Intelligent task scheduling method and system combined with artificial intelligence optimization

Abstract

The present invention discloses an intelligent task scheduling method and system combined with artificial intelligence optimization, which relates to the technical field of task scheduling. The present invention calculates the correlation coefficient between the execution data and the priority value of each task in the historical tasks, and classifies the execution data of each task according to the correlation coefficient. Then, based on the classification results, i.e., the first and second categories of data, the first and second priority indexes are calculated in a data-driven manner, thereby obtaining a more objective and accurate priority index. The first and second priority indexes are then used to compensate the original priority values, thereby obtaining a more comprehensive and reasonable final priority index. Finally, the corrected priority index is used as a training label. For the prediction model, this can learn a more accurate task priority pattern, thereby improving the reliability of the prediction results and the effectiveness and quality of the scheduling strategy.

Description

Technical Field

[0001] The present invention relates to the field of task scheduling technology, and in particular to an intelligent task scheduling method and system combined with artificial intelligence optimization. Background Art

[0002] With the growing demand for cloud computing and big data processing, how to efficiently manage computing resources has become an important research area. Intelligent task scheduling has become one of the key solutions in this field by rationally allocating computing resources to improve efficiency and reduce costs. In recent years, the development of artificial intelligence technology has brought new possibilities to intelligent task scheduling, enabling it to handle complex scheduling problems more accurately and flexibly.

[0003] Existing intelligent task scheduling methods combined with artificial intelligence optimization mainly rely on historical data to train prediction models. During this process, expert experience is used to set the priority value of each task, and each task priority value is used as a label for subsequent model training. By leveraging the knowledge and experience of experts, this method can quickly assign priorities to tasks, thereby achieving preliminary task scheduling optimization. Specifically, the advantages of existing methods include: because expert experience can be directly used to set priorities, the task scheduling system can be deployed in a short time. At the same time, for some highly structured application scenarios with relatively simple task types, expert experience can provide relatively reasonable priority settings, thereby achieving better scheduling results in the early stages.

[0004] However, although these methods excel in some aspects, they also have some inherent flaws and limitations:

[0005] Existing methods rely primarily on expert experience to set task priorities. This approach is inevitably highly subjective, and different experts may give different priority assessments for the same task, making it difficult to ensure the consistency and accuracy of priority values. Furthermore, because existing priority values ​​are set based on subjective judgment, this can easily lead to inaccurate labels used during model training, which in turn affects the model's generalization ability and the quality of the final scheduling decision. For example, when resources are limited, incorrect priority setting can easily result in critical tasks not being processed in a timely manner, causing unnecessary delays and wasted resources.

[0006] Therefore, the existing technology urgently needs a technical solution for an intelligent task scheduling method and system that combines artificial intelligence optimization. Summary of the Invention

[0007] In order to solve the above technical problems, the present invention provides an intelligent task scheduling method combined with artificial intelligence optimization, which specifically includes the following steps:

[0008] Step S1: collect the execution data and priority value of each task in the historical tasks and perform preprocessing;

[0009] Step S2: analyzing the execution data and priority value of each task in the pre-processed historical tasks, compensating the priority value of each task in the historical tasks based on the analysis results, and obtaining a final priority index of each task in the historical tasks;

[0010] Step S2a, calculating the correlation coefficient between the execution data of each task in the pre-processed historical tasks and the priority value of each task in the historical tasks;

[0011] Step S2a1: Obtain all execution data of each task in the pre-processed historical tasks and take the average;

[0012] Step S2a2: Calculate the average priority value of all tasks in the pre-processed historical tasks;

[0013] Step S2a3: Calculate the correlation coefficient between the execution data of each task in the preprocessed historical tasks and the priority value of each task in the preprocessed historical tasks based on the mean of all execution data of each task in the preprocessed historical tasks and the mean of the priority values ​​of all tasks in the preprocessed historical tasks;

[0014] The calculation formula for the correlation coefficient between the execution data of each task in the historical tasks and the priority value of each task in the historical tasks is:

[0015] ;

[0016] Where, Represents the correlation coefficient between the jth execution data of the i-th task in the historical tasks and the priority value of the i-th task in the historical tasks; Represents the jth execution data of the i-th task in the historical tasks; Represents the mean of all execution data of the i-th task in the historical tasks; Represents the priority value of the i-th task in the history tasks; Represents the average priority value of all tasks in the historical tasks; Represents the total number of execution data of the i-th task in the historical tasks;

[0017] Step S2b: Calculate a first correlation coefficient threshold and a second correlation coefficient threshold based on the correlation coefficient;

[0018] Step S2b1, obtaining the absolute value of the correlation coefficient, and sorting the correlation coefficients from large to small according to the absolute value;

[0019] Step S2b2: Obtain the correlation coefficients of the top N items according to the sorting results;

[0020] Step S2b3: Calculate the average value of the top N correlation coefficients to obtain a first correlation coefficient threshold;

[0021] Step S2b4: Count the number K of all correlation coefficients, and remove the top N correlation coefficients according to the sorting result to obtain the correlation coefficients ranked KN;

[0022] Step S2b5: Calculate the average value of the correlation coefficients ranked as KN to obtain a second correlation coefficient threshold;

[0023] Step S2c, filtering the execution data of each task in the historical tasks according to the first correlation coefficient threshold and the second correlation coefficient threshold, and obtaining the first category data and the second category data of each task in the historical tasks according to the filtering results;

[0024] Step S2c1: if the absolute value of the correlation coefficient between the current execution data of the current task in the historical tasks and the priority value of the current task in the historical tasks is greater than or equal to a first correlation coefficient threshold and less than 1, classifying the current execution data of the current task in the historical tasks as first category data;

[0025] Step S2c2: if the absolute value of the correlation coefficient between the current execution data of the current task in the historical tasks and the priority value of the current task in the historical tasks is less than the first correlation coefficient threshold and greater than or equal to the second correlation coefficient threshold, classifying the current execution data of the current task in the historical tasks as the second category of data;

[0026] Step S2c3: if the absolute value of the correlation coefficient between the current execution data of the current task in the historical tasks and the priority value of the current task in the historical tasks is less than the second correlation coefficient threshold and greater than or equal to 0, then the current execution data of the current task in the historical tasks is eliminated;

[0027] Step S2d, calculating the first priority index and the second priority index of each task in the historical tasks according to the first category data and the second category data of each task in the historical tasks;

[0028] Step S2d1: Calculate the mean of all execution data in the first category of data for each task in the historical tasks, and calculate the standard deviation of all execution data in the first category of data for each task in the historical tasks based on the mean;

[0029] Step S2d2: Calculate the mean of all execution data in the second category data of each task in the historical tasks, and calculate the standard deviation of all execution data in the second category data of each task in the historical tasks based on the mean;

[0030] Step S2d3, calculating the first priority index and the second priority index of each task in the historical tasks based on the mean of all execution data in the first category of data of each task in the historical tasks, the standard deviation of all execution data in the first category of data of each task in the historical tasks, and the mean of all execution data in the second category of data of each task in the historical tasks, and the standard deviation of all execution data in the second category of data of each task in the historical tasks;

[0031] The calculation formula for the first priority index and the second priority index of each task in the historical tasks is:

[0032] ;

[0033] ;

[0034] Where, Represents the first priority index of the i-th task in the history tasks; Represents the second priority index of the i-th task in the history tasks; Represents the jth execution data in the first category of data of the i-th task in the historical tasks; Represents the mean of all execution data in the first category of data for the i-th task in the historical tasks; Represents the standard deviation of all execution data in the first category of data for the i-th task in the historical tasks; Represents the jth execution data in the second category of data of the i-th task in the historical tasks; Represents the mean of all execution data in the second category of data for the i-th task in the historical tasks; Represents the standard deviation of all execution data in the second category of data for the i-th task in the historical tasks; Represents the total number of execution data in the first category of data; Represents the total number of execution data in the second category of data;

[0035] Step S2e, combining the first priority index and the second priority index of each task in the historical tasks and the priority value of each task in the historical tasks to calculate a final priority index of each task in the historical tasks;

[0036] The calculation formula for obtaining the final priority index of each task in the historical tasks is:

[0037] ;

[0038] Where, Represents the final priority index of the i-th task in the history tasks; Represents the first priority index of the i-th task in the history tasks; Represents the second priority index of the i-th task in the history tasks; Represents the priority value of the i-th task in the history tasks;

[0039] Step S3: calling the pre-trained prediction model, and using the final priority index of each task in the historical tasks as the training label of the pre-trained prediction model to obtain an optimized pre-trained prediction model;

[0040] Step S4: Input the execution data of the task to be predicted into the optimized pre-trained prediction model, output the prediction result, and formulate the optimal scheduling strategy based on the prediction result.

[0041] The intelligent task scheduling system combined with artificial intelligence optimization implements the intelligent task scheduling method combined with artificial intelligence optimization as described above, including the following modules:

[0042] Data collection and preprocessing module: used to collect the execution data and priority value of each task in the historical tasks and perform preprocessing;

[0043] Final priority index analysis module: connected to the data collection and preprocessing module, used to analyze the execution data and priority value of each task in the preprocessed historical tasks, compensate the priority value of each task in the historical tasks based on the analysis results, and obtain the final priority index of each task in the historical tasks;

[0044] Prediction model optimization module: This module is connected to the final priority index analysis module to call the pre-trained prediction model and use the final priority index of each task in the historical tasks as the training label of the pre-trained prediction model to obtain an optimized pre-trained prediction model.

[0045] Optimal scheduling strategy formulation module: connected to the prediction model optimization module, used to input the execution data of the task to be predicted into the optimized pre-trained prediction model, output the prediction results, and formulate the optimal scheduling strategy based on the prediction results.

[0046] The embodiments of the present invention have the following technical effects:

[0047] The present invention calculates the correlation coefficient between the execution data and the priority value of each task in the historical tasks, and classifies the execution data of each task according to the correlation coefficient. In the classification process, the first and second correlation coefficient thresholds are obtained respectively in a scientific calculation manner, and the execution data of each task are classified using the first and second correlation coefficient thresholds. Then, based on the classified results, i.e., the first and second categories of data, the first and second priority indexes are calculated in a data-driven manner, thereby obtaining a more objective and accurate priority index. The first and second priority indexes are then used to compensate the original priority value, thereby obtaining a more comprehensive and reasonable final priority index. Finally, the corrected priority index is used as a training label. For the prediction model, this can learn a more accurate task priority pattern, thereby improving the reliability of the prediction results and the effectiveness and quality of the scheduling strategy. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] In order to more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the specific embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0049] Figure 1 This is a flow chart of an intelligent task scheduling method combined with artificial intelligence optimization provided by an embodiment of the present invention;

[0050] Figure 2 This is a framework diagram of an intelligent task scheduling system combined with artificial intelligence optimization provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0051] To make the objectives, technical solutions, and advantages of the present invention more clear, the technical solutions of the present invention are described clearly and completely below. Obviously, the embodiments described are only some of the embodiments of the present invention, not all of them. All other embodiments derived by persons of ordinary skill in the art based on the embodiments of the present invention without inventive effort are also within the scope of protection of the present invention.

[0052] Example 1: Figure 1 As shown, the present invention provides an intelligent task scheduling method combined with artificial intelligence optimization, comprising the following steps:

[0053] Step S1: collect the execution data and priority value of each task in the historical tasks and perform preprocessing;

[0054] It is worth noting that when the application scenario is a task scheduling system in a data center, historical tasks may include different types of tasks such as data backup, data analysis, and machine learning model training. For each task, the following execution data is usually included: resource usage, i.e., CPU utilization, memory usage, disk I / O, etc.; task completion time, i.e., the time from start to finish; error rate, i.e., the number or proportion of errors that occurred during execution; dependencies, i.e., whether the task depends on the completion of other tasks; for example, a "data backup" task may need to record the CPU time used, the storage space occupied, and whether it was successfully completed as execution data;

[0055] It is worth further explaining that in the current implementation process of setting priority values ​​based on expert experience, experts usually set priority values ​​based on the importance and urgency of the task and its impact on the business. For example, if a "data backup" task involves the security of critical data, it will be assigned a higher priority value, and the priority value is usually set in the range of 1 to 10. During the setting process, experts will consider factors such as the impact of task failure on the business, the time window required to complete the task, and whether there are external dependencies, such as waiting for responses from other systems.

[0056] It is worth further explaining that after collecting the above data, including execution data and priority values, data preprocessing should be performed. This preprocessing process usually includes: missing value processing, that is, for missing data points, the mean is filled or other methods are used to fill them; standardization / normalization, that is, converting data of different scales to the same range, such as scaling all values ​​to between 0 and 1; outlier detection and processing, that is, identifying and correcting outliers to avoid their adverse effects on the analysis results.

[0057] Step S2: analyzing the execution data and priority value of each task in the pre-processed historical tasks, compensating the priority value of each task in the historical tasks based on the analysis results, and obtaining a final priority index of each task in the historical tasks;

[0058] Step S2a, calculating the correlation coefficient between the execution data of each task in the pre-processed historical tasks and the priority value of each task in the historical tasks;

[0059] It is worth noting that by calculating the correlation coefficient, we can quantify the strength of the linear relationship between each execution data and the task priority value. This helps identify which execution data has a significant impact on the task priority value, thus providing a basis for subsequent classification. For example, if the correlation coefficient between the CPU utilization and priority of a task is high, it means that high CPU utilization is an important factor leading to the high priority value of the task.

[0060] Step S2a1: Obtain all execution data of each task in the pre-processed historical tasks and take the average;

[0061] Step S2a2: Calculate the average priority value of all tasks in the pre-processed historical tasks;

[0062] Step S2a3: Calculate the correlation coefficient between the execution data of each task in the preprocessed historical tasks and the priority value of each task in the preprocessed historical tasks based on the mean of all execution data of each task in the preprocessed historical tasks and the mean of the priority values ​​of all tasks in the preprocessed historical tasks;

[0063] The calculation formula for the correlation coefficient between the execution data of each task in the historical tasks and the priority value of each task in the historical tasks is:

[0064] ;

[0065] Where, Represents the correlation coefficient between the jth execution data of the i-th task in the historical tasks and the priority value of the i-th task in the historical tasks; Represents the jth execution data of the i-th task in the historical tasks; Represents the mean of all execution data of the i-th task in the historical tasks; Represents the priority value of the i-th task in the history tasks; Represents the average priority value of all tasks in the historical tasks; Represents the total number of execution data of the i-th task in the historical tasks;

[0066] It is worth noting that the correlation coefficient formula can help understand how each execution data is related to the overall priority of the task. It not only reveals the importance of individual execution data, but also allows this information to be used for further classification and priority adjustment.

[0067] Step S2b: Calculate a first correlation coefficient threshold and a second correlation coefficient threshold based on the correlation coefficient;

[0068] It is worth noting that setting the first and second correlation coefficient thresholds can help distinguish the execution data with the greatest impact on priority (the first category) from the data with the second greatest impact (the second category). The benefits of this method are: First, by screening out the most relevant execution data (the first category), the actual priority of the task can be more accurately reflected. These data usually contain the most important information and can significantly improve the accuracy of the first priority index. Second, the second category of data provides additional information support. Although not as important as the first category, they still help to comprehensively assess the priority of the task. This can combine the priority index of the two types of data to obtain a more stable, reliable and scientific priority index.

[0069] Step S2b1, obtaining the absolute value of the correlation coefficient, and sorting the correlation coefficients from large to small according to the absolute value;

[0070] Step S2b2: Obtain the correlation coefficients of the top N items according to the sorting results;

[0071] Step S2b3: Calculate the average value of the top N correlation coefficients to obtain a first correlation coefficient threshold;

[0072] Step S2b4: Count the number K of all correlation coefficients, and remove the top N correlation coefficients according to the sorting result to obtain the correlation coefficients ranked KN;

[0073] Step S2b5: Calculate the average value of the correlation coefficients ranked as KN to obtain a second correlation coefficient threshold;

[0074] Step S2c, filtering the execution data of each task in the historical tasks according to the first correlation coefficient threshold and the second correlation coefficient threshold, and obtaining the first category data and the second category data of each task in the historical tasks according to the filtering results;

[0075] It is worth noting that by classifying the execution data, the priority index of the task can be adjusted more accurately. The first category of execution data represents the factors that have the greatest impact on the priority, while the second category of execution data provides supplementary information. This classification method makes the evaluation and calculation of the final priority index more detailed and accurate.

[0076] Step S2c1: if the absolute value of the correlation coefficient between the current execution data of the current task in the historical tasks and the priority value of the current task in the historical tasks is greater than or equal to a first correlation coefficient threshold and less than 1, classifying the current execution data of the current task in the historical tasks as first category data;

[0077] Step S2c2: if the absolute value of the correlation coefficient between the current execution data of the current task in the historical tasks and the priority value of the current task in the historical tasks is less than the first correlation coefficient threshold and greater than or equal to the second correlation coefficient threshold, classifying the current execution data of the current task in the historical tasks as the second category of data;

[0078] Step S2c3: if the absolute value of the correlation coefficient between the current execution data of the current task in the historical tasks and the priority value of the current task in the historical tasks is less than the second correlation coefficient threshold and greater than or equal to 0, then the current execution data of the current task in the historical tasks is eliminated;

[0079] Step S2d, calculating the first priority index and the second priority index of each task in the historical tasks according to the first category data and the second category data of each task in the historical tasks;

[0080] It is worth noting that calculating the priority index of the two types of execution data separately can better reflect the different contributions of different types of execution data to the task priority. This not only improves the accuracy of priority evaluation, but also lays a good foundation for subsequent calculations to obtain a more scientific and reasonable final priority index.

[0081] Step S2d1: Calculate the mean of all execution data in the first category of data for each task in the historical tasks, and calculate the standard deviation of all execution data in the first category of data for each task in the historical tasks based on the mean;

[0082] Step S2d2: Calculate the mean of all execution data in the second category data of each task in the historical tasks, and calculate the standard deviation of all execution data in the second category data of each task in the historical tasks based on the mean;

[0083] Step S2d3, calculating the first priority index and the second priority index of each task in the historical tasks based on the mean of all execution data in the first category of data of each task in the historical tasks, the standard deviation of all execution data in the first category of data of each task in the historical tasks, and the mean of all execution data in the second category of data of each task in the historical tasks, and the standard deviation of all execution data in the second category of data of each task in the historical tasks;

[0084] The calculation formula for the first priority index and the second priority index of each task in the historical tasks is:

[0085] ;

[0086] ;

[0087] Where, Represents the first priority index of the i-th task in the history tasks; Represents the second priority index of the i-th task in the history tasks; Represents the jth execution data in the first category of data of the i-th task in the historical tasks; Represents the mean of all execution data in the first category of data for the i-th task in the historical tasks; Represents the standard deviation of all execution data in the first category of data for the i-th task in the historical tasks; Represents the jth execution data in the second category of data of the i-th task in the historical tasks; Represents the mean of all execution data in the second category of data for the i-th task in the historical tasks; Represents the standard deviation of all execution data in the second category of data for the i-th task in the historical tasks; Represents the total number of execution data in the first category of data; Represents the total number of execution data in the second category of data;

[0088] It is worth noting that the above calculation formula group ensures that the calculation of the priority index is based on the statistical characteristics of the data itself, such as the mean and standard deviation, rather than a simple subjective assignment. Therefore, this can significantly improve the scientificity and objectivity of the calculation of the first and second priority indices.

[0089] Step S2e, combining the first priority index and the second priority index of each task in the historical tasks and the priority value of each task in the historical tasks to calculate a final priority index of each task in the historical tasks;

[0090] The calculation formula for obtaining the final priority index of each task in the historical tasks is:

[0091] ;

[0092] Where, Represents the final priority index of the i-th task in the history tasks; Represents the first priority index of the i-th task in the history tasks; Represents the second priority index of the i-th task in the history tasks; Represents the priority value of the i-th task in the history tasks;

[0093] It is worth noting that comprehensive consideration of the original priority values ​​and the new priority index, namely the first and second priority indexes, can effectively balance the empirical advantages of the traditional method with the data-driven advantages of the new method, thereby ensuring that the final priority index reflects both historical experience and the latest data analysis results.

[0094] Step S3: calling the pre-trained prediction model, and using the final priority index of each task in the historical tasks as the training label of the pre-trained prediction model to obtain an optimized pre-trained prediction model;

[0095] It is worth noting that the pre-trained prediction model called is the prediction model currently used, and its training process is as follows: first, a large amount of historical task execution data and its corresponding expert-set priority values ​​are collected, and after preprocessing, the corresponding historical task execution data is used as the data set for model training, and the priority value set by the expert is used as the label during model training. Among them, the selected prediction model usually adopts a neural network model. During the training process, the model attempts to learn the mapping relationship between the input execution data and the output priority value. Common loss functions include mean square error, and evaluation indicators include accuracy or F1 score, etc. Finally, the model performance is evaluated by cross-validation technology, and the model parameters are adjusted according to the verification results to optimize the performance, thereby obtaining a trained prediction model. However, since the training labels used in the previous model training, that is, the priority values ​​set by experts, need to be replaced with the final priority index obtained based on data-driven analysis in this application, the prediction model needs to be retrained. The training process is basically the same as the existing one. The only difference is that the labels used during training are inconsistent, so I will not go into details here.

[0096] Step S4: Input the execution data of the task to be predicted into the optimized pre-trained prediction model, output the prediction results, and formulate the optimal scheduling strategy based on the prediction results;

[0097] It is worth noting that based on the prediction results output by the above-mentioned optimized prediction model, a more reasonable optimal scheduling strategy can be formulated; for example, if the prediction results show that certain tasks will face higher load demands in the future, then the optimal scheduling strategy formulated can be to allocate more resources in advance to prevent performance bottlenecks; or, if the prediction results show that the priority of certain tasks is expected to increase, then the optimal scheduling strategy formulated can be to dynamically adjust their execution order to ensure that critical tasks are processed first.

[0098] Example 2: Figure 2 As shown, the present invention also proposes an intelligent task scheduling system combined with artificial intelligence optimization, which executes the above-mentioned intelligent task scheduling method combined with artificial intelligence optimization, including the following modules:

[0099] Data collection and preprocessing module: used to collect the execution data and priority value of each task in the historical tasks and perform preprocessing;

[0100] Final priority index analysis module: connected to the data collection and preprocessing module, used to analyze the execution data and priority value of each task in the preprocessed historical tasks, compensate the priority value of each task in the historical tasks based on the analysis results, and obtain the final priority index of each task in the historical tasks;

[0101] Prediction model optimization module: This module is connected to the final priority index analysis module to call the pre-trained prediction model and use the final priority index of each task in the historical tasks as the training label of the pre-trained prediction model to obtain an optimized pre-trained prediction model.

[0102] Optimal scheduling strategy formulation module: connected to the prediction model optimization module, used to input the execution data of the task to be predicted into the optimized pre-trained prediction model, output the prediction results, and formulate the optimal scheduling strategy based on the prediction results.

[0103] It should be noted that the terms used in the present invention are only for describing specific embodiments and are not intended to limit the scope of this application. As shown in the present specification, unless the context clearly indicates an exception, the words "one", "a", "a kind of" and / or "the" do not specifically refer to the singular and may also include the plural. The terms "comprise", "include" or any other variants thereof are intended to cover non-exclusive inclusion, so that the process, method or device comprising a series of elements includes not only those elements, but also includes other elements not explicitly listed, or also includes elements inherent to such process, method or device. In the absence of further restrictions, the elements defined by the sentence "comprise a..." do not exclude the presence of other identical elements in the process, method or device comprising the elements.

[0104] It should also be noted that the terms "center", "up", "down", "left", "right", "vertical", "horizontal", "inside", "outside", etc., indicating orientations or positional relationships, are based on the orientations or positional relationships shown in the accompanying drawings. They are only for the convenience of describing the present invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation. Therefore, they cannot be understood as limitations on the present invention. Unless otherwise clearly specified and limited, the terms "installed", "connected", "connected", etc. should be understood in a broad sense. For example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be a direct connection, or an indirect connection through an intermediate medium, or it can be a communication between the internal parts of two elements. For those of ordinary skill in the art, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.

[0105] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the above embodiments, or replace some or all of the technical features therein with equivalents. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the technical solutions of the embodiments of the present invention.

Claims

1. An intelligent task scheduling method combined with artificial intelligence optimization is characterized by: The following steps are involved: Step S1: collect the execution data and priority value of each task in the historical tasks and perform preprocessing; Step S2: analyzing the execution data and priority value of each task in the pre-processed historical tasks, compensating the priority value of each task in the historical tasks based on the analysis results, and obtaining a final priority index of each task in the historical tasks; Step S2a, calculating the correlation coefficient between the execution data of each task in the pre-processed historical tasks and the priority value of each task in the historical tasks; Step S2b: Calculate a first correlation coefficient threshold and a second correlation coefficient threshold based on the correlation coefficient; Step S2c, filtering the execution data of each task in the historical tasks according to the first correlation coefficient threshold and the second correlation coefficient threshold, and obtaining the first category data and the second category data of each task in the historical tasks according to the filtering results; Step S2c1: if the absolute value of the correlation coefficient between the current execution data of the current task in the historical tasks and the priority value of the current task in the historical tasks is greater than or equal to a first correlation coefficient threshold and less than 1, classifying the current execution data of the current task in the historical tasks as first category data; Step S2c2: if the absolute value of the correlation coefficient between the current execution data of the current task in the historical tasks and the priority value of the current task in the historical tasks is less than the first correlation coefficient threshold and greater than or equal to the second correlation coefficient threshold, classifying the current execution data of the current task in the historical tasks as the second category of data; Step S2c3: if the absolute value of the correlation coefficient between the current execution data of the current task in the historical tasks and the priority value of the current task in the historical tasks is less than the second correlation coefficient threshold and greater than or equal to 0, then the current execution data of the current task in the historical tasks is eliminated; Step S2d, calculating the first priority index and the second priority index of each task in the historical tasks according to the first category data and the second category data of each task in the historical tasks; Step S2e, combining the first priority index and the second priority index of each task in the historical tasks and the priority value of each task in the historical tasks to calculate a final priority index of each task in the historical tasks; Step S3: calling the pre-trained prediction model, and using the final priority index of each task in the historical tasks as the training label of the pre-trained prediction model to obtain an optimized pre-trained prediction model; Step S4: Input the execution data of the task to be predicted into the optimized pre-trained prediction model, output the prediction result, and formulate the optimal scheduling strategy based on the prediction result.

2. The intelligent task scheduling method combined with artificial intelligence optimization according to claim 1 is characterized in that: The calculation of the correlation coefficient between the execution data of each task in the preprocessed historical tasks and the priority value of each task in the historical tasks includes: Step S2a1: Obtain all execution data of each task in the pre-processed historical tasks and take the average; Step S2a2: Calculate the average priority value of all tasks in the pre-processed historical tasks; Step S2a3: Based on the mean of all execution data of each task in the preprocessed historical tasks and the mean of the priority values ​​of all tasks in the preprocessed historical tasks, calculate the correlation coefficient between the execution data of each task in the preprocessed historical tasks and the priority value of each task in the historical tasks.

3. The intelligent task scheduling method combined with artificial intelligence optimization according to claim 1 is characterized in that: The first correlation coefficient threshold and the second correlation coefficient threshold are calculated based on the correlation coefficient, including: Step S2b1, obtaining the absolute value of the correlation coefficient, and sorting the correlation coefficients from large to small according to the absolute value; Step S2b2: Obtain the correlation coefficients of the top N items according to the sorting results; Step S2b3: Calculate the average value of the top N correlation coefficients to obtain a first correlation coefficient threshold; Step S2b4: Count the number K of all correlation coefficients, and remove the top N correlation coefficients according to the sorting result to obtain the correlation coefficients ranked KN; Step S2b5: Take the average value of the correlation coefficients ranked as KN to obtain a second correlation coefficient threshold.

4. The intelligent task scheduling method combined with artificial intelligence optimization according to claim 1 is characterized in that: The first priority index and the second priority index of each task in the historical tasks are calculated based on the first category data and the second category data of each task in the historical tasks, respectively, including: Step S2d1: Calculate the mean of all execution data in the first category of data for each task in the historical tasks, and calculate the standard deviation of all execution data in the first category of data for each task in the historical tasks based on the mean; Step S2d2: Calculate the mean of all execution data in the second category data of each task in the historical tasks, and calculate the standard deviation of all execution data in the second category data of each task in the historical tasks based on the mean; Step S2d3: Calculate the first priority index and the second priority index of each task in the historical tasks based on the mean of all execution data in the first category of data of each task in the historical tasks, the standard deviation of all execution data in the first category of data of each task in the historical tasks, and the mean of all execution data in the second category of data of each task in the historical tasks, and the standard deviation of all execution data in the second category of data of each task in the historical tasks.

5. An intelligent task scheduling system combined with artificial intelligence optimization is applied to the intelligent task scheduling method combined with artificial intelligence optimization according to any one of claims 1 to 4, characterized in that: The system comprises: Data collection and preprocessing module: used to collect the execution data and priority value of each task in the historical tasks and perform preprocessing; Final priority index analysis module: connected to the data collection and preprocessing module, used to analyze the execution data and priority value of each task in the preprocessed historical tasks, compensate the priority value of each task in the historical tasks based on the analysis results, and obtain the final priority index of each task in the historical tasks; Prediction model optimization module: This module is connected to the final priority index analysis module to call the pre-trained prediction model and use the final priority index of each task in the historical tasks as the training label of the pre-trained prediction model to obtain an optimized pre-trained prediction model. Optimal scheduling strategy formulation module: connected to the prediction model optimization module, used to input the execution data of the task to be predicted into the optimized pre-trained prediction model, output the prediction results, and formulate the optimal scheduling strategy based on the prediction results.

Images