Dynamic load balancing regulation and control system and method for energy cloud platform based on AI large model

Abstract

The invention discloses an energy cloud platform dynamic load balancing regulation and control system and method based on an AI large model, and the method comprises the steps: obtaining the operation state, energy consumption and task request data of each node in real time through a data collection module, generating a load evaluation value through an AI large model analysis module, and predicting the change trend; different from a traditional method based on historical data or simple rules, the real-time data enables the system to timely sense the dynamic change of the nodes and better adapt to a complex environment; besides, a regulation and control strategy is formulated by integrating multiple factors through a load balancing decision module, energy efficiency, task priority and the like are considered, so that the decision is scientific and reasonable, and optimal allocation of energy resources is realized; in addition, the task scheduling module allocates tasks in time according to a real-time strategy, different from static configuration, dynamic adjustment can flexibly cope with actual conditions, the utilization rate is increased by allocating multiple tasks when the node load is light and energy is sufficient, allocation is reduced to avoid overload when the load is close to the upper limit, and the operation efficiency and stability of the platform are effectively improved.

Description

Technical Field

[0001] This invention relates to the field of load balancing and control technology, and in particular to a dynamic load balancing and control system and method for an energy cloud platform based on an AI large model. Background Technology

[0002] As an important means of integrating multiple energy resources and achieving efficient energy management and distribution, energy cloud platforms are being used more and more widely. In the actual operation of energy cloud platforms, load balancing is a crucial issue. Since the operating status and energy demand of each node in the energy cloud platform (including energy production nodes, energy storage nodes, and consumption nodes) change over time, how to rationally allocate tasks and energy resources based on real-time operating data and task requests to ensure that the load of each node is at a reasonable level and to avoid situations where some nodes are overloaded while others are idle is directly related to the stability, reliability, and operating efficiency of the energy cloud platform.

[0003] Traditional load balancing methods are often based on simple rules or static configurations, which are difficult to adapt to the complex and ever-changing operating environment of energy cloud platforms. For example, some methods may simply distribute loads evenly based on the historical load data of nodes, which cannot accurately predict future load changes of nodes, resulting in load imbalance problems in actual operation. In addition, traditional methods usually lack comprehensive consideration of energy consumption and task characteristics, and cannot achieve optimal allocation of energy resources. Summary of the Invention

[0004] In view of this, the present invention proposes a dynamic load balancing control system and method for energy cloud platforms based on AI large models, which can effectively solve the defects of existing technologies that are difficult to adapt to the complex and ever-changing operating environment of energy cloud platforms and cannot achieve optimal allocation of energy resources.

[0005] The technical solution of this invention is implemented as follows:

[0006] A dynamic load balancing and control system for an energy cloud platform based on an AI large-scale model includes:

[0007] The data acquisition module is used to collect real-time operating status data, energy consumption data, and task request data of each node in the energy cloud platform.

[0008] The AI ​​large model analysis module is used to generate load assessment values ​​for each node based on the operating status data, energy consumption data, and task request data of each node, and predict the load change trend of each node based on the load assessment values.

[0009] The load balancing decision module is used to formulate dynamic load balancing control strategies based on the load assessment values ​​and load change trends of each node.

[0010] The task scheduling module is used to allocate task requests to various nodes in the energy cloud platform based on dynamic load balancing control strategies, so as to achieve dynamic load balancing of the energy cloud platform.

[0011] As a further optional solution to the aforementioned AI-based large-scale model-based dynamic load balancing control system for energy cloud platforms, the AI ​​large-scale model analysis module includes:

[0012] The load assessment value calculation unit is used to generate the load assessment value of each node based on the dynamic load assessment formula, combined with the operating status data, energy consumption data and task request data of each node.

[0013] The load change trend prediction unit is used to predict the load change trend of each node by using a long short-term memory network algorithm, combining historical load assessment data and the current load assessment values ​​of each node.

[0014] As a further optional solution to the aforementioned AI-based large-scale energy cloud platform dynamic load balancing and control system, the dynamic load assessment formula is specifically as follows:

[0015] ;

[0016] in, This represents the load assessment value of node i at time t. This represents the CPU utilization of node i at time t. This indicates the node's maximum CPU utilization. This represents the memory usage of node i at time t. This indicates the maximum memory usage of the node. This represents the network bandwidth utilization of node i at time t. This indicates the node's maximum network bandwidth utilization. This represents the consumption rate of node i for the j-th energy source at time t. This represents the maximum consumption rate of the node for the j-th energy source. This represents the weighting coefficient of the j-th energy source. , , , This is a weighting parameter used to adjust the degree of influence of different factors on the load assessment value.

[0017] As a further optional solution to the aforementioned AI-based large-scale energy cloud platform dynamic load balancing control system, the load balancing decision module includes:

[0018] The receiving unit is used to receive the load assessment values ​​and load change trends of each node;

[0019] The scheduling priority calculation unit is used to calculate the scheduling priority of each task to be assigned to each node based on the load balancing control formula.

[0020] The formulation unit is used to formulate dynamic load balancing control strategies based on the scheduling priority of each task to be assigned to each node.

[0021] As a further optional solution to the aforementioned AI-based large-scale energy cloud platform dynamic load balancing control system, the load balancing control formula is specifically as follows:

[0022] ;

[0023] in, This represents the scheduling priority at time t for assigning task j to node i. This represents the load assessment value of node i at time t. This indicates the priority of task j at time t. This represents the sum of priorities of all tasks to be assigned at time t. This represents the average load assessment value of all nodes in the energy cloud platform at time t. This is an adjustment factor used to adjust the range of scheduling priorities.

[0024] As a further optional solution to the AI-based large-scale energy cloud platform dynamic load balancing control system, the task scheduling module includes:

[0025] The acquisition unit is used to acquire the scheduling priority of each task to be assigned to each node;

[0026] The selection unit is used to select the node with the highest scheduling priority as the target allocation node for the task.

[0027] The allocation and update unit is used to allocate tasks to target allocation nodes and update the node's operating status data, energy consumption data, and task request data.

[0028] A dynamic load balancing control method for energy cloud platforms based on AI large-scale models, specifically including:

[0029] Real-time collection of operational status data, energy consumption data, and task request data from each node in the energy cloud platform;

[0030] Based on the operating status data, energy consumption data, and task request data of each node, a load assessment value for each node is generated, and the load change trend of each node is predicted based on the load assessment value.

[0031] Based on the load assessment values ​​and load change trends of each node, a dynamic load balancing control strategy is formulated.

[0032] Based on the dynamic load balancing control strategy, task requests are reasonably distributed to each node in the energy cloud platform to achieve dynamic load balancing of the energy cloud platform.

[0033] A computing device includes a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the steps of the above-described dynamic load balancing control method for an energy cloud platform based on an AI large model.

[0034] A computer-readable storage medium storing a computer program, which, when executed by a processor, implements the steps of the above-described dynamic load balancing control method for an energy cloud platform based on an AI large model.

[0035] The beneficial effects of this invention are as follows: The data acquisition module collects real-time operating status data, energy consumption data, and task request data from each node in the energy cloud platform. The AI ​​large-scale model analysis module generates load assessment values ​​for each node based on this real-time data and predicts load change trends. Unlike traditional methods based on historical data or simple rules, the use of real-time data enables the system to promptly perceive the dynamic changes of each node in the platform. Whether it's fluctuations in the production capacity of energy production nodes, changes in the charging and discharging status of energy storage nodes, or increases or decreases in demand at consumption nodes, these can be quickly reflected in load assessment and trend prediction, thus better adapting to the complex and ever-changing operating environment of the energy cloud platform. The load balancing decision module formulates dynamic load balancing control strategies based on the load assessment values ​​and load change trends of each node, comprehensively considering... Multiple factors, not just a single load metric, are considered when formulating strategies. For example, energy consumption efficiency, task priority, and urgency are taken into account. This comprehensive evaluation method makes decision-making more scientific and reasonable, enabling the rational allocation of task requests to the most suitable nodes, thereby achieving optimal allocation of energy resources. In addition, the task scheduling module can promptly allocate task requests to each node based on real-time generated dynamic load balancing control strategies. Compared with traditional static configuration methods, dynamic adjustment strategies can flexibly allocate tasks according to actual conditions, ensuring that energy resources are always used most effectively. For example, when a node has a light load and sufficient energy, the system can allocate more tasks to that node to improve energy utilization efficiency; while when the node load is close to the limit, task allocation is reduced to avoid overload. Attached Figure Description

[0036] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0037] Figure 1 This is a schematic diagram of the composition of a dynamic load balancing control system for an energy cloud platform based on an AI large model, according to the present invention.

[0038] Figure 2 This is a flowchart of a dynamic load balancing control method for an energy cloud platform based on an AI large model, according to the present invention.

[0039] Figure 3 This is a schematic diagram of the composition of a computing device according to the present invention. Detailed Implementation

[0040] The technical solutions in the embodiments of the present invention will be clearly and completely described below. Obviously, the described embodiments are only a part of the embodiments of the present invention, and not all of them. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0041] refer to Figures 1 to 3 A dynamic load balancing control system for an energy cloud platform based on an AI big data model includes a data acquisition module, an AI big data model analysis module, a load balancing decision module, and a task scheduling module, wherein:

[0042] The data acquisition module is used to collect real-time operational status data, energy consumption data, and task request data from each node in the energy cloud platform.

[0043] Specifically, the data acquisition module has multi-source data access capability and can collect the operating status data of each node in real time and accurately through communication connections established with each energy production node, energy storage node and consumption node in the energy cloud platform. The operating status data includes the node's equipment temperature, running time and fault status information.

[0044] Meanwhile, for energy consumption data, a combination of high-precision metering sensors and intelligent algorithms is used to accurately measure and record the energy input and output of each node at different time periods in order to obtain detailed energy consumption data.

[0045] For task request data, a dedicated interface is set up to receive task request information from inside and outside the energy cloud platform, and to parse and classify the request content to extract key information such as task type, priority, and estimated execution time.

[0046] Furthermore, the data acquisition module has a built-in data preprocessing unit, which can perform preliminary verification, cleaning, and format conversion on the acquired raw data, remove noisy data and outliers, and convert the data into a unified format for efficient processing by the subsequent AI large model analysis module.

[0047] In addition, the data acquisition module also has data caching and transmission management functions. In the event of network instability, the collected data can be temporarily stored in the local cache. After the network is restored to normal, the data can be reliably transmitted to the AI ​​large model analysis module according to the set transmission strategy.

[0048] The AI ​​large-scale model analysis module is used to generate load assessment values ​​for each node based on the operating status data, energy consumption data, and task request data of each node, and to predict the load change trend of each node based on the load assessment values; in some embodiments, the AI ​​large-scale model analysis module includes:

[0049] The load assessment value calculation unit is used to generate the load assessment value of each node based on the dynamic load assessment formula, combined with the operating status data, energy consumption data and task request data of each node.

[0050] The load change trend prediction unit is used to predict the load change trend of each node by using a long short-term memory network algorithm, combining historical load assessment data and the current load assessment values ​​of each node.

[0051] Specifically, the load assessment calculation unit generates load assessment values ​​based on a dynamic load assessment formula, integrating the operating status data of each node (such as equipment temperature, runtime, fault status, etc.), energy consumption data, and task request data (task type, priority, estimated execution time, etc.). This fusion of multi-source data makes the assessment of node load more comprehensive and accurate. For example, excessively high equipment temperature may affect node performance, and runtime may be related to the wear and tear of the equipment. Combining these factors can more realistically reflect the actual load capacity of the node, avoiding the limitations of assessment based on a single data dimension. The dynamic load assessment formula can calculate based on real-time changing data and update the load assessment values ​​of each node in a timely manner. The node status and task requests in the energy cloud platform are constantly changing, and dynamic assessment can keenly capture these changes, enabling the system to better adapt to complex and ever-changing operating environments.

[0052] The load change trend prediction unit employs the Long Short-Term Memory (LSTM) algorithm, a deep learning algorithm with significant advantages in processing time-series data. It effectively captures long-term dependencies and short-term fluctuation patterns in historical load assessment data. Combined with the current load assessment values ​​of each node, it accurately predicts the load change trend of each node. For example, by analyzing the periodic changes in node load over a period of time, it predicts potential future load peaks or troughs, providing a basis for developing response strategies in advance. Accurate load change trend prediction provides forward-looking information for load balancing decisions. The system can adjust task allocation and resource scheduling strategies in advance based on the prediction results, avoiding node overload or idleness. For instance, if it is predicted that a node is about to enter a high-load state, the system can migrate some tasks to other nodes with lighter loads in advance, ensuring the stable operation of the entire energy cloud platform and improving system reliability and response speed.

[0053] It should be noted that the Long Short-Term Memory (LSTM) network algorithm, which combines historical load assessment data with the current load assessment values ​​of each node, is used to predict the load change trend of each node. Specifically, this includes:

[0054] The load change trend prediction unit first constructs a long short-term memory network model, which includes an input layer, multiple hidden layers, and an output layer. The input layer is used to receive historical load assessment data and the current load assessment values ​​of each node, wherein the historical load assessment data is arranged in chronological order to form a time series input.

[0055] During the model training phase, a large amount of historical load assessment data accumulated by the energy cloud platform is used as training samples. The weight parameters in the long short-term memory network model are continuously adjusted through the backpropagation algorithm to minimize the error between the predicted value and the actual load change value, so that the model can learn the inherent law of load change.

[0056] During the prediction process, the load assessment values ​​of each node obtained at the current moment and the historical load assessment values ​​over a recent continuous period are input into the pre-trained Long Short-Term Memory network model. The model selectively memorizes and forgets the input information through its unique gating mechanism (input gate, forget gate, output gate), effectively capturing the long-term dependencies and short-term fluctuation characteristics in the time series data.

[0057] Finally, the Long Short-Term Memory (LSTM) network model outputs the predicted load change trends of each node at multiple future time points. The prediction results are presented in the form of numerical values ​​or visualized trend curves, providing an accurate basis for the subsequent load balancing decision module to formulate dynamic load balancing control strategies.

[0058] In some embodiments, the dynamic load assessment formula is specifically:

[0059] ;

[0060] in, This represents the load assessment value of node i at time t. This represents the CPU utilization of node i at time t. This indicates the node's maximum CPU utilization. This represents the memory usage of node i at time t. This indicates the maximum memory usage of the node. This represents the network bandwidth utilization of node i at time t. This indicates the node's maximum network bandwidth utilization. This represents the consumption rate of node i for the j-th energy source at time t. This represents the maximum consumption rate of the node for the j-th energy source. Denotes the weighting coefficient of the j-th energy source, and , , , , These are weight parameters, and This is used to adjust the degree of influence of different factors on the load assessment value.

[0061] Specifically, the formula comprehensively considers the CPU utilization of the node at time t. Memory usage Network bandwidth utilization and energy consumption rate Several key indicators are used. These indicators reflect the node's operating status and resource usage from different perspectives. For example, CPU utilization reflects the node's computing power, memory utilization reflects the use of data storage and caching, network bandwidth utilization is related to the node's data transmission capability with the outside world, and energy consumption rate considers the node's energy usage characteristics. By integrating these indicators into a single formula, the node's load can be comprehensively and holistically evaluated, avoiding the one-sidedness of evaluation by a single indicator.

[0062] The formula includes weight parameters. , , , and energy weighting coefficient and satisfy and These weight parameters can be flexibly adjusted based on the actual operation and business needs of the energy cloud platform. For example, if computing tasks account for a large proportion of the platform and the demand for CPU resources is more critical, then the weights can be appropriately increased. The value of this parameter makes CPU utilization take on a more significant role in load assessment; if the supply or consumption of a certain energy source is of particular importance in the platform, the corresponding parameter can be adjusted. This flexibility allows load assessments to better align with the actual needs of different platforms, improving the accuracy of the assessments.

[0063] A load balancing decision module is used to formulate dynamic load balancing control strategies based on the load assessment values ​​and load change trends of each node; in some embodiments, the load balancing decision module includes:

[0064] The receiving unit is used to receive the load assessment values ​​and load change trends of each node;

[0065] The scheduling priority calculation unit is used to calculate the scheduling priority of each task to be assigned to each node based on the load balancing control formula.

[0066] The formulation unit is used to formulate dynamic load balancing control strategies based on the scheduling priority of each task to be assigned to each node.

[0067] In some embodiments, the load balancing control formula is specifically:

[0068] ;

[0069] in, This represents the scheduling priority at time t for assigning task j to node i. This represents the load assessment value of node i at time t. This indicates the priority of task j at time t. This represents the sum of priorities of all tasks to be assigned at time t. This represents the average load assessment value of all nodes in the energy cloud platform at time t. This is an adjustment factor used to adjust the range of scheduling priorities.

[0070] Specifically, the scheduling priority calculation unit calculates the scheduling priority of each task to be assigned to each node based on the load balancing control formula. The formula considers not only the load assessment value of the node at time t, but also... It also incorporates the task's priority at time t. and the sum of priorities of all tasks to be assigned at time t. This multi-factor consideration makes the calculation of scheduling priorities more comprehensive and scientific. For example, for some urgent and important tasks (high priority), the priority calculation can be more precise. Even if the corresponding node has a relatively high load, it can obtain a suitable scheduling priority in the comprehensive calculation, ensuring that critical tasks can be processed in a timely manner; because the formula uses the real-time load assessment value of the node. The calculation of scheduling priorities can dynamically reflect the current load status of nodes. When the load of a node changes, its corresponding scheduling priority will also be adjusted accordingly. This allows task allocation to be optimized in real time according to the actual load of the nodes, avoiding the allocation of tasks to overloaded nodes and improving the flexibility and adaptability of task scheduling.

[0071] The system formulates dynamic load balancing control strategies based on the scheduling priority of each task to be assigned to each node. This ensures that tasks are rationally allocated to the most suitable nodes. For example, tasks with high scheduling priority are allocated first, thereby making full use of node resources and improving task execution efficiency. The formulation of the control strategy comprehensively considers the node load and task priority, and can find the optimal balance between balancing node load and meeting task requirements. On the one hand, it avoids node overload leading to performance degradation or even failure; on the other hand, it ensures that important tasks can be completed in a timely manner, thereby improving the overall service quality and stability of the energy cloud platform.

[0072] A task scheduling module is used to rationally distribute task requests to various nodes in the energy cloud platform based on a dynamic load balancing control strategy, so as to achieve dynamic load balancing of the energy cloud platform; in some embodiments, the task scheduling module includes:

[0073] The acquisition unit is used to acquire the scheduling priority of each task to be assigned to each node;

[0074] The selection unit is used to select the node with the highest scheduling priority as the target allocation node for the task.

[0075] The allocation and update unit is used to allocate tasks to target allocation nodes and update the node's operating status data, energy consumption data, and task request data.

[0076] Specifically, the acquisition unit is responsible for obtaining the scheduling priority of each task to be assigned to each node. This process is based on the scheduling priority information calculated by the preceding modules, making task allocation decisions based on data. By accurately obtaining the scheduling priority, it is possible to clearly understand the suitability of each task for execution on each node, avoiding resource waste and load imbalance caused by blindly allocating tasks. As the operating status of the energy cloud platform is constantly changing, the load of nodes and the characteristics of tasks will also change accordingly. The acquisition unit can obtain the latest scheduling priority information in real time, enabling task allocation to adapt to the dynamic changes of the platform in a timely manner. For example, when the load of a node suddenly increases, the scheduling priority of related tasks will be adjusted accordingly to ensure that tasks can be reasonably allocated to other more suitable nodes.

[0077] The selection unit chooses the node with the highest scheduling priority as the target node for task allocation based on the obtained scheduling priority. This selection method follows the principle of resource optimization and can allocate tasks to the nodes that can complete the tasks most efficiently. For example, for computationally intensive tasks, they will be preferentially allocated to nodes with strong computing power and light current load, thereby improving task execution efficiency and shortening task completion time. By selecting the node with the highest scheduling priority for task allocation, it helps to balance the load of the entire energy cloud platform, avoiding situations where some nodes are overloaded due to task accumulation while other nodes are idle, keeping the load of each node at a relatively reasonable level, and improving the overall stability and reliability of the system.

[0078] The allocation and update unit assigns tasks to target allocation nodes and quickly updates the node's operating status data, energy consumption data, and task request data. This efficient allocation and update mechanism ensures that the platform can respond to task allocation operations in a timely manner, keeping the system status synchronized with the actual operating conditions. For example, after task allocation is completed, the node's operating status data is updated in a timely manner so that subsequent task allocation decisions can be made based on the latest information.

[0079] A dynamic load balancing control method for energy cloud platforms based on AI large-scale models, specifically including:

[0080] Real-time collection of operational status data, energy consumption data, and task request data from each node in the energy cloud platform;

[0081] Based on the operating status data, energy consumption data, and task request data of each node, a load assessment value for each node is generated, and the load change trend of each node is predicted based on the load assessment value.

[0082] Based on the load assessment values ​​and load change trends of each node, a dynamic load balancing control strategy is formulated.

[0083] Based on the dynamic load balancing control strategy, task requests are reasonably distributed to each node in the energy cloud platform to achieve dynamic load balancing of the energy cloud platform.

[0084] A computing device includes a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the steps of the above-described dynamic load balancing control method for an energy cloud platform based on an AI large model.

[0085] A computer-readable storage medium storing a computer program, which, when executed by a processor, implements the steps of the above-described dynamic load balancing control method for an energy cloud platform based on an AI large model.

[0086] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A dynamic load balancing and control system for an energy cloud platform based on an AI large-scale model, characterized in that, include: The data acquisition module is used to collect real-time operating status data, energy consumption data, and task request data of each node in the energy cloud platform. The AI ​​large model analysis module is used to generate load assessment values ​​for each node based on the operating status data, energy consumption data, and task request data of each node, and predict the load change trend of each node based on the load assessment values. The load balancing decision module is used to formulate dynamic load balancing control strategies based on the load assessment values ​​and load change trends of each node. The task scheduling module is used to allocate task requests to various nodes in the energy cloud platform based on dynamic load balancing control strategies, so as to achieve dynamic load balancing of the energy cloud platform.

2. The dynamic load balancing and control system for an energy cloud platform based on an AI large-scale model as described in claim 1, characterized in that, The AI ​​large-scale model analysis module includes: The load assessment value calculation unit is used to generate the load assessment value of each node based on the dynamic load assessment formula, combined with the operating status data, energy consumption data and task request data of each node. The load change trend prediction unit is used to predict the load change trend of each node by using a long short-term memory network algorithm, combining historical load assessment data and the current load assessment values ​​of each node.

3. The dynamic load balancing and control system for an energy cloud platform based on an AI large-scale model as described in claim 2, characterized in that, The dynamic load assessment formula is as follows: ； in, This represents the load assessment value of node i at time t. This represents the CPU utilization of node i at time t. This indicates the node's maximum CPU utilization. This represents the memory usage of node i at time t. This indicates the maximum memory usage of the node. This represents the network bandwidth utilization of node i at time t. This indicates the node's maximum network bandwidth utilization. This represents the consumption rate of node i for the j-th energy type at time t. This represents the maximum consumption rate of the node for the j-th energy source. This represents the weighting coefficient of the j-th energy source. , , , This is a weighting parameter used to adjust the degree of influence of different factors on the load assessment value.

4. The dynamic load balancing and control system for an energy cloud platform based on an AI large model as described in claim 3, characterized in that, The load balancing decision module includes: The receiving unit is used to receive the load assessment values ​​and load change trends of each node; The scheduling priority calculation unit is used to calculate the scheduling priority of each task to be assigned to each node based on the load balancing control formula. The formulation unit is used to formulate dynamic load balancing control strategies based on the scheduling priority of each task to be assigned to each node.

5. The dynamic load balancing and control system for an energy cloud platform based on an AI large model as described in claim 4, characterized in that, The load balancing control formula is as follows: ； in, This represents the scheduling priority at time t for assigning task j to node i. This represents the load assessment value of node i at time t. This indicates the priority of task j at time t. This represents the sum of priorities of all tasks to be assigned at time t. This represents the average load assessment value of all nodes in the energy cloud platform at time t. This is an adjustment factor used to adjust the range of scheduling priorities.

6. The dynamic load balancing and control system for an energy cloud platform based on an AI large model as described in claim 5, characterized in that, The task scheduling module includes: The acquisition unit is used to acquire the scheduling priority of each task to be assigned to each node; The selection unit is used to select the node with the highest scheduling priority as the target allocation node for the task. The allocation and update unit is used to allocate tasks to target allocation nodes and update the node's operating status data, energy consumption data, and task request data.

7. A dynamic load balancing control method for an energy cloud platform based on an AI large-scale model, characterized in that, Specifically, it includes: Real-time collection of operational status data, energy consumption data, and task request data from each node in the energy cloud platform; Based on the operating status data, energy consumption data, and task request data of each node, a load assessment value for each node is generated, and the load change trend of each node is predicted based on the load assessment value. Based on the load assessment values ​​and load change trends of each node, a dynamic load balancing control strategy is formulated. Based on the dynamic load balancing control strategy, task requests are reasonably distributed to each node in the energy cloud platform to achieve dynamic load balancing of the energy cloud platform.

8. A computing device, characterized in that, It includes a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the steps of the dynamic load balancing control method for an energy cloud platform based on an AI large model as described in claim 7.

9. A computer-readable storage medium, characterized in that, The storage medium stores a computer program, which, when executed by a processor, implements the steps of the dynamic load balancing control method for an energy cloud platform based on an AI large model as described in claim 7.

Images