Cloud edge hierarchical collaborative information system integrated resource scheduling system

Abstract

The invention discloses a cloud edge hierarchical collaborative information system integrated resource scheduling system. The system comprises an acquisition module. Data preliminary cleaning and feature extraction are locally executed through the edge end, only key feature data are uploaded, network transmission quantity is reduced, cloud computing pressure is relieved, real-time tasks can respond quickly, task timeout caused by transmission delay is avoided, feature vectors of the edge end and the cloud end are dynamically analyzed based on a scheduling decision model of reinforcement learning, and the real-time performance of the edge end is improved. Task requirements and resource states are intelligently matched, the scheduling accuracy is improved, model parameters are dynamically updated through task execution state data (such as duration and resource utilization rate deviation), the model is adaptively optimized along with environment changes (such as edge end hardware aging and cloud load fluctuation), the efficient scheduling capacity is kept for a long time, and the scheduling efficiency is improved. Layered processing and intelligent scheduling avoid edge end overload or cloud end resource idling, resource allocation of a CPU, a memory and the like is optimized, resource waste is reduced, and the system collaboration efficiency is integrally improved.

Description

Technical Field

[0001] This invention relates to the field of cloud computing technology, specifically to a cloud-edge layered collaborative information system integration resource scheduling system. Background Technology

[0002] With the rapid development of information technology, cloud computing and edge computing collaborative architectures have become a key infrastructure supporting the integration of complex information systems. In cloud-edge collaborative scenarios, edge nodes (such as industrial gateways and edge servers) are responsible for real-time data acquisition and local computing, while cloud servers provide massive data storage and complex analysis capabilities. However, due to the limited resources at the edge (such as computing power and storage space) and their dynamic changes in state (such as fluctuations in CPU utilization and changes in task queue length), and the uneven distribution of cloud resources (such as differences in cluster load and distribution of remaining storage space on storage nodes), traditional resource scheduling methods face the following technical challenges: First, existing cloud-edge collaborative scheduling methods mostly adopt a centralized data processing model, where raw data from the edge is directly uploaded to the cloud for unified analysis. This model not only increases network transmission pressure (especially in bandwidth-constrained industrial scenarios), but also leads to excessive cloud computing load, making it unable to quickly respond to real-time task demands. For example, in smart manufacturing scenarios, equipment status monitoring tasks require millisecond-level responses; if all data is uploaded to the cloud for processing, transmission delays may cause task timeouts. Secondly, traditional scheduling models are mostly based on static rules (such as round-robin and minimum connection count) or simple heuristic algorithms (such as genetic algorithms and particle swarm optimization), lacking the ability to accurately match dynamic resource states with task requirements. For example, when a sudden surge in high-load tasks occurs at the edge, the static model cannot offload the tasks to the cloud in time, leading to resource overload at the edge; conversely, if computationally intensive tasks are still allocated when cloud resources are scarce, it will exacerbate the cloud load and reduce the overall system efficiency. Furthermore, existing methods are insufficient for processing data integration from heterogeneous systems. In information system integration scenarios, edge devices and the cloud often involve multi-source heterogeneous data. Traditional data mapping and transformation methods rely on manual configuration, which is inefficient and prone to errors, resulting in an inaccurate data foundation for scheduling decisions and further affecting the effectiveness of scheduling strategies. Summary of the Invention

[0003] To address the aforementioned technical problems, a cloud-edge layered collaborative information system integration resource scheduling system is provided. This technical solution resolves the issues raised in the background section.

[0004] To achieve the above objectives, the technical solution adopted by the present invention is as follows: In a first aspect of the present invention, a cloud-edge layered collaborative information system integration resource scheduling system is provided, comprising: The acquisition module is used to acquire real-time status data of edge nodes and resource status data of cloud servers. The edge status data includes the CPU utilization, memory usage, remaining storage space, current task queue length, and task type identifier of the edge nodes. The cloud status data includes the average CPU load of the cloud server cluster, the total available memory, the distribution of remaining storage space of distributed nodes, and the task priority distribution of the cloud task queue. The data processing module is used to perform hierarchical processing on the collected edge status data and cloud status data. The edge data is preliminarily cleaned and feature extracted locally at the edge node to obtain the edge feature vector set. The cloud data is uploaded to the cloud and then subjected to deep analysis and feature fusion to obtain the cloud feature vector set. The matching module is used to generate a resource scheduling strategy based on the feature vector set of the edge end and the feature vector set of the cloud end through a pre-trained scheduling decision model. The scheduling decision model is used to characterize the resource demand matching relationship between the edge end and the cloud end. The allocation module is used to allocate tasks to edge nodes or cloud servers for execution according to resource scheduling strategies, and to monitor task execution status data in real time during task execution. An update module is used to dynamically update the parameters of the scheduling decision model based on task execution status data, so as to optimize the generation of subsequent resource scheduling strategies.

[0005] Preferably, the hierarchical processing of the collected edge status data and cloud status data includes: A preliminary cleaning operation is performed on the edge state data. The preliminary cleaning operation includes removing noisy data, filling missing values, and correcting outliers. Noisy data is identified by a sliding window filtering algorithm, missing values ​​are filled by linear interpolation between adjacent time points, and outliers are detected by the 3σ criterion and replaced with the mean. The formula for calculating the 3σ criterion is as follows: ; in, These are the data sample points to be detected. The mean of the data sample. The standard deviation of the data sample; Feature extraction is performed on the cleaned edge state data to obtain an edge feature vector set. The feature extraction includes extracting task type feature codes from task type identifiers and extracting resource usage trend features from resource state data. The resource usage trend features are calculated based on the resource utilization rate change rate within a preset time period in the past. Deep analysis and feature fusion are performed on cloud status data. The deep analysis includes predicting the availability of cloud resources within a preset time period through a time series analysis model. The feature fusion includes concatenating the feature vector set from the edge end with the feature vector set from the cloud end to obtain a fused feature vector set.

[0006] Preferably, the step of generating a resource scheduling strategy based on the edge feature vector set and the cloud feature vector set through a pre-trained scheduling decision model includes: The fused feature vector set is input into the scheduling decision model, which is a decision model trained based on reinforcement learning algorithm. The state space of reinforcement learning includes the feature vector set of the edge end and the feature vector set of the cloud end, the action space includes the action of assigning the task to the edge end or the cloud end, and the reward function is calculated based on the task execution time, resource utilization and energy consumption. The formula for calculating the reward function R is as follows: ; in, For task execution time, For resource utilization, For energy consumption, For the corresponding weight coefficients, and ; The scheduling decision model outputs a resource scheduling strategy, which includes the target node identifier for the tasks to be processed, the priority order of task execution, and the amount of resources allocated.

[0007] Preferably, the training method for the scheduling decision model includes: Obtain a historical scheduling dataset, which includes historical status data, historical scheduling strategies, and historical task execution effect data; Based on the historical state data and historical scheduling strategy, the model parameters are iteratively updated using Q-learning of the reinforcement learning algorithm to maximize the cumulative reward value, which is calculated based on historical task execution performance data. The trained model is validated, and when the validation is successful, the pre-trained scheduling decision model is obtained.

[0008] Preferably, the step of allocating tasks to be processed to edge nodes or cloud servers for execution according to a resource scheduling strategy includes: The resource scheduling strategy is parsed to obtain the task attributes and target node identifiers of the tasks to be processed. The task attributes include the task data volume, computational complexity, latency sensitivity, and dependencies. Based on the task attributes, determine the resource configuration parameters required for task execution. These parameters include the number of CPU cores, memory size, and storage space. The task to be processed is sent to the target node, and the resource configuration parameters corresponding to the task are allocated on the target node.

[0009] Preferably, the task attribute includes latency sensitivity, and determining the resource configuration parameters required for task execution based on the task attribute includes: When the latency sensitivity of a task is higher than a preset threshold, the task is determined to be a real-time task, and resource configuration parameters such as the number of CPU cores ≥ preset value and the memory size ≥ preset value are allocated to the edge nodes. When the latency sensitivity of a task is lower than or equal to a preset threshold, the task is determined to be a non-real-time task, and resource configuration parameters with storage space ≥ preset value are allocated in the cloud server.

[0010] Preferably, the step of dynamically updating the parameters of the scheduling decision model based on the task execution status data includes: Real-time acquisition of task execution status data, including task start time, task end time, peak CPU utilization during task execution, peak memory usage, and task execution result identifier; Calculate task execution performance indicators, which include task execution time, resource utilization deviation rate, and task failure rate. The formula for calculating the resource utilization deviation rate ΔU is as follows: ; in, The actual resource utilization rate of the task. The amount of resources allocated to the task; Based on the task execution performance indicators, the parameters of the scheduling decision model are updated using the gradient descent algorithm so that the task execution performance indicators corresponding to the subsequently generated resource scheduling strategies meet the preset conditions.

[0011] Preferably, before collecting the real-time status data of the edge nodes and the resource status data of the cloud server, the method further includes: Acquire information system integration requirement data, which includes data format descriptions, data transmission protocols, and data interaction frequencies of multiple heterogeneous systems; Based on the data format description, a data mapping relationship table is constructed, which is used to convert the data format of heterogeneous systems into a unified scheduling data format; Based on the data transmission protocol and data interaction frequency, configure the bandwidth allocation strategy and data transmission priority of the data acquisition link.

[0012] Preferably, the step of constructing a data mapping relationship table based on the data format description includes: Parse the data format descriptions of each heterogeneous system to obtain the data fields and field attributes of each heterogeneous system; Determine the mapping relationship between the data fields of each heterogeneous system and the fields of the unified scheduling data format, wherein the unified scheduling data format includes edge status data fields, cloud status data fields and task execution status data fields; Based on the field mapping relationship, a data mapping relationship table is generated, which includes heterogeneous system data fields, unified scheduling data format fields, and field conversion algorithms.

[0013] In a second aspect of the invention, an electronic device is also provided. The electronic device includes at least one processor; and a memory communicatively connected to the at least one processor; the memory stores instructions executable by the at least one processor to enable the at least one processor to perform the system of the first aspect of the invention.

[0014] Compared with existing technologies, the present invention provides a cloud-edge layered collaborative information system integration resource scheduling system, which has the following beneficial effects: This invention performs preliminary data cleaning and feature extraction locally at the edge, uploading only key feature data to reduce network transmission volume, alleviate cloud computing pressure, enable real-time tasks to respond quickly, and avoid task timeouts caused by transmission delays. Based on a reinforcement learning-based scheduling decision model, it dynamically analyzes feature vectors from the edge and cloud, intelligently matches task requirements with resource status, and improves scheduling accuracy. By dynamically updating model parameters through task execution status data (such as duration and resource utilization deviation), the model can adaptively optimize with environmental changes (such as edge hardware aging and cloud load fluctuations), maintaining high-efficiency scheduling capabilities over the long term. Furthermore, layered processing and intelligent scheduling avoid edge overload or cloud resource idleness, optimize the allocation of resources such as CPU and memory, reduce resource waste, and improve overall system collaboration efficiency. Attached Figure Description

[0015] Figure 1 This is a schematic diagram of the method flow for S101-S105 in this invention; Figure 2 This is a schematic diagram of the method flow for S201-S203 in this invention; Figure 3 This is a schematic diagram of the method flow for S301-S302 in this invention; Figure 4 This is a schematic diagram of the method flow for S401-S403 in this invention; Figure 5 This is a schematic diagram of the method flow for S501-S503 in this invention; Figure 6This is a schematic diagram of the method flow for S601-S602 in this invention. Detailed Implementation

[0016] The following description is intended to disclose the invention and enable those skilled in the art to implement it. The preferred embodiments described below are merely examples, and other obvious variations will occur to those skilled in the art.

[0017] Example 1 Please refer to Figure 1 As shown, in a first aspect of the present invention, a cloud-edge layered collaborative information system integration resource scheduling method is provided, comprising: S101. Collect real-time status data of edge nodes and resource status data of cloud servers. The edge status data includes the CPU utilization, memory usage, remaining storage space, current task queue length and task type identifier of edge nodes. The cloud status data includes the average CPU load of cloud server cluster, total available memory, distribution of remaining storage space of distributed nodes and task priority distribution of cloud task queue. S102. The collected edge status data and cloud status data are processed in layers. The edge data is preliminarily cleaned and feature extracted locally at the edge node to obtain the edge feature vector set. The cloud data is uploaded to the cloud and then subjected to deep analysis and feature fusion to obtain the cloud feature vector set. S103. Based on the feature vector sets of the edge and the cloud, a resource scheduling strategy is generated through a pre-trained scheduling decision model. The scheduling decision model is used to characterize the resource demand matching relationship between the edge and the cloud. S104. According to the resource scheduling strategy, the tasks to be processed are allocated to edge nodes or cloud servers for execution, and the task execution status data is monitored in real time during the task execution process. S105. Based on task execution status data, dynamically update the parameters of the scheduling decision model to optimize the generation of subsequent resource scheduling strategies.

[0018] As will be understood by those skilled in the art, this invention performs preliminary data cleaning and feature extraction locally at the edge, uploading only key feature data, reducing network transmission volume, alleviating cloud computing pressure, enabling real-time tasks to respond quickly, and avoiding task timeouts caused by transmission delays. Based on a reinforcement learning-based scheduling decision model, it dynamically analyzes feature vectors at the edge and in the cloud, intelligently matching task requirements and resource status to improve scheduling accuracy. By dynamically updating model parameters through task execution status data (such as duration and resource utilization deviation), the model can adaptively optimize with environmental changes (such as edge hardware aging and cloud load fluctuations), maintaining high-efficiency scheduling capabilities over the long term. Furthermore, layered processing and intelligent scheduling avoid edge overload or cloud resource idleness, optimize the allocation of resources such as CPU and memory, reduce resource waste, and improve overall system collaboration efficiency.

[0019] Please refer to Figure 2 As shown, the collected edge status data and cloud status data are processed in layers, including: S201. Perform preliminary cleaning operations on the edge state data. The preliminary cleaning operations include removing noisy data, filling missing values, and correcting outliers. Noisy data is identified by a sliding window filtering algorithm, missing values ​​are filled by linear interpolation between adjacent time points, and outliers are detected by the 3σ criterion and replaced with the mean. The formula for calculating the 3σ criterion is: ; in, These are the data sample points to be detected. The mean of the data sample. The standard deviation of the data sample; S202. Extract features from the cleaned edge state data to obtain an edge feature vector set. Feature extraction includes extracting task type feature codes from task type identifiers and extracting resource usage trend features from resource state data. Resource usage trend features are calculated based on the resource utilization rate change rate within a preset time period in the past. S203. Perform deep analysis and feature fusion on the cloud status data. The deep analysis includes predicting the availability of cloud resources within a preset time period through a time series analysis model. The feature fusion includes concatenating the feature vector set from the edge end with the feature vector set from the cloud end to obtain the fused feature vector set.

[0020] Please refer to Figure 3 As shown, based on the feature vector sets from the edge and the cloud, a resource scheduling strategy is generated through a pre-trained scheduling decision model, including: S301. Input the fused feature vector set into the scheduling decision model. The scheduling decision model is a decision model trained based on the reinforcement learning algorithm. The state space of the reinforcement learning includes the edge feature vector set and the cloud feature vector set. The action space includes the action of assigning the task to the edge or the cloud. The reward function is calculated based on the task execution time, resource utilization and energy consumption. The formula for calculating the reward function R is: ; in, For task execution time, For resource utilization, For energy consumption, For the corresponding weight coefficients, and ; S302. Output resource scheduling strategy through scheduling decision model. Resource scheduling strategy includes the target node identifier of the task to be processed, the priority order of task execution and the amount of resources allocated.

[0021] Please refer to Figure 4 As shown, the training methods for the scheduling decision model include: S401. Obtain the historical scheduling dataset, which includes historical status data, historical scheduling strategies, and historical task execution effect data. S402. Based on historical state data and historical scheduling strategies, the model parameters are iteratively updated through Q-learning of the reinforcement learning algorithm to maximize the cumulative reward value, which is calculated based on historical task execution performance data. S403. Validate the trained model. When the validation is successful, the pre-trained scheduling decision model is obtained.

[0022] Please refer to Figure 5 As shown, according to the resource scheduling strategy, the tasks to be processed are allocated to edge nodes or cloud servers for execution, including: S501. Parse the resource scheduling strategy to obtain the task attributes and target node identifier of the task to be processed. The task attributes include the task data volume, computational complexity, latency sensitivity and dependency relationship. S502. Based on the task attributes, determine the resource configuration parameters required for task execution. The resource configuration parameters include the number of CPU cores, memory size, and storage space. S503. Send the task to be processed to the target node and allocate the resources corresponding to the resource configuration parameters for the task on the target node.

[0023] Please refer to Figure 6 As shown, task attributes include latency sensitivity. Based on these attributes, the resource configuration parameters required for task execution are determined, including: S601. When the latency sensitivity of a task is higher than a preset threshold, the task is determined to be a real-time task, and resource configuration parameters of CPU cores ≥ preset value and memory size ≥ preset value are allocated to the edge nodes. S602. When the latency sensitivity of a task is lower than or equal to a preset threshold, the task is determined to be a non-real-time task, and resource configuration parameters with storage space ≥ preset value are allocated in the cloud server.

[0024] Based on task execution status data, the parameters of the scheduling decision model are dynamically updated, including: Real-time acquisition of task execution status data, including task start time, task end time, peak CPU utilization during task execution, peak memory usage, and task execution result identifier; Calculate task execution performance indicators, which include task execution time, resource utilization deviation rate, and task failure rate. The formula for calculating the resource utilization deviation rate ΔU is: ; in, The actual resource utilization rate of the task. The amount of resources allocated to the task; Based on the task execution performance metrics, the parameters of the scheduling decision model are updated using the gradient descent algorithm so that the task execution performance metrics corresponding to the subsequently generated resource scheduling strategies meet the preset conditions.

[0025] Before collecting real-time status data from edge nodes and resource status data from cloud servers, the following steps are also included: Acquire information system integration requirements data, which includes data format descriptions, data transmission protocols, and data interaction frequencies of multiple heterogeneous systems; Based on the data format description, a data mapping relationship table is constructed. The data mapping relationship table is used to convert the data format of heterogeneous systems into a unified scheduling data format. Configure the bandwidth allocation strategy and data transmission priority of the data acquisition link based on the data transmission protocol and data interaction frequency.

[0026] Based on the data format description, construct a data mapping relationship table, including: Parse the data format descriptions of each heterogeneous system to obtain the data fields and field attributes of each heterogeneous system; Determine the mapping relationship between the data fields of each heterogeneous system and the fields of the unified scheduling data format. The unified scheduling data format includes edge status data fields, cloud status data fields, and task execution status data fields. Based on the field mapping relationship, a data mapping relationship table is generated. The data mapping relationship table includes heterogeneous system data fields, unified scheduling data format fields, and field conversion algorithms.

[0027] In a second aspect, the present invention also provides a cloud-edge layered collaborative information system integration resource scheduling system, comprising: The data acquisition module is used to collect real-time status data of edge nodes and resource status data of cloud servers. The edge status data includes the CPU utilization, memory usage, remaining storage space, current task queue length and task type identifier of the edge nodes. The cloud status data includes the average CPU load of the cloud server cluster, total available memory, distribution of remaining storage space of distributed nodes and task priority distribution of the cloud task queue. The data processing module is used to perform hierarchical processing on the collected edge status data and cloud status data. The edge data undergoes preliminary cleaning and feature extraction locally at the edge node to obtain the edge feature vector set. The cloud data is uploaded to the cloud and then subjected to deep analysis and feature fusion to obtain the cloud feature vector set. The matching module is used to generate resource scheduling strategies based on the feature vector sets of the edge and the feature vector sets of the cloud through a pre-trained scheduling decision model. The scheduling decision model is used to represent the resource demand matching relationship between the edge and the cloud. The allocation module is used to allocate tasks to edge nodes or cloud servers for execution according to resource scheduling strategies, and to monitor task execution status data in real time during task execution. The update module is used to dynamically update the parameters of the scheduling decision model based on task execution status data, so as to optimize the generation of subsequent resource scheduling strategies.

[0028] In a third aspect of the invention, an electronic device is also provided. The electronic device includes at least one processor; and a memory communicatively connected to the at least one processor; the memory stores instructions executable by the at least one processor, which, when executed by the at least one processor, enable the at least one processor to perform the system of the first aspect of the invention.

[0029] The foregoing has shown and described the basic principles, main features, and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The embodiments and descriptions in the specification are merely principles of the invention. Various changes and modifications can be made to the invention without departing from its spirit and scope, and all such changes and modifications fall within the scope of the claimed invention. The scope of protection claimed by the appended claims and their equivalents is defined.

Claims

1. A cloud-edge layered collaborative information system integration resource scheduling system, characterized in that, include: The acquisition module is used to acquire real-time status data of edge nodes and resource status data of cloud servers. The edge status data includes the CPU utilization, memory usage, remaining storage space, current task queue length, and task type identifier of the edge nodes. The cloud status data includes the average CPU load of the cloud server cluster, the total available memory, the distribution of remaining storage space of distributed nodes, and the task priority distribution of the cloud task queue. The data processing module is used to perform hierarchical processing on the collected edge status data and cloud status data. The edge data is preliminarily cleaned and feature extracted locally at the edge node to obtain the edge feature vector set. The cloud data is uploaded to the cloud and then subjected to deep analysis and feature fusion to obtain the cloud feature vector set. The matching module is used to generate a resource scheduling strategy based on the feature vector set of the edge end and the feature vector set of the cloud end through a pre-trained scheduling decision model. The scheduling decision model is used to characterize the resource demand matching relationship between the edge end and the cloud end. The allocation module is used to allocate tasks to edge nodes or cloud servers for execution according to resource scheduling strategies, and to monitor task execution status data in real time during task execution. An update module is used to dynamically update the parameters of the scheduling decision model based on task execution status data, so as to optimize the generation of subsequent resource scheduling strategies.

2. The cloud-edge layered collaborative information system integration resource scheduling system according to claim 1, characterized in that, The layered processing of the collected edge-end status data and cloud-based status data includes: A preliminary cleaning operation is performed on the edge state data. The preliminary cleaning operation includes removing noisy data, filling missing values, and correcting outliers. Noisy data is identified by a sliding window filtering algorithm, missing values ​​are filled by linear interpolation between adjacent time points, and outliers are detected by the 3σ criterion and replaced with the mean. The formula for calculating the 3σ criterion is as follows: ； in, These are the data sample points to be detected. The mean of the data sample. The standard deviation of the data sample; Feature extraction is performed on the cleaned edge state data to obtain an edge feature vector set. The feature extraction includes extracting task type feature codes from task type identifiers and extracting resource usage trend features from resource state data. The resource usage trend features are calculated based on the resource utilization rate change rate within a preset time period in the past. Deep analysis and feature fusion are performed on cloud status data. The deep analysis includes predicting the availability of cloud resources within a preset time period through a time series analysis model. The feature fusion includes concatenating the feature vector set from the edge end with the feature vector set from the cloud end to obtain a fused feature vector set.

3. The cloud-edge layered collaborative information system integration resource scheduling system according to claim 2, characterized in that, The process of generating resource scheduling strategies based on edge feature vector sets and cloud feature vector sets through a pre-trained scheduling decision model includes: The fused feature vector set is input into the scheduling decision model, which is a decision model trained based on reinforcement learning algorithm. The state space of reinforcement learning includes the feature vector set of the edge end and the feature vector set of the cloud end, the action space includes the action of assigning the task to the edge end or the cloud end, and the reward function is calculated based on the task execution time, resource utilization and energy consumption. The formula for calculating the reward function R is as follows: ； in, For task execution time, For resource utilization, For energy consumption, For the corresponding weight coefficients, and ; The scheduling decision model outputs a resource scheduling strategy, which includes the target node identifier for the tasks to be processed, the priority order of task execution, and the amount of resources allocated.

4. The cloud-edge layered collaborative information system integration resource scheduling system according to claim 3, characterized in that, The training method for the scheduling decision model includes: Obtain a historical scheduling dataset, which includes historical status data, historical scheduling strategies, and historical task execution effect data; Based on the historical state data and historical scheduling strategy, the model parameters are iteratively updated using Q-learning of the reinforcement learning algorithm to maximize the cumulative reward value, which is calculated based on historical task execution performance data. The trained model is validated, and when the validation is successful, the pre-trained scheduling decision model is obtained.

5. A cloud-edge layered collaborative information system integration resource scheduling system according to claim 4, characterized in that, The step of allocating tasks to be processed to edge nodes or cloud servers for execution according to resource scheduling strategies includes: The resource scheduling strategy is parsed to obtain the task attributes and target node identifiers of the tasks to be processed. The task attributes include the task data volume, computational complexity, latency sensitivity, and dependencies. Based on the task attributes, determine the resource configuration parameters required for task execution. These parameters include the number of CPU cores, memory size, and storage space. The task to be processed is sent to the target node, and the resource configuration parameters corresponding to the task are allocated on the target node.

6. A cloud-edge layered collaborative information system integration resource scheduling system according to claim 5, characterized in that, The task attributes include latency sensitivity, and determining the resource configuration parameters required for task execution based on the task attributes includes: When the latency sensitivity of a task is higher than a preset threshold, the task is determined to be a real-time task, and resource configuration parameters such as the number of CPU cores ≥ preset value and the memory size ≥ preset value are allocated to the edge nodes. When the latency sensitivity of a task is lower than or equal to a preset threshold, the task is determined to be a non-real-time task, and resource configuration parameters with storage space ≥ preset value are allocated in the cloud server.

7. A cloud-edge layered collaborative information system integration resource scheduling system according to claim 6, characterized in that, The step of dynamically updating the parameters of the scheduling decision model based on the task execution status data includes: Real-time acquisition of task execution status data, including task start time, task end time, peak CPU utilization during task execution, peak memory usage, and task execution result identifier; Calculate task execution performance indicators, which include task execution time, resource utilization deviation rate, and task failure rate. The formula for calculating the resource utilization deviation rate ΔU is: ； in, The actual resource utilization rate of the task. The amount of resources allocated to the task; Based on the task execution performance indicators, the parameters of the scheduling decision model are updated using the gradient descent algorithm so that the task execution performance indicators corresponding to the subsequently generated resource scheduling strategies meet the preset conditions.

8. A cloud-edge layered collaborative information system integration resource scheduling system according to claim 7, characterized in that, Before collecting real-time status data from edge nodes and resource status data from cloud servers, the process also includes: Acquire information system integration requirement data, which includes data format descriptions, data transmission protocols, and data interaction frequencies of multiple heterogeneous systems; Based on the data format description, a data mapping relationship table is constructed, which is used to convert the data format of heterogeneous systems into a unified scheduling data format; Configure the bandwidth allocation strategy and data transmission priority of the data acquisition link based on the data transmission protocol and data interaction frequency.

9. A cloud-edge layered collaborative information system integration resource scheduling system according to claim 8, characterized in that, The process of constructing a data mapping relationship table based on the data format description includes: Parse the data format descriptions of each heterogeneous system to obtain the data fields and field attributes of each heterogeneous system; Determine the mapping relationship between the data fields of each heterogeneous system and the fields of the unified scheduling data format, wherein the unified scheduling data format includes edge status data fields, cloud status data fields and task execution status data fields; Based on the field mapping relationship, a data mapping relationship table is generated, which includes heterogeneous system data fields, unified scheduling data format fields, and field conversion algorithms.

10. An electronic device, comprising at least one processor; and a memory communicatively connected to said at least one processor; characterized in that, The memory stores instructions that can be executed by the at least one processor to enable the at least one processor to perform the system according to any one of claims 1-8.

Images