Resource allocation calling method and system based on nuclear power calculation large model
Through an intelligent scheduling method based on the large nuclear power computing model, combined with neural networks and naive Bayes algorithm, the problems of uneven resource allocation and low utilization rate of high-performance computing platforms are solved, and efficient resource utilization and dynamic scheduling capabilities are achieved.
Patent Information
- Application Number
- CN202411160063.3
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-08-22
- Publication Date
- 2024-11-29
AI Technical Summary
The existing high-performance computing platforms have uneven resource allocation, low utilization rate, and lack intelligent scheduling capabilities. They cannot dynamically respond to sudden resource needs and fail to effectively combine external supercomputing center resources.
Based on the nuclear power computing big model, combined with neural networks, k-nearest neighbors and naive Bayes algorithms, a nuclear power computing data model is built, and through software licensing, resource quota and hybrid cloud scheduling strategies, intelligent scheduling and efficient utilization of resources are achieved.
It effectively improves the utilization efficiency of simulation computing resources, improves the utilization rate of computing resources by about 20%, and realizes intelligent and efficient resource scheduling.
Smart Images

Figure CN119048279A_ABST
Abstract
Description
Technical Field
[0001] The present disclosure relates to the technical field of nuclear power computing resource scheduling, and in particular to a resource allocation and calling method and system based on a nuclear power computing large model. Background Art
[0002] The statements in this section merely provide background information related to the present disclosure and do not necessarily constitute prior art.
[0003] With the continuous iteration and development of nuclear power computing resources, there are generally many computing resource pools with different configuration types. Traditional high-performance computing platforms generally use manual selection of static allocation to allocate and schedule computing resources, which will cause problems such as uneven resource allocation and low computing resource utilization. . Secondly, the existing high-performance computing scheduling is only from the perspective of software licensing and hardware resource status, and is not related to the computing characteristics and software characteristics of nuclear power numerical simulation calculations. Resource scheduling is not intelligent enough. Thirdly, the existing high-performance computing scheduling is not combined with external supercomputing center resources, and cannot achieve dynamic and elastic scheduling of high-performance computing resources, and lacks the ability to respond to sudden resource demands. Summary of the invention
[0004] In order to solve the above problems, the present disclosure proposes a resource allocation and calling method and system based on a nuclear power computing big model. Based on the computing characteristics of nuclear power research and development, computing software characteristics and accumulated historical computing data information, a nuclear power computing big model is constructed in combination with the performance configuration of high-performance servers. According to the model, through software license scheduling strategy, resource quota strategy, resource load strategy mechanism, and combined with hybrid cloud resource scheduling strategy, intelligent scheduling and efficient utilization of nuclear power research and development computing resources are realized.
[0005] According to some embodiments, the present disclosure adopts the following technical solutions:
[0006] The resource allocation calling method based on the nuclear power computing large model includes:
[0007] Obtain basic data on nuclear power system operation;
[0008] Automatically conduct data comparison and evaluation based on the bench test data of nuclear power equipment and nuclear power plant operation data in the basic data;
[0009] For the evaluated data, the accumulated historical computing operation data is analyzed from multiple dimensions including computing users, computing types, computing devices, software and versions, operating systems, and the number of cores through neural network algorithms, k-nearest neighbor algorithms, and naive Bayes algorithms to build a nuclear power computing data model.
[0010] The job information submitted by the user is obtained. The nuclear power computing data model aims to minimize the time consumption, solves the matching optimal resource specifications and computing scale, and sets the job as a forwarding job according to the hybrid cloud scheduling strategy to realize the scheduling of the forwarding job status.
[0011] According to some embodiments, the present disclosure adopts the following technical solutions:
[0012] The resource allocation and calling system based on the nuclear power computing model includes:
[0013] Data acquisition module, used to obtain basic data of nuclear power system operation;
[0014] A data evaluation module is used to automatically compare and evaluate data based on the bench test data of nuclear power equipment and the nuclear power plant operation data in the basic data;
[0015] The model building module is used to calculate and analyze the accumulated historical computing operation data from multiple dimensions such as computing users, computing types, computing devices, software and versions, operating systems, and the number of cores based on the evaluated data through neural network algorithms, k-nearest neighbor algorithms, and naive Bayes algorithms to build a nuclear power computing data model.
[0016] The solution scheduling module is used to obtain the job information submitted by the user. The nuclear power calculation data model aims to minimize the time consumption, solves the optimal matching resource specifications and computing scale, and sets the job as a forwarding job according to the hybrid cloud scheduling strategy to realize the scheduling of the forwarding job status.
[0017] According to some embodiments, the present disclosure adopts the following technical solutions:
[0018] A non-transitory computer-readable storage medium is used to store computer instructions. When the computer instructions are executed by a processor, the resource allocation and calling method based on the nuclear power computing large model is implemented.
[0019] According to some embodiments, the present disclosure adopts the following technical solutions:
[0020] An electronic device comprises: a processor, a memory and a computer program; wherein the processor is connected to the memory, the computer program is stored in the memory, and when the electronic device is running, the processor executes the computer program stored in the memory so that the electronic device executes the resource allocation calling method based on the nuclear power computing large model.
[0021] Compared with the prior art, the present invention has the following beneficial effects:
[0022] The present invention discloses a resource allocation and calling method based on a nuclear power computing big model. The nuclear power computing big model is constructed based on the computing characteristics of nuclear power research and development, computing software characteristics and accumulated historical computing data information, combined with the performance configuration of high-performance servers. According to the model, through software license scheduling strategy, resource quota strategy, resource load and other policy mechanisms, and combined with hybrid cloud resource scheduling strategy, intelligent scheduling and efficient utilization of nuclear power research and development computing resources are realized.
[0023] The resource allocation and calling method based on the nuclear power computing large model disclosed in the present invention can intelligently and efficiently schedule simulation computing resources, effectively improve the utilization efficiency of simulation computing, and achieve an increase of about 20% in the interest rate of high-performance computing resources. BRIEF DESCRIPTION OF THE DRAWINGS
[0024] The accompanying drawings constituting a part of the present disclosure are used to provide a further understanding of the present disclosure. The illustrative embodiments of the present disclosure and their descriptions are used to explain the present disclosure and do not constitute an improper limitation on the present disclosure.
[0025] Figure 1 A resource allocation call flow chart based on a nuclear power computing large model according to an embodiment of the present disclosure;
[0026] Figure 2 A schematic diagram of a computing cluster according to an embodiment of the present disclosure;
[0027] Figure 3 A resource allocation call architecture diagram based on a nuclear power computing large model according to an embodiment of the present disclosure;
[0028] Figure 4 This is a diagram showing the effect of improving cluster utilization according to an embodiment of the present disclosure. DETAILED DESCRIPTION
[0029] The present disclosure is further described below in conjunction with the accompanying drawings and embodiments.
[0030] It should be noted that the following detailed descriptions are all illustrative and are intended to provide further explanation of the present disclosure. Unless otherwise specified, all technical and scientific terms used herein have the same meanings as those commonly understood by those skilled in the art to which the present disclosure belongs.
[0031] It should be noted that the terms used herein are only for describing specific embodiments and are not intended to limit the exemplary embodiments according to the present disclosure. As used herein, unless the context clearly indicates otherwise, the singular form is also intended to include the plural form. In addition, it should be understood that when the terms "comprising" and / or "including" are used in this specification, it indicates the presence of features, steps, operations, devices, components and / or combinations thereof.
[0032] Explanation of terms
[0033] Neural Network Algorithms
[0034] Neural network algorithm: It is an algorithm based on neuroscience theory, and its basic unit is a neuron. Each neuron receives input signals and generates output signals after processing through an activation function. Multiple neurons are combined to form a neural network, which automatically adjusts weights and thresholds according to input data and learning algorithms to optimize output results. The main types of neural network algorithms include perceptrons, convolutional neural networks, and recurrent neural networks.
[0035] Computing power scheduling: When computing resources are limited, computing power scheduling is the core process of rationally allocating and managing available computing resources. This process includes AI-optimized allocation methods, scientific configuration of computing resources and task allocation to achieve efficient utilization and balanced execution, as well as consideration of task management, resource monitoring, resource allocation, load balancing, and resource recycling. AI plays a key role in forecasting and planning, optimizing resource allocation, load balancing management, automated decision-making, and anomaly detection and processing.
[0036] k-nearest neighbor algorithm: k-nearest neighbor algorithm is a kind of instance-based learning, or lazy learning with local approximation and deferring all computation until after classification. The nearest neighbors (k) are used to predict unknown data points. The k value is a key factor in prediction accuracy. It is very useful to weigh the weights of neighbors, whether for classification or regression. The weights of closer neighbors are greater than those of farther neighbors.
[0037] Naive Bayes algorithm: Naive Bayes is based on the Bayesian theorem of probability theory and is widely used, from text classification, spam filters, medical diagnosis, etc. Naive Bayes is suitable for scenarios where features are independent of each other, such as using the length and width of petals to predict the type of flower.
[0038] Support Vector Machine: Support Vector Machines (SVM) is a binary classification model. Its purpose is to find a hyperplane to segment samples. The principle of segmentation is to maximize the interval, which is ultimately converted into a convex quadratic programming problem to solve.
[0039] Example 1
[0040] In one embodiment of the present disclosure, a resource allocation and calling method based on a nuclear power computing large model is provided, comprising:
[0041] Step 1: Obtain basic data on nuclear power system operation;
[0042] Step 2: Automatically conduct data comparison and evaluation based on the bench test data of nuclear power equipment and nuclear power plant operation data in the basic data;
[0043] Step 3: For the evaluated data, the accumulated historical computing operation data is analyzed from multiple dimensions such as computing users, computing types, computing devices, software and versions, operating systems, and the number of cores through the neural network algorithm, k-nearest neighbor algorithm, and naive Bayes algorithm to build a nuclear power computing data model.
[0044] Step 4: Obtain the job information submitted by the user. The nuclear power computing data model aims to minimize the time consumption, solves the optimal matching resource specifications and computing scale, and sets the job as a forwarding job according to the hybrid cloud scheduling strategy to realize the scheduling of the forwarding job status.
[0045] As an embodiment, the specific implementation process of a resource allocation calling method based on a nuclear power computing large model disclosed in the present invention includes:
[0046] Step 1: Obtain basic data on nuclear power system operation;
[0047] Specifically, data collection and aggregation include:
[0048] 1) Resource type specification data collection: Collect specification attribute data of each high-performance computing resource group, including specification name, number of cores per node, CPU main frequency, memory size, memory frequency, hard disk size, operating system, whether IB is available, total number of resource cores, whether GPU card is available, etc.
[0049] 2) Computing software feature data collection: Collect feature information data of each nuclear power high-performance computing software, including software name, software version, whether it supports multi-node parallelism, whether it is memory sensitive, whether it is CPU frequency sensitive, the optimal operating system, whether it supports GPU acceleration, whether local hard disk support is required, etc.
[0050] 3) Historical computing data collection: Collect the attributes and performance data information of historical high-performance computing jobs, including software name, software version, computing type, computing device, computing scale, resource specifications, computing time, whether GPU is used, memory usage, hard disk usage, etc., and intelligently identify and extract indicator data of simulation input and output parameters of computing jobs.
[0051] 4) Bench test data collection: Based on the experimental data management system, the bench test data of nuclear power equipment is collected, including the experimental design data, experimental operation parameters and experimental result analysis data of nuclear power equipment.
[0052] 5) Nuclear power plant equipment operation data collection: Through the nuclear power plant equipment monitoring and management system, the real-time operation data of the nuclear power plant equipment is collected, including the equipment design parameters, operating environment parameters, equipment operating performance parameters, abnormal operation data, etc.
[0053] Step 2: Automatically conduct data comparison and evaluation based on the bench test data of nuclear power equipment and nuclear power plant operation data in the basic data;
[0054] Specifically, in order to ensure the correctness and reliability of the data in the nuclear power calculation data model, automatic evaluation and manual evaluation are added before building the nuclear power calculation data model, including:
[0055] Automatic evaluation is to automatically compare and evaluate the bench test data of nuclear power equipment and the collected nuclear power plant operation data. The data comparison and evaluation are input parameter data and output result data, including shape, material, material properties, boundary conditions, temperature, mass, thickness, quantity, current, voltage, torque, angle, stiffness, speed, pressure, acceleration, frequency, resistance, stress, strain, strain energy, displacement, support reaction force, offset, etc. By classifying and sub-classifying, and integrating the information of the relationship between numerical simulation analysis data, bench test data of nuclear power equipment and collected nuclear power plant operation data, the three The correlation coupling of square data is carried out. For the numerical simulation analysis data, the input and output parameter data of shape, material, material properties, boundary conditions, temperature, mass, thickness, quantity, current, voltage, torque, angle, stiffness, speed, pressure, acceleration, frequency, resistance, stress, strain, strain energy, displacement, support reaction force, offset, etc. are compared with the bench test data of nuclear power equipment and the input and output data of collected nuclear power plant operation data. Only when the input parameter data and the output result data match, the data is considered valid in the subsequently constructed nuclear power calculation data model, otherwise it is invalid and discarded.
[0056] Manual evaluation is associated with the project review management system. The calculation data that passes the simulation calculation review can be marked as valid in the subsequent nuclear power calculation data model, otherwise it will be invalid and discarded.
[0057] Step 3: Construct a nuclear power calculation data model;
[0058] Specifically, a historical computing data model for nuclear power computing operations is constructed through data collection and data evaluation. The accumulated historical computing operation data are summarized and analyzed from multiple dimensions such as computing users, computing types, computing devices, software and versions, operating systems, and number of cores through neural network algorithms, naive Bayesian algorithms, etc., to form a nuclear power computing data model, providing data decision-making basis for subsequent intelligent scheduling.
[0059] First, the Naive Bayes algorithm is used to classify all jobs in the past month by taking information such as computing users, computing types, computing devices, software and versions, operating systems, number of cores, and computing time as feature values. The probability of each resource group appearing under this sample condition is solved, and the one with the largest probability is considered to be the optimal resource group.
[0060] In order to improve the accuracy of the computing data model, a neural network algorithm was also introduced. The existing neural network algorithm was used to take parameters such as computing users, computing types, computing devices, software and versions, operating systems, number of cores, and computing time as the input layer of the neural network algorithm. The neural network only has a single layer, and the weight values of the above inputs are continuously adjusted during the entire training process to finally obtain the optimal resource group.
[0061] Step 4: Obtain the job information submitted by the user. The nuclear power computing data model aims to minimize the time consumption, solves the optimal matching resource specifications and computing scale, and sets the job as a forwarding job according to the hybrid cloud scheduling strategy to realize the scheduling of the forwarding job status.
[0062] Specifically, the intelligent matching scheduling algorithm is implemented. Users do not need to select fixed information such as which resource pool to choose on the job submission page. Users select the software and version of the calculation, set the input file, and submit the job. The system automatically and intelligently submits the job to the matching resource pool for solution calculation based on the nuclear power calculation data model and the busy status of each resource pool in the current cluster, including:
[0063] 4.1) According to the software name, software version, computing type, computing device and other attributes of the job submitted by the user, the nuclear power computing data model is used to obtain the optimal (the shortest time consumption is the optimal) resource specifications and computing scale that can be matched, including:
[0064] 4.1.1) Obtain all jobs submitted in the cluster within the last month based on input information such as software name, software version, operating system, computing type, and computing device. Return information includes software name, software version, operating system, computing type, computing device, resource name, computing scale, computing time, etc.
[0065] 4.1.2) The returned job list is grouped according to software name, software version, operating system, computing type, computing device, computing scale and resource name. For each group, only the record with the shortest time consumption is taken. This record is the nuclear power computing data model.
[0066] 4.1.3) Based on the software name, software version, and operating system of the job submitted by the user, the software features can be read from the computing software feature data, including whether it supports multi-node parallelism and the sensitivity type (CPU sensitive, memory sensitive, GPU sensitive).
[0067] 4.1.4) Filter the nuclear power computing data model according to the acquired software features and the computing scale selected when submitting the job, and only retain the resources that meet the requirements.
[0068] 4.1.5) Filter the nuclear power calculation data model based on whether the resource is downgraded when the user submits the job, and only retain the resources that meet the requirements.
[0069] 4.1.6) Obtain the resource name with the shortest time consumption from the filtered nuclear power calculation data model as the matched optimal resource.
[0070] 4.2) If there are idle resources in the cluster resource specification and the idle resources meet the matching computing scale, the job will be automatically scheduled and allocated to the specification resources. If the idle resources do not meet the matching computing scale, the computing scale will be downgraded and then scheduled and allocated to the specification resources.
[0071] 4.3) If there are no idle resources in the cluster resource specifications, the suboptimal (the one with the shortest time is the best) resource specifications and computing scale that can be matched are obtained through the nuclear power computing data model based on the software name, software version, computing type, computing device and other attributes of the job submitted by the user.
[0072] 4.4) Execute 4.2) logical judgment. If 4.2) fails, continue
[0073] 4.5) If there are no idle resources in the cluster resource specifications, the third-best (the one with the shortest time consumption) resource specifications and computing scale that can be matched are obtained through the nuclear power computing data model based on the software name, software version, computing type, computing device and other attributes of the job submitted by the user.
[0074] 4.6) Execute 4.2) logical judgment. If 4.2) fails, continue
[0075] 4.7) The optimal match only matches the first N resource specifications. If none of them pass, the resource specifications that have not been matched and contain idle resources are obtained, and automatic matching is performed according to the resource specification attributes and software version attributes. After matching the optimal resource, it is scheduled and allocated to the resource with this specification.
[0076] 4.8) If there is no matching resource specification, the job is set as a forwarding job according to the hybrid cloud scheduling strategy, and the job is assigned to the supercomputer or public cloud resource platform pool for calculation. After the calculation is completed, the data results are pulled into local storage.
[0077] 4.8.1) When the local computing cluster has no idle resources, the local computing job is automatically set to the forwarding job state. After the state is set, the local cluster will no longer schedule the job, but will schedule it through the forwarding job scheduler.
[0078] Example 2
[0079] In one embodiment of the present disclosure, a resource allocation and calling system based on a nuclear power computing large model is provided, including:
[0080] Data acquisition module, used to obtain basic data of nuclear power system operation;
[0081] A data evaluation module is used to automatically compare and evaluate data based on the bench test data of nuclear power equipment and the nuclear power plant operation data in the basic data;
[0082] The model building module is used to calculate and analyze the accumulated historical computing operation data from multiple dimensions such as computing users, computing types, computing devices, software and versions, operating systems, and the number of cores based on the evaluated data through neural network algorithms, k-nearest neighbor algorithms, and naive Bayes algorithms to build a nuclear power computing data model.
[0083] The solution scheduling module is used to obtain the job information submitted by the user. The nuclear power calculation data model aims to minimize the time consumption, solves the optimal matching resource specifications and computing scale, and sets the job as a forwarding job according to the hybrid cloud scheduling strategy to realize the scheduling of the forwarding job status.
[0084] Example 3
[0085] In one embodiment of the present disclosure, a non-transitory computer-readable storage medium is provided, wherein the non-transitory computer-readable storage medium is used to store computer instructions. When the computer instructions are executed by a processor, the resource allocation and calling method based on the nuclear power computing large model is implemented.
[0086] Example 4
[0087] In one embodiment of the present disclosure, an electronic device is provided, including: a processor, a memory and a computer program; wherein the processor is connected to the memory, the computer program is stored in the memory, and when the electronic device is running, the processor executes the computer program stored in the memory, so that the electronic device executes the resource allocation calling method based on the nuclear power computing large model.
[0088] The present disclosure is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present disclosure. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.
[0089] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.
[0090] Although the above describes the specific implementation methods of the present disclosure in conjunction with the accompanying drawings, it is not intended to limit the scope of protection of the present disclosure. Technical personnel in the relevant field should understand that on the basis of the technical solution of the present disclosure, various modifications or variations that can be made by those skilled in the art without creative work are still within the scope of protection of the present disclosure.
Claims
1. A resource allocation and calling method based on a large nuclear power computing model, characterized in that: include: Obtain basic data on nuclear power system operation; Automatically conduct data comparison and evaluation based on the bench test data of nuclear power equipment and nuclear power plant operation data in the basic data; For the evaluated data, the accumulated historical computing operation data is analyzed from multiple dimensions including computing users, computing types, computing devices, software and versions, operating systems, and the number of cores through neural network algorithms, k-nearest neighbor algorithms, and naive Bayes algorithms to build a nuclear power computing data model. The job information submitted by the user is obtained. The nuclear power computing data model aims to minimize the time consumption, solves the matching optimal resource specifications and computing scale, and sets the job as a forwarding job according to the hybrid cloud scheduling strategy to realize the scheduling of the forwarding job status.
2. The resource allocation and calling method based on the nuclear power computing large model according to claim 1 is characterized in that: The basic data of nuclear power system operation include: resource type specification attribute data, software feature data, historical operation attribute data and performance data, nuclear power equipment bench test data and nuclear power equipment real-time operation data.
3. The resource allocation and calling method based on the nuclear power computing large model according to claim 2 is characterized in that: The software feature data includes software name, software version, whether it supports multi-node parallelism, whether it is memory sensitive, whether it is CPU frequency sensitive, optimal operating system, whether it supports GPU acceleration, and whether local hard disk support is required.
4. The resource allocation and calling method based on the nuclear power computing large model according to claim 2 is characterized in that: The historical job attribute data and performance data include software name, software version, computing type, computing device, computing scale, resource specifications, computing time, whether GPU is used, memory occupancy, hard disk occupancy, and indicator data of simulation input and output parameters.
5. The resource allocation and calling method based on the nuclear power computing large model according to claim 1 is characterized in that: The bench test data of nuclear power equipment and the collected nuclear power plant operation data are automatically compared and evaluated, and the input parameter data and output result data are matched respectively. When the input parameter data and the output result data match, the data is considered valid in the nuclear power calculation data model, otherwise it is invalid and discarded.
6. The resource allocation and calling method based on the nuclear power computing large model according to claim 1 is characterized in that: The user selects the computing software and version on the job submission page, sets the input file, and submits the job. The nuclear power computing data model automatically and intelligently submits the job to the matching resource pool for solution and calculation based on the busy status of each resource pool in the current cluster. The nuclear power computing data model obtains the optimal matching resource specifications and computing scale based on the software name, software version, computing type, and computing device properties of the job submitted by the user.
7. The resource allocation and calling method based on the nuclear power computing large model according to claim 1 is characterized in that: According to the software name, software version, operating system, computing type and computing device input information, all jobs submitted in the cluster within the set historical time are obtained and a job list is returned; the returned job list is grouped according to the software name, software version, operating system, computing type, computing device, computing scale and resource name, and only the record with the shortest time consumption is taken from each group; the software features are read from the computing software feature data according to the software name, software version and operating system of the job submitted by the user; the nuclear power computing data model is filtered according to the obtained software features and the computing scale selected when the job is submitted, and only the resources that meet the requirements are retained; the nuclear power computing data model is filtered according to whether the resource is downgraded when the user submits the job, and only the resources that meet the requirements are retained. The resource name with the shortest time consumption is obtained from the filtered nuclear power computing data model as the optimal resource matched.
8. The resource allocation and calling system based on the nuclear power computing model is characterized by: include: Data acquisition module, used to obtain basic data of nuclear power system operation; A data evaluation module is used to automatically compare and evaluate data based on the bench test data of nuclear power equipment and the nuclear power plant operation data in the basic data; The model building module is used to calculate and analyze the accumulated historical computing operation data from multiple dimensions such as computing users, computing types, computing devices, software and versions, operating systems, and the number of cores based on the evaluated data through neural network algorithms, k-nearest neighbor algorithms, and naive Bayes algorithms to build a nuclear power computing data model. The solution scheduling module is used to obtain the job information submitted by the user. The nuclear power calculation data model aims to minimize the time consumption, solves the optimal matching resource specifications and computing scale, and sets the job as a forwarding job according to the hybrid cloud scheduling strategy to realize the scheduling of the forwarding job status.
9. A non-transitory computer-readable storage medium, characterized in that: The non-transitory computer-readable storage medium is used to store computer instructions. When the computer instructions are executed by the processor, the resource allocation and calling method based on the nuclear power computing model as described in any one of claims 1-7 is implemented.
10. An electronic device, characterized in that: include: A processor, a memory and a computer program; wherein the processor is connected to the memory, the computer program is stored in the memory, and when the electronic device is running, the processor executes the computer program stored in the memory to enable the electronic device to implement the resource allocation and calling method based on the nuclear power computing model as described in any one of claims 1-7.
