Intelligent scheduling method and system in a media content production process

Abstract

The application relates to the technical field of manufacturing process management, in particular to an intelligent scheduling method and system in a media content production process, which comprises the following steps: collecting historical and current editing data in the media content production process, and preprocessing the editing data; constructing an editing time length prediction model by a statistical analysis method and introducing dynamic editing efficiency parameters, a team efficiency synchronization index and an efficiency trend index, calculating the predicted editing processing time length corresponding to each editing task and the predicted completion time of the editing task based on the preprocessed editing data; judging the scheduling constraint conditions and optimization objectives in the media content production process based on the predicted editing processing time length; and performing editing task scheduling calculation based on the scheduling constraint conditions and optimization objectives. The application realizes accurate time length prediction of editing tasks and efficient allocation of resources by fusing the static and dynamic efficiency parameters of editing personnel, team cooperation states and timing constraints.

Description

Technical Field

[0001] This invention relates to the field of manufacturing process management technology, specifically to an intelligent scheduling method and system for media content production. Background Technology

[0002] In the context of continuously expanding media content production scale and increasingly compressed production cycles, traditional editing task scheduling methods mainly rely on manual experience for allocation, lacking the ability to quantitatively model and dynamically respond to individual efficiency differences among editors, team collaboration status, task complexity, and resource consumption. This extensive scheduling mechanism, when faced with real-world constraints such as multi-tasking, limited workstation resources, and complex process dependencies, easily leads to resource conflicts, task queuing, rework delays, and other problems, severely restricting the efficiency and reliability of media content production process management. Especially in scenarios with extremely high timeliness requirements, such as short videos, live news broadcasts, and variety shows, traditional methods struggle to accurately predict editing task completion times and optimally match personnel and workstation resources, resulting in significant hidden waste and scheduling blind spots in production process management. Summary of the Invention

[0003] The purpose of this invention is to address the problems existing in the background technology by proposing an intelligent scheduling method and system for the media content production process.

[0004] The technical solution of this invention: an intelligent scheduling method in the media content production process, comprising: S1. Collect historical and current editing data during the media content production process, and preprocess the editing data; S2. By using statistical analysis methods and introducing dynamic editing efficiency parameters, team efficiency synchronization index and efficiency trend index, a editing duration prediction model is constructed. Based on the pre-processed editing data, the expected editing processing time and the expected completion time of each editing task are calculated. S3. Determine the scheduling constraints and optimization objectives in the media content production process based on the expected editing processing time; S4. Based on scheduling constraints and optimization objectives, perform editing task scheduling calculations to intelligently schedule the allocation order of editing tasks between editing personnel and editing workstations.

[0005] As a further improvement to this technical solution, in S1, the editing data includes: the duration of the editing task and the identifier of the editor who performed the editing task.

[0006] As a further improvement to this technical solution, in step S2, a editing duration prediction model is constructed to calculate the estimated editing processing time and the estimated completion time of each editing task based on the preprocessed editing data. This includes the following steps: S2.1 Collect the actual completion time of historical editing tasks. Based on the preprocessed historical editing data, establish a mapping relationship between the material duration of historical editing tasks and the actual completion time of historical editing tasks for each editor, and calculate the individual's static editing efficiency parameters. By combining the collaborative network relationship among editors with a dynamic efficiency change model, dynamic editing efficiency parameters reflecting their processing capabilities under actual working conditions are calculated, and team efficiency synchronization index and efficiency trend index at the start of the task are generated. S2.2, using the duration of footage from historical editing tasks and dynamic editing efficiency parameters. Using the team efficiency synchronization index and efficiency trend index at the start of the task as input features, and the actual completion time of the editing task as the target output, an editing time prediction model is constructed. S2.3 Based on the preprocessed editing data, the estimated editing processing time of the editing task is calculated using the editing duration prediction model; S2.4. Calculate the estimated completion time of the editing task based on its start time.

[0007] As a further improvement to this technical solution, in step S2.1, by combining the collaborative network relationship among editors with a dynamic efficiency change model, dynamic editing efficiency parameters reflecting their processing capabilities under actual working conditions are calculated, and team efficiency synchronization index and efficiency trend index at the start of the task are generated, including the following steps: S2.11. Collect collaboration records of editors within the time window, and construct a collaboration weight matrix among editors based on these records. ; S2.12. For the current editing task, extract the task feature vector and calculate the task similarity. S2.13, Based on the Collaboration Weight Matrix The cooperative coupling coefficient is calculated by combining the time decay factor and task similarity. ; S2.14, Based on Cooperative Coupling Coefficient Build dynamic editing efficiency parameters for each editor The dynamic equations that change with time; The trajectory of the editor's efficiency over time is generated by iterative calculation at discrete time steps. S2.15. Based on the trajectory of editors' efficiency changes over time, calculate the team efficiency synchronization index and efficiency trend index.

[0008] As a further improvement to this technical solution, in S2.13, based on the cooperative weight matrix... The cooperative coupling coefficient is calculated by combining the time decay factor and task similarity. This includes the following steps: For editors and The most recent collaborative task was analyzed, and the collaboration time interval was calculated. Simultaneously, the number of editors within the time window was recorded. and Task association strength ; A time decay factor is constructed based on the collaboration time interval and the strength of task association. ; Collaboration weight matrix With time decay factor Combined, calculate the basic synergistic coupling coefficient. : If the task similarity exceeds a set threshold This enhances the basic synergistic coupling coefficient.

[0009] As a further improvement to this technical solution, in S2.14, the dynamic editing efficiency parameter for each editor... The time-varying dynamic equation consists of individual self-decay and baseline regression terms, cooperative coupling terms, task load influence terms, and random disturbance terms.

[0010] As a further improvement to this technical solution, step S3, which determines the scheduling constraints and optimization objectives in the media content production process based on the expected editing processing time, includes the following steps: S3.1 Based on the estimated editing processing time and the estimated completion time of the editing task, combined with the predetermined release time of the corresponding video content, an adaptive correction method for time uncertainty based on a penalty mechanism is used to calculate the time margin of the editing task. And according to the time margin and urgency evaluation indicators Determine the time feasibility status of the editing task within the current scheduling cycle; S3.2 Based on the estimated editing processing time of each editing task, construct a task occupancy interval for each editor and each editing workstation on the future timeline to generate resource occupancy constraints; S3.3 Based on resource occupancy constraints, calculate the idle time segments of each editor under the current scheduling scheme and the waiting time caused by the unavailability of editors; S3.4 For multiple editing tasks corresponding to the same video content, identify the order of pre- and post-processing in the production process, and determine the earliest executable time of each task based on the expected editing processing time. Use the order of pre- and post-processing and the earliest executable time as the order dependency constraint between tasks. S3.5. Based on resource space constraints and sequential dependency constraints, the goal of scheduling optimization is to minimize the maximum completion time of all editing tasks so that all videos are completed before the predetermined release time.

[0011] As a further improvement to this technical solution, in step S3.1, an adaptive correction method for time uncertainty based on a penalty mechanism is used to calculate the time margin of the editing task. This includes the following steps: S3.11, Video release time The time difference between the estimated completion time of the editing task and the original time margin is used as the raw time margin. ; S3.12, For editing tasks Based on the completion records of similar editing tasks in the past, the fluctuation intensity of the expected completion time is calculated to generate the asymmetric completion time fluctuation range of the editing task. S3.13, Based on the asymmetric completion time fluctuation range and video release time Calculate editing tasks Overdue risk coefficient At the same time, an urgency evaluation index was introduced. ; S3.14. Adjusting the original time margin using a risk adjustment coefficient. Make corrections to generate the corrected time margin. .

[0012] As a further improvement to this technical solution, step S4, which involves performing editing task scheduling calculations based on scheduling constraints and optimization objectives, includes the following steps: S4.1. Based on resource space constraints, sequential dependency constraints, and scheduling optimization objectives, establish an editing task scheduling model; S4.2, Based on the time leeway for each editing task Combining urgency evaluation indicators Calculate the urgency weight of the editing task; and generate a priority queue of tasks to be scheduled based on the release time of the video to which the editing task belongs. S4.3, Based on dynamic editing efficiency parameters Based on the current workload of personnel, preliminary screening of candidate editors suitable for the current editing task is conducted. Based on the occupancy status of editing workstations and available time windows, a set of candidate editing workstations that can perform the editing task within the schedulable time range is selected to form a resource combination. S4.4 For the editing tasks in the priority queue of tasks to be scheduled, calculate the resource idle and waiting cost factors under different resource combinations; S4.5 Construct a comprehensive cost function based on resource idle and waiting cost factors to calculate the comprehensive cost for all resource combinations, and select the resource combination with the minimum comprehensive cost as the scheduling decision result for the current editing task; S4.6 For tasks with a sequential relationship under the same video content, determine the completion status of its predecessor task before executing the scheduling decision; S4.7. During the task execution process, collect real-time information on the editing task progress, personnel status changes, and workstation anomalies. S4.8 Output the scheduling results of the final editing tasks, which are allocated between the editors and editing stations.

[0013] On the other hand, the present invention provides an intelligent scheduling system for the media content production process, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the intelligent scheduling method for the media content production process described above.

[0014] Compared with the prior art, the above-mentioned technical solution of the present invention has the following beneficial technical effects: by integrating the static and dynamic efficiency parameters of editing personnel, team collaboration status and timing constraints, it realizes accurate duration prediction and efficient resource allocation for editing tasks, significantly improving the scientific nature and timeliness of scheduling decisions; it effectively optimizes the resource allocation and task collaboration mechanism in the media content production process management, reduces waiting and idle losses caused by fluctuations in personnel efficiency, workstation conflicts or process dependencies, thereby ensuring timely video release while improving overall production efficiency and the level of intelligence in the manufacturing process management. Attached Figure Description

[0015] Figure 1 This is a flowchart of the overall method of the present invention. Detailed Implementation

[0016] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of the present invention.

[0017] Example 1: Please refer to Figure 1 As shown, this embodiment provides an intelligent scheduling method in the media content production process, including the following steps: S1. Collect historical and current editing data during the media content production process (i.e., collect historical and current editing data) and preprocess the editing data (preprocessing includes data cleaning, format unification, and outlier removal to ensure data quality and consistency, providing reliable and standardized data input for subsequent construction of editing duration prediction models and optimization scheduling decisions). In this embodiment, the editing data includes: the duration of the footage for the editing task and the identifier of the editor performing the editing task.

[0018] S2. By using statistical analysis methods and introducing dynamic editing efficiency parameters, team efficiency synchronization index and efficiency trend index, a editing duration prediction model is constructed. Based on the pre-processed editing data, the expected editing processing time and the expected completion time of each editing task are calculated. In this embodiment, an editing duration prediction model is constructed to calculate the estimated editing processing time and the estimated completion time of each editing task based on the preprocessed editing data. This includes the following steps: S2.1. Collect the actual completion time of historical editing tasks. Based on the preprocessed historical editing data, for each editor (the editor is identified by the editor identifier who performed the editing task), establish a mapping relationship between the material duration of historical editing tasks and the actual completion time of historical editing tasks, and calculate the individual's static editing efficiency parameter (as an individual efficiency benchmark). Specifically: for each editor, summarize the historical editing task records completed within the preset historical time window K, and group the historical editing task data based on the editor identifier; for each group of historical editing task data, extract... For each editing task, a sample set is constructed with the footage duration as the independent variable and the actual completion time as the dependent variable. After removing abnormal samples with missing footage, abnormal completion times, or significant deviations from the overall distribution, statistical analysis is performed on the sample set to establish a mapping relationship between the footage duration and the actual completion time of historical editing tasks (the mapping relationship is a linear proportional mapping, i.e., the ratio between the footage duration and the actual completion time). Based on this mapping relationship, the average editing processing time per unit footage duration for the editor is calculated and used as the editor's static editing efficiency parameter. By combining the collaborative network relationship among editors with the dynamic efficiency change model, dynamic editing efficiency parameters reflecting their processing capabilities under actual working conditions are calculated, and team efficiency synchronization index and efficiency trend index at the start of the task are generated. The editing efficiency parameters are used to characterize the average editing processing time required per unit length of footage, or the degree of influence of changes in footage length on the editing completion time, thereby quantifying the differences in editing speed and processing capabilities among different editors. In this embodiment, by combining the collaborative network relationships among editors with a dynamic efficiency change model, dynamic editing efficiency parameters are calculated, and a team efficiency synchronization index and efficiency trend index are generated. This addresses the problem in traditional media content production scheduling where "personnel processing capabilities are statically modeled, individuals are isolated, and the model fails to reflect the true collaborative state and the time-varying nature of efficiency." In actual editing production environments, the efficiency of editors is not a fixed constant but is influenced by multiple factors, including recent collaboration partners, task type matching, team synchronization, individual workload, and state fluctuations. Simultaneously, the overall team efficiency exhibits a phased convergence or divergence trend. If scheduling is still based on a single historical average efficiency or static parameters solely based on personal experience, problems such as distorted duration predictions, unreasonable task ordering, resource waiting, and frequent rework can easily occur when there is high collaboration dependence, changes in task complexity, or fluctuations in team state. This invention integrates the collaborative network of editors, time decay, task similarity, and efficiency dynamics into the pre-scheduling duration prediction and constraint construction stage. It achieves controllable transmission of efficiency among personnel through a collaborative coupling coefficient, avoiding the simplistic equation of collaborative experience with static weighting. A team efficiency synchronization index characterizes whether a team is in a state of high collaboration or imbalance, while an efficiency trend index anticipates rising or falling team efficiency trends, making scheduling no longer a "post-event correction" but a proactive risk avoidance and resource allocation capability. This method significantly improves the stability and robustness of editing duration prediction, reduces waiting and interruptions caused by collaboration mismatches, and enhances overall on-time delivery rate and resource continuous utilization efficiency in scenarios with multiple parallel tasks and dense deadlines. This involves combining the collaborative network relationships among editors with a dynamic efficiency change model to calculate dynamic editing efficiency parameters that reflect their actual processing capabilities under working conditions, and generating team efficiency synchronization index and efficiency trend index at the start of the task. The process includes the following steps: S2.11. Editing personnel collaboration records within the acquisition time window (past T days), including the number of task handovers. Overlapping time between co-editing and number of technical support sessions (For any two editors) and Statistics were compiled by the editors within the stated time window. To the editors The number of times editing tasks are handed over or continued is counted as the number of task handovers. ; Statistical editors With the editors The length of the time overlap interval in the same editing task, and the cumulative length of the overlap interval, are used to obtain the degree of joint editing time overlap. Meanwhile, the statistical editors During the editing process, to the editors The number of technical support requests initiated or the number of assistance interactions received is counted as the number of technical support sessions. ), and construct a collaboration weight matrix among editors based on their collaboration records. Used to describe the strength of collaboration among people: ; ; In the formula, , For editors' indexing, For editors The total number of task handovers with all personnel within the statistical time window is used for normalization. For editors The sum of all overlapping edit times within the statistical time window. For editors The total number of technical support interactions within the statistical time window. As a weight for the number of task handovers, Weighting for the degree of overlap in co-editing time. Weighting based on the number of times technical support was provided; S2.12. To reflect the impact of different task types on collaboration efficiency, extract task feature vectors for the current editing task. (Task feature vector here) (Normalization has been performed to eliminate differences in the dimensions of different features and ensure the stability of similarity calculation), and task similarity is calculated. (In the formula, , For two editing tasks to be compared, For editing tasks The task feature vector, For editing tasks The task feature vector, (For editing tasks), used to determine the similarity between historical collaborative tasks and the current task, thereby introducing adjustments for task relevance in the collaborative coupling coefficient; S2.13, Based on the Collaboration Weight Matrix The cooperative coupling coefficient is calculated by combining the time decay factor and task similarity. (The collaborative coupling coefficient is calculated, taking into account historical collaboration strength, time decay effect, and task similarity, to dynamically adjust the strength of the efficiency impact among personnel; this coefficient characterizes the actual effectiveness of efficiency transmission among personnel, taking into account collaboration novelty and task matching.) Furthermore, based on the collaboration weight matrix The cooperative coupling coefficient is calculated by combining the time decay factor and task similarity. This includes the following steps: For editors and Calculate the collaboration time interval of the most recent collaborative task. (In the formula, For the current calculation time, For editors With the editors (The most recent point in time when collaboration occurred), and also, statistics on the number of editors within the time window. and Task association strength (Task correlation strength) To allow editors to work within the time window With the editors The number of times a team or team members participates in the same editing task together or consecutively, as a percentage of the total number of tasks within the time window. A time decay factor is constructed based on the collaboration time interval and the strength of task association. (In the formula, This is the time decay coefficient, used to control the rate at which the cooperative relationship weakens over time. Its value ranges from 0.05 to 0.2 (unit: reciprocal of time), and is determined through expert experience. If measured in days... =0.1 means that the weight of the collaborative effect decays to about 37% (e⁻¹) after about 7 days (1 / 0.1). Collaboration weight matrix With time decay factor Combined, calculate the basic synergistic coupling coefficient. (This basic collaborative coupling coefficient reflects the strength of the collaborative efficiency transmission among editors without considering differences in task content.) The unit is h⁻¹): If the task similarity exceeds a set threshold (In this embodiment, a threshold is set) The value ranges from 0.6 to 0.8 (judged based on historical data), thus enhancing the basic synergistic coupling coefficient: (In the formula, This is the task similarity enhancement coefficient, used to adjust the amplification effect of task similarity on the strength of collaborative coupling. Its value ranges from 0.1 to 0.5, and it is calibrated by the statistical average of the efficiency improvement ratio of similar tasks in historical data. (This is the enhanced collaborative coupling coefficient) if the task similarity does not exceed a set threshold. Then keep: This step amplifies the positive impact of task content matching on collaboration efficiency, enabling the model to more accurately reflect the applicability of historical collaboration experience in similar task contexts, and ensuring that the intensity of efficiency transmission among personnel matches the actual task requirements. This dynamic adjustment based on task characteristics can improve the prediction accuracy of the collaborative coupling coefficient, making the generated dynamic editing efficiency parameters, team efficiency synchronization index, and trend index more closely match the actual working state, thereby providing a more reliable decision-making basis for task allocation and team optimization, and ultimately enhancing the overall editing efficiency. S2.14, Based on Cooperative Coupling Coefficient Build dynamic editing efficiency parameters for each editor Dynamic equations that change over time (establish dynamic equations for dynamic editing efficiency parameters to simulate the change of individual efficiency over time; the equations combine static benchmarks, interpersonal coupling, task load and random disturbances to generate efficiency trajectories to reflect the dynamic evolution of processing capacity under actual working conditions). The trajectory of editor efficiency as a function of time is iteratively calculated at discrete time steps; specifically: at discrete time steps, for each editor... From the start time of the mission Starting from this point, time is divided into several continuous discrete time steps. At each discrete time step Above, based on the editor's static efficiency parameters Cooperative coupling coefficient Current task load and random disturbances The dynamic editing efficiency parameter is updated iteratively using the dynamic equation. That is, at each time step, the increments of self-decay and baseline regression terms, collaborative coupling terms, and task load impact terms are calculated, and the increments are accumulated into the dynamic efficiency value of the previous step, and then sequentially advanced to the next time step; through iterative calculation of the entire time window or task execution cycle, a discrete trajectory of the efficiency of each editor over time is generated, which is used for the subsequent calculation of the team efficiency synchronization index and efficiency trend index, and reflects the actual work efficiency evolution under different task loads and collaborative states; Furthermore, the dynamic editing efficiency parameters for each editor. The time-varying dynamic equation consists of individual self-decay and baseline regression terms ( ), Cooperative coupling terms ( ), Task load impact items ( ) and random disturbance term ( )constitute; The specific dynamic equations are as follows: ; In the formula, For time, This is a parameter for static editing efficiency. For editors The efficiency regression coefficient is used to characterize the strength of the regression from dynamic efficiency to static efficiency, with a value range of 0.1 to 0.5 (h⁻¹). It is individually calibrated based on the autocorrelation analysis of each individual's historical efficiency data. For editors In time The efficiency of dynamic editing. For editors The task load impact coefficient is used to adjust the degree of influence of task load on efficiency, with a value range of -0.2 to 0.2 (unit: standard task load / h²). This task load impact coefficient characterizes the amount of gain or suppression change in dynamic editing efficiency parameters per unit of task load per unit time. It is obtained by fitting historical efficiency and concurrent load data. For editors In time The workload, This is a task load mapping function used to convert task load into efficiency gains or suppression terms. , For reference load, This is a random disturbance term used to characterize the impact of sudden events, state fluctuations, or uncontrollable factors on editing efficiency (unit: standard workload / h²). S2.15. Based on the trajectory of editors' efficiency changes over time, calculate the team efficiency synchronization index and efficiency trend index (the team efficiency synchronization index measures the degree of synergy among team members, and the efficiency trend index reflects the future direction of team efficiency changes); specifically: first, at each time step, statistically analyze the efficiency distribution of all editors and calculate the deviation between individual efficiency and the team average efficiency to measure the consistency of efficiency among team members; by accumulating and standardizing the individual deviations at each time step, obtain the team efficiency synchronization index to characterize the degree of synergy in the overall team efficiency; simultaneously, perform regression or smoothing fitting on the trend of team average efficiency changes over time to generate the efficiency trend index, which reflects the direction and trend of team efficiency changes in the future time period; the team efficiency synchronization index and efficiency trend index can serve as reference indicators for editor scheduling decisions, task allocation optimization, and dynamic adjustment of collaboration efficiency. S2.2, using the duration of footage from historical editing tasks and dynamic editing efficiency parameters. Using the team efficiency synchronization index and efficiency trend index at the start of the task as input features, and the actual completion time of the editing task as the target output, an editing time prediction model is constructed. The editing duration prediction model employs a hierarchical structure, comprising an input layer, an intermediate feature modeling layer, and an output layer. The input layer receives multi-dimensional feature information related to the editing task. These input features include at least the duration of footage from historical editing tasks, dynamic editing efficiency parameters of the editors associated with the task, and the team efficiency synchronization index and efficiency trend index at the task's start time. The intermediate feature modeling layer jointly models the input features through nonlinear mapping and weight learning of multiple layers of neurons. This aims to characterize the comprehensive nonlinear impact of footage size, personnel processing capabilities, and team collaboration on the editing processing duration, generating a high-dimensional intermediate feature representation. The output layer generates the predicted editing processing duration based on this intermediate representation and further calculates the predicted completion time of the editing task by combining the task's start time. This provides a time parameter basis for subsequent scheduling constraint construction and task allocation decisions. S2.3 Based on the preprocessed editing data, the estimated editing processing time of the editing task is calculated using the editing duration prediction model; S2.4. Calculate the estimated completion time of the editing task based on its start time. (The start time of the editing task is defined as the moment it is scheduled and actually enters the execution state. If the editing personnel and editing workstations remain continuously occupied and uninterrupted during the execution of the editing task, the estimated editing processing time is added to the start time of the task to obtain the estimated completion time of the editing task; the estimated completion time calculated at this time is...) To predict the completion time, let it be denoted as This is a predicted value based on the current personnel status and task characteristics, and does not yet consider resource conflicts and task queuing in actual scheduling; in contrast, the planned completion time generated after the scheduling calculation is completed is denoted as... The predicted completion time is a planned value determined based on the actual resource allocation and scheduling order; Used for initial assessment of task urgency, and planning completion time. (Used for actual scheduling and execution tracking).

[0019] S3. Determine the scheduling constraints and optimization objectives in the media content production process based on the expected editing processing time; In this embodiment, determining the scheduling constraints and optimization objectives in the media content production process based on the expected editing processing time includes the following steps: S3.1, Based on the estimated editing processing time and the estimated completion time of the editing task (here, the estimated completion time is the predicted completion time). Based on the predetermined release time of the corresponding video content, an adaptive correction method for time uncertainty based on a penalty mechanism is used to calculate the time margin for the editing task. (The time difference between the video release time and the estimated completion time of the editing task), and based on the time leeway. and urgency evaluation indicators Determine the time feasibility status of the editing task within the current scheduling cycle: when or ( When the emergency threshold is greater than 0 (e.g., 1.0 or 2.0), the editing task is determined to be a risk-critical, rigid-deadline task and must be scheduled with priority; when... When a task is deemed risk-adjustable, sequential optimization is allowed provided that the overall scheduling objectives are met. Through time feasibility assessment, all parallel editing tasks are divided into different urgency levels, providing a foundation for the construction of subsequent scheduling constraints. In this embodiment, the time uncertainty adaptive correction method based on the penalty mechanism addresses the problem that the completion time of editing tasks in the media content production process has significant uncertainty, and that this uncertainty is underestimated or symmetrically treated in traditional scheduling. In actual production scenarios, different types of editing tasks vary significantly in terms of special effects complexity, rework probability, and personnel status fluctuations. Their completion time often exhibits a "right-skewed distribution" characteristic, meaning that the risk of delay is far greater than the benefit of early completion. If only the simple difference between the predicted completion time and the release time is used as the time margin, it is easy to misjudge the feasibility of the task when the prediction is too optimistic or the fluctuation is amplified, resulting in risky tasks being incorrectly classified as tasks that can be postponed, and thus a concentrated outbreak of delays in the later stages of scheduling. This invention transforms time uncertainty into an adjustable, quantifiable, and comparable risk factor through a penalty mechanism, and directionally corrects the time margin: It employs an asymmetric completion time fluctuation range, amplifying the risk of delay in the time margin calculation rather than treating it as equal to early completion, aligning with the practical reality in editing production where "delay costs are far higher than early completion costs"; it introduces a fluctuation adjustment coefficient based on historical prediction deviations, enabling adaptive time margin corrections for different task types and avoiding a "one-size-fits-all" safety factor; and it links the overdue risk coefficient with the urgency evaluation index, allowing the corrected time margin to directly participate in task classification and priority ranking. This achieves a shift from "static margin judgment" to "risk-aware time constraints," significantly improving the scheduling system's robustness to time uncertainty and its ability to identify high-risk editing tasks in advance, thus better ensuring on-time video release under complex production rhythms. Among them, an adaptive correction method for time uncertainty based on a penalty mechanism is used to calculate the time margin of the editing task. This includes the following steps: S3.11, Video release time The estimated completion time of the editing task (here, estimated completion time is the predicted completion time). The time difference between them is used as the original time margin. ; S3.12, For editing tasks Based on the completion records of similar editing tasks in the past, the fluctuation intensity of their estimated completion time is calculated (here, estimated completion time refers to predicted completion time). This generates asymmetric completion time fluctuation ranges for editing tasks (it can dynamically consider the completion time fluctuations of different task types, enhance the adaptability of the scheduling scheme to time uncertainties, avoid resource conflicts or failure to complete tasks as planned due to task time fluctuations, and improve the robustness of scheduling). Within the historical window, calculate the standard deviation of the actual completion time for this task type. ; Among them, the standard deviation of the actual completion time is adjusted by the historical completion time fluctuation adjustment factor. Optimize: ; The above formula is essentially a standard deviation of the actual completion time. The optimization performed, in the formula, , is the historical prediction bias coefficient, used to quantify the systematic bias in predictions for a specific type of editing task. This represents the average actual completion time for this type of task within the historical window. The historical average prediction time. The final standard deviation of the optimized actual completion time. This is a historical completion time fluctuation adjustment factor, used to amplify or reduce the uncertainty impact of historical task deviations on the current task. Its value ranges from 0.2 to 0.5 and is determined through expert experience. Predicted completion time of editing tasks based on the editing duration prediction model. Construct an asymmetric completion time fluctuation range: ; ; In the formula, This is an asymmetric factor used to control the amplification factor of the upper bound volatility coefficient relative to the basic volatility coefficient, with a value ranging from 1.2 to 2.0. The basic volatility coefficient, used to define the initial width of the completion time interval, is determined through expert experience and is fixed at 1.0–1.5. The upper bound fluctuation coefficient is used to determine the upper limit of the asymmetric completion time interval through expert experience. S3.13, Based on the asymmetric completion time fluctuation range and video release time Calculate editing tasks Overdue risk coefficient (Overdue risk factor) Used to quantify the time elapsed beyond the release date. The risk level is set to 0 if the deadline has not passed. By identifying high-risk tasks in advance, risk alerts are provided to the scheduling system, enabling the system to prioritize urgent tasks or adjust resource allocation, thereby reducing the possibility of task delays and video release delays and improving overall production efficiency. Simultaneously, an urgency evaluation index is introduced. ( To prevent the denominator from being zero for extremely small positive numbers, (to adjust the weights); S3.14. Adjusting the original time margin using a risk adjustment coefficient. Make corrections to generate the corrected time margin. (In the formula, (This is the risk adjustment coefficient, ranging from 0.5 to 1.0, determined through expert experience). S3.2. Based on the feasibility assessment of the editing task time, and based on the estimated editing processing time of each editing task, a task occupancy interval is constructed for each editor and each editing workstation on the future timeline to generate resource occupancy constraints. Specifically: based on the estimated editing processing time of each editing task, the estimated start time of each task is first determined, and the estimated occupancy interval of the task is delineated along the timeline starting from this time. Subsequently, for each editor, the occupancy interval of the editing task assigned to them at each time step is marked on the future timeline to ensure that each editor occupies only one task at the same time. Similarly, for each editing workstation, the occupancy interval of the task assigned to them is marked on the timeline to ensure that each editing workstation is occupied by only one task at the same time. Within the task occupancy interval, the continuous occupancy of the editor and the editing workstation is maintained to avoid unnecessary switching and interruption. By summarizing the task occupancy intervals of all editors and editing workstations, resource occupancy constraints can be generated, providing basic data for subsequent editing task scheduling calculations and ensuring that resource conflicts can be accurately quantified and avoided during task allocation. S3.3. Based on resource occupancy constraints, calculate the idle time segments of each editor under the current scheduling scheme and the waiting time caused by the unavailability of workstations or editors (based on the constructed resource occupancy constraints, first scan the task occupancy interval of each editor along the future time axis, identify the idle time segments between adjacent tasks or before the start of a task, and record them as available time resources; at the same time, for each editing task, when assigning it to an editor or editing workstation, if it is found that the editor or workstation is unavailable at the start of the task plan, calculate the delay time caused by waiting for resources and accumulate it as waiting time). Among them, idle time is used to characterize the efficiency of personnel resource utilization, and waiting time is used to characterize the invalid delay caused by resource conflicts or unreasonable time arrangements in the scheduling scheme. The above quantitative results of personnel idle time and waiting time are used as important indicators for evaluating the quality of the scheduling scheme and input into the subsequent optimization process. S3.4 For multiple editing tasks corresponding to the same video content, identify their pre- and post-processing order relationships in the production process (e.g., fine editing or special effects processing can only be performed after the rough cut is completed), and determine the earliest executable time of each task based on the estimated editing processing time. Use the pre- and post-processing order relationship and the earliest executable time as the order dependency constraint between tasks. Specifically: For multiple editing tasks of the same video content, first identify the pre- and post-processing order relationships of each task in the production process. For example, the rough cut must be completed before fine editing or special effects processing can be performed. Calculate the earliest executable time for each task based on its estimated editing processing time. Then, combine the pre- and post-processing order relationship of each task with its corresponding earliest executable time to construct the order dependency constraint between tasks. That is, when scheduling multiple tasks in parallel, ensure that the start time of a subsequent task is not earlier than the estimated completion time of its predecessor task (here, the estimated completion time is the predicted completion time). And maintain order constraints within the task occupancy range to prevent violations of video production workflow logic during scheduling, thereby avoiding rework or workflow conflicts, and providing clear constraints for optimizing task allocation; S3.5. Based on resource space constraints and sequential dependency constraints, the goal of scheduling optimization is to minimize the maximum completion time of all editing tasks so that all videos are completed before the predetermined release time.

[0020] S4. Based on scheduling constraints and optimization objectives, perform editing task scheduling calculations to intelligently schedule the allocation order of editing tasks between editing personnel and editing workstations. In this embodiment, the editing task scheduling calculation is performed based on scheduling constraints and optimization objectives, including the following steps: S4.1. Based on resource occupancy constraints, sequential dependency constraints, and scheduling optimization objectives, establish an editing task scheduling model; wherein, the input layer of the editing task scheduling model includes the estimated editing processing time and estimated completion time of each editing task (here, the estimated completion time is the predicted completion time). ), revised time margin Task priority tags, resource occupancy information for editors and editing workstations, sequential dependency constraints, and team efficiency parameters (such as dynamic editing efficiency). (and team efficiency synchronization index); the middle layer comprehensively analyzes the input information through constraint processing units, priority weighted calculation, and resource combination evaluation, including calculating the availability of candidate editors and workstations, the idle and waiting costs under different allocation schemes, task sequence dependency verification, and weighted integration of scheduling optimization objectives; the output layer generates the final scheduling results, including the allocated editors and editing workstations for each editing task, the start time and expected completion time of the task, as well as the task execution order table for each editor and the task occupancy time table for each editing workstation, thereby achieving continuous resource occupancy and optimal overall scheduling efficiency; S4.2, Based on the time leeway for each editing task Combining urgency evaluation indicators Calculating the urgency weight of editing tasks: In the pre-scheduling phase, the time margin is based on the predicted completion time. Calculated from the predetermined release time, denoted as This is used for an initial assessment of the task's urgency; in the post-scheduling phase, it will be based on the planned completion time. Recalculate actual time margin Used to monitor execution progress; Risk-related rigid deadline tasks are given the highest priority; Less than the preset threshold For editing tasks with a value greater than 0, the urgency evaluation index will be used. The size of the urgency weight is used to calculate the urgency weight (at this point, the urgency weight is...). Using the formula: calculate, For time margin decay weight, For risk factor weights, Based on the weights, and All are non-negative real numbers, and their values ​​satisfy... The adjustment range shall not exceed the maximum adjustment range of the urgency weight to ensure the boundedness of the urgency weight and the stability of scheduling. (Emergency evaluation indicators) The higher the urgency level, the higher its weight; for Greater than the preset threshold For editing tasks, a lower urgency weight or a weight that allows for delayed execution is assigned. At the same time, the urgency weight is further adjusted based on the release time of the video to which the task belongs. In addition, a priority queue of tasks to be scheduled is generated based on the release time of the video to which the editing task belongs (the urgency weight of all editing tasks is used to generate a queue of tasks to be scheduled according to priority, so as to ensure that high-urgency tasks enter the scheduling calculation process first in the resource allocation process, thereby improving the on-time completion rate of tasks and optimizing the overall scheduling efficiency). S4.3, Based on dynamic editing efficiency parameters Based on the current workload of personnel, a preliminary screening of candidate editors suitable for the current editing task is conducted. Then, based on the occupancy status of editing workstations and available time windows, a set of candidate editing workstations that can execute the editing task within the task's schedulable time frame is selected, forming a resource combination. Specifically, during task scheduling, for each editor, their current time... Dynamic editing efficiency parameters Based on the current number of tasks and the time window of each task, a set of candidate editors with high efficiency and manageable load is selected. At the same time, the occupancy status and future available time period of each editing workstation are traversed to select a set of workstations available within the expected start and end time range of the editing task. Subsequently, each candidate editor is combined with each available workstation to generate a resource combination, and information such as personnel availability, workstation occupancy time, and task start and end time windows for each combination is recorded. This provides a set of feasible solutions for subsequent resource allocation calculation based on cost functions, ensuring that the scheduling generates the optimal allocation candidate while guaranteeing personnel efficiency, load balancing, and workstation availability. S4.4 For editing tasks in the priority queue of tasks to be scheduled, calculate the resource idle and waiting cost factors under different resource combinations (all time predictions are based on the predicted completion time under different resource combinations). The process involves several steps: First, the combination of each candidate editor and available editing workstation, along with their schedulable time window, is obtained. Then, based on the current task occupancy status of the workstation, the waiting time is predicted. If the editor still has unfinished tasks before the expected start time, the waiting time is the difference between the end time of the previous task and the start time of the current task; otherwise, it is zero. Simultaneously, the waiting time for personnel at the workstation is predicted based on its current occupancy status. If the workstation is still occupied before the expected start time of the task, the waiting time is the difference between the available time of the workstation and the expected start time of the task; otherwise, it is zero. Finally, combining the queue of tasks to be scheduled and the task sequence dependencies, the queuing time that a task might experience due to incomplete preceding tasks is predicted. These factors combined form the idle and waiting cost factors for each resource combination, providing basic data for subsequent comprehensive cost function calculations. S4.5 Constructing a comprehensive cost function based on resource idle and waiting cost factors (the process of constructing the comprehensive cost function is as follows: for each task to be scheduled and its candidate resource combination, the various resource idle and waiting cost factors (including the waiting time of editing personnel, the waiting time of editing personnel, and the waiting time of task queuing, etc.) calculated in step S4.4 are normalized to eliminate the differences in the dimensions and numerical ranges of different factors, so that each factor is mapped to a uniform scale range (such as 0~1); then, weights are assigned to each normalized factor to reflect its relative importance to the overall scheduling efficiency; then, the various normalized resource idle and waiting cost factors are combined according to their weights to form a comprehensive cost function), quantifying the editing task. Assigned to editors With editing station The idle loss is used to calculate the comprehensive cost of all resource combinations, and the resource combination with the minimum comprehensive cost is selected as the scheduling decision result for the current editing task, so as to achieve continuous occupation of editing personnel and editing workstations and optimal overall scheduling efficiency (after selecting the scheduling decision result, the planned completion time of the task is calculated based on the actual start time and the estimated editing processing time determined by the combination). This time will be output as the scheduling result and used to update the schedulable time window for subsequent tasks. S4.6 For tasks with a sequential relationship within the same video content, before executing the scheduling decision, determine the completion status of its predecessor task (if not completed, postpone the current task scheduling; if completed, allow it to enter the execution queue), and base the decision on the estimated completion time of the edit (here, the estimated completion time of the edit is the planned completion time). Update the schedulable time window for subsequent tasks to prevent workflow conflicts; specifically: when scheduling editing tasks with dependencies on each other within the same video content, first identify the set of predecessor tasks for each task (all the necessary predecessor tasks for the current task), and obtain the planned completion time of each predecessor task. The current task is either in its current state or its actual completion status. Before making a scheduling decision for the current task, the completion status of its predecessor tasks is checked: if all predecessor tasks have been completed, the current task is allowed to enter the scheduling execution queue, and the earliest schedulable start time window of the current task is updated according to the completion time of the predecessor tasks; if any predecessor task has not been completed, the current task is temporarily postponed, and its schedulable time window is dynamically adjusted to avoid conflicts. Through this mechanism, it is ensured that the sequential dependency constraints between tasks are strictly followed when making scheduling decisions, preventing process conflicts or rework caused by the incomplete completion of predecessor tasks due to premature scheduling, thereby maintaining the internal process logic and overall production efficiency of video content production. S4.7. During task execution, collect real-time information on editing task progress, personnel status changes, and workstation anomalies: If the task is completed ahead of schedule, release resources and trigger rescheduling; if the task is delayed, re-estimate the estimated completion time based on the current progress (referred to as re-estimated completion time, i.e., the re-predicted completion time), and determine whether a recalculation of the allocation plan is necessary; if personnel or workstations are unavailable, trigger emergency rescheduling. Only when the re-estimated completion time exceeds the planned completion time... Rescheduling is only triggered when the duration of the task reaches 15% of the total estimated duration or when a device crashes, in order to avoid frequent adjustments due to minor progress fluctuations. S4.8 Output the final editing task allocation order among editors and editing stations: including a task execution order table for each editor; a task time schedule for each editing station; and the start time and planned completion time for each editing task. The overall completion time prediction for each video content; specifically: after completing the resource allocation decision for all scheduled task priority queues, based on the scheduling decision result of step S4.5: first, establish a task execution sequence table for each editor, recording the start time and planned completion time of each task within the scheduling cycle. The system generates the corresponding video content; secondly, it generates a task occupancy schedule for each editing workstation, specifying the time period and order of each task at the workstation; simultaneously, it outputs the designated editor, editing workstation, task start time, and planned completion time for each editing task. Based on this, the completion times of all tasks under the same video content are integrated to form an overall completion time prediction for the video content. This result not only clarifies the task allocation order for personnel and workstations, but also provides an operable time arrangement basis for scheduling, progress tracking and subsequent process optimization, ensuring that editing tasks are completed efficiently in the predetermined order.

[0021] Example 2: This example provides an intelligent scheduling system for the media content production process, including a memory, a processor, and a computer program stored in the memory and executable on the processor. The processor executes the computer program to implement the intelligent scheduling method for the media content production process described in Example 1.

[0022] The embodiments of the present invention have been described in detail above with reference to the accompanying drawings. However, the present invention is not limited thereto. Various changes can be made within the scope of knowledge possessed by those skilled in the art without departing from the spirit of the present invention.

Claims

1. An intelligent scheduling method for media content production, characterized in that, include: S1. Collect historical and current editing data during the media content production process, and preprocess the editing data; S2. By using statistical analysis methods and introducing dynamic editing efficiency parameters, team efficiency synchronization index and efficiency trend index, a editing duration prediction model is constructed. Based on the pre-processed editing data, the expected editing processing time and the expected completion time of each editing task are calculated. S3. Determine the scheduling constraints and optimization objectives in the media content production process based on the expected editing processing time; S4. Based on scheduling constraints and optimization objectives, perform editing task scheduling calculations to intelligently schedule the allocation order of editing tasks between editing personnel and editing workstations.

2. The intelligent scheduling method in the media content production process according to claim 1, characterized in that, In S1, the editing data includes: the duration of the footage for the editing task and the identifier of the editor who performed the editing task.

3. The intelligent scheduling method in the media content production process according to claim 2, characterized in that, In step S2, a editing duration prediction model is constructed to calculate the estimated editing processing time and the estimated completion time of each editing task based on the preprocessed editing data. This includes the following steps: S2.1 Collect the actual completion time of historical editing tasks. Based on the preprocessed historical editing data, establish a mapping relationship between the material duration of historical editing tasks and the actual completion time of historical editing tasks for each editor, and calculate the individual's static editing efficiency parameters. By combining the collaborative network relationship among editors with a dynamic efficiency change model, dynamic editing efficiency parameters reflecting their processing capabilities under actual working conditions are calculated, and team efficiency synchronization index and efficiency trend index at the start of the task are generated. S2.2, using the duration of footage from historical editing tasks and dynamic editing efficiency parameters. Using the team efficiency synchronization index and efficiency trend index at the start of the task as input features, and the actual completion time of the editing task as the target output, an editing time prediction model is constructed. S2.3 Based on the preprocessed editing data, the estimated editing processing time of the editing task is calculated using the editing duration prediction model; S2.

4. Calculate the estimated completion time of the editing task based on the start time of the editing task.

4. The intelligent scheduling method in the media content production process according to claim 3, characterized in that, In step S2.1, by combining the collaborative network relationship among editors with a dynamic efficiency change model, dynamic editing efficiency parameters reflecting their processing capabilities under actual working conditions are calculated, and team efficiency synchronization index and efficiency trend index at the start of the task are generated, including the following steps: S2.

11. Collect collaboration records of editors within the time window, and construct a collaboration weight matrix among editors based on these records. ; S2.

12. For the current editing task, extract the task feature vector and calculate the task similarity. S2.13, Based on the Collaboration Weight Matrix The cooperative coupling coefficient is calculated by combining the time decay factor and task similarity. ; S2.14, Based on Cooperative Coupling Coefficient Build dynamic editing efficiency parameters for each editor The dynamic equations that change with time; The trajectory of the editor's efficiency over time is generated by iterative calculation at discrete time steps. S2.

15. Based on the trajectory of editors' efficiency changes over time, calculate the team efficiency synchronization index and efficiency trend index.

5. The intelligent scheduling method in the media content production process according to claim 4, characterized in that, In S2.13, based on the cooperative weight matrix The cooperative coupling coefficient is calculated by combining the time decay factor and task similarity. This includes the following steps: For editors and The most recent collaborative task was analyzed, and the collaboration time interval was calculated. Simultaneously, the number of editors within the time window was recorded. and Task association strength ; A time decay factor is constructed based on the collaboration time interval and the strength of task association. ; Cooperative weight matrix With time decay factor Combined, calculate the basic synergistic coupling coefficient. : If the task similarity exceeds a set threshold This enhances the basic synergistic coupling coefficient.

6. The intelligent scheduling method in the media content production process according to claim 4, characterized in that, In S2.14, the dynamic editing efficiency parameter for each editor. The time-varying dynamic equation consists of individual self-decay and baseline regression terms, cooperative coupling terms, task load influence terms, and random disturbance terms.

7. The intelligent scheduling method in the media content production process according to claim 1, characterized in that, In step S3, determining the scheduling constraints and optimization objectives during the media content production process based on the expected editing processing time includes the following steps: S3.1 Based on the estimated editing processing time and the estimated completion time of the editing task, combined with the predetermined release time of the corresponding video content, an adaptive correction method for time uncertainty based on a penalty mechanism is used to calculate the time margin of the editing task. And according to the time margin and urgency evaluation indicators Determine the time feasibility status of the editing task within the current scheduling cycle; S3.2 Based on the estimated editing processing time of each editing task, construct a task occupancy interval for each editor and each editing workstation on the future timeline to generate resource occupancy constraints; S3.3 Based on resource occupancy constraints, calculate the idle time segments of each editor under the current scheduling scheme and the waiting time caused by the unavailability of editors; S3.4 For multiple editing tasks corresponding to the same video content, identify the order of pre- and post-processing in the production process, and determine the earliest executable time of each task based on the expected editing processing time. Use the order of pre- and post-processing and the earliest executable time as the order dependency constraint between tasks. S3.

5. Based on resource space constraints and sequential dependency constraints, the goal of scheduling optimization is to minimize the maximum completion time of all editing tasks so that all videos are completed before the predetermined release time.

8. The intelligent scheduling method in the media content production process according to claim 7, characterized in that, In step S3.1, an adaptive correction method for time uncertainty based on a penalty mechanism is used to calculate the time margin for the editing task. This includes the following steps: S3.11, Video release time The time difference between the estimated completion time of the editing task and the original time margin is used as the raw time margin. ; S3.12, For editing tasks Based on the completion records of similar editing tasks in the past, the fluctuation intensity of the expected completion time is calculated to generate the asymmetric completion time fluctuation range of the editing task. S3.13, Based on the asymmetric completion time fluctuation range and video release time Calculate editing tasks Overdue risk coefficient At the same time, an urgency evaluation index was introduced. ; S3.

14. Adjusting the original time margin using a risk adjustment coefficient. Make corrections to generate the corrected time margin. .

9. The intelligent scheduling method in the media content production process according to claim 1, characterized in that, In step S4, the editing task scheduling calculation is performed based on scheduling constraints and optimization objectives, including the following steps: S4.

1. Based on resource space constraints, sequential dependency constraints, and scheduling optimization objectives, establish an editing task scheduling model; S4.2, Based on the time leeway for each editing task Combining urgency evaluation indicators Calculate the urgency weight of the editing task; and generate a priority queue of tasks to be scheduled based on the release time of the video to which the editing task belongs. S4.3, Based on dynamic editing efficiency parameters Based on the current workload of personnel, preliminary screening of candidate editors suitable for the current editing task is conducted. Based on the occupancy status of editing workstations and available time windows, a set of candidate editing workstations that can perform the editing task within the schedulable time range is selected to form a resource combination. S4.4 For the editing tasks in the priority queue of tasks to be scheduled, calculate the resource idle and waiting cost factors under different resource combinations; S4.5 Construct a comprehensive cost function based on resource idle and waiting cost factors to calculate the comprehensive cost for all resource combinations, and select the resource combination with the minimum comprehensive cost as the scheduling decision result for the current editing task; S4.6 For tasks with a sequential relationship under the same video content, determine the completion status of its predecessor task before executing the scheduling decision; S4.

7. During the task execution process, collect real-time information on the editing task progress, personnel status changes, and workstation anomalies. S4.8 Output the scheduling results of the final editing tasks, which are allocated between the editors and editing stations.

10. An intelligent scheduling system for media content production, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: The processor executes a computer program to implement the intelligent scheduling method in the media content production process as described in any one of claims 1-9.

Images