization.

[0037] In this embodiment, core parameters include a task capability parameter vector, a technician capability parameter vector, a task time constraint, a capability matching threshold, a task priority value, a priority adjustment factor, a greedy task assignment algorithm order, a time conflict judgment parameter, and a task-personnel mapping matching score. The task capability parameter vector represents the skill requirements for each task. Each skill ranges from 0 to 1 and is typically determined through task modeling or historical data mining, reflecting the skill requirements of the technician. The technician capability parameter vector describes the technician's ability level in different skill dimensions. The parameter value is typically dynamically adjusted based on performance evaluations, training records, and task completion feedback. The task time constraint defines the start and end times of the task, typically provided by the alarm system or customer SLA, and influences the priority of task scheduling. The capability matching threshold represents the minimum match between the task requirements and the technician's capabilities. It is typically set by the system or obtained through historical data regression analysis to ensure the rationality of task allocation. The task priority value measures the urgency of the task and is typically determined by factors such as the alarm level and customer demand. Tasks with higher priority are scheduled first. The priority adjustment factor controls the degree to which a task's priority decreases when a task can't be fully matched. This factor can be dynamically adjusted based on the task's urgency and matching gaps. Task allocation uses a greedy algorithm to assign tasks to the appropriate personnel based on the adjusted priority. The time conflict judgment parameter is used to detect conflicts in task time intervals. If there is overlap in task time, a conflict is considered to exist. Finally, the task-personnel mapping match score is used to measure the match between tasks and personnel. The system evaluates the match by calculating the similarity between the task capability requirement vector and the assigned personnel capability vector, ensuring that tasks are assigned to the most suitable personnel on time and efficiently.

[0038] This implementation integrates three scheduling dimensions: task capability requirements, time resource constraints, and task priority, achieving precise matching of tasks and personnel under multiple constraints. The introduction of a capability mismatch optimization mechanism and a greedy scheduling algorithm improves the speed and feasibility of task scheduling. Time conflict detection and dynamic scheduling significantly reduce duplicate and conflicting scheduling, improving overall scheduling success rates and on-time task completion rates.

[0039] For example, in an intelligent manufacturing equipment monitoring platform, hundreds of alarm events need to be responded to every day. The tasks have significant differences in skills and are time-sensitive. After the system deploys the method of the present invention, it can automatically select the most suitable and free personnel from multiple technical teams to perform the tasks based on the required technical type and response time window for each alarm task, and automatically postpone or reallocate some tasks based on scheduling conflicts. Through the capability matching and conflict handling mechanism, the task matching success rate has increased from 78% to 91%, while avoiding the previous situation of more than 10 tasks being delayed per day, greatly improving operation and maintenance efficiency and system stability.

[0040] Obtaining the load change trend of the technicians after assignment, and reallocating tasks to generate an adjusted fifth assignment plan includes obtaining the load change trend of the technicians from the task mapping table, calculating the deviation of the daily task duration of each technician from the historical average, and obtaining the load change trend. According to the load change trend, combined with the preset workload upper limit, the distribution ratio of the current workload of each technician is calculated, and the distribution ratio data is determined. If the distribution ratio of a technician exceeds the dynamic threshold, the task assignment is adjusted through a linear programming algorithm to generate a temporary task assignment plan. The load data of the technicians after task reallocation is extracted from the temporary task assignment plan, the load balance index is calculated, and a balance evaluation value is obtained. The balance evaluation value is compared with the balance of the historical assignment plan. If the current balance is higher than the historical average, the current plan is used to generate the fifth assignment plan, and the task assignment details of each technician are extracted from the fifth assignment plan. The task mapping table is updated to obtain the optimized assignment result. Through the optimized assignment result, the load change trend of each technician is calculated to verify whether the dynamic threshold constraint is met and confirm the final assignment plan.

[0041] This implementation dynamically analyzes technician workload trends, combining task priorities, time constraints, and personnel capacity allocation to optimize task allocation. The system's core goal is to ensure each technician's workload is appropriate, avoiding overload while ensuring tasks are completed on time and efficiently. First, the system retrieves each technician's workload change data from the task mapping table. A technician's workload change is calculated by comparing the deviation between the technician's current task duration and their historical average. This analysis determines whether a technician's workload has significantly changed. If a technician's task duration deviates significantly from their historical average, this indicates a workload change, which the system tracks and takes necessary adjustments. Next, the system calculates each technician's workload ratio based on their current task duration. The workload ratio represents the ratio of a technician's current workload to their maximum workload capacity. If a technician's workload ratio exceeds a set dynamic threshold (for example, exceeding 80% of their maximum workload), the system triggers an adjustment mechanism to prevent that technician from taking on additional tasks. If a technician's workload exceeds the threshold, the system uses a linear programming algorithm to reallocate tasks. The linear programming algorithm aims to optimize task allocation and maximize task completion efficiency while ensuring that each technician's load does not exceed the required limits. This adjustment algorithm ensures that all tasks are appropriately assigned to qualified technicians, maximizing overall scheduling efficiency. After task reallocation, the system generates a temporary task allocation plan and reassesses the load balance based on each technician's workload. The load balance index measures the balance between tasks assigned to each technician; a value closer to 1 indicates a more balanced load distribution. The system calculates the current load balance index and compares it with the load balance index of historical task allocation plans. If the current plan's load balance index exceeds the historical average, indicating a more balanced allocation, the system adopts this plan as the new task allocation plan. Finally, the system generates a fifth allocation plan based on the optimized schedule and the adjusted task allocation plan. This plan incorporates the execution time and performer for each task, ensuring that tasks can be completed within the specified timeframe and that each technician's load remains within an acceptable range. The system also updates the task mapping table to reflect the latest mapping between tasks and technicians. Finally, the system continues to monitor the technicians' workload trends based on the adjusted allocation plan, ensuring that the technicians' workload consistently meets dynamic thresholds during task execution. If the workload exceeds the preset limit, the system automatically adjusts again to ensure balanced and sustainable task allocation.

[0042] This implementation ensures the rational and efficient allocation of tasks to appropriate technicians by dynamically adjusting task allocation, optimizing load balancing, and adjusting task allocation plans using linear programming, thus avoiding issues such as overload and task delays. This method ensures the stability and efficiency of the scheduling system and optimizes resource utilization by analyzing technician load trends, historical task data, and load balancing in real time.

[0043] For example, at a global technical support company, multiple technical teams needed to respond to customer equipment troubleshooting within varying timeframes. The system intelligently adjusted the workload based on each technician's task duration and load trends, avoiding overload on individual teams during peak periods and successfully improving timely task completion rates. By calculating load balancing, the system ensured balanced task distribution, resulting in an 18% increase in overall task completion rates and a reduction in task delays.

[0044] Predicting the stability of task allocation in future time periods, regenerating a list of substitute personnel through batch comparison operations, and obtaining a final allocation plan includes using a time series analysis algorithm to extract features from historical task allocation data, generating a prediction result of task allocation stability in future time periods, and obtaining an allocation stability trend. If the allocation stability trend shows that a task has a risk of failing time constraint verification, then the risk identification mechanism is used to extract the failed task identifier from the prediction result to obtain a risk task set. For the risk task set, pre-established personnel resource pool data is obtained, and batch comparison operations are used to screen candidate personnel that meet the task requirements to generate a list of substitute personnel. According to the task priority sorting rules, personnel attributes are extracted from the substitute personnel list, and dynamic weights are calculated based on the priority to obtain an adjusted weight allocation matrix. The dynamic weight allocation matrix is used to recalculate the matching degree between tasks and personnel, generate a set of candidate allocation plans, and obtain an optimized allocation plan set. If there are multiple plans in the candidate allocation plan set, the time constraint verification algorithm is used to verify the time feasibility of the plans one by one to determine the final allocation plan. The final allocation plan is adopted, the task allocation database is updated, and a stable task allocation result is generated and output to the system log.

[0045] First, the system analyzes historical task allocation data using a time series analysis algorithm to extract characteristic information such as task execution time, task type, task completion time, and the personnel assigned to the task. Using these characteristics, the system predicts the stability of task allocation in future time periods. This stability prediction reflects the trend of task allocation and helps the system identify potentially risky tasks. Specifically, the system uses regression analysis on historical data to identify regularities and changing trends in tasks across different time periods. Based on this historical data, the system builds a predictive model to predict whether task allocation will be stable within a specific time period. If the prediction results indicate that a task may not be completed on time or faces the risk of a time constraint violation, the system marks the task as "risky." If the prediction results indicate that a task's time constraints pose a risk of failure, the system uses a risk identification mechanism to extract a failure indicator for the task. These indicators are typically derived from time constraint verification and task priority matching analysis. A task is marked as risky if its time requirement overlaps with the personnel's available time or if its priority is poorly aligned with the currently assigned personnel's capabilities. The system collects all tasks marked as "risky" into a risky task set. This collection includes information such as the task ID, time constraints, priority, performer, and predicted risk of failure. Next, for each risky task, the system selects candidate candidates from the personnel resource pool that meet the task requirements. The personnel resource pool contains detailed information about each technician, such as their skills, availability, historical workload, and task completion status. The system uses batch comparisons to select candidates based on the following criteria: skill match (whether the technician's skills meet the task requirements); availability (whether the technician has free time during the task execution period); and workload (whether the technician is currently engaged in other tasks and can complete the new task within the specified timeframe). Using these metrics, the system selects technicians who meet the task requirements and generates a list of candidate candidates. For each selected candidate, the system extracts personnel attributes from the list based on task priority sorting rules and combines them with the task priority to calculate a dynamic weight for each candidate. This dynamic weight calculation takes into account the match between the task priority and the technician's capabilities. Specifically, the higher the match between the task priority and the technician's capabilities, the greater the dynamic weight assigned to the technician. For example, if a task has a high priority and is highly compatible with a technician's skills, that technician will be assigned a higher weight. If a task has a low priority or the technician's skills don't fully meet the task requirements, the technician's weight will be lowered accordingly. By calculating dynamic weights, the system generates a weight allocation matrix and uses it to calculate the match between tasks and personnel. This matrix shows each candidate's priority in task allocation and reflects how well the technician's workload matches the task requirements.Each task is assigned to a suitable individual based on their matching degree, generating a set of candidate assignments. If multiple candidate assignments exist, the system will perform a time constraint check on each one. This process verifies whether the execution time of each task matches the technician's available time. The system verifies each assignment individually to ensure that each task can be completed on time. If a task's time constraint cannot be met, the system adjusts the assignment and searches for a new suitable technician until all tasks' time constraints are met. After time constraint verification, the system determines a final assignment that includes the final matching of all tasks and technicians. This final assignment selects the most suitable individual for each task by analyzing factors such as task priority, technician skill matching, time constraints, and workload. Finally, the system updates the task assignment database with the final assignment, ensuring accurate recording of task execution information. This information is also output to the system log for easy tracking and querying.

[0046] This implementation combines time series analysis with risk identification mechanisms to predict the stability of task allocation and automatically identify potentially risky tasks, effectively reducing conflicts and delays in task allocation. The system dynamically adjusts the allocation weights of backup personnel to ensure that tasks are appropriately assigned to the most appropriate technicians, improving the efficiency and accuracy of task allocation. Automated time constraint verification and prioritization optimize resource utilization, avoid overload and task mismatches, and enhance the overall system's responsiveness and stability.

[0047] For example, the scheduling system of a global technical support company handles thousands of customer requests daily, including equipment failures, network issues, and system maintenance. One day, an urgent equipment repair task was marked as the highest priority and required specific skills, such as network troubleshooting. However, the technician assigned to this task was already assigned to other tasks and could not complete it within the stipulated six hours. Using time series analysis and risk identification mechanisms, the system predicted potential instability in task allocation, automatically marked the task as risky, and initiated a search for a replacement. The system then conducted a batch comparison operation to select several suitable candidates from the resource pool based on the technician's skills, availability, and current workload. Ultimately, the system selected a technician who possessed the necessary skills and was available within the task's execution timeframe. After prioritization and dynamic weighting adjustments, the technician was assigned to the task and successfully completed it within six hours. This system not only prevented task delays but also optimized the technician's workload and improved task response speed. Customer satisfaction has significantly improved, and through this intelligent scheduling, technical support companies are able to manage resources more efficiently, reduce labor costs and task backlogs, and ensure service quality.

[0048] Combined with the task priority weighting factor and the capability parameter vector, a real-time task execution log is generated to obtain system status data, including the task priority and weighting factor calculation, obtaining the task sorting result, and determining the task allocation plan. The calculation formula is: S=P×W+C, where S is the task score, P is the task priority, W is the priority weight, and C is the capability parameter. After sorting, the task allocation plan is obtained. According to the task allocation plan, a real-time task execution log is generated. The task identifier, execution time, and technician identifier are extracted from the allocation plan. A log generation algorithm is used to output a structured execution log and store it in a multi-user schedule database. If an execution log is generated, the time constraint data in the multi-user schedule database is updated. The task start time and end time are obtained from the execution log, the technician time occupancy is calculated, and the time constraint field corresponding to the technician in the database is updated to obtain the updated time constraint data. The technician capability parameter vector is adjusted according to the time constraint data, and the occupied time ratio is extracted from the time constraint data. Combined with the task complexity, vector operation is used to update the capability parameter vector. The specific formula is: C'=C×(1-T), where C' is the adjusted capability parameter, C is the capability parameter, and T is the time occupancy ratio. The adjusted capability parameter vector is obtained. If the capability parameter vector is adjusted, the task allocation plan is recalculated, and the technician skill data is extracted from the adjusted capability parameter vector. Combined with the task priority and weighting factor, the task scoring formula is re-executed to obtain a new task allocation plan. According to the new task allocation plan, the system status data is updated, and the task status, technician status and time constraint data are obtained from the task allocation plan. The state aggregation algorithm is used to generate system status data and store it in a multi-user schedule database to obtain the latest system status data. Real-time task monitoring records are generated through the system status data, and the task execution progress and technician occupancy status are extracted from the system status data. The log generation algorithm is used to output real-time task monitoring records and store them in a multi-user schedule database to obtain task monitoring results.

[0049] This implementation combines task priorities, weighting factors, and technician capability parameter vectors to generate real-time task execution logs and system status data to optimize task scheduling and resource management. The entire process, consisting of task scoring, task assignment, execution log generation, and capability parameter adjustment, ensures accurate and timely task assignment, improving the system's responsiveness and task monitoring capabilities. The system first calculates a score for each task based on the task priority, weighting factors, and capability parameters. It then generates a task assignment plan based on the task scores. The task scoring formula is: S = P × W + C, where S is the task score, P is the task priority, W is the priority weight, and C is the capability parameter. Task priority P indicates the urgency of the task, typically determined by customer requirements or alarm levels. Priority weight W adjusts the influence of task priority in the scoring. It is typically set by the system, and its value range can be dynamically adjusted based on the task type. Capability parameter C represents the technician's skill level and reflects their ability to perform a particular task. Using the calculated task scores, the system ranks tasks and determines a task assignment plan based on the ranking results. This ranking ensures that the most urgent and appropriate tasks are assigned first to the appropriate technicians.

[0050] Based on the task allocation plan, the system generates a real-time task execution log. This log records the execution status of the task, including information such as task ID, execution time, and technician ID. The algorithm that generates the log extracts the task ID, technician ID, and time information and stores them in a structured manner in a multi-user schedule database. Task ID: A unique identifier for each task, used to track the progress and execution of the task; Execution Time: Records the start and end time of the task, used for subsequent calculation of the time occupancy and progress of the task; Technician ID: Identifies the technician who executed the task, used for subsequent calculation of the technician's load and time occupancy. After the log is generated, the system updates the time constraint data in the database based on the execution status, recording the task completion time and the technician's time occupancy.

[0051] The system extracts task start and end times from real-time task execution logs and calculates technician time usage. By updating the technician's time constraint field in the database, the system accurately reflects each technician's time resources and workload. This time usage information influences subsequent task allocation and the adjustment of technician capacity parameters.

[0052] Based on the technician's time utilization ratio, the system adjusts the technician's capacity parameter vector. The adjustment formula is: C'=C×(1-T), where C' is the adjusted capacity parameter, C is the capacity parameter, and T is the time utilization ratio. The capacity parameter C represents the technician's initial capacity value and is usually set based on their historical performance and the complexity of the current task. The time utilization ratio T is calculated as the ratio of the technician's current load to their total available time. This ratio is usually calculated based on the ratio of the technician's time spent on tasks to their available time. Through this adjustment, the system can accurately reflect changes in the technician's load and dynamically update their capacity parameters. Adjustments to capacity parameters directly influence the allocation of subsequent tasks, ensuring that tasks are always assigned to the most appropriate technician.

[0053] After adjusting the capability parameter vector, the system recalculates the task allocation plan. Based on the new capability parameter vector, the system recalculates the score of each task and generates a new task allocation plan based on the updated task score ranking. Task status: The progress and status of the task, which determines whether the task has been completed or needs adjustment. Technician status: The workload and time usage of each technician, which determines their availability for new tasks. Time constraint data: Records the time requirements of each task to ensure timely completion.

[0054] By recalculating the task allocation plan, the system updates system status data. This system status data includes information such as task status, technician status, and time constraints. A status aggregation algorithm aggregates all this information to form the overall system status. The status aggregation algorithm consolidates the status data of multiple tasks and technicians into a unified system status report, providing a global view. This report helps administrators quickly understand the overall status of task allocation and the operating status of the system.

[0055] Based on updated system status data, the system generates real-time task monitoring records and stores them in a multi-user schedule database. These records provide information such as task progress and technician occupancy, enabling real-time tracking of task execution. These records provide timely feedback on task progress, helping managers adjust task allocation and resource scheduling to ensure all tasks are completed on time and efficiently.

[0056] For example, a large enterprise's operations and maintenance management system processes hundreds or even thousands of tasks daily, each with varying priorities and technician workloads. The system calculates a score for each task by combining task priority with technician capability parameters. High-priority tasks are assigned to technicians with the highest matching scores. During task execution, the system generates real-time execution logs, recording task progress and technician time usage. The system also updates technician capability parameters based on these logs to ensure continuous optimization of resource allocation. This real-time feedback mechanism allows the system to adjust task allocation at any time, avoiding task delays and technician overload, thereby improving the efficiency and quality of task execution.

[0057] While embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions, and variations may be made to these embodiments without departing from the principles and spirit of the invention, and that the scope of the invention is defined by the appended claims and their equivalents.

Claims

1. A hierarchical alarm processing and task scheduling method for a multi-user environment, characterized in that: include: Obtain the technical personnel's capability parameter vector, time constraint data and workload upper limit to obtain candidate allocation plans; Calculate the technician's task priority weighting factor in different time periods. If the technician's time constraint data conflicts with the task time requirement, mark it as unavailable, and obtain the first allocation plan after eliminating the conflict; Obtain the capability parameter vectors and workload distribution ratios of other technicians, generate a list of substitute personnel closest to the capability parameter vectors of the unavailable personnel, and obtain a second allocation plan; Combining the task priority weighting factor and time constraint check, the substitute personnel are sorted to obtain the third allocation plan; According to the time constraint verification and capability parameter vector matching requirements of each task, determine the task and personnel mapping table; Obtain the load change trend of technicians after assignment, reallocate tasks and generate an adjusted fifth assignment plan; Predict the stability of task allocation in the future time period, perform batch comparison operations to regenerate the list of substitute personnel, and obtain the final allocation plan; Combining the task priority weighting factor and capability parameter vector, a real-time task execution log is generated to obtain the system status data.

2. The hierarchical alarm processing and task scheduling method for a multi-user environment according to claim 1, characterized in that: The candidate allocation schemes obtained by obtaining the technician capability parameter vector, time constraint data, and workload upper limit include: The technical personnel's ability parameter vectors, time constraint data and workload upper limit are extracted from the schedule database. The multi-user data is parsed using structured query language to obtain the initial personnel data set. The ability parameter vectors in the initial personnel data set are processed by vector normalization. Combined with the task priority list, a personnel and task matching matrix is generated to obtain an intermediate allocation plan that includes time availability and ability score. If the workload of a technician in the intermediate allocation plan exceeds the upper limit, the task allocation is adjusted through a linear programming algorithm to obtain an optimized candidate allocation plan. Based on the optimized candidate allocation plan, the time availability of each technician is obtained, and the time window partitioning method is used to determine the task execution time period. If there is a conflict between the task execution time period and the time constraint data, the task order is adjusted through priority sorting to obtain the allocation plan after the conflict is resolved. By comparing the allocation plan after the conflict is resolved with the ability score, the weighted average method is used to determine the final task allocation result. The final task allocation result is obtained, and an allocation record including technicians, tasks and time periods is generated and stored in the schedule database.

3. The hierarchical alarm processing and task scheduling method for a multi-user environment according to claim 1, characterized in that: The calculation of the task priority weighting factors of technicians in different time periods, if the technician's time constraint data conflicts with the task time requirement, marks it as unavailable, and obtains a first allocation plan after eliminating the conflict, including: Obtain the initial personnel and task matching matrix, extract technicians, task priorities and time period information, generate preliminary allocation data including time constraints, use dynamic integration algorithm to calculate the task priority weighting factors in each time period for the preliminary allocation data, and obtain a set of weighting factors. If the technician time constraint conflicts with the task time requirement, the corresponding technician will be marked as unavailable according to the conflict marking rule to obtain the allocation data after eliminating the conflict. By eliminating the conflict and combining the allocation data with the weighting factor set, a first allocation plan is generated. For the first allocation plan, obtain the matching degree of each task priority and time period to determine whether it meets the preset allocation balance threshold and obtain the allocation balance evaluation result. According to the allocation balance evaluation result, if the matching degree is lower than the preset threshold, adjust the weighting factor and recalculate the allocation data to obtain the optimized allocation plan. Through the optimized allocation plan, determine the final personnel and task matching relationship and generate the allocation result including time period and task priority.

4. The hierarchical alarm processing and task scheduling method for a multi-user environment according to claim 1, characterized in that: The step of obtaining the capability parameter vectors and workload distribution ratios of other technicians, generating a list of substitute personnel closest to the capability parameter vectors of the unavailable personnel, and obtaining a second distribution plan includes: The capability parameter vectors of unavailable technicians and the capability parameter vectors of all technicians and the workload distribution ratio are extracted from the multi-user schedule database to obtain the initial personnel data set. The similarity between the capability parameter vectors of each technician in the initial personnel data set and the capability parameter vectors of unavailable technicians is calculated by cosine distance to obtain a similarity value set, where the cosine distance formula is: cos(θ)=(A·B) / (||A||·||B||), A is the capability parameter vector of unavailable technicians, B is the capability parameter vector of other technicians, ||A|| and ||B|| represent the modulus of vector A and vector B respectively, and the similarity value set is screened according to the preset threshold. If the similarity value is large, the similarity value set is filtered. If the workload distribution ratio is lower than the preset maximum load threshold, the technician will be retained to obtain an optimized substitute list. The optimized substitute list is sorted in descending order according to the similarity value in the optimized substitute list to obtain a sorted substitute sequence. The sorted substitute sequence is combined with the task data corresponding to the unavailable technician to generate a second allocation plan, determine the final substitute allocation result, and update the workload distribution ratio of the technician in the second allocation plan from the multi-user schedule database to obtain an updated schedule data set.

5. The hierarchical alarm processing and task scheduling method for a multi-user environment according to claim 1, characterized in that: The third allocation scheme obtained by combining the task priority weighting factor and the time constraint check to sort the substitute personnel includes: Obtain the list of substitute personnel, task priority weighting factors and time constraints from the second allocation plan, use the weighted sorting algorithm to calculate the personnel priority and obtain the initial priority ranking. Based on the initial priority ranking, obtain the workload data of each substitute personnel and calculate the load ratio. If the load ratio exceeds the upper limit threshold, eliminate the corresponding substitute personnel to obtain the updated priority ranking. From the updated priority ranking, obtain the personnel's time constraints to determine whether the task time requirements are met. If not, adjust the personnel order to obtain the adjusted priority ranking. Based on the adjusted priority ranking, use the allocation algorithm to generate the third allocation plan. Obtain the allocation result from the third allocation plan to determine whether there is a load conflict. If so, recalculate the load ratio and eliminate overloaded personnel to obtain the optimized allocation plan. Based on the optimized allocation plan, generate the final personnel allocation list and determine the third allocation plan.

6. The hierarchical alarm processing and task scheduling method for a multi-user environment according to claim 1, characterized in that: The time constraint check and capability parameter vector matching requirements for each task and determining the task-personnel mapping table include: By parsing the third allocation plan, obtaining the task time constraints and capability parameter vectors, and determining the basic allocation data, if the task time constraints do not match the capability parameter vectors, adjusting the task priority, and obtaining the verified task sequence, a greedy algorithm is used to allocate technicians based on the verified task sequence and capability parameter vectors, and a preliminary allocation result is determined. Based on the preliminary allocation result, a time schedule is generated, and time conflicts are determined to obtain an optimized schedule. Based on the optimized schedule and the preliminary allocation result, a fourth allocation plan is generated to determine the task allocation details, and the correspondence between tasks and technicians is extracted from the fourth allocation plan to generate a task-personnel mapping table. Based on the task-personnel mapping table, the matching degree between the allocation plan and the capability parameter vector is verified to determine the final allocation result.

7. The hierarchical alarm processing and task scheduling method for a multi-user environment according to claim 1, characterized in that: The obtaining of the load change trend of the technicians after assignment and reallocating the tasks to generate the adjusted fifth assignment plan includes: Obtain the technician load change trend from the task mapping table, calculate the deviation between the daily task duration of each technician and the historical average, and obtain the load change trend. According to the load change trend and the preset workload upper limit, calculate the distribution ratio of each technician's current workload and determine the distribution ratio data. If the distribution ratio of a technician exceeds the dynamic threshold, adjust the task distribution through the linear programming algorithm to generate a temporary task distribution plan. Extract the technician load data after task reallocation from the temporary task distribution plan, calculate the load balance index, and obtain the balance evaluation value. Compare the balance evaluation value with the balance of the historical distribution plan. If the current balance is higher than the historical average, use the current plan to generate the fifth distribution plan, extract the task distribution details of each technician from the fifth distribution plan, update the task mapping table, and obtain the optimized distribution result. Based on the optimized distribution result, calculate the load change trend of each technician, verify whether the dynamic threshold constraint is met, and confirm the final distribution plan.

8. The hierarchical alarm processing and task scheduling method for a multi-user environment according to claim 1, characterized in that: The prediction of task allocation stability in the future time period, batch comparison operation to regenerate the substitute personnel list, and obtain the final allocation plan include: A time series analysis algorithm is used to extract features from historical task allocation data, generate task allocation stability prediction results for future time periods, and obtain allocation stability trends. If the allocation stability trend shows that a task has a risk of time constraint verification failure, the risk identification mechanism is used to extract the failed task identifier from the prediction results to obtain a risk task set. For the risk task set, pre-established personnel resource pool data is obtained, and batch comparison operations are performed to screen candidate personnel that meet the task requirements. A list of substitute personnel is generated, and according to the task priority sorting rules, personnel attributes are extracted from the substitute personnel list. The dynamic weight is calculated based on the priority to obtain an adjusted weight allocation matrix.

9. The hierarchical alarm processing and task scheduling method for a multi-user environment according to claim 8, characterized in that: The prediction of task allocation stability in the future time period, batch comparison operation to regenerate the substitute personnel list, and obtaining the final allocation plan also include: Through the dynamic weight allocation matrix, the matching degree between tasks and personnel is recalculated to generate a set of candidate allocation plans and obtain an optimized allocation plan set. If there are multiple plans in the candidate allocation plan set, the time constraint verification algorithm is used to verify the time feasibility of each plan one by one, determine the final allocation plan, adopt the final allocation plan, update the task allocation database, generate a stable task allocation result, and output it to the system log.

10. The hierarchical alarm processing and task scheduling method for a multi-user environment according to claim 1, characterized in that: The system status data obtained by combining the task priority weighting factor and the capability parameter vector to generate a real-time task execution log includes: By calculating the task priority and weighting factor, the task sorting result is obtained and the task allocation plan is determined. The calculation formula is: S=P×W+C, where S is the task score, P is the task priority, W is the priority weight, and C is the capability parameter. After sorting, the task allocation plan is obtained. According to the task allocation plan, a real-time task execution log is generated. The task identifier, execution time and technician identifier are extracted from the allocation plan. The log generation algorithm is used to output a structured execution log and store it in the multi-user schedule database. If the execution log is generated, the time constraint data in the multi-user schedule database is updated. The task start time and end time are obtained from the execution log, and the technician time occupancy is calculated. The time constraint field of the corresponding technician in the database is updated to obtain the updated time constraint data. The technician capability parameter vector is adjusted through the time constraint data, and the occupied time ratio is extracted from the time constraint data. Combined with the task complexity, vector operations are used to update the capability parameters. Vector, the specific formula is: C'=C×(1-T), where C' is the adjusted capacity parameter, C is the capacity parameter, and T is the time occupancy ratio. The adjusted capacity parameter vector is obtained. If the capacity parameter vector is adjusted, the task allocation plan is recalculated, and the technician skill data is extracted from the adjusted capacity parameter vector. Combined with the task priority and weighting factor, the task scoring formula is re-executed to obtain a new task allocation plan. According to the new task allocation plan, the system status data is updated, and the task status, technician status and time constraint data are obtained from the task allocation plan. The state aggregation algorithm is used to generate system status data and store it in a multi-user schedule database to obtain the latest system status data. Through the system status data, real-time task monitoring records are generated, and the task execution progress and technician occupancy are extracted from the system status data. The log generation algorithm is used to output real-time task monitoring records and store them in a multi-user schedule database to obtain task monitoring results.

Images