A method and apparatus for power grid resource scheduling based on distributed artificial intelligence agents.

Abstract

This application relates to a method and apparatus for power grid resource scheduling based on distributed artificial intelligence agents. Applied to artificial intelligence agents, the method includes: deploying an artificial intelligence agent on each resource scheduling node of the target power grid to acquire real-time power grid operating parameters, external meteorological data, and equipment status data; optimizing the node's power grid resource scheduling model and uploading it to the power grid resource scheduling network. The scheduling network merges the optimized models of each node into a global power grid resource scheduling calculation model, and inputs the latest operating parameters, meteorological data, and equipment status into the global model to generate scheduling information for each node. Subsequently, the global scheduling information is sent to each node through the network, dynamically adjusting the node's transmission parameters and equipment parameters to achieve load balancing and resource optimization, thereby ensuring the efficient and stable operation of the overall power grid. This method can flexibly respond to the uncertainty of power demand and supply, effectively reducing the risk of power grid operation failures.

Description

Technical Field

[0001] This application relates to the field of smart grid technology, and in particular to a method and apparatus for power grid resource scheduling based on distributed artificial intelligence agents. Background Technology

[0002] With the development of new energy and renewable energy generation technologies, traditional power grid resource dispatch relies on centralized control systems. This approach often suffers from slow response times, inability to adjust loads in real time, and vulnerability to single-point failures when dealing with complex dynamic changes in the power grid. With the integration of new and renewable energy sources and the distributed development of power grid structures, the operating environment of the power grid has become more complex and dynamic. Existing resource dispatch systems struggle to flexibly cope with the uncertainties in electricity demand and supply. Summary of the Invention

[0003] Based on this, it is necessary to provide a power grid resource scheduling method, device, computer equipment, computer-readable storage medium, and computer program product based on distributed artificial intelligence agents that can improve the flexibility of resource scheduling systems in responding to power demand and eliminate supply uncertainties, in order to address the aforementioned technical problems.

[0004] Firstly, this application provides a power grid resource scheduling method based on a distributed artificial intelligence agent. The method includes:

[0005] The system acquires the power grid operation parameters, external meteorological parameters, and equipment status parameters corresponding to the AI ​​agents; each AI agent is deployed on the power grid resource scheduling node corresponding to the target power grid.

[0006] Based on the power grid operating parameters, the external meteorological parameters, and the equipment status parameters, the model parameters of the node power grid resource scheduling calculation model are optimized and sent to the power grid resource scheduling network.

[0007] The global power grid resource scheduling calculation model sent by the power grid resource scheduling network is obtained; the global power grid resource scheduling calculation model is obtained by merging various optimized power grid resource scheduling calculation models through the power grid resource scheduling network;

[0008] The power grid operating parameters, the external meteorological parameters, and the equipment status parameters are input into the global power grid resource scheduling calculation model to obtain node power grid resource scheduling information, which is then sent to the power grid resource scheduling network.

[0009] Based on the global power grid resource scheduling information sent by the power grid resource scheduling network, the power grid transmission parameters and power grid equipment parameters of the power grid resource scheduling nodes are adjusted; the global power grid resource scheduling information is obtained by fusing the power grid resource scheduling information of each node through the power grid resource scheduling network.

[0010] Secondly, this application also provides a power grid resource scheduling device based on a distributed artificial intelligence agent. The device includes:

[0011] The parameter acquisition module is used to acquire the power grid operation parameters, external meteorological parameters, and equipment status parameters corresponding to the artificial intelligence agents; each of the artificial intelligence agents is deployed on the power grid resource scheduling node corresponding to the target power grid.

[0012] The model optimization module is used to optimize the model parameters of the node power grid resource scheduling calculation model based on the power grid operation parameters, the external meteorological parameters, and the equipment status parameters, and then send the optimized model parameters to the power grid resource scheduling network.

[0013] The model acquisition module is used to acquire the global power grid resource scheduling calculation model sent by the power grid resource scheduling network; the global power grid resource scheduling calculation model is obtained by merging various optimized power grid resource scheduling calculation models through the power grid resource scheduling network;

[0014] The parameter calculation module is used to input the power grid operating parameters, the external meteorological parameters, and the equipment status parameters into the global power grid resource scheduling calculation model to obtain node power grid resource scheduling information and send it to the power grid resource scheduling network.

[0015] The resource scheduling module is used to adjust the power grid transmission parameters and power grid equipment parameters of the power grid resource scheduling nodes according to the global power grid resource scheduling information sent by the power grid resource scheduling network; the global power grid resource scheduling information is obtained by fusing the power grid resource scheduling information of each node through the power grid resource scheduling network.

[0016] Thirdly, this application also provides a computer device. The computer device includes a memory and a processor, the memory storing a computer program, and the processor executing the computer program to perform the following steps:

[0017] The system acquires the power grid operation parameters, external meteorological parameters, and equipment status parameters corresponding to the AI ​​agents; each AI agent is deployed on the power grid resource scheduling node corresponding to the target power grid.

[0018] Based on the power grid operating parameters, the external meteorological parameters, and the equipment status parameters, the model parameters of the node power grid resource scheduling calculation model are optimized and sent to the power grid resource scheduling network.

[0019] The global power grid resource scheduling calculation model sent by the power grid resource scheduling network is obtained; the global power grid resource scheduling calculation model is obtained by merging various optimized power grid resource scheduling calculation models through the power grid resource scheduling network;

[0020] The power grid operating parameters, the external meteorological parameters, and the equipment status parameters are input into the global power grid resource scheduling calculation model to obtain node power grid resource scheduling information, which is then sent to the power grid resource scheduling network.

[0021] Based on the global power grid resource scheduling information sent by the power grid resource scheduling network, the power grid transmission parameters and power grid equipment parameters of the power grid resource scheduling nodes are adjusted; the global power grid resource scheduling information is obtained by fusing the power grid resource scheduling information of each node through the power grid resource scheduling network.

[0022] Fourthly, this application also provides a computer-readable storage medium. The computer-readable storage medium stores a computer program thereon, which, when executed by a processor, performs the following steps:

[0023] The system acquires the power grid operation parameters, external meteorological parameters, and equipment status parameters corresponding to the AI ​​agents; each AI agent is deployed on the power grid resource scheduling node corresponding to the target power grid.

[0024] Based on the power grid operating parameters, the external meteorological parameters, and the equipment status parameters, the model parameters of the node power grid resource scheduling calculation model are optimized and sent to the power grid resource scheduling network.

[0025] The global power grid resource scheduling calculation model sent by the power grid resource scheduling network is obtained; the global power grid resource scheduling calculation model is obtained by merging various optimized power grid resource scheduling calculation models through the power grid resource scheduling network;

[0026] The power grid operating parameters, the external meteorological parameters, and the equipment status parameters are input into the global power grid resource scheduling calculation model to obtain node power grid resource scheduling information, which is then sent to the power grid resource scheduling network.

[0027] Based on the global power grid resource scheduling information sent by the power grid resource scheduling network, the power grid transmission parameters and power grid equipment parameters of the power grid resource scheduling nodes are adjusted; the global power grid resource scheduling information is obtained by fusing the power grid resource scheduling information of each node through the power grid resource scheduling network.

[0028] Fifthly, this application also provides a computer program product. The computer program product includes a computer program that, when executed by a processor, performs the following steps:

[0029] The system acquires the power grid operation parameters, external meteorological parameters, and equipment status parameters corresponding to the AI ​​agents; each AI agent is deployed on the power grid resource scheduling node corresponding to the target power grid.

[0030] Based on the power grid operating parameters, the external meteorological parameters, and the equipment status parameters, the model parameters of the node power grid resource scheduling calculation model are optimized and sent to the power grid resource scheduling network.

[0031] The global power grid resource scheduling calculation model sent by the power grid resource scheduling network is obtained; the global power grid resource scheduling calculation model is obtained by merging various optimized power grid resource scheduling calculation models through the power grid resource scheduling network;

[0032] The power grid operating parameters, the external meteorological parameters, and the equipment status parameters are input into the global power grid resource scheduling calculation model to obtain node power grid resource scheduling information, which is then sent to the power grid resource scheduling network.

[0033] Based on the global power grid resource scheduling information sent by the power grid resource scheduling network, the power grid transmission parameters and power grid equipment parameters of the power grid resource scheduling nodes are adjusted; the global power grid resource scheduling information is obtained by fusing the power grid resource scheduling information of each node through the power grid resource scheduling network.

[0034] The aforementioned method, apparatus, computer equipment, storage medium, and computer program product for power grid resource scheduling based on distributed artificial intelligence agents involve: acquiring power grid operating parameters, external meteorological parameters, and equipment status parameters corresponding to the artificial intelligence agents; deploying each artificial intelligence agent on a power grid resource scheduling node corresponding to the target power grid; optimizing the model parameters of the node power grid resource scheduling calculation model based on the power grid operating parameters, external meteorological parameters, and equipment status parameters, and sending the optimized model to the power grid resource scheduling network; acquiring the global power grid resource scheduling calculation model sent by the power grid resource scheduling network; merging the optimized power grid resource scheduling calculation models through the power grid resource scheduling network to obtain the global power grid resource scheduling calculation model; inputting the power grid operating parameters, external meteorological parameters, and equipment status parameters into the global power grid resource scheduling calculation model to obtain node power grid resource scheduling information, and sending this information to the power grid resource scheduling network; adjusting the power grid transmission parameters and power grid equipment parameters of the power grid resource scheduling nodes based on the global power grid resource scheduling information sent by the power grid resource scheduling network; and fusing the power grid resource scheduling information of each node through the power grid resource scheduling network to obtain the global power grid resource scheduling information.

[0035] By acquiring real-time power grid operating parameters, external meteorological data, and equipment status information through intelligent agents, the parameters of the scheduling calculation model are optimized at the power grid resource scheduling nodes to form a more adaptable local scheduling strategy. Each node integrates the optimized model into a global power grid resource scheduling model through the power grid resource scheduling network, achieving information sharing and overall scheduling coordination among nodes. Ultimately, the intelligent adjustment of power grid operating parameters and equipment status greatly enhances adaptability to fluctuations in meteorological and equipment status, enabling flexible responses to uncertainties in power demand and supply, effectively reducing the risk of power grid failures, and improving the optimal allocation of power grid resources. This further improves the operational efficiency and overall security of the power grid, supporting more stable energy supply and scheduling management. Attached Figure Description

[0036] Figure 1 This is an application environment diagram of a power grid resource scheduling method based on a distributed artificial intelligence agent in one embodiment;

[0037] Figure 2 This is a flowchart illustrating a power grid resource scheduling method based on a distributed artificial intelligence agent in one embodiment.

[0038] Figure 3 This is a flowchart illustrating a model parameter optimization method in one embodiment;

[0039] Figure 4 This is a flowchart illustrating the first model parameter optimization method in one embodiment;

[0040] Figure 5This is a flowchart illustrating a method for adjusting power grid parameters in one embodiment;

[0041] Figure 6 This is a flowchart illustrating the second method for adjusting power grid parameters in one embodiment;

[0042] Figure 7 This is a flowchart illustrating a computational model merging method in one embodiment;

[0043] Figure 8 This is a flowchart illustrating the third model parameter optimization method in one embodiment;

[0044] Figure 9 This is a structural block diagram of a power grid resource scheduling device based on a distributed artificial intelligence agent in one embodiment;

[0045] Figure 10 This is an internal structural diagram of a computer device in one embodiment. Detailed Implementation

[0046] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0047] This application provides a power grid resource scheduling method based on a distributed artificial intelligence agent, which can be applied to, for example... Figure 1In the application environment shown, the AI ​​agent 102 communicates with the power grid resource scheduling network 104 via a network. The data storage system can store the data that the power grid resource scheduling network 104 needs to process. The data storage system can be integrated into the power grid resource scheduling network 104, or it can be placed in the cloud or on other network power grid resource scheduling networks. The system acquires the power grid operation parameters, external meteorological parameters, and equipment status parameters corresponding to the AI ​​agent 102. Each AI agent 102 is deployed on a power grid resource scheduling node corresponding to the target power grid. Based on the power grid operation parameters, external meteorological parameters, and equipment status parameters, the model parameters of the node power grid resource scheduling calculation model are optimized and sent to the power grid resource scheduling network. The system acquires the global power grid resource scheduling calculation model sent by the power grid resource scheduling network. The global power grid resource scheduling calculation model is obtained by merging the optimized power grid resource scheduling calculation models through the power grid resource scheduling network. The power grid operation parameters, external meteorological parameters, and equipment status parameters are input into the global power grid resource scheduling calculation model to obtain the node power grid resource scheduling information, which is then sent to the power grid resource scheduling network. Based on the global power grid resource scheduling information sent by the power grid resource scheduling network, the power grid transmission parameters and power grid equipment parameters of the power grid resource scheduling nodes are adjusted. The global power grid resource scheduling information is obtained by fusing the power grid resource scheduling information of each node through the power grid resource scheduling network. The power grid resource scheduling network 104 can be implemented using an independent power grid resource scheduling network or a power grid resource scheduling network cluster composed of multiple power grid resource scheduling networks.

[0048] In one embodiment, such as Figure 2 As shown, a power grid resource scheduling method based on distributed artificial intelligence agents is provided, which is applied to... Figure 1 Taking the AI ​​agent in the example, the following steps are included:

[0049] Step 202: Obtain the power grid operation parameters, external meteorological parameters, and equipment status parameters corresponding to the artificial intelligence agent.

[0050] Among them, the artificial intelligence agent can be an intelligent computing unit deployed on the power grid node, which has the ability to collect data, analyze models and perform adaptive optimization. It is responsible for acquiring data such as power grid operation, weather and equipment status in real time and generating corresponding local scheduling strategies.

[0051] Among them, power grid operating parameters can be data that reflects the actual operating status of the power grid, including voltage, current, load and frequency.

[0052] Among them, external meteorological parameters can be meteorological data of the area where the power grid is located, such as temperature, wind speed, humidity and precipitation, which can affect the power grid load demand and equipment operating efficiency.

[0053] Among them, the equipment status parameters can be information such as the operating status, temperature, and load level of various equipment in the power grid (such as transformers, circuit breakers, etc.).

[0054] Among them, the power grid resource scheduling node can be a distributed management unit of the power grid system. Each node is responsible for the power resource scheduling of a specific area and collects data and performs local scheduling through intelligent agents.

[0055] Specifically, each AI agent deployed at a power grid resource scheduling node is responsible for acquiring real-time power grid operating parameters, external meteorological parameters, and equipment status information for the node's region. This includes meteorological data such as load demand, current and voltage data, wind speed, and temperature, as well as the operating status of equipment such as transformers and switches. The agent uses edge computing technology to preprocess the collected data to reduce data transmission latency and network load, ensuring the timeliness and accuracy of the information.

[0056] Step 204: Optimize the model parameters of the node power grid resource scheduling calculation model based on the power grid operation parameters, external meteorological parameters, and equipment status parameters, and send the optimization results to the power grid resource scheduling network.

[0057] Among them, the node power grid resource scheduling calculation model can be a preset intelligent model inside the power grid scheduling node, which can generate resource scheduling strategies for this node through real-time data, but has not yet been adapted to the current situation.

[0058] Among them, the power grid resource scheduling network can be a communication and coordination platform connecting various nodes, responsible for data transmission, scheduling information sharing, and server computing functions.

[0059] Specifically, each AI agent inputs collected power grid operating parameters, external meteorological parameters, and equipment status parameters into the node's power grid resource scheduling calculation model, adaptively optimizing the model's core parameters. For example, by updating key parameters such as load regulation coefficients, wind speed and temperature influence factors, and equipment load limits in real time, the model can automatically adjust its output scheduling strategy under different environmental and load conditions. Simultaneously, the agent uses a feedback loop to compare and learn from the node's scheduling results with actual operating conditions, continuously correcting model parameters to improve its prediction and scheduling accuracy. The optimized power grid resource scheduling calculation models are then encapsulated, encrypted, and uploaded to the power grid resource scheduling network to ensure data transmission security and efficiency.

[0060] Step 206: Obtain the global power grid resource scheduling calculation model sent by the power grid resource scheduling network.

[0061] Among them, the global power grid resource scheduling calculation model can be a comprehensive scheduling model generated by aggregating the optimized power grid resource scheduling calculation models of each node, providing a scheduling decision scheme from the perspective of the entire network, and ensuring the balance and consistency of load and resource allocation among different nodes.

[0062] Among them, the optimized power grid resource scheduling calculation model can be a node scheduling model with updated parameters through an artificial intelligence agent. The parameters of this model are adaptively adjusted based on real-time data, reflecting the current optimal scheduling strategy of the nodes.

[0063] Specifically, the power grid resource scheduling network receives and integrates optimized power grid resource scheduling calculation models from various nodes, using aggregation algorithms and collaborative optimization methods to calibrate and integrate these models from a global perspective. For example, the network analyzes the correlations, load complementarity, and environmental impact differences between the models of each node, extracts global optimization parameters, and ultimately forms a global power grid resource scheduling calculation model applicable to the entire power grid. This global model is then sent to various AI agents, enabling them to access the global power grid resource scheduling calculation model sent by the power grid resource scheduling network. Because the global model employs a collaborative optimization method, comprehensively considering the correlation of power grid parameters between nodes, load complementarity relationships, and meteorological linkages, it ensures that the optimized parameters of each node are compatible across the entire network.

[0064] Step 208: Input the power grid operating parameters, external meteorological parameters, and equipment status parameters into the global power grid resource scheduling calculation model to obtain node power grid resource scheduling information, and send it to the power grid resource scheduling network.

[0065] Among them, the node power grid resource scheduling information can be scheduling data generated by the node based on the global power grid resource scheduling calculation model and the current operating parameters, including information such as the current node's resource configuration, load demand and equipment status.

[0066] Specifically, after the global power grid resource scheduling calculation model is established, the AI ​​agents at each node input the real-time collected power grid operating parameters, external meteorological data, and equipment status information into the global power grid resource scheduling model. Based on this input data, the global power grid resource scheduling calculation model performs multi-level analysis of the current node's load demand, meteorological impact, and equipment operating status, and generates node resource scheduling information, including the current optimal load allocation, equipment start-up and shutdown, and operating parameter adjustment suggestions. Simultaneously, the node's power grid resource scheduling information is fed back to the power grid resource scheduling network, providing data support for the next round of network-wide optimization.

[0067] Step 210: Adjust the power grid transmission parameters and power grid equipment parameters of the power grid resource scheduling nodes according to the global power grid resource scheduling information sent by the power grid resource scheduling network.

[0068] Among them, the global power grid resource scheduling information can be a comprehensive scheduling scheme generated by the power grid resource scheduling network after integrating the scheduling information of each node. It provides consistent resource allocation and load adjustment suggestions for the entire network and guides each node to carry out coordinated scheduling.

[0069] Among them, power grid transmission parameters can be physical parameters that affect the power transmission process, such as transmission voltage, frequency, and phase. Adjusting these parameters can optimize the efficiency and stability of power transmission and adapt to changes in demand under different load conditions.

[0070] Among them, the power grid equipment parameters can be the configuration and operating status data of key equipment in the power grid, such as the output power of transformers and the opening and closing status of switches. By adjusting these parameters, the safe operation of the equipment can be ensured and it can adapt to real-time changes in power demand.

[0071] Specifically, after the AI ​​agents at each node generate local resource scheduling information based on a global model, the power grid resource scheduling network collects and comprehensively processes this local scheduling information. Through techniques such as cross-node data aggregation, priority ranking, and weight allocation, the independent scheduling results of each node are integrated. The network first identifies the relationship between load demand, weather impacts, and equipment status at each node. It then uses algorithms such as weighted averaging and distance priority to evaluate the scheduling priority of each node and performs balanced allocation based on the overall power grid's target load and operating efficiency. For example, for areas with high loads, the network will allocate more resources appropriately, while for nodes with high equipment operating pressure, its load pressure will be reduced accordingly, generating global power grid resource scheduling information. The power grid resource scheduling network sends the integrated global power grid resource scheduling information back to each node. The node's AI agent adjusts local power grid transmission parameters and power grid equipment parameters based on the feedback global information. For example, a node may adjust transmission voltage, switch backup power sources, increase or decrease load, or change equipment power output according to scheduling instructions. Through this coordinated adjustment, nodes can respond quickly to sudden load fluctuations or weather changes, ensuring the stability and efficiency of the power grid. In addition, the real-time adjustment data of each node will be further uploaded to the power grid resource scheduling network, providing closed-loop support for continuous updating and optimization of the global model, enabling the power grid to achieve self-adaptation and efficiency in more refined dynamic control.

[0072] In the aforementioned power grid resource scheduling method based on distributed artificial intelligence agents, the following steps are taken: Power grid operating parameters, external meteorological parameters, and equipment status parameters corresponding to the artificial intelligence agents are acquired; each artificial intelligence agent is deployed on a power grid resource scheduling node corresponding to the target power grid; the model parameters of the node power grid resource scheduling calculation model are optimized based on the power grid operating parameters, external meteorological parameters, and equipment status parameters, and then sent to the power grid resource scheduling network; a global power grid resource scheduling calculation model sent by the power grid resource scheduling network is acquired; the global power grid resource scheduling calculation model is obtained by merging the optimized power grid resource scheduling calculation models through the power grid resource scheduling network; the power grid operating parameters, external meteorological parameters, and equipment status parameters are input into the global power grid resource scheduling calculation model to obtain node power grid resource scheduling information, which is then sent to the power grid resource scheduling network; based on the global power grid resource scheduling information sent by the power grid resource scheduling network, the power grid transmission parameters and power grid equipment parameters of the power grid resource scheduling nodes are adjusted; the global power grid resource scheduling information is obtained by fusing the power grid resource scheduling information of each node through the power grid resource scheduling network.

[0073] By acquiring real-time power grid operating parameters, external meteorological data, and equipment status information through intelligent agents, the parameters of the scheduling calculation model are optimized at the power grid resource scheduling nodes to form a more adaptable local scheduling strategy. Each node integrates the optimized model into a global power grid resource scheduling model through the power grid resource scheduling network, achieving information sharing and overall scheduling coordination among nodes. Ultimately, the intelligent adjustment of power grid operating parameters and equipment status greatly enhances adaptability to fluctuations in meteorological and equipment status, enabling flexible responses to uncertainties in power demand and supply, effectively reducing the risk of power grid failures, and improving the optimal allocation of power grid resources. This further improves the operational efficiency and overall security of the power grid, supporting more stable energy supply and scheduling management.

[0074] In one embodiment, such as Figure 3 As shown, based on power grid operating parameters, external meteorological parameters, and equipment status parameters, the model parameters of the node power grid resource scheduling calculation model are optimized, including:

[0075] Step 302: Initialize the node power grid resource scheduling calculation model based on the historical power grid operation parameters, historical external meteorological parameters and historical equipment status parameters corresponding to the artificial intelligence agent, and obtain the initial resource scheduling calculation model.

[0076] Among them, historical power grid operation parameters, historical external meteorological parameters, and historical equipment status parameters are power grid operation parameters, external meteorological parameters, and equipment status parameters obtained by the artificial intelligence agent at different times or time periods in the past.

[0077] The initial resource scheduling calculation model can be a basic model established based on the node's historical power grid operation data, meteorological conditions, and equipment status information. By analyzing this historical data, the initial scheduling parameters and configurations are defined, providing a reference scheduling benchmark for the node under normal conditions.

[0078] Specifically, in the model initialization phase, the AI ​​agent analyzes the long-term trends of grid operation and equipment load based on the node's historical grid operation data, past meteorological conditions, and equipment operation history, using data backtracking and pattern recognition methods. Simultaneously, historical grid operation data, past meteorological conditions, and equipment operation history are input into the node resource scheduling model. Through the model's analysis of this data, the agent can infer the long-term trends of grid operation and equipment load under historical conditions. The difference between the long-term trends of grid operation and equipment load obtained from the data backtracking and pattern recognition methods and the long-term trends of grid operation and equipment load predicted by the model is used to initialize the model parameters of the node grid resource scheduling calculation model, i.e., to adjust the bias of the model parameters to meet typical load response patterns and optimal equipment operating conditions under different environments, thus obtaining the initial resource scheduling calculation model.

[0079] Step 304: Based on the power grid operating parameters, external meteorological parameters, and equipment status parameters, update the calculation objective of the initial resource scheduling calculation model to obtain the updated target resource scheduling calculation model.

[0080] The calculation objective can be a specific optimization goal or result that the power grid resource scheduling model needs to achieve during operation, such as load balancing, maximizing equipment utilization, or minimizing energy consumption.

[0081] The updated target resource scheduling calculation model can be an adjusted model based on the initial model, according to real-time power grid operating parameters, meteorological conditions and equipment status, with the latest scheduling target settings, so as to more accurately reflect the current node's operating needs, thereby generating a scheduling strategy that meets the current power system requirements.

[0082] Specifically, in actual operation, the AI ​​agent updates the calculation objectives of the initial resource scheduling model in real time based on the latest collected real-time power grid operating parameters, external meteorological parameters, and equipment status parameters. For example, when load demand increases or weather conditions change drastically, the agent readjusts the model's objective function, shifting the focus of resource scheduling from conventional load balancing to precise management of fluctuating loads. Through this real-time objective update, the model resets its scheduling objectives based on the current needs of the nodes, such as prioritizing reducing load peaks, optimizing equipment power output, or adjusting equipment operating sequences, generating an updated objective resource scheduling calculation model to adapt to the current operating environment and power demand.

[0083] Step 306: Optimize the model parameters of the updated target resource scheduling calculation model based on power grid operating parameters, external meteorological parameters, equipment status parameters, and the first calculation objective.

[0084] The first calculation objective can be the core optimization objective in the updated target resource scheduling model, which usually represents the highest priority scheduling requirement under the current operating conditions, such as prioritizing load distribution balance under high load conditions, or ensuring safe operation of equipment under extreme weather changes.

[0085] Specifically, after obtaining the updated computational target, the AI ​​agent inputs the latest grid operation data, meteorological data, equipment status, and adjusted computational target into the updated target resource scheduling model. It then applies optimization algorithms such as deep learning or reinforcement learning to iteratively adjust the model's core parameters. For example, based on current load demand and weather changes, the model will re-optimize parameters such as weighting coefficients and equipment priorities, while dynamically adjusting power allocation and equipment switching strategies. Through this refined parameter optimization process, the model can generate an optimal scheduling scheme adapted to the current node conditions, achieving more efficient and reliable resource scheduling and ensuring the stable operation of the grid under high load or extreme weather conditions.

[0086] In this embodiment, an artificial intelligence agent is used to initialize the node resource scheduling model using historical power grid operating parameters, meteorological data, and equipment status. This method enables the scheduling model to possess basic adaptability to node load and environment from the initial stage. As real-time power grid parameters and the external environment change, the model's computational objectives can be dynamically updated, allowing it to quickly adapt to actual operating conditions such as load demand and equipment status changes. Furthermore, by optimizing model parameters guided by the primary computational objective, the scheduling model can continuously self-adjust, achieving a sensitive response to sudden load fluctuations and equipment status changes, ensuring the efficiency, stability, and accurate allocation of resources in power grid operation.

[0087] In one embodiment, such as Figure 4 As shown, based on power grid operating parameters, external meteorological parameters, equipment status parameters, and the first calculation objective, the model parameters of the updated target resource scheduling calculation model are optimized, including:

[0088] Step 402: Input the power grid operating parameters, external meteorological parameters, and equipment status parameters into the updated target resource scheduling calculation model to obtain the test node scheduling information.

[0089] Among them, the test node scheduling information can be a preliminary scheduling scheme generated by the updated target resource scheduling calculation model based on real-time grid parameters and the current calculation target. It includes node resource allocation suggestions, load adjustment plans and equipment operation strategies, and is used to simulate and evaluate the effect of the scheduling scheme in actual application.

[0090] Specifically, the AI ​​agent inputs the latest power grid operating parameters (such as voltage, current, and load data), external meteorological data (such as wind speed and temperature), and equipment status (such as transformer load and switch status) into the updated target resource scheduling calculation model. Based on this input information and the set first calculation objective (such as load balancing or optimal energy efficiency), the model generates test node scheduling information, including preliminary resource allocation suggestions, load adjustment strategies, and equipment start-up and shutdown schemes, which are used to evaluate the ideal scheduling scheme under the current conditions.

[0091] Step 404: Perform virtual scheduling of the power grid resources of the power grid resource scheduling node based on the test node scheduling information to obtain the virtual scheduling result.

[0092] The virtual scheduling result can be generated by applying the test node scheduling information to a virtual environment to simulate the scheduling of node resource allocation and equipment operation. It reflects the expected performance of the current scheduling scheme in the actual environment and is used to compare the effect with the ideal scheduling.

[0093] Specifically, the AI ​​agent uses the generated test node scheduling information to simulate the actual scheduling operations of power grid resource scheduling nodes. By virtually scheduling the operating status and resource configuration of the test nodes, it predicts their potential performance in a real-world environment. The virtual scheduling considers factors such as the load capacity, resource utilization efficiency, and scheduling response speed of each device within the node under this scheme. Through computational simulation, a virtual scheduling result is derived. This result provides a benchmark for comparing whether the current performance of the scheduling model meets the real-time needs of the nodes, providing clear feedback for subsequent optimization.

[0094] Step 406: If the difference between the virtual scheduling result and the preset scheduling result is greater than a threshold, adjust and update the calculation target of the target resource scheduling calculation model according to the virtual scheduling result to obtain the second calculation target.

[0095] Among them, the preset scheduling result can be an ideal scheduling state set based on node operation requirements, load balancing and safety standards. It is used as a target benchmark to measure the scheduling effect of the model and is used to compare with the virtual scheduling result to judge the effectiveness of the current scheduling scheme.

[0096] Specifically, the AI ​​agent analyzes the differences between the virtual scheduling results and the preset ideal scheduling results. If the difference exceeds a preset threshold (e.g., load allocation deviation, excessive equipment overload rate, etc.), the calculation objective of the current model is re-evaluated and adjusted. At this time, the calculation objective of the updated target resource scheduling model is redefined to better approximate the overall grid scheduling needs, resulting in a redefined second calculation objective that replaces the first calculation objective in the updated target resource scheduling calculation model. This second objective places greater emphasis on coping measures under different environmental conditions, such as power supply security under extreme weather conditions or equipment protection under load fluctuations, in order to generate a scheduling strategy that better meets actual needs.

[0097] Step 408: Optimize the model parameters of the updated target resource scheduling calculation model based on power grid operating parameters, external meteorological parameters, equipment status parameters, and the second calculation objective.

[0098] The second calculation objective can be a further optimization objective set in the updated target resource scheduling calculation model. It is usually based on the specific needs or load pressure of the current power grid operation (such as equipment protection under extreme weather conditions) and is used to guide the further optimization of model parameters to ensure that the scheduling strategy achieves higher adaptability and efficiency under the adjusted objective.

[0099] Specifically, the AI ​​agent inputs updated grid operating parameters, meteorological data, equipment status data, and a second calculation objective into the updated target resource scheduling calculation model, and initiates a further model parameter optimization process. This optimization process refines key parameters such as scheduling weights, load allocation ratios, equipment priorities, and resource scheduling thresholds through deep learning or adaptive algorithms. The model dynamically adjusts these parameters according to the new calculation objectives, enabling the scheduling results to better meet the real-time operational needs and safety standards of grid nodes, ultimately generating a highly accurate and responsive scheduling scheme to achieve optimal utilization and reliable operation of grid resources.

[0100] In this embodiment, by inputting grid operating parameters, meteorological data, and equipment status into an updated resource scheduling model, test node scheduling information is generated, and the model's expected performance in a real-world scenario is evaluated through virtual scheduling. When the virtual scheduling result deviates significantly from the preset target, the system automatically adjusts the model's calculation target, generating a second calculation target to replace the initial target, thus more accurately reflecting actual needs. Subsequently, the system performs in-depth optimization of the model parameters, enabling the model to dynamically adapt to real-time load and environmental changes, thereby significantly improving the accuracy and response speed of grid resource scheduling, ensuring the efficient and stable operation of the grid, and reducing the risks caused by load deviations.

[0101] In one embodiment, such as Figure 5As shown, based on the global power grid resource scheduling information sent by the power grid resource scheduling network, the power grid transmission parameters and power grid equipment parameters of the power grid resource scheduling nodes are adjusted, including:

[0102] Step 502: Analyze the global power grid resource scheduling information to determine the scheduling strategy type and adjustment target state of the power grid resource scheduling node.

[0103] The scheduling strategy type can be a category of scheduling schemes set in the power grid resource scheduling model for specific operational needs, such as load balancing, energy consumption optimization, or equipment protection. It determines the main scheduling direction of nodes under different operating conditions to ensure that the power grid maintains its optimal state under changing demands or environments.

[0104] Among them, the target state of adjustment can be the specific operating indicators or conditions that the power grid resource scheduling node needs to achieve when executing the scheduling strategy, such as maintaining voltage stability at a certain level, ensuring that the equipment temperature does not exceed a specific upper limit, or achieving a specific load distribution.

[0105] Specifically, after receiving global power grid resource scheduling information, the AI ​​agents of each power grid resource scheduling node use the information parsing module to read the information content one by one, identify the core strategy direction that needs to be executed, and further convert it into scheduling strategy type (such as load balancing, energy saving or equipment protection) and the specific adjustment target status of the node (such as voltage level, equipment temperature limit or load distribution requirements).

[0106] Step 504: Calculate the equipment parameter adjustment amount and the power transmission parameter adjustment amount according to the scheduling strategy type and the adjustment target state.

[0107] Among them, the equipment parameter adjustment amount can be a specific adjustment value for the operating parameters of various equipment in the power grid (such as transformers, circuit breakers, etc.), used to adjust the output power, operating temperature, load-bearing capacity, etc. of the equipment.

[0108] Among them, the transmission parameter adjustment amount can be the parameter adjustment value of the power grid transmission line to realize the dispatch strategy, such as adjusting the transmission parameters such as voltage, current or phase; it can optimize power transmission efficiency, maintain load balance and adapt to the real-time demand changes of the power grid.

[0109] Specifically, after parsing the scheduling strategy type and adjustment target state, the AI ​​agents at each node input these target parameters into the equipment parameter and transmission parameter calculation modules. The calculation modules, combining the actual operating data of the current equipment and the grid transmission parameters, determine the specific adjustment amount for each piece of equipment and transmission line based on differential calculation and optimization algorithms. For example, if the scheduling target is load balancing, the calculation module will adjust the power output of the transformer or redistribute the load between different areas to ensure voltage and current stability and that equipment temperature does not exceed safe thresholds. Therefore, the calculated equipment parameter and transmission parameter adjustment amounts ensure that changes in each parameter adapt to the current scheduling strategy and avoid over- or under-adjustment.

[0110] Step 506: Adjust the power grid transmission parameters and power grid equipment parameters according to the equipment parameter adjustment amount and the power transmission parameter adjustment amount, respectively.

[0111] Specifically, after determining the adjustment amounts for equipment parameters and transmission parameters, the AI ​​agents at each node execute adjustment operations on the equipment and transmission system. Specifically, nodes optimize current grid transmission parameters (such as voltage and current) and equipment operating states (such as temperature and output power) by controlling transformer output power, adjusting voltage settings, changing switch states, or adjusting equipment load configurations. These adjustment operations, based on a real-time feedback correction mechanism, gradually approach the set adjustment target state during execution, ensuring that the operation of grid nodes meets global scheduling requirements and optimizes resource allocation, thereby achieving safe, stable, and efficient grid operation.

[0112] In this embodiment, by parsing global power grid resource scheduling information, the system can clearly identify the scheduling strategy type and required target state of each power grid node, thereby formulating a precise resource scheduling direction for the node. Based on this strategy, the system calculates specific equipment parameters and transmission parameters adjustment amounts, enabling nodes to finely control changes in transmission and equipment parameters. This precise adjustment not only improves the response speed of power grid nodes to load and environmental changes but also significantly optimizes the allocation and utilization efficiency of power grid resources, ensuring the stability and efficient operation of the power grid under different load conditions, while effectively reducing energy waste and equipment operation risks.

[0113] In one embodiment, such as Figure 6 As shown, based on the equipment parameter adjustment amount and the transmission parameter adjustment amount, the power grid transmission parameters and power grid equipment parameters are adjusted respectively, including:

[0114] Step 602: Based on the equipment parameter adjustment amount and the transmission parameter adjustment amount, adjust the power grid transmission parameters and power grid equipment parameters respectively to obtain the initial adjusted transmission parameters and the initial adjusted equipment parameters.

[0115] The initial adjustment of transmission parameters can be the parameters obtained after preliminary adjustment of the power grid's voltage, current, and other transmission system parameters based on the transmission demand in the dispatch strategy, in order to adapt to the load demand and resource allocation objectives of the current node.

[0116] The initial adjustment of equipment parameters can be the parameters obtained after adjusting the preliminary operating parameters of power grid equipment (such as transformers, switches, etc.), such as power output or switch status, in order to achieve the preliminary equipment configuration required by the dispatch strategy.

[0117] Specifically, the AI ​​agent of the power grid resource dispatch node performs preliminary adjustments to the current power grid transmission parameters and power grid equipment parameters based on calculated equipment parameter adjustments (such as transformer power output regulation and switch status changes) and transmission parameter adjustments (such as voltage and current adjustment magnitudes). This adjustment operation enables the node's operating state to quickly adapt to the current dispatch strategy. By adjusting the output power, load, voltage, and current parameters of each device, the target state in the dispatch strategy is initially achieved. After this step, the node obtains the adjusted transmission parameters and equipment parameters, laying the foundation for further network-wide dispatch balancing.

[0118] Step 604: Generate equipment parameter fine-tuning data and transmission parameter fine-tuning data based on the dynamic load balance data of the target power grid.

[0119] Among them, dynamic load balance data can be calculated by the power grid resource dispatch network through the whole network load balance algorithm after summarizing the initial adjustment results of all nodes, reflecting the load distribution and resource utilization among the nodes.

[0120] Among them, the equipment parameter fine-tuning data can be further adjustment data generated based on the dynamic load balance results, and detailed adjustment suggestions are made for the equipment operating parameters of each node.

[0121] Among them, the transmission parameter fine-tuning data can be fine-tuning suggestions based on the load balance calculation of the entire network, which are used to further optimize the voltage, current and other parameters of the transmission system.

[0122] Specifically, after collecting the initial adjustment parameters of each node, the power grid resource dispatch network, based on the network-wide load balance model, calculates the impact of the initial adjustment of transmission parameters and equipment parameters on global operation. A global dynamic balancing algorithm identifies nodes requiring fine-tuning and their parameters to eliminate resource imbalances caused by the initial adjustments, thus obtaining dynamic load balance data. Furthermore, based on the dynamic load balance data, an AI agent identifies remaining resource imbalance points and load deviations, and generates more detailed adjustment data based on the identification results, namely, equipment parameter fine-tuning data and transmission parameter fine-tuning data.

[0123] Step 606: Based on the equipment parameter fine-tuning data and the power transmission parameter fine-tuning data, fine-tune the initial power transmission parameters and the initial equipment parameters respectively.

[0124] Specifically, after calculating the fine-tuning data for equipment parameters and transmission parameters, grid nodes perform precise fine-tuning of the initially adjusted parameters. Based on the fine-tuning data, nodes gradually adjust equipment output, load distribution, and voltage and current settings to better align with the requirements of overall grid balance. For example, if the load on a node is still too high, the fine-tuning further reduces its power output or redistributes the load. This fine-tuning operation ensures that nodes achieve refined resource allocation, ultimately reaching the overall grid scheduling goals, making grid resource allocation more efficient and balanced, and ensuring stable operation and reliability.

[0125] In this embodiment, by initially adjusting equipment and transmission parameters, grid nodes can quickly respond to current load demands and optimize initial resource allocation. However, to ensure load balance across the entire network, the grid resource scheduling network analyzes the initial adjustment results of all nodes based on dynamic load balance data, generating more refined equipment and transmission fine-tuning data. Subsequently, through further fine-tuning of the initial adjustment parameters, each node achieves more precise resource scheduling and load balancing. This method not only improves the overall stability and resource utilization of the grid but also effectively prevents risks caused by load deviations, ensuring the efficient and reliable operation of the grid in dynamic environments.

[0126] In one embodiment, such as Figure 7 As shown, the global power grid resource scheduling calculation model is obtained by merging various optimized power grid resource scheduling calculation models through a power grid resource scheduling network, including:

[0127] Step 702: Determine the model weights corresponding to each optimized power grid resource scheduling calculation model based on each power grid resource scheduling node.

[0128] Specifically, after collecting the optimized computational models of each power grid resource dispatching node, the power grid resource dispatching network first calculates the model weight of each model in the global dispatching based on factors such as the operating conditions, historical load demand, geographical characteristics, climate impact, and equipment stability of each node. For example, nodes with frequent load fluctuations and high equipment criticality are assigned higher weights to ensure that their demands are prioritized in the overall network resource dispatching. The weight calculation uses multivariate analysis to ensure that the influence of different nodes in the global dispatching model matches their actual operating needs and contribution to the overall stability of the power grid.

[0129] Step 704: Based on the weights of each model, merge the optimized power grid resource scheduling calculation models to obtain the initial power grid resource scheduling calculation model.

[0130] The initial power grid resource scheduling calculation model can be a unified model generated by the power grid resource scheduling network after merging the optimized scheduling models of each node. This model is based on the weighted integration of different nodes and initially has the function of global power grid resource scheduling, which can integrate the load demand and scheduling strategies of different regions.

[0131] Specifically, after determining the model weights for each model, the power grid resource scheduling network performs a weighted merging of the optimization models of each node. Using methods such as arithmetic averaging and weighted summation, the scheduling strategies and demands of each node are unified and integrated to generate an initial power grid resource scheduling calculation model. The merging process retains the core scheduling parameters of high-weight nodes and incorporates the characteristics of other nodes, forming a comprehensive model with a global perspective. This initial power grid resource scheduling calculation model integrates the optimization strategies of each node to adapt to the demand and load characteristics of different regions across the network, providing a foundation for subsequent fault-tolerant optimization.

[0132] Step 706: Optimize the fault tolerance parameters of the initial power grid resource scheduling calculation model based on each power grid resource scheduling node to obtain the global power grid resource scheduling calculation model.

[0133] Among these, fault-tolerance parameters can be key parameters set in the power grid resource scheduling calculation model to cope with sudden load fluctuations and equipment failures, such as load fluctuation response thresholds or fault tolerance. Setting and optimizing fault-tolerance parameters can improve the robustness of the model, enabling it to automatically adjust resource allocation under emergencies and ensure the stable operation of the power grid.

[0134] Specifically, after generating the initial computational model, the power grid resource scheduling network further adjusts the model's fault tolerance parameters to adapt to the unique needs of each node. For example, for nodes whose equipment is susceptible to weather conditions, fault tolerance optimization increases the sensitivity to load fluctuations, while for nodes with a higher risk of equipment failure, it increases fault tolerance. During the adjustment of the model's fault tolerance parameters, a dynamic fault tolerance adjustment algorithm is used to optimize the fault tolerance range and response speed of the initial model based on the actual operating records and risk coefficients of each node, ultimately forming a global power grid resource scheduling computational model with higher robustness and adaptability. This model can efficiently allocate resources under complex or sudden conditions, ensuring the overall stability and operational efficiency of the power grid.

[0135] In this embodiment, by assigning weights to the optimized computational models of each power grid resource scheduling node, the system ensures that the influence of key nodes on global scheduling is fully reflected, enabling the merged initial power grid resource scheduling model to more accurately reflect the actual needs of the entire network. Subsequently, based on the optimization of fault-tolerant parameters in the initial model, the system further enhances the robustness and adaptability of the global scheduling model, allowing it to maintain effective scheduling capabilities even under load fluctuations or equipment failures. This method improves the scheduling accuracy and risk resistance of the power grid in complex environments, ensuring the efficient allocation of network resources and the long-term stability of the power system.

[0136] In one embodiment, such as Figure 8 As shown, the method also includes:

[0137] Step 802: Determine the evolution and adjustment information of the global power grid resource scheduling calculation model based on the global power grid resource scheduling information for different scenarios.

[0138] Among them, evolution adjustment information can be parameter adjustment guidance information generated by the power grid resource scheduling network after analyzing the global scheduling needs under different operating scenarios (such as peak load, extreme weather, equipment maintenance, etc.). It includes the special requirements of various scenarios for the scheduling model and the corresponding parameter adjustment directions, such as load response sensitivity, fault tolerance threshold, equipment priority, etc.

[0139] Specifically, the power grid resource scheduling network conducts in-depth analysis of global power grid resource scheduling information collected under different scenarios to determine the specific scheduling needs and load distribution characteristics of the power grid under each scenario (such as peak hours, sudden load increases, extreme weather conditions, equipment maintenance periods, etc.). Combining the requirements for resource allocation, load response, and security and stability in these scenarios, it identifies scheduling differences between scenarios and generates "evolutionary adjustment information." This information includes the adjustment directions and objectives of key parameters for each scenario condition that the global model needs to adapt to, covering load allocation, fault tolerance thresholds, equipment priorities, and response sensitivity under different situations, to ensure that the model can effectively balance power grid resource allocation under different conditions.

[0140] Step 804: Adjust the model parameters of the global power grid resource scheduling calculation model based on the evolution adjustment information.

[0141] Specifically, based on evolutionary adjustment information, the power grid resource scheduling network refines the model parameters of the global power grid resource scheduling calculation model. For example, in high-load scenarios, the system improves the sensitivity of load response and fault tolerance to adapt to high-demand environments; in equipment maintenance scenarios, the system reallocates loads, prioritizing the protection of critical equipment. Specific adjustments include meticulous optimization of resource allocation weights, fault-tolerance parameters, response speed, and dynamic scheduling thresholds in the model, enabling it to adapt more flexibly and accurately to operational needs in different scenarios. Through such parameter adjustments, the global power grid resource scheduling calculation model can adaptively optimize resource scheduling, thereby improving the robustness and efficiency of the power grid in complex and variable environments.

[0142] In this embodiment, by analyzing global power grid resource scheduling information under different scenarios, the system can identify the specific needs of power grid resource allocation for each operating condition (such as peak load, extreme weather, equipment maintenance, etc.) and generate corresponding evolutionary adjustment information. Subsequently, based on this information, the parameters of the global scheduling model are dynamically adjusted, enabling the model to adaptively optimize resource allocation in various scenarios. This method significantly enhances the adaptability and scheduling flexibility of the power grid in variable environments, ensures more accurate resource allocation, thereby improving the stability and efficiency of the power grid and reducing operational risks under extreme conditions.

[0143] It should be understood that although the steps in the flowcharts of the above embodiments are shown sequentially according to the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless explicitly stated herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least some steps in the flowcharts of the above embodiments may include multiple steps or multiple stages. These steps or stages are not necessarily completed at the same time, but can be executed at different times. The execution order of these steps or stages is not necessarily sequential, but can be performed alternately or in turn with other steps or at least some of the steps or stages of other steps.

[0144] Based on the same inventive concept, this application also provides a power grid resource scheduling device based on a distributed artificial intelligence agent for implementing the aforementioned power grid resource scheduling method based on a distributed artificial intelligence agent. The solution provided by this device is similar to the implementation described in the above method. Therefore, the specific limitations of one or more embodiments of the power grid resource scheduling device based on a distributed artificial intelligence agent provided below can be found in the above-described limitations of the power grid resource scheduling method based on a distributed artificial intelligence agent, and will not be repeated here.

[0145] In one embodiment, such as Figure 9 As shown, a power grid resource scheduling device based on a distributed artificial intelligence agent is provided, including: a parameter acquisition module 902, a model optimization module 904, a model acquisition module 906, a parameter calculation module 908, and a resource scheduling module 910, wherein:

[0146] The parameter acquisition module 902 is used to acquire the power grid operation parameters, external meteorological parameters, and equipment status parameters corresponding to the artificial intelligence agents; each artificial intelligence agent is deployed on the power grid resource scheduling node corresponding to the target power grid.

[0147] The model optimization module 904 is used to optimize the model parameters of the node power grid resource scheduling calculation model based on power grid operating parameters, external meteorological parameters and equipment status parameters, and send the optimized model parameters to the power grid resource scheduling network.

[0148] The model acquisition module 906 is used to acquire the global power grid resource scheduling calculation model sent by the power grid resource scheduling network; the global power grid resource scheduling calculation model is obtained by merging various optimized power grid resource scheduling calculation models through the power grid resource scheduling network;

[0149] The parameter calculation module 908 is used to input power grid operating parameters, external meteorological parameters and equipment status parameters into the global power grid resource scheduling calculation model, obtain node power grid resource scheduling information, and send it to the power grid resource scheduling network.

[0150] The resource scheduling module 910 is used to adjust the power grid transmission parameters and power grid equipment parameters of the power grid resource scheduling nodes according to the global power grid resource scheduling information sent by the power grid resource scheduling network; the global power grid resource scheduling information is obtained by fusing the power grid resource scheduling information of each node through the power grid resource scheduling network.

[0151] In one embodiment, the model optimization module 904 is further configured to initialize the node power grid resource scheduling calculation model based on the historical power grid operating parameters, historical external meteorological parameters, and historical equipment status parameters corresponding to the artificial intelligence agent, to obtain an initial resource scheduling calculation model; update the calculation objective of the initial resource scheduling calculation model based on the power grid operating parameters, external meteorological parameters, and equipment status parameters, to obtain an updated target resource scheduling calculation model; the updated target resource scheduling calculation model includes a first calculation objective; and optimize the model parameters of the updated target resource scheduling calculation model based on the power grid operating parameters, external meteorological parameters, equipment status parameters, and the first calculation objective.

[0152] In one embodiment, the model optimization module 904 is further configured to input power grid operating parameters, external meteorological parameters, and equipment status parameters into the updated target resource scheduling calculation model to obtain test node scheduling information; perform virtual scheduling of power grid resources of power grid resource scheduling nodes based on the test node scheduling information to obtain virtual scheduling results; if the difference between the virtual scheduling results and the preset scheduling results is greater than a threshold, adjust the calculation objective of the updated target resource scheduling calculation model based on the virtual scheduling results to obtain a second calculation objective; the second calculation objective is used to replace the first calculation objective in the updated target resource scheduling calculation model; and optimize the model parameters of the updated target resource scheduling calculation model based on the power grid operating parameters, external meteorological parameters, equipment status parameters, and the second calculation objective.

[0153] In one embodiment, the resource scheduling module 910 is further configured to parse the global power grid resource scheduling information, determine the scheduling strategy type and adjustment target state of the power grid resource scheduling node; calculate the equipment parameter adjustment amount and the transmission parameter adjustment amount according to the scheduling strategy type and the adjustment target state; and adjust the power grid transmission parameters and power grid equipment parameters respectively according to the equipment parameter adjustment amount and the transmission parameter adjustment amount.

[0154] In one embodiment, the resource scheduling module 910 is further configured to adjust the power grid transmission parameters and power grid equipment parameters according to the equipment parameter adjustment amount and the transmission parameter adjustment amount, respectively, to obtain the initial adjusted transmission parameters and the initial adjusted equipment parameters; generate equipment parameter fine-tuning data and transmission parameter fine-tuning data according to the dynamic load balance data of the target power grid; the dynamic load balance data is obtained by the power grid resource scheduling network performing global dynamic balance calculations on the initial adjusted transmission parameters and the initial adjusted equipment parameters of each power grid resource scheduling node; and fine-tune the initial adjusted transmission parameters and the initial adjusted equipment parameters according to the equipment parameter fine-tuning data and the transmission parameter fine-tuning data.

[0155] In one embodiment, the model acquisition module 906 is further configured to determine the model weights corresponding to each optimized power grid resource scheduling calculation model based on each power grid resource scheduling node; merge each optimized power grid resource scheduling calculation model according to each model weight to obtain an initial power grid resource scheduling calculation model; and optimize the fault tolerance parameters of the initial power grid resource scheduling calculation model according to each power grid resource scheduling node to obtain a global power grid resource scheduling calculation model.

[0156] In one embodiment, the resource scheduling module 910 is further configured to determine the evolution adjustment information of the global power grid resource scheduling calculation model based on the global power grid resource scheduling information for different scenarios; and adjust the model parameters of the global power grid resource scheduling calculation model based on the evolution adjustment information.

[0157] The modules in the aforementioned power grid resource scheduling device based on distributed artificial intelligence agents can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in or independent of the processor in a computer device, or stored in the memory of a computer device as software, so that the processor can invoke and execute the corresponding operations of each module.

[0158] In one embodiment, a computer device is provided, which may be an artificial intelligence agent, and its internal structure diagram may be as follows: Figure 10 As shown, the computer device includes a processor, memory, communication interface, display screen, and input devices connected via a system bus. The processor provides computing and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system and computer programs. The internal memory provides an environment for the operation of the operating system and computer programs stored in the non-volatile storage media. The communication interface is used for wired or wireless communication with an external artificial intelligence agent; wireless communication can be achieved through Wi-Fi, mobile cellular networks, NFC (Near Field Communication), or other technologies. When the computer program is executed by the processor, it implements a power grid resource scheduling method based on a distributed artificial intelligence agent. The display screen can be an LCD screen or an e-ink screen. The input devices can be a touch layer covering the display screen, buttons, a trackball, or a touchpad on the computer device's casing, or an external keyboard, touchpad, or mouse.

[0159] Those skilled in the art will understand that Figure 10 The structure shown is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the computer device to which the present application is applied. Specific computer devices may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0160] In one embodiment, a computer device is also provided, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the steps in the above method embodiments.

[0161] In one embodiment, a computer-readable storage medium is provided storing a computer program that, when executed by a processor, implements the steps in the above method embodiments.

[0162] In one embodiment, a computer program product or computer program is provided, the computer program product or computer program including computer instructions stored in a computer-readable storage medium. A processor of a computer device reads the computer instructions from the computer-readable storage medium, and executes the computer instructions, causing the computer device to perform the steps in the above method embodiments.

[0163] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, data stored, data displayed, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties.

[0164] Those skilled in the art will understand that all or part of the processes in the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium. When executed, the computer program can include the processes of the embodiments described above. Any references to memory, databases, or other media used in the embodiments provided in this application can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM can take many forms, such as Static Random Access Memory (SRAM) or Dynamic Random Access Memory (DRAM). The databases involved in the embodiments provided in this application may include at least one type of relational database and non-relational database. Non-relational databases may include, but are not limited to, blockchain-based distributed databases. The processors involved in the embodiments provided in this application may be general-purpose processors, central processing units, graphics processing units, digital signal processors, programmable logic devices, quantum computing-based data processing logic devices, etc., and are not limited to these.

[0165] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0166] The embodiments described above are merely illustrative of several implementation methods of this application, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of this patent application. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application, and these all fall within the protection scope of this application. Therefore, the protection scope of this application should be determined by the appended claims.

Claims

1. A power grid resource scheduling method based on distributed artificial intelligence agents, applied to artificial intelligence agents, characterized in that, The method includes: The system acquires the power grid operation parameters, external meteorological parameters, and equipment status parameters corresponding to the AI ​​agents; each AI agent is deployed on the power grid resource scheduling node corresponding to the target power grid. Based on the power grid operating parameters, the external meteorological parameters, and the equipment status parameters, the model parameters of the node power grid resource scheduling calculation model are optimized and sent to the power grid resource scheduling network. The global power grid resource scheduling calculation model sent by the power grid resource scheduling network is obtained; the global power grid resource scheduling calculation model is obtained by merging various optimized power grid resource scheduling calculation models through the power grid resource scheduling network; The power grid operating parameters, the external meteorological parameters, and the equipment status parameters are input into the global power grid resource scheduling calculation model to obtain node power grid resource scheduling information, which is then sent to the power grid resource scheduling network. Based on the global power grid resource scheduling information sent by the power grid resource scheduling network, the power grid transmission parameters and power grid equipment parameters of the power grid resource scheduling nodes are adjusted; the global power grid resource scheduling information is obtained by fusing the power grid resource scheduling information of each node through the power grid resource scheduling network.

2. The method according to claim 1, characterized in that, The optimization of the model parameters of the node power grid resource scheduling calculation model based on the power grid operating parameters, the external meteorological parameters, and the equipment status parameters includes: Based on the historical power grid operation parameters, historical external meteorological parameters and historical equipment status parameters corresponding to the artificial intelligence agent, the node power grid resource scheduling calculation model is initialized to obtain the initial resource scheduling calculation model. Based on the power grid operating parameters, the external meteorological parameters, and the equipment status parameters, the calculation objective of the initial resource scheduling calculation model is updated to obtain the updated target resource scheduling calculation model; the updated target resource scheduling calculation model includes a first calculation objective; The model parameters of the updated target resource scheduling calculation model are optimized based on the power grid operation parameters, the external meteorological parameters, the equipment status parameters, and the first calculation target.

3. The method according to claim 2, characterized in that, The step of optimizing the model parameters of the updated target resource scheduling calculation model based on the power grid operating parameters, the external meteorological parameters, the equipment status parameters, and the first calculation target includes: The power grid operating parameters, the external meteorological parameters, and the equipment status parameters are input into the updated target resource scheduling calculation model to obtain the test node scheduling information; Based on the test node scheduling information, the power grid resources of the power grid resource scheduling node are virtually scheduled to obtain the virtual scheduling result; If the difference between the virtual scheduling result and the preset scheduling result is greater than a threshold, the calculation objective of the updated target resource scheduling calculation model is adjusted according to the virtual scheduling result to obtain a second calculation objective; the second calculation objective is used to replace the first calculation objective in the updated target resource scheduling calculation model. The model parameters of the updated target resource scheduling calculation model are optimized based on the power grid operation parameters, the external meteorological parameters, the equipment status parameters, and the second calculation objective.

4. The method according to claim 1, characterized in that, The step of adjusting the power grid transmission parameters and power grid equipment parameters of the power grid resource scheduling node based on the global power grid resource scheduling information sent by the power grid resource scheduling network includes: The global power grid resource scheduling information is parsed to determine the scheduling strategy type and adjustment target state of the power grid resource scheduling node; Calculate the equipment parameter adjustment amount and the power transmission parameter adjustment amount according to the scheduling strategy type and the adjustment target state; The power grid transmission parameters and the power grid equipment parameters are adjusted according to the equipment parameter adjustment amount and the power transmission parameter adjustment amount, respectively.

5. The method according to claim 4, characterized in that, The step of adjusting the power grid transmission parameters and power grid equipment parameters according to the equipment parameter adjustment amount and the transmission parameter adjustment amount respectively includes: Based on the equipment parameter adjustment amount and the power transmission parameter adjustment amount, the power grid transmission parameters and the power grid equipment parameters are adjusted respectively to obtain the initial adjusted power transmission parameters and the initial adjusted equipment parameters; Based on the dynamic load balance data of the target power grid, fine-tuning data for equipment parameters and fine-tuning data for transmission parameters are generated; the dynamic load balance data is obtained by the power grid resource scheduling network performing global dynamic balance calculations on the initial adjustments of transmission parameters and equipment parameters of each power grid resource scheduling node. Based on the equipment parameter fine-tuning data and the power transmission parameter fine-tuning data, the initial power transmission parameters and the initial equipment parameters are fine-tuned respectively.

6. The method according to any one of claims 1 to 5, applied to the power grid resource scheduling network, characterized in that, The global power grid resource scheduling calculation model is obtained by merging various optimized power grid resource scheduling calculation models through the power grid resource scheduling network, including: Based on each of the power grid resource scheduling nodes, determine the model weights corresponding to each of the optimized power grid resource scheduling calculation models; Based on the model weights, the optimized power grid resource scheduling calculation models are merged to obtain the initial power grid resource scheduling calculation model. Based on each of the power grid resource scheduling nodes, the fault tolerance parameters of the initial power grid resource scheduling calculation model are optimized to obtain the global power grid resource scheduling calculation model.

7. The method according to any one of claims 1 to 5, applied to the power grid resource scheduling network, characterized in that, The method further includes: Based on the global power grid resource scheduling information in different scenarios, determine the evolution and adjustment information of the global power grid resource scheduling calculation model; Based on the evolution adjustment information, the model parameters of the global power grid resource scheduling calculation model are adjusted.

8. A power grid resource scheduling device based on a distributed artificial intelligence agent, applied to an artificial intelligence agent, characterized in that, The device includes: The parameter acquisition module is used to acquire the power grid operation parameters, external meteorological parameters, and equipment status parameters corresponding to the artificial intelligence agents; each of the artificial intelligence agents is deployed on the power grid resource scheduling node corresponding to the target power grid. The model optimization module is used to optimize the model parameters of the node power grid resource scheduling calculation model based on the power grid operation parameters, the external meteorological parameters, and the equipment status parameters, and then send the optimized model parameters to the power grid resource scheduling network. The model acquisition module is used to acquire the global power grid resource scheduling calculation model sent by the power grid resource scheduling network; the global power grid resource scheduling calculation model is obtained by merging various optimized power grid resource scheduling calculation models through the power grid resource scheduling network; The parameter calculation module is used to input the power grid operating parameters, the external meteorological parameters, and the equipment status parameters into the global power grid resource scheduling calculation model to obtain node power grid resource scheduling information and send it to the power grid resource scheduling network. The resource scheduling module is used to adjust the power grid transmission parameters and power grid equipment parameters of the power grid resource scheduling nodes according to the global power grid resource scheduling information sent by the power grid resource scheduling network; the global power grid resource scheduling information is obtained by fusing the power grid resource scheduling information of each node through the power grid resource scheduling network.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that, When the processor executes the computer program, it implements the steps of the method according to any one of claims 1 to 7.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1 to 7.

Images