Cross-platform data task scheduling system and method based on authorization chain

Abstract

The invention discloses a cross-platform data task scheduling system and method based on an authorization chain, and relates to the technical field of data task scheduling. When a task processing platform receives a task, task data is preprocessed, a task database is constructed according to task features obtained through preprocessing, and the task processing platform starts self-inspection; the method comprises the steps of selecting a preset task processing platform according to a real-time operation load, judging a scheduling demand of a task by considering a task matching degree, analyzing a transmission condition between the platforms in real time, selecting a to-be-selected task processing platform according to the transmission condition between the platforms, and comprehensively considering the matching degree and the transmission condition of the to-be-selected task processing platform to the task. Analyzing the predicted completion rate of the task, analyzing the real-time task execution state of the optimal task processing platform, and selecting the optimal task processing platform to execute the task, the task load of each platform can be reasonably balanced by considering the dynamic allocation mode of the platform load and the operation state, and partial task allocation imbalance is avoided.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of data task scheduling, in particular to a cross-platform data task scheduling system and method based on an authorization chain. BACKGROUND

[0002] In the existing cross-platform data task scheduling technology, the mining of the characteristics of the task itself is often not deep enough, the key attributes of the task are not fully extracted, and the system task information reserve is not established, and the past task processing experience and effect of each processing platform are also less combined. The scheduling process depends on simple task type division or platform basic function parameters for matching, which cannot accurately judge the processing capacity of the platform for a specific task, resulting in that the task is often allocated to a platform that is not good at processing such demand, and then frequent processing deviation, result error and other problems occur, which seriously affect the overall quality and reliability of cross-platform task processing. The existing cross-platform data task scheduling mode lacks dynamic perception and flexible deployment mechanism for the real-time running state of each processing platform. When allocating tasks, the current load of different platforms is often not fully considered, and the tasks are allocated according to fixed rules or priorities, which easily causes some platforms to be in a running congestion state due to excessive task accumulation, while another part of the platform is in a resource idle state for a long time, which cannot effectively play its processing potential. This uneven resource allocation makes it difficult to improve the overall resource utilization efficiency of the cross-platform system, resulting in a waste of a large amount of processing resources. The existing cross-platform task scheduling technology has obvious shortcomings in the data transmission and task execution links. On the one hand, the stability difference of data transmission between different processing platforms is ignored, and the transmission effect between platforms is not analyzed and selected, which easily leads to data transmission failure due to unstable transmission link; on the other hand, after task allocation, real-time monitoring of the running state of the execution platform is lacking, if the execution platform suddenly appears running abnormity or load increases suddenly, the task arrangement cannot be adjusted in time, which leads to task delay or even interruption, and it is difficult to ensure the smooth progress of the whole process from transmission to execution of cross-platform data task. SUMMARY

[0003] The present application aims to provide a cross-platform data task scheduling system and method based on an authorization chain to solve the problems raised in the background art.

[0004] In order to solve the above technical problems, the present application provides the following technical scheme: a cross-platform data task scheduling method based on an authorization chain, comprising the following steps: S1, when receiving a task, the task processing platform pre-processes the task data, and constructs a task database according to the task characteristics obtained by pre-processing; S2, when receiving a task, the task processing platform starts self-checking, and selects a preset task processing platform according to the real-time running load; S3. Analyze the matching degree of the preset task processing platform with the task, and determine the task scheduling requirements based on the task matching degree. S4. Analyze the transmission status between platforms in real time and select the candidate task processing platform based on the transmission status between platforms; S5. Taking into account the matching degree and transmission status of the candidate task processing platforms, analyze and consider the predicted completion rate of the candidate task processing platforms for the tasks. S6. Analyze the real-time task execution status of the preferred task processing platform and select the preferred task processing platform to execute the task.

[0005] Furthermore, in step S1, after authorization, any task processing platform begins receiving tasks. When a task processing platform receives task α, it collects task data, preprocesses and normalizes the task data to obtain task features, which are {A1, A2, ..., A...}. m ,…,A M}, where A m Let M represent the m-th task feature. Task features include: the ratio of the task's data volume to the historical average data volume, and the ratio of the amount of non-repeating data in the task to the total task data volume. M represents the number of task features. After obtaining the task features, the task processing platform transmits them to a cloud repository to establish a task database based on these features. By preprocessing and normalizing the task data, redundant and messy information can be effectively removed, allowing the extracted task features to more accurately reflect the core attributes of the task and avoiding deviations in subsequent task analysis due to raw data quality issues. Simultaneously, transmitting task features to the cloud repository and establishing a dedicated task database systematically accumulates feature information for various tasks, providing comprehensive and reliable data support for task-platform matching and scheduling in subsequent steps. This breaks through the limitations of fragmented and unreusable task information in traditional scheduling, ensuring the quality of early-stage cross-platform task scheduling from the source and laying the foundation for accurate scheduling later.

[0006] Furthermore, in step S2, when any task processing platform receives a task, the task processing platform begins a self-check. If the real-time operating load of the task processing platform exceeds a preset self-check load threshold, it is determined that the task processing platform cannot process the task, and the task processing platform is classified as a non-preset task processing platform. The real-time operating load is set to the CPU real-time operating utilization rate. Otherwise, it is determined that the task processing platform can process the task, and the task processing platform is classified as a preset task processing platform, thus obtaining N preset task processing platforms, where the N preset task processing platforms are {B1, B2, ..., B...}. n ,…,B N}, where B nThe nth preset task processing platform; through a self-checking mechanism, preset task processing platforms that meet the load requirements are selected. This mechanism accurately identifies platforms that are currently overloaded and unable to process tasks normally, excluding them from the initial scheduling scope. This operation effectively avoids assigning tasks to platforms with excessive operating pressure, preventing task processing delays, crashes, or incorrect results due to platform overload. Simultaneously, it ensures that all platforms participating in subsequent scheduling have basic task processing capabilities, acting as the first line of defense for the rationality and security of subsequent task allocation, reducing invalid scheduling attempts, and improving overall scheduling efficiency.

[0007] Furthermore, in step S3, when the preset task processing platform B... n Upon receiving task α, analyze the preset task processing platform B. n Based on the matching degree of task α, the preset task processing platform B is invoked. n Historical task processing data, preset task processing platform B n The m-th feature in the historical task completion is A m The number of tasks is D n_m The number of tasks that were successfully processed is d. n_m This leads to the pre-defined task processing platform B. n The matching degree C of the m-th feature of task α n_m C n_m For d n_m With D n_m The ratio is used to obtain the preset task processing platform B. n Task matching degree E for task α n E n Preset task processing platform B n The average value of the matching degree of the M features for task α; Substituting each value into the input n=1,2,…,N, we obtain the matching degree {E1,E2,…,E} of the N task processing platforms for task α. n ,…,E N Set a matching degree self-check threshold. If the preset task processing platform B... n Task matching degree E for task α n The match degree is greater than the self-check threshold and the task processing platform B n Task matching degree E for task α n If the matching degree of task α is greater than the average of N task processing platforms, then task α will be assigned to task processing platform B. nIf a task is not successfully processed, it is processed; otherwise, task α is scheduled based on task matching degree. By analyzing the platform's historical task processing data and calculating the matching degree in combination with the successful processing of various task characteristics, the processing capability of each preset platform for the current task can be accurately evaluated, rather than being roughly matched based solely on the platform's surface functions or task type. This matching method based on historical success experience allows tasks to be preferentially assigned to platforms that are truly good at handling this type of task, significantly reducing processing errors caused by inappropriate platform-task compatibility. At the same time, through the dual judgment of matching degree threshold and average value, platforms with better overall adaptability are further screened out, which not only avoids low-compatibility platforms from participating in task processing, but also reduces the workload of subsequent complex scheduling, improving the accuracy and efficiency of the initial task allocation.

[0008] Furthermore, in step S4, the X task processing platforms with the highest matching degree to task α are selected as candidate task processing platforms, and the X candidate task processing platforms are {Z1, Z2, ..., Z...} x ,…,Z X}, where Z x Let x be the candidate task processing platform. Analyze the preset task processing platform B. n For the preset task processing platform Z x The transmission stability, if the preset task processing platform B n and the preset task processing platform Z x If they are on the same task processing platform, then determine the preset task processing platform B. n For the preset task processing platform Z x The predicted transmission success rate is 1; otherwise, the preset task processing platform B is used for data acquisition. n For the preset task processing platform Z x The data transmission characteristics are {F1, F2, ..., F...}. x ,…,F X}, where X represents the number of data transmission features, F x Indicates the default task processing platform B n For the preset task processing platform Z x The xth data transmission characteristic calls the historical database, where task processing platform B... n For the preset task processing platform Z x The xth data transmission feature is F x The number of times is H x The number of successful transmissions is h. x This leads to the task processing platform B. n For the preset task processing platform Z x The transmission feature adaptation rate J of the xth data transmission feature x J x =h x / H x Substitute x=1,2,…,X one by one to obtain the task processing platform B. n For the preset task processing platform Z x The adaptation rate of the transmission features of X data transmission features is then used to obtain the task processing platform B. n For the preset task processing platform Z x Predicted transmission success rate L x Y pre-defined task processing platforms are selected as candidate platforms based on their predicted transmission success rates, ranked from highest to lowest. When selecting candidate platforms, not only is the platform's suitability for the task considered, but the stability of data transmission between platforms is also analyzed, overcoming the shortcomings of traditional scheduling that neglect the reliability of the transmission link. For the same platform, directly determining the high transmission success rate is both realistic and simplifies calculations. For different platforms, the adaptation rate and predicted transmission success rate are calculated using historical transmission data, accurately selecting platforms with stable transmission and low risk of data loss or failure. This operation ensures that the candidate platforms not only adapt to task processing but also guarantee smooth data flow between platforms, reducing the risk of task failure at the transmission level and providing stable data flow support for subsequent task progress.

[0009] Furthermore, in step S5, the matching degree {R1, R2, ..., R...} of task α among the Y candidate task processing platforms is called. y ,…,R Y}, where R y This represents the matching degree of the y-th candidate task processing platform for task α, and the task processing platform B is called. n Predicted transmission success rates {S1,S2,…,S} for Y candidate task processing platforms y ,…,S Y}, where S y Indicates task processing platform B n For the predicted transmission success rate of the y-th candidate task processing platform, calculate the comprehensive task prediction completion rate T of task α on the y-th candidate task processing platform. y Comprehensive task prediction completion rate T y Let the matching degree between the y-th candidate task processing platform and task processing platform B be the degree of matching between task α and task processing platform B. n The product of the predicted transmission success rates of the y-th candidate task processing platform is used to obtain the comprehensive predicted task completion rate {T1,T2,…,T} of task α across the y candidate task processing platforms. y ,…,T Yfurthermore, the Y candidate task processing platforms are sorted in descending order of the comprehensive task predicted completion rate, and the Q candidate task processing platforms with the highest comprehensive task predicted completion rate are selected as the preferred task processing platforms; the task matching degree and the inter-platform predicted transmission success rate are combined to calculate the comprehensive task predicted completion rate, so that the processing capacity and the transmission capacity of the platforms are considered. This comprehensive evaluation method avoids the selection bias caused by single dependence on the matching degree or the transmission rate, and neither selects the platform that is good at processing but has difficulty in stably receiving data, nor selects the platform that has stable transmission but is not good at task processing. The preferred platform selected through the comprehensive completion rate sorting has advantages in both task processing and data transmission, and can maximize the overall reliability of the task from data reception to processing completion, providing a scientific and accurate selection basis for the final task authorization.

[0010] Further, in step S6, the preferred task processing platform with the highest comprehensive task predicted completion rate is monitored. If the preferred task processing platform with the highest comprehensive task predicted completion rate is not executing a task, the preferred task processing platform with the highest predicted completion rate is marked with a task a execution authorization tag; otherwise, the next preferred task processing platform is monitored until a preferred task processing platform that is not executing a task is found, and the preferred task processing platform that is not executing a task is selected and marked with a task a execution authorization tag. If the Q preferred task processing platforms are all executing tasks, the preferred task processing platform with the smallest real-time running load among the Q preferred task processing platforms is selected and marked with a task a execution authorization tag, and the task a is added to the to-be-executed task list of the preferred task processing platform with the smallest real-time running load. By monitoring the running state of the preferred platform, the idle platform is preferentially granted the task execution right, so that the task can quickly enter the execution phase, the task waiting time caused by platform busy is reduced, and the task execution efficiency is improved. When all the preferred platforms are in a busy state, the platform with the smallest load is selected and the task is added to the to-be-executed list, which not only avoids task omission, but also balances the running pressure of each platform, preventing some platforms from being excessively congested due to task accumulation, and some platforms from wasting resources due to low load. This flexible authorization and arrangement method takes into account the timeliness of task execution and the balanced use of platform resources, ensuring that the task is orderly promoted while maximizing the processing potential of the preferred platform.

[0011] A cross-platform data task scheduling system based on an authorization chain, the system comprising: a task database establishment module, a task processing platform self-checking module, a platform matching degree judgment module, a platform transmission status analysis module, a predicted completion rate analysis module, and a task execution platform confirmation module; The task data database establishment module is used for pre-processing task data when a task processing platform receives a task, and constructing a task database according to the task characteristics obtained by pre-processing; The task processing platform self-checking module is used for starting self-checking of the task processing platform when the task processing platform receives a task, and selecting a preset task processing platform according to real-time running load; The platform matching degree judging module is used for analyzing the matching degree of the preset task processing platform to the task, and judging the scheduling demand of the task by considering the matching degree of the task; The platform transmission condition analyzing module is used for analyzing the transmission condition between platforms in real time, and selecting a candidate task processing platform according to the transmission condition between the platforms; The predicted completion rate analyzing module is used for comprehensively considering the matching degree of the candidate task processing platform to the task and the transmission condition, and analyzing the predicted completion rate of the candidate task processing platform to the task; The task execution platform confirming module is used for analyzing the real-time task execution state of the preferred task processing platform, and selecting the preferred task processing platform to execute the task.

[0012] Compared with the prior art, the present application has the following beneficial effects: on the one hand, the task data is preprocessed and normalized to extract key features, and a dedicated task database is established, and the matching degree of the platform to the current task is analyzed by combining the historical task processing data of each task processing platform. Subsequently, the matching degree of the platform to the task and the data transmission stability between the platforms are comprehensively considered, and the preferred platform more suitable for processing the current task is screened out. This multi-dimensional adaptive analysis can accurately match each task to the platform skilled in processing the task, effectively reduces the processing deviation or error caused by improper matching of the platform and the task, and significantly improves the accuracy and reliability of the overall task processing; On the one hand, in the task scheduling process, the platform with load meeting the requirements is first screened out through load self-checking, so as to avoid assigning the task to the platform already running overload; after entering the preferred platform selection stage, the platform in idle state is preferentially considered, and if all the preferred platforms are in busy state, the platform with the smallest load is selected and the task is put into the execution list thereof. This dynamic allocation mode based on the platform load and running state can reasonably balance the task amount of each platform, avoid the situation that part of the platform resources is long-term idle while part of the platform is excessively congested due to task accumulation, fully tap the processing potential of each platform, and maximize the overall utilization efficiency of cross-platform resources; On the other hand, for the historical data transmission between different platforms, the transmission success rate between platforms is analyzed and predicted, and the platform with stable transmission is preferentially selected for data interaction to reduce the risk of data transmission failure from the source. In the task execution link, the running state of the preferred platform is monitored in real time to ensure that the task can be allocated to the platform with immediate execution conditions. Even if all preferred platforms are in a busy state, the platform with the smallest load will be selected to arrange subsequent execution. This double protection of the transmission link and the execution link can effectively reduce the task delay caused by transmission failure or platform running congestion, and provide stable support for the smooth progress of cross-platform data tasks. BRIEF DESCRIPTION OF DRAWINGS

[0013] The accompanying drawings are included to provide a further understanding of the application, and constitute a part of the specification, which together with the embodiments of the application are used to explain the application, and do not constitute a limitation on the application. In the drawings: Figure 1 is a structural diagram of a cross-platform data task scheduling system based on an authorization chain of the application; Figure 2 is a flowchart of a cross-platform data task scheduling method based on an authorization chain of the application. DETAILED DESCRIPTION

[0014] The technical solutions in the embodiments of the application will be described in detail below with reference to the drawings of the embodiments of the application. Obviously, the described embodiments are only a part of the embodiments of the application, not all the embodiments. Based on the embodiments in the application, all other embodiments obtained by those skilled in the art without creative labor are within the scope of protection of the application.

[0015] Please refer to Figure 1 and Figure 2 The application provides a technical solution: a cross-platform data task scheduling method based on an authorization chain, comprising the following steps: S1, when the task processing platform receives a task, the task data is preprocessed, and a task database is constructed according to the task characteristics obtained by preprocessing; S2, when the task processing platform receives a task, the task processing platform starts self-checking, and selects a preset task processing platform according to the real-time running load; S3, the matching degree of the preset task processing platform to the task is analyzed, and the scheduling demand of the task is judged considering the task matching degree; S4, the transmission status between platforms is analyzed in real time, and the standby task processing platform is selected according to the transmission status between platforms; S5, the matching degree and transmission status of the standby task processing platform to the task are comprehensively considered, and the predicted completion rate of the standby task processing platform to the task is analyzed; S6, analyze the real-time task execution state of the preferred task processing platform, and select the preferred task processing platform to execute the task.

[0016] In step S1, any task processing platform is authorized to start receiving tasks. When the task processing platform receives a task a, the task data is collected, preprocessed and normalized to obtain task features. The task features are {A1, A2, …, A m ,…,A M} where A m represents the mth task feature of the task. The task features include: the ratio of the task data volume to the average historical task data volume, the ratio of the non-redundant data volume in the task to the task data volume, and M represents the number of task features. After obtaining the task features, the task processing platform transmits the task features to the cloud storage library to establish a task database for the task features. By pre-processing and normalizing the task data, redundant and chaotic information in the task data can be effectively eliminated, and the extracted task features can more accurately reflect the core attributes of the task, avoiding deviations in subsequent task analysis due to original data quality problems. At the same time, the task features are transmitted to the cloud storage library and a dedicated task database is established, which can systematically accumulate feature information of various tasks, providing comprehensive and reliable data support for the matching and scheduling of tasks and platforms in subsequent steps, breaking the limitations of scattered task information and difficult reuse in traditional scheduling, ensuring the data quality of cross-platform task scheduling from the source, and laying a foundation for subsequent accurate scheduling.

[0017] In step S2, when any task processing platform receives a task, the task processing platform starts self-checking. When the real-time running load of the task processing platform is greater than the preset self-checking load threshold, it is judged that the task processing platform cannot process the task, and the task processing platform is classified as a non-pre-set task processing platform. The real-time running load is set as the CPU real-time running occupancy rate. Otherwise, it is judged that the task processing platform can process the task, and the task processing platform is classified as a preset task processing platform, and then N preset task processing platforms are obtained. The N preset task processing platforms are {B1, B2, …, B n ,…,B N} where B n is the nth preset task processing platform. By using the self-checking mechanism to select the preset task processing platform that meets the load requirement, the platform that is currently overloaded and cannot normally process the task can be accurately identified and excluded from the initial scheduling range. This operation can effectively avoid assigning tasks to platforms with excessive running pressure, prevent task processing delay, crash or result error caused by platform overload, and at the same time ensure that the platforms participating in subsequent scheduling all have basic task processing capability, which is the first step to ensure the rationality and safety of subsequent task allocation, reduces invalid scheduling attempts, and improves the overall scheduling efficiency.

[0018] In step S3, when the preset task processing platform B... n Upon receiving task α, analyze the preset task processing platform B. n Based on the matching degree of task α, the preset task processing platform B is invoked. n Historical task processing data, preset task processing platform B n The m-th feature in the historical task completion is A m The number of tasks is D n_m The number of tasks that were successfully processed is d. n_m This leads to the pre-defined task processing platform B. n The matching degree C of the m-th feature of task α n_m C n_m For d n_m With D n_m The ratio is used to obtain the preset task processing platform B. n Task matching degree E for task α n E n Preset task processing platform B n The average value of the matching degree of the M features for task α; Substituting each value into the input n=1,2,…,N, we obtain the matching degree {E1,E2,…,E} of the N task processing platforms for task α. n ,…,E N Set a matching degree self-check threshold. If the preset task processing platform B... n Task matching degree E for task α n The match degree is greater than the self-check threshold and the task processing platform B n Task matching degree E for task α n If the matching degree of task α is greater than the average of N task processing platforms, then task α will be assigned to task processing platform B. n If a task is not successfully processed, it is processed; otherwise, task α is scheduled based on task matching degree. By analyzing the platform's historical task processing data and calculating the matching degree in combination with the successful processing of various task characteristics, the processing capability of each preset platform for the current task can be accurately evaluated, rather than being roughly matched based solely on the platform's surface functions or task type. This matching method based on historical success experience allows tasks to be preferentially assigned to platforms that are truly good at handling this type of task, significantly reducing processing errors caused by inappropriate platform-task compatibility. At the same time, through the dual judgment of matching degree threshold and average value, platforms with better overall adaptability are further screened out, which not only avoids low-compatibility platforms from participating in task processing, but also reduces the workload of subsequent complex scheduling, improving the accuracy and efficiency of the initial task allocation.

[0019] In step S4, the X task processing platforms with the highest matching degree to task α are selected as candidate task processing platforms. The X candidate task processing platforms are {Z1, Z2, ..., Z...} x ,…,Z X}, where Z x Let x be the candidate task processing platform. Analyze the preset task processing platform B. n For the preset task processing platform Z x The transmission stability, if the preset task processing platform B n and the preset task processing platform Z x If they are on the same task processing platform, then determine the preset task processing platform B. n For the preset task processing platform Z x The predicted transmission success rate is 1; otherwise, the preset task processing platform B is used for data acquisition. n For the preset task processing platform Z x The data transmission characteristics are {F1, F2, ..., F...}. x ,…,F X}, where X represents the number of data transmission features, F x Indicates the default task processing platform B n For the preset task processing platform Z x The xth data transmission characteristic calls the historical database, where task processing platform B... n For the preset task processing platform Z x The xth data transmission feature is F x The number of times is H x The number of successful transmissions is h. x This leads to the task processing platform B. n For the preset task processing platform Z x The transmission feature adaptation rate J of the xth data transmission feature x J x =h x / H x Substitute x=1,2,…,X one by one to obtain the task processing platform B. n For the preset task processing platform Z x The adaptation rate of the transmission features of X data transmission features is then used to obtain the task processing platform B. n For the preset task processing platform Z x Predicted transmission success rate L x, Y preset task processing platforms with high to low predicted transmission success rate are selected as candidate task processing platforms; when screening the candidate platforms, not only the matching degree of the platform to the task is considered, but also the data transmission stability between the platforms is analyzed, which makes up for the defect of traditional scheduling ignoring the reliability of the transmission link. For the same platform, a high transmission success rate is directly determined, which not only conforms to the actual situation but also simplifies the calculation; for different platforms, the fitting rate and the predicted transmission success rate are calculated through historical transmission data, which can accurately screen out platforms with stable transmission and less data loss or failure. This operation ensures that the candidate platforms not only adapt to the task processing, but also guarantee the smooth flow of data between the platforms, reduces the risk of task failure from the transmission layer, and provides stable data flow support for subsequent task promotion.

[0020] In step S5, the matching degrees {R1, R2, …, RY} of the Y candidate task processing platforms to the task a are called, where RY represents the matching degree of the yth candidate task processing platform to the task a. y ,…,R Y} are called, where S y represents the predicted transmission success rate of the task processing platform B n to the yth candidate task processing platform, and Y is an integer greater than or equal to 1. y ,…,S Y} are called, where S y represents the predicted transmission success rate of the task processing platform B n to the yth candidate task processing platform, and Y is an integer greater than or equal to 1. y The comprehensive task prediction completion rate T y of the task a to the yth candidate task processing platform is calculated, which is the product of the matching degree of the yth candidate task processing platform to the task a and the predicted transmission success rate of the task processing platform B n to the yth candidate task processing platform, and Y is an integer greater than or equal to 1. y ,…,T YY candidate task processing platforms are sorted from high to low according to the comprehensive task predicted completion rate, and the Q candidate task processing platforms with the highest comprehensive task predicted completion rate are selected as the preferred task processing platforms; the task matching degree of the candidate platform is combined with the inter-platform predicted transmission success rate to calculate the comprehensive task predicted completion rate, so that the "processing capacity" and "transmission capacity" of the platform are considered. This comprehensive evaluation method avoids the selection bias caused by single dependence on matching degree or transmission rate, and neither selects the platform that is good at processing but difficult to stably receive data, nor selects the platform that has stable transmission but is not good at task processing. The preferred platform selected through the comprehensive completion rate sorting has advantages in both task processing and data transmission, which can maximize the overall reliability of the task from data reception to processing completion, and provides a scientific and accurate selection basis for the final task authorization.

[0021] In step S6, the preferred task processing platform with the highest comprehensive task predicted completion rate is monitored. If the preferred task processing platform with the highest predicted completion rate is not executing a task, the preferred task processing platform with the highest predicted completion rate is marked with a task a execution authorization tag; otherwise, the next preferred task processing platform is monitored until a preferred task processing platform that is not executing a task is found, and the preferred task processing platform that is not executing a task is selected and marked with a task a execution authorization tag. If Q preferred task processing platforms are all executing tasks, the preferred task processing platform with the smallest real-time running load among the Q preferred task processing platforms is selected and marked with a task a execution authorization tag, and the task a is added to the to-be-executed task list of the preferred task processing platform with the smallest real-time running load. By monitoring the running state of the preferred platform, the task execution right is preferentially granted to the idle platform, which can make the task quickly enter the execution phase, reduce the task waiting time caused by platform busy, and improve the task execution efficiency. When all the preferred platforms are in a busy state, the platform with the smallest load is selected and the task is added to the to-be-executed list, which not only avoids task omission, but also balances the running pressure of each platform, preventing some platforms from being excessively congested due to task accumulation, and some platforms from wasting resources due to low load. This flexible authorization and arrangement method takes into account the timeliness of task execution and the balanced use of platform resources, ensuring that the task is orderly promoted while maximizing the processing potential of the preferred platform.

[0022] A cross-platform data task scheduling system based on an authorization chain, the system comprising: a task database establishment module, a task processing platform self-checking module, a platform matching degree judgment module, a platform transmission status analysis module, a predicted completion rate analysis module, and a task execution platform confirmation module; The task data database establishment module is used for pre-processing task data when a task processing platform receives a task, and constructing a task database according to the task characteristics obtained by pre-processing; The task processing platform self-checking module is configured to start self-checking of the task processing platform when the task processing platform receives a task, and to select a preset task processing platform according to a real-time running load; The platform matching degree judgment module is configured to analyze the matching degree of the preset task processing platform to the task, and to consider the scheduling demand of the task according to the matching degree of the task; The platform transmission condition analysis module is configured to analyze the transmission condition between platforms in real time, and to select a candidate task processing platform according to the transmission condition between the platforms; The predicted completion rate analysis module is configured to comprehensively consider the matching degree of the candidate task processing platform to the task and the transmission condition, and to analyze the predicted completion rate of the candidate task processing platform to the task; The task execution platform confirmation module is configured to analyze the real-time task execution state of the preferred task processing platform, and to select the preferred task processing platform to execute the task.

[0023] In embodiment 1, a task feature collection and database establishment stage is entered. After being authorized, any task processing platform starts a task receiving function. When a target task is received, the platform first collects relevant data of the task comprehensively, and then carries out data preprocessing work, including removing invalid data, completing missing information, and then unifying data format and magnitude through normalization processing. Finally, the key features of the task are extracted. The platform transmits all the features to a cloud storage library, and gradually accumulates to form a task database that can be used for subsequent query and analysis.

[0024] Then, task processing platform self-checking and classification are carried out. After receiving a task, the platform automatically starts real-time running load self-checking. A self-checking load threshold is preset in the platform, and the threshold is determined according to hardware performance and running stability requirements. If the current real-time running load of the platform exceeds this threshold, it means that the platform cannot temporarily undertake a new task, and will be classified as a platform without processing capacity; if the load does not exceed the threshold, it is determined that the platform has processing capacity and is classified into the range of candidate platforms that can participate in subsequent scheduling.

[0025] Then, the matching degree of the platform and the task is evaluated. For each candidate platform, the historical task processing data of the platform in the task database is called. The situation of the platform in processing tasks containing each feature of the current task in the past is analyzed, the proportion of successful task processing under each feature is calculated, and the overall matching degree of the platform to the current task is obtained by averaging. Then, the matching degree of the single platform is compared with the preset matching degree self-checking threshold, and compared with the average matching degree of all candidate platforms, and the platform with higher adaptability is preliminarily screened out.

[0026] Then, the candidate platform screening and transmission stability analysis are performed. First, a batch of platforms with high matching degree are selected from the platforms screened initially as candidate platforms. Then, the data transmission stability between the original platform and each candidate platform is analyzed. If the two platforms are the same, the transmission success rate is directly determined as the highest. If the two platforms are different, the characteristics of the past data transmission between the two platforms are collected, and the success rate of the predicted transmission is calculated according to the proportion of the success number to the total number of each characteristic in the historical transmission. Then, a new round of candidate platforms is selected according to the success rate from high to low.

[0027] Then, the comprehensive task prediction completion rate calculation and optimal platform determination are performed. The matching degree of the new round of candidate platforms to the current task and the predicted transmission success rate between the original platform and each candidate platform are obtained respectively, and the comprehensive task prediction completion rate of each candidate platform is obtained by multiplying the two. The optimal platforms with the highest comprehensive completion rate are selected.

[0028] Finally, the authorized tag issuing and task allocation are implemented. The optimal platforms are monitored one by one. If a certain optimal platform is not currently executing a task, it is marked with the execution authorization tag of the current task. If the platform is currently executing a task, the next optimal platform is monitored. If all the optimal platforms are executing tasks, the platform with the smallest real-time running load is selected, marked with the authorization tag, and the current task is added to the to-be-executed task list of the platform, waiting for subsequent processing.

[0029] It is obvious to those skilled in the art that the present application is not limited to the details of the above exemplary sensor device embodiments, and can be implemented in other specific forms without departing from the spirit or essential characteristics of the present application. Therefore, the embodiments should be regarded as exemplary and non-limiting, and the scope of the present application is defined by the appended claims rather than the above description, and all changes falling within the meaning and scope of the equivalent elements of the claims are intended to be included in the present application. Any reference signs in the claims should not be regarded as limiting the claims involved.

Claims

1. A method for cross-platform data task scheduling based on an authorization chain, characterized in that: The method comprises the following steps: S1, when the task processing platform receives a task, pre-processing the task data, and constructing a task database according to the task characteristics obtained by pre-processing; S2, when the task processing platform receives a task, the task processing platform starts self-checking, and selects a preset task processing platform according to the real-time running load; S3, analyzing the matching degree of the preset task processing platform to the task, and considering the task matching degree to judge the scheduling demand of the task; S4, real-time analysis of the transmission status between platforms, and selecting a candidate task processing platform according to the transmission status between platforms; S5, comprehensively considering the matching degree and transmission status of the candidate task processing platform to the task, and analyzing the predicted completion rate of the candidate task processing platform to the task; S6, analyzing the real-time task execution state of the preferred task processing platform, and selecting the preferred task processing platform to execute the task. 2.The cross-platform data task scheduling method based on an authorization chain according to claim 1, wherein: In step S1, any task processing platform is authorized to start receiving tasks. When the task processing platform receives task a, it collects task data, pre-processes and normalizes the task data to obtain task features, and the task features are {A1, A2, …, A m ,…,A M}, wherein A m represents the mth task feature of the task, M represents the number of task features, and after obtaining the task features, the task processing platform transmits the task features to the cloud storage library to establish a task database about the task features.

3. The cross-platform data task scheduling method based on an authorization chain according to claim 2, characterized in that: In step S2, when any task processing platform receives a task, the task processing platform starts self-checking, and when the real-time running load of the task processing platform is greater than the preset self-checking load threshold, it is judged that the task processing platform cannot process the task, and the task processing platform is classified as a non-pre-set task processing platform; Otherwise, if the task processing platform is deemed capable of processing the task, it is categorized as a preset task processing platform, resulting in N preset task processing platforms, which are {B1, B2, ..., B...}. n ,…,B N }, where B n The nth preset task processing platform.

4. The cross-platform data task scheduling method based on an authorization chain according to claim 3, characterized in that: In step S3, when the preset task processing platform B n receives the task α, it analyzes the preset task processing platform B n the matching degree of the task α, calls the historical task processing data of the preset task processing platform B n , and the preset task processing platform B n the number of tasks with the mth feature A m in the historical completed tasks is D n_m , wherein the number of tasks with successful task processing is d n_m , and the preset task processing platform B n the matching degree of the mth feature of the task α is C n_m , C n_m is the ratio of d n_m and D n_m , and the preset task processing platform B n the task matching degree E of the task α is obtained n , E n is the average value of the M feature matching degrees of the preset task processing platform B n to the task α.

5. The cross-platform data task scheduling method based on an authorization chain according to claim 4, characterized in that: Substitute n = 1, 2, …, N, to obtain N task processing platform matching degrees {E1, E2, …, EN} of the task α. n ,…,EN} N , set a matching degree self-check threshold, if the preset task processing platform B n The task matching degree E n of the task α is greater than the matching degree self-check threshold and the task processing platform B n The task matching degree E n of the task α is greater than the average of the matching degrees of the N task processing platforms to the task α, then the task α is handed over to the task processing platform B n for processing; otherwise, the task matching degree is considered to schedule the task α.

6. The cross-platform data task scheduling method based on an authorization chain according to claim 5, characterized in that: In step S4, the X task processing platforms with the highest matching degree for task a are selected as candidate task processing platforms, and the X candidate task processing platforms are {Z1, Z2, …, Zx}. x ,…,Z X}, where Z x represents the xth candidate task processing platform, the transmission stability of the preset task processing platform B n is analyzed. x The transmission stability of the preset task processing platform Z n is analyzed. x If the preset task processing platform B n and the preset task processing platform Z x are the same task processing platform, it is determined that the prediction transmission success rate of the preset task processing platform B n for the preset task processing platform Z x is 1; otherwise, the data transmission characteristics of the preset task processing platform B x for the preset task processing platform Z X are collected, and the data transmission characteristics are {F1, F2, …, FX}. x , where X represents the number of data transmission characteristics, F n represents the xth data transmission characteristic of the preset task processing platform B x for the preset task processing platform Z n , the historical database is called, and the number of times of the xth data transmission characteristic of the preset task processing platform Z x for the preset task processing platform B x is H x , where the number of transmission successes is h x , and the transmission characteristic adaptation rate J n of the xth data transmission characteristic of the preset task processing platform Z x for the preset task processing platform B x is obtained, J x =h x / H x , and x=1, 2, …, X is substituted one by one to obtain the transmission characteristic adaptation rate of the X data transmission characteristics of the preset task processing platform Z n for the preset task processing platform B x , and the prediction transmission success rate L n of the preset task processing platform Z x for the preset task processing platform B x is obtained, and Y preset task processing platforms are selected as candidate task processing platforms according to the prediction transmission success rate from high to low.

7. The cross-platform data task scheduling method based on an authorization chain according to claim 6, characterized in that: In step S5, the matching degrees {R1, R2, …, RY} of Y candidate task processing platforms to task a are called, where RY represents the matching degree of the yth candidate task processing platform to task a. y ,…,R Y} are called, where S y represents the predicted transmission success rate of the yth candidate task processing platform by task processing platform B n . y ,…,S Y} are called, where S y represents the predicted transmission success rate of the yth candidate task processing platform by task processing platform B n . y The comprehensive task prediction completion rate T y of task a to the yth candidate task processing platform is calculated, which is the product of the matching degree of the yth candidate task processing platform to task a and the predicted transmission success rate of the yth candidate task processing platform by task processing platform B n . y ,…,T Y} are obtained, and then the Y candidate task processing platforms are sorted in descending order of the comprehensive task prediction completion rate, and the Q candidate task processing platforms with the highest comprehensive task prediction completion rate are obtained as the preferred task processing platforms.

8. The cross-platform data task scheduling method based on an authorization chain according to claim 6, characterized in that: In step S6, the preferred task processing platform with the highest comprehensive task predicted completion rate is monitored, if the preferred task processing platform with the highest comprehensive task predicted completion rate is not executing a task, the preferred task processing platform with the highest predicted completion rate is marked with a task α execution authorization tag; otherwise, the next preferred task processing platform is monitored until a preferred task processing platform that is not executing a task is monitored, the preferred task processing platform that is not executing a task is selected and marked with a task α execution authorization tag, and if Q preferred task processing platforms are monitored and are all executing a task, the preferred task processing platform with the smallest real-time running load among the Q preferred task processing platforms is selected and marked with a task α execution authorization tag, and task α is added to the to-be-executed task list of the preferred task processing platform with the smallest real-time running load.

9. A cross-platform data task scheduling system based on an authorization chain, the system being applied to the cross-platform data task scheduling method based on the authorization chain in any one of claims 1-8, characterized in that: The system comprises a task database establishment module, a task processing platform self-checking module, a platform matching degree judgment module, a platform transmission status analysis module, a predicted completion rate analysis module, and a task execution platform confirmation module; The task data database establishment module is used for pre-processing task data when a task processing platform receives a task, and constructing a task database according to the task characteristics obtained by pre-processing; The task processing platform self-checking module is used for starting self-checking of the task processing platform when the task processing platform receives a task, and selecting a preset task processing platform according to the real-time running load; The platform matching degree judgment module is used for analyzing the matching degree of the preset task processing platform to the task, and considering the task matching degree to judge the scheduling demand of the task; The platform transmission status analysis module is used for real-time analysis of the transmission status between platforms, and selecting a candidate task processing platform according to the transmission status between platforms; The predicted completion rate analysis module is used for comprehensively considering the matching degree and transmission status of the candidate task processing platform to the task, and analyzing the predicted completion rate of the candidate task processing platform to the task; The task execution platform confirmation module is configured to analyze the real-time task execution state of the preferred task processing platform, and select the preferred task processing platform to execute the task.

Images