Distributed computing resource dynamic scheduling method suitable for new energy scene

Abstract

The invention belongs to the technical field of distributed computing, and relates to a distributed computing resource dynamic scheduling method and system suitable for a new energy scene. According to the method, the priorities of the calculation tasks are obtained by carrying out priority division on the calculation tasks according to the task characteristics. And a distributed computing node resource pool is established, so that centralized management and flexible scheduling of computing resources can be conveniently realized. Performance indexes of all nodes in the resource pool are collected and monitored, and reasonable allocation and efficient utilization of resources can be achieved. And scheduling the calculation tasks based on the priorities of the calculation tasks and the performance indexes of the nodes, generating a scheduling strategy, and ensuring that the tasks with high priorities preferentially obtain the calculation resources so as to meet the requirements on the task execution sequence in a new energy scene. And the scheduling strategy is adaptively optimized according to the performance index of each node, so that the utilization rate of resources and the stability of the system are improved. According to the method, the priority of the calculation task can be dynamically adjusted according to the characteristics of a new energy scene, and flexible scheduling of calculation resources is realized.

Description

Technical Field

[0001] The present invention belongs to the field of distributed computing technology and relates to a distributed computing resource dynamic scheduling method and system suitable for new energy scenarios. Background Art

[0002] With the rapid development of the new energy industry, the scale of wind farms continues to expand, and various data acquisition devices generate massive amounts of data in real time. These data need to be processed in real time through distributed computing systems to achieve functions such as wind turbine fault warning, efficiency analysis, and equipment diagnosis. The scheduling of distributed computing resources is of great significance to ensure the effective execution of computing tasks in new energy scenarios.

[0003] At present, mainstream distributed computing resource scheduling methods are mainly designed for traditional industrial scenarios, such as traditional thermal power, hydropower and other fields. These methods usually adopt fixed task priority division strategies and static resource allocation mechanisms. When processing computing tasks in traditional industrial scenarios, such scheduling methods can meet basic needs because the load is relatively stable and the amount of data does not change much. However, in new energy scenarios, due to the drastic changes in wind conditions, the amount of data collected fluctuates significantly, and the urgency of different types of tasks varies greatly. Fixed task priority division strategies and static resource allocation mechanisms can no longer meet the actual needs of resource allocation in new energy scenarios, resulting in unreasonable resource allocation in new energy scenarios.

[0004] In summary, the existing distributed computing resource scheduling methods have the problem of unreasonable resource allocation in new energy scenarios. Summary of the invention

[0005] The purpose of the present invention is to provide a distributed computing resource dynamic scheduling method and system suitable for new energy scenarios, so as to solve the technical problem of unreasonable resource allocation in existing distributed computing resource scheduling methods in new energy scenarios. The present invention can dynamically adjust the priority of computing tasks according to the characteristics of new energy scenarios, and realize flexible scheduling of computing resources.

[0006] In order to achieve the above object, the present invention adopts the following technical solutions: In a first aspect, the present invention provides a method for dynamically scheduling distributed computing resources applicable to new energy scenarios, comprising the following steps: Prioritize the computing tasks according to their characteristics to obtain the priorities of the computing tasks; Establish a distributed computing node resource pool, collect and monitor the performance indicators of each node in the resource pool; The computing tasks are scheduled based on their priorities and the performance indicators of each node. During the scheduling process, the computing tasks are preferentially assigned to the nodes where the data is located, and a scheduling strategy is generated. The scheduling strategy is adaptively optimized according to the performance indicators of each node.

[0007] Furthermore, the priorities of the computing tasks are divided according to the task characteristics to obtain the priorities of the computing tasks, which are as follows: The task characteristics include urgency, timeliness and complexity; Extract the urgency, timeliness and complexity of computing tasks; The priority of the computing task is determined according to the urgency, timeliness and complexity of the computing task.

[0008] Furthermore, the performance indicators of each node in the resource pool are collected and monitored as follows: The performance indicators of the node include CPU utilization, memory utilization and network bandwidth occupancy; Use sliding window method to regularly collect node performance indicators; Evaluate the health status of the node based on its performance indicators; If the performance indicators of a node are abnormal, the node offline mechanism will be automatically triggered.

[0009] Furthermore, the scheduling of computing tasks is performed based on the priority division of computing tasks and the performance indicators of each node. In the scheduling process of computing tasks, computing tasks are preferentially allocated to the nodes where the data is located, as follows: Obtain the priority of computing tasks, performance indicators of each node, resource requirements and available resources of the node; Tasks are assigned to the nodes where the data is located based on the priority of the computing task and the matching degree of the node's performance indicators, the matching degree of resource requirements and the available resources of the node, and the data locality factor.

[0010] Furthermore, the scheduling strategy is adaptively optimized according to the performance indicators of each node, as follows: Get the performance indicators of each node; Analyze and evaluate the scheduling strategy based on the performance indicators of each node and generate an optimization strategy; Dynamically adjust the parameters of the scheduling strategy according to the optimization strategy and update the optimization strategy; Verify the effectiveness of the updated optimization strategy.

[0011] Furthermore, it also includes adjusting the dispatching strategy based on wind forecast data, as follows: Receive wind forecast data; Obtain the relationship between wind conditions and calculated loads from historical data; Estimate future computing resource requirements based on wind forecast data and the relationship between wind conditions and computing load in historical data; Adjust the scheduling strategy based on computing resource requirements.

[0012] Furthermore, before obtaining the priority of the computing task by dividing the priority of the computing task according to the task characteristics, the computing task is dynamically scaled and adjusted, as follows: Monitor the queue length of computing tasks; When the queue length of computing tasks exceeds the first threshold, capacity expansion is triggered; The capacity is reduced when the queue length of the computing task is less than a second threshold, and the second threshold is less than the first threshold.

[0013] Furthermore, it also includes the classification of computing tasks, specifically classified into: real-time warning, performance analysis, and equipment diagnosis; Different timeliness requirements are set for different types of computing tasks, as follows: Real-time warning tasks are set to respond within seconds; Performance analysis tasks are set to respond in minutes; Response time of device diagnosis tasks.

[0014] Furthermore, the real-time warning category includes fan fault warning and performance warning tasks; The efficiency analysis category includes wind turbine power curve analysis and power generation efficiency assessment tasks; The equipment diagnosis category includes equipment health assessment and life prediction tasks.

[0015] In a second aspect, the present invention provides a distributed computing resource dynamic scheduling system applicable to new energy scenarios, comprising: Task analysis and priority management module: used to prioritize computing tasks according to task characteristics to obtain the priority of computing tasks; Resource monitoring module: used to establish a distributed computing node resource pool, collect and monitor the performance indicators of each node in the resource pool; Resource scheduling decision module: used to schedule computing tasks based on the priority of computing tasks and the performance indicators of each node. During the scheduling process of computing tasks, computing tasks are preferentially assigned to the nodes where the data is located, and a scheduling strategy is generated; Optimization feedback module: used to adaptively optimize the scheduling strategy based on the performance indicators of each node.

[0016] Compared with the prior art, the present invention has the following beneficial effects: 1. The present invention divides the priority of computing tasks according to the task characteristics to obtain the priority of computing tasks, which can ensure that critical tasks or urgent tasks are given priority. A distributed computing node resource pool is established to facilitate the centralized management and flexible scheduling of computing resources. The performance indicators of each node in the resource pool are collected and monitored to facilitate real-time understanding of the load status and performance bottleneck of the node, providing a key basis for subsequent task scheduling, and helping to achieve reasonable allocation and efficient utilization of resources. Based on the priority of the computing task and the performance indicators of each node, the computing task is scheduled, and a scheduling strategy is generated to ensure that high-priority tasks obtain computing resources first, thereby meeting the requirements for task execution order in the new energy scenario. It is conducive to reducing data transmission delays and bandwidth consumption, and improving the efficiency of task execution; the scheduling strategy is adaptively optimized according to the performance indicators of each node, and the scheduling strategy is continuously adjusted according to the performance indicators (actual operating status) to achieve optimal resource allocation and task execution efficiency, which is conducive to improving resource utilization and system stability. The present invention can dynamically adjust the priority of computing tasks according to the characteristics of the new energy scenario, and realize flexible scheduling of computing resources.

[0017] 2. The system includes: a task analysis and priority management module, a resource monitoring module, a resource scheduling decision module and an optimization feedback module. The task analysis and priority management module is used to prioritize computing tasks according to task characteristics to obtain the priority of computing tasks. The resource monitoring module is used to establish a distributed computing node resource pool, collect and monitor the performance indicators of each node in the resource pool. The resource scheduling decision module is used to schedule computing tasks based on the priority of computing tasks and the performance indicators of each node. During the scheduling process of computing tasks, computing tasks are preferentially assigned to the nodes where the data is located, and a scheduling strategy is generated. The optimization feedback module is used to adaptively optimize the scheduling strategy according to the performance indicators of each node. The various modules cooperate with each other and can dynamically adjust the priority of computing tasks according to the characteristics of new energy scenarios, thereby realizing flexible scheduling of computing resources. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] Figure 1 It is a schematic diagram of the system architecture of the present invention; Figure 2 A flowchart for task priority division of the present invention; Figure 3 It is a schematic diagram of resource pool management of the present invention; Figure 4 It is a scheduling decision flow chart of the present invention; Figure 5 It is the adaptive optimization flow chart of the present invention; Figure 6 The present invention is a flow chart of the method. DETAILED DESCRIPTION

[0019] In order to enable those skilled in the art to better understand the scheme of the present invention, the technical scheme in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of the present invention.

[0020] It should be noted that the terms "first", "second", etc. in the specification of the present invention and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchanged where appropriate, so that the embodiments of the present invention described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device that includes a series of steps or units is not necessarily limited to those steps or units that are clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0021] The present invention is further described in detail below in conjunction with the accompanying drawings: See also Figure 6 The present invention discloses a distributed computing resource dynamic scheduling method applicable to new energy scenarios, comprising the following steps: S1, the priority of computing tasks is divided according to the task characteristics to obtain the priority of computing tasks, which can ensure that critical tasks or urgent tasks are processed first. It helps to improve the efficiency of overall task processing and reduce the waiting time of important tasks, thereby meeting the requirements of time sensitivity in new energy scenarios.

[0022] S2, establishes a distributed computing node resource pool to facilitate centralized management and flexible scheduling of computing resources. Collect and monitor the performance indicators of each node in the resource pool to facilitate real-time understanding of the node's load status and performance bottlenecks, providing a key basis for subsequent task scheduling and helping to achieve reasonable allocation and efficient use of resources.

[0023] S3, based on the priority of computing tasks and the performance indicators of each node, the computing tasks are scheduled to ensure that high-priority tasks have priority in obtaining computing resources, thereby meeting the requirements for task execution order in new energy scenarios. During the scheduling of computing tasks, the computing tasks are preferentially assigned to the nodes where the data is located, and a scheduling strategy is generated, which is conducive to reducing data transmission delays and bandwidth consumption, and improving the efficiency of task execution.

[0024] S4, adaptively optimize the scheduling strategy according to the performance indicators of each node, and continuously adjust the scheduling strategy according to the performance indicators (actual operating status) to achieve the best resource allocation and task execution efficiency, ensuring that the scheduling strategy is always consistent with the actual state of the system, which is conducive to improving resource utilization and system stability. The present invention can dynamically adjust the priority of computing tasks according to the characteristics of new energy scenarios, and realize flexible scheduling of computing resources.

[0025] Embodiment 1: See also Figure 6 The present invention discloses a distributed computing resource dynamic scheduling method applicable to new energy scenarios, comprising the following steps: S1, the priority of computing tasks is divided according to the task characteristics to obtain the priority of computing tasks, as follows: The task characteristics include urgency, timeliness and complexity; Extract the urgency, timeliness and complexity of computing tasks; The priority of the computing task is determined according to the urgency, timeliness and complexity of the computing task.

[0026] Preferably, before obtaining the priority of the computing task by performing priority division of the computing task according to the task characteristics, the computing task is dynamically scaled and adjusted, as follows: Monitor the queue length of computing tasks; When the queue length of computing tasks exceeds the first threshold, capacity expansion is triggered; The capacity is reduced when the queue length of the computing task is less than a second threshold, and the second threshold is less than the first threshold.

[0027] S2, establish a distributed computing node resource pool, collect and monitor the performance indicators of each node in the resource pool, as follows: The performance indicators of the node include CPU utilization, memory utilization and network bandwidth occupancy; Use sliding window method to regularly collect node performance indicators; Evaluate the health status of the node based on its performance indicators; If the performance indicators of a node are abnormal, the node offline mechanism will be automatically triggered.

[0028] S3, based on the priority of computing tasks and the performance indicators of each node, the computing tasks are scheduled. During the scheduling process, the computing tasks are preferentially assigned to the nodes where the data is located, and the scheduling strategy is generated. The details are as follows: Obtain the priority of computing tasks, performance indicators of each node, resource requirements and available resources of the node; Tasks are assigned to the nodes where the data is located based on the priority of the computing task and the matching degree of the node's performance indicators, the matching degree of resource requirements and the available resources of the node, and the data locality factor.

[0029] S4, adaptively optimize the scheduling strategy according to the performance indicators of each node, as follows: Get the performance indicators of each node; Analyze and evaluate the scheduling strategy based on the performance indicators of each node and generate an optimization strategy; Dynamically adjust the parameters of the scheduling strategy according to the optimization strategy and update the optimization strategy; Verify the effectiveness of the updated optimization strategy.

[0030] Preferably, the method further includes adjusting the dispatching strategy based on the wind condition forecast data, as follows: Receive wind forecast data; Obtain the relationship between wind conditions and calculated loads from historical data; Estimate future computing resource requirements based on wind forecast data and the relationship between wind conditions and computing load in historical data; Adjust the scheduling strategy based on computing resource requirements.

[0031] Preferably, the computing tasks are classified into: real-time warning, performance analysis and equipment diagnosis; Different timeliness requirements are set for different types of computing tasks, as follows: Real-time warning tasks are set to respond within seconds; Performance analysis tasks are set to respond in minutes; Response time of device diagnosis tasks.

[0032] Preferably, the real-time warning category includes fan fault warning and performance warning tasks; The efficiency analysis category includes wind turbine power curve analysis and power generation efficiency assessment tasks; The equipment diagnosis category includes equipment health assessment and life prediction tasks.

[0033] To sum up, the present invention can reasonably classify computing tasks and dynamically adjust priorities according to the characteristics of new energy scenarios, and at the same time realize a method for flexible scheduling of computing resources to solve the technical problem of unreasonable resource allocation in existing distributed computing resource scheduling methods under new energy scenarios.

[0034] The present invention can better meet the timeliness requirements of different types of tasks in new energy scenarios by dividing computing tasks into real-time warning, performance analysis, and equipment diagnosis categories, and dynamically adjusting priorities based on task characteristics. The performance monitoring module is used to regularly collect node status, and the offline mechanism is automatically triggered when the node performance index is abnormal, thereby improving the reliability of the distributed computing node resource pool. By giving priority to assigning tasks to the scheduling strategy of the node where the data is located, the data transmission overhead is reduced. At the same time, an adaptive optimization mechanism is introduced to enable the system to continuously optimize the scheduling strategy according to the operating conditions.

[0035] In addition, the present invention also realizes resource pre-allocation in combination with wind condition prediction, and realizes dynamic resource scaling through task queue monitoring, further enhancing the system's adaptability to new energy scenario characteristics.

[0036] The present invention not only solves the problem of unreasonable resource allocation in distributed computing resource scheduling under new energy scenarios, but also can effectively cope with fluctuations in data processing demand caused by changes in wind conditions, so that computing resources can be used more reasonably and efficiently.

[0037] The present invention classifies and prioritizes computing tasks into real-time warning, performance analysis, and equipment diagnosis categories; establishes a distributed computing node resource pool, collects and monitors the CPU utilization, memory usage, network bandwidth occupancy, and other performance indicators of each node through a performance monitoring module; makes intelligent scheduling decisions based on task characteristics and resource status, and preferentially assigns tasks to nodes where data is located; and adaptively optimizes scheduling strategies based on system operation conditions. The present invention also includes a resource pre-allocation mechanism based on wind condition prediction and a dynamic scaling mechanism for computing tasks. This method solves the problem of unreasonable resource allocation in distributed computing resource scheduling under new energy scenarios, and improves the utilization efficiency of computing resources.

[0038] Based on the above method, the present invention also discloses a distributed computing resource dynamic scheduling system suitable for new energy scenarios, see Figure 1 ,include: Task analysis and priority management module: used to prioritize computing tasks according to task characteristics to obtain the priority of computing tasks; Resource monitoring module: used to establish a distributed computing node resource pool, collect and monitor the performance indicators of each node in the resource pool; Resource scheduling decision module: used to schedule computing tasks based on the priority of computing tasks and the performance indicators of each node. During the scheduling process of computing tasks, computing tasks are preferentially assigned to the nodes where the data is located, and a scheduling strategy is generated; Optimization feedback module: used to adaptively optimize the scheduling strategy based on the performance indicators of each node.

[0039] The various modules of the present invention cooperate with each other and can dynamically adjust the priority of computing tasks according to the characteristics of new energy scenarios, thereby realizing flexible scheduling of computing resources.

[0040] Embodiment 2: See also Figure 1 This embodiment discloses a distributed computing resource dynamic scheduling system suitable for new energy scenarios, which is used to solve the technical problems of unreasonable computing resource allocation and low scheduling efficiency in new energy scenarios in the prior art. The method realizes efficient scheduling and utilization of computing resources through technical means such as task classification and priority division, dynamic management of resource pools, intelligent scheduling decisions and adaptive optimization, as follows: like Figure 1 As shown, the distributed computing resource dynamic scheduling system architecture of the present invention includes a task access layer, a task analysis and priority management module, a resource scheduling decision module, a resource monitoring module, a task execution module and an optimization feedback module. The task access layer receives and pre-processes computing tasks, and passes the processed tasks to the task analysis and priority management module; the resource monitoring module collects resource status (performance indicators of each node) information in real time and feeds it back to the resource scheduling decision module, and the resource status is the performance indicator of each node; the task execution module is responsible for the specific execution of the task and passes the execution status to the optimization feedback module; the optimization feedback module optimizes the scheduling strategy according to the execution situation and feeds it back to the resource scheduling decision module. Each module exchanges information through the data bus to ensure the real-time and reliability of the system.

[0041] like Figure 2 As shown, the present invention first classifies and prioritizes computing tasks. The system receives tasks through the task input module and divides computing tasks into three categories: real-time warning, performance analysis, and equipment diagnosis. The real-time warning category mainly includes tasks such as fan fault warning and performance warning; the performance analysis category includes tasks such as fan power curve analysis and power generation efficiency evaluation; the equipment diagnosis category includes tasks such as equipment health assessment and life prediction. The feature extraction module analyzes the task, extracts the features of the task such as urgency, timeliness, and complexity, and finally determines the final priority of the task through the priority calculation module.

[0042] like Figure 3 As shown, the present invention establishes a distributed computing node resource pool to monitor the resource status of each computing node (the performance indicators of each node) in real time. Each computing node in the resource pool is equipped with a performance monitoring module, which uses a sliding window method to regularly collect the node's CPU utilization, memory usage, network bandwidth occupancy and other performance indicators. By comprehensively evaluating the various performance indicators of the node, the system can accurately grasp the health status of each node. When the performance indicators of the node are abnormal, the system automatically triggers the node offline mechanism to ensure the reliability of the resource pool.

[0043] like Figure 4 As shown, the present invention makes intelligent scheduling decisions based on the priority of computing tasks and the performance indicators of each node. After the task arrives, the system first checks the resource status, then performs node matching calculations, then performs data locality evaluation, and finally completes the task allocation execution. During the node matching calculation process, the system comprehensively considers factors such as task priority and node performance matching, resource requirements and node available resource matching, and data locality. The system preferentially assigns tasks to the nodes where the data is located to reduce data transmission overhead. When the load difference between nodes is too large, the load balancing mechanism is triggered to achieve dynamic migration of tasks.

[0044] like Figure 5 As shown in the figure, the present invention realizes an adaptive optimization mechanism for scheduling strategies. The mechanism includes five key links: performance indicator collection, data analysis and evaluation, optimization strategy generation, dynamic parameter adjustment and effect verification, forming a closed-loop optimization process. The system continuously collects performance indicator data, evaluates the effect of the current scheduling strategy through data analysis, generates an optimization strategy and dynamically adjusts relevant parameters, and finally ensures the effectiveness of the optimization through effect verification. This cyclic process enables the system to continuously adapt to load changes and maintain optimal performance.

[0045] The present invention also implements a resource pre-allocation mechanism in combination with wind forecasting. The system receives the wind forecast data for the next 24 hours, estimates the future computing resource requirements by analyzing the relationship between wind conditions and computing load in historical data, and adjusts resource configuration in advance. For computing tasks, the system implements a dynamic scaling mechanism. By monitoring the length of the task queue, it triggers expansion when there is a backlog of tasks and shrinks capacity when there is a low load, always maintaining the optimal resource pool size.

[0046] In terms of data timeliness, the present invention sets strict timeliness requirements for different types of computing tasks: real-time warning tasks require a response in seconds, performance analysis tasks require a response in minutes, and equipment diagnosis tasks can accept a longer response time. The system ensures the timeliness requirements of data processing through mechanisms such as task priority sorting and data degradation processing.

[0047] The present invention solves the technical problems faced by distributed computing resource scheduling in new energy scenarios through the organic combination of task classification and priority division mechanism, resource pool dynamic management mechanism, intelligent scheduling decision mechanism and adaptive optimization mechanism. Compared with the prior art, the present invention has the following advantages: First, through detailed task classification and dynamic priority management, the timely processing of key businesses is ensured; second, based on the dynamic management of the resource pool and intelligent scheduling decisions, the utilization efficiency of computing resources is improved; third, through the adaptive optimization mechanism, the system can continuously optimize the scheduling strategy according to the actual operating conditions; finally, the resource pre-allocation mechanism combined with wind condition prediction further improves the system's adaptability to the characteristics of new energy scenarios. The present invention can effectively solve the problems of unreasonable computing resource allocation and low scheduling efficiency in new energy scenarios, and has important practical value.

[0048] The above contents are only for explaining the technical idea of ​​the present invention and cannot be used to limit the protection scope of the present invention. Any changes made on the basis of the technical solution in accordance with the technical idea proposed by the present invention shall fall within the protection scope of the present invention.

Claims

1. A distributed computing resource dynamic scheduling method suitable for new energy scenarios, characterized in that: The following steps are involved: Prioritize the computing tasks according to their characteristics to obtain the priorities of the computing tasks; Establish a distributed computing node resource pool, collect and monitor the performance indicators of each node in the resource pool; The computing tasks are scheduled based on their priorities and the performance indicators of each node. During the scheduling process, the computing tasks are preferentially assigned to the nodes where the data is located, and a scheduling strategy is generated. The scheduling strategy is adaptively optimized according to the performance indicators of each node.

2. The method for dynamic scheduling of distributed computing resources applicable to new energy scenarios according to claim 1, characterized in that: The priority of the computing task is obtained by dividing the priority of the computing task according to the task characteristics, which is as follows: The task characteristics include urgency, timeliness and complexity; Extract the urgency, timeliness and complexity of computing tasks; The priority of the computing task is determined according to the urgency, timeliness and complexity of the computing task.

3. The method for dynamic scheduling of distributed computing resources applicable to new energy scenarios according to claim 1, characterized in that: The performance indicators of each node in the resource pool are collected and monitored as follows: The performance indicators of the node include CPU utilization, memory utilization and network bandwidth occupancy; Use sliding window method to regularly collect node performance indicators; Evaluate the health status of the node based on its performance indicators; If the performance indicators of a node are abnormal, the node offline mechanism will be automatically triggered.

4. The method for dynamic scheduling of distributed computing resources applicable to new energy scenarios according to claim 1, characterized in that: The scheduling of computing tasks is performed based on the priority division of computing tasks and the performance indicators of each node. In the scheduling process of computing tasks, computing tasks are preferentially allocated to the nodes where the data is located. The details are as follows: Obtain the priority of computing tasks, performance indicators of each node, resource requirements and available resources of the node; Tasks are assigned to the nodes where the data is located based on the priority of the computing task and the matching degree of the node's performance indicators, the matching degree of resource requirements and the available resources of the node, and the data locality factor.

5. The method for dynamic scheduling of distributed computing resources applicable to new energy scenarios according to claim 1, characterized in that: The scheduling strategy is adaptively optimized according to the performance indicators of each node, as follows: Get the performance indicators of each node; Analyze and evaluate the scheduling strategy based on the performance indicators of each node and generate an optimization strategy; Dynamically adjust the parameters of the scheduling strategy according to the optimization strategy and update the optimization strategy; Verify the effectiveness of the updated optimization strategy.

6. The method for dynamic scheduling of distributed computing resources applicable to new energy scenarios according to claim 1, characterized in that: It also includes adjusting the dispatch strategy based on wind forecast data, as follows: Receive wind forecast data; Obtain the relationship between wind conditions and calculated loads from historical data; Estimate future computing resource requirements based on wind forecast data and the relationship between wind conditions and computing load in historical data; Adjust the scheduling strategy based on computing resource requirements.

7. The method for dynamic scheduling of distributed computing resources applicable to new energy scenarios according to claim 1, characterized in that: Before obtaining the priority of the computing task by dividing the priority of the computing task according to the task characteristics, the computing task is dynamically scaled and adjusted, as follows: Monitor the queue length of computing tasks; When the queue length of computing tasks exceeds the first threshold, capacity expansion is triggered; The capacity is reduced when the queue length of the computing task is less than a second threshold, and the second threshold is less than the first threshold.

8. The method for dynamic scheduling of distributed computing resources applicable to new energy scenarios according to claim 1, characterized in that: It also includes the classification of computing tasks, specifically into: real-time warning, performance analysis, and equipment diagnosis; Different timeliness requirements are set for different types of computing tasks, as follows: Real-time warning tasks are set to respond within seconds; Performance analysis tasks are set to respond in minutes; Response time of device diagnosis tasks.

9. The method for dynamic scheduling of distributed computing resources applicable to new energy scenarios according to claim 8, characterized in that: The real-time warning category includes fan fault warning and performance warning tasks; The efficiency analysis category includes wind turbine power curve analysis and power generation efficiency assessment tasks; The equipment diagnosis category includes equipment health assessment and life prediction tasks.

10. A distributed computing resource dynamic scheduling system suitable for new energy scenarios, characterized in that: include: Task analysis and priority management module: used to prioritize computing tasks according to task characteristics to obtain the priority of computing tasks; Resource monitoring module: used to establish a distributed computing node resource pool, collect and monitor the performance indicators of each node in the resource pool; Resource scheduling decision module: used to schedule computing tasks based on the priority of computing tasks and the performance indicators of each node. During the scheduling process of computing tasks, computing tasks are preferentially assigned to the nodes where the data is located, and a scheduling strategy is generated; Optimization feedback module: used to adaptively optimize the scheduling strategy based on the performance indicators of each node.

Images