Spatiotemporal-scale coordinated control method for multi-energy power system

Abstract

Provided in the present application is a spatiotemporal-scale coordinated control method for a multi-energy power system. The method comprises: identifying a first energy node topology network of a multi-energy power system, wherein the first energy node topology network is acquired on the basis of spatial positions of energy nodes in a target area; performing real-time monitoring on the first energy node topology network to acquire a real-time energy dataset; performing training on the basis of a historical energy dataset to construct a multi-energy joint distribution model; acquiring a preset temporal scale, inputting the real-time energy dataset and the preset temporal scale into the multi-energy joint distribution model, and outputting complementary probabilities among a plurality of energy sources on the basis of the preset temporal scale to obtain a complementary probability set; on the basis of the size of the complementary probability set, performing complementary path identification on the first energy node topology network, and outputting a second energy node topology network; and on the basis of the second energy node topology network, performing coordinated power control.

Description

Spatiotemporal Coordination Control Methods for Multi-Energy Power Systems

[0001] This application claims priority to Chinese Patent Application No. 202411939574.5, filed with the Chinese Patent Office on December 26, 2024, the entire contents of which are incorporated herein by reference. Technical Field

[0002] This application relates to the field of power dispatching technology, and for example to a spatiotemporal scale coordinated control method for multi-energy power systems. Background Technology

[0003] Multi-energy power dispatch refers to the scientific dispatching methods used in power systems with multiple energy sources (including traditional energy sources such as thermal power, nuclear power, and natural gas, and renewable energy sources such as wind power, solar power, and hydropower) to optimize the proportion of various energy sources in order to achieve a balance between power supply and demand, stable system operation, and maximum energy efficiency. Multi-energy power dispatch not only needs to consider the real-time demand for power load, but also needs to comprehensively consider the characteristics of multiple energy sources and flexibly respond to changing power load and energy supply conditions.

[0004] Traditional multi-energy power dispatching methods have certain limitations in practical applications, especially when dealing with complex multi-energy systems. They cannot fully leverage the complementary advantages of different energy sources, making it difficult to achieve efficient and stable coordinated dispatching. Summary of the Invention

[0005] This application provides a spatiotemporal scale coordinated control method for multi-energy power systems to solve the technical problem in traditional multi-energy power dispatching methods that cannot achieve efficient energy coordinated dispatching due to the inability to accurately quantify the complementary characteristics of multiple energy sources in the spatiotemporal scale.

[0006] This application provides a spatiotemporal scale coordinated control method for a multi-energy power system, comprising: identifying a first energy node topology network of the multi-energy power system, wherein the first energy node topology network is obtained based on the spatial location of energy nodes within a target area; performing real-time monitoring of the first energy node topology network to obtain a real-time energy dataset; training based on a historical energy dataset to construct a multi-energy joint distribution model; obtaining a preset time-series scale, inputting the real-time energy dataset and the preset time-series scale into the joint energy distribution model, and outputting the complementarity probabilities between multiple energy sources based on the preset time-series scale to obtain a set of complementarity probabilities; identifying complementary paths in the first energy node topology network according to the magnitude of the set of complementarity probabilities, and outputting a second energy node topology network with complementary path identifiers, wherein the complementary path is a path composed of energy type pairs with complementary probabilities greater than a preset complementary probability; and performing power coordinated control based on the second energy node topology network.

[0007] In one embodiment, a multi-energy joint distribution model is constructed by training on a historical energy dataset, including: acquiring multiple time-series scale samples, which include short-term, medium-term, and long-term time-series scales; performing feature convolution on the historical energy dataset based on the multiple time-series scale samples to output an energy feature dataset corresponding to each time-series scale sample; constructing a multi-energy joint distribution sub-model corresponding to each time-series scale sample using the energy feature dataset corresponding to each time-series scale sample; and integrating the multi-energy joint distribution sub-models corresponding to each time-series scale sample to construct a multi-energy joint distribution model.

[0008] In one embodiment, a multi-energy joint distribution sub-model is constructed based on the energy feature dataset corresponding to each time-series sample. This includes: constructing marginal distributions for each energy type based on the energy feature dataset corresponding to each time-series sample, and outputting multiple marginal distributions; identifying the correlation between each energy type and obtaining correlation identification results; and connecting the multiple marginal distributions according to the correlation identification results to output a multi-energy joint distribution sub-model. The model parameters of the multi-energy joint distribution sub-model are obtained through gradient descent optimization.

[0009] In one embodiment, the formula for calculating the complementarity probability is: ;in, For energy based on the preset time scale and energy The probability of complementarity between them To preset the timing scale, For energy and energy The joint edge distribution, where a and b are energy sources respectively. The upper and lower limits of the electricity index, c and d are energy... The upper limit and lower limit of electricity indicators.

[0010] In one embodiment, after outputting the second energy node topology network with complementary path identifiers, the method further includes: identifying the number of path identifiers for multiple nodes in the second energy node topology network, and obtaining power output data and power input data for multiple nodes; performing power output and power input load analysis on multiple nodes based on the power output and power input data to obtain power output and power input load indicators; if the power output and power input load indicators are greater than preset power output and power input load indicators, updating the complementary path identifiers of the second energy node topology network based on the preset power output and power input load indicators, and outputting the third energy node topology network.

[0011] In one embodiment, the method further includes: performing anomaly detection on the real-time energy dataset and returning anomaly detection results; if the anomaly detection results are empty, performing power coordination control based on the second energy node topology network; if the anomaly detection results are not empty, updating the second energy node topology network in real time based on the anomaly detection results.

[0012] In one embodiment, obtaining a preset time-series scale includes: obtaining a power load index of the target area based on the multi-energy power system; determining the power status of the target area based on the power load index; if the power status of the target area is in a peak power state, adjusting the preset time-series scale to a short-term scale; if the power status of the target area is in a mid-peak power state, adjusting the preset time-series scale to a medium-term scale; and if the power status of the target area is in a low-peak power state, adjusting the preset time-series scale to a long-term scale.

[0013] In one embodiment, the spatiotemporal scale coordinated control method for a multi-energy power system further includes: the second energy node topology network includes the energy type of each node, the transmission path between multiple nodes, and control parameters, wherein the transmission path includes the transmission direction.

[0014] This application also provides a spatiotemporal scale coordinated control device for a multi-energy power system, comprising: a first module configured to identify a first energy node topology network of the multi-energy power system, wherein the first energy node topology network is obtained based on the spatial location of energy nodes within a target area; a second module configured to monitor the first energy node topology network in real time and obtain a real-time energy dataset; a third module configured to train a multi-energy joint distribution model based on a historical energy dataset; a fourth module configured to obtain a preset time-series scale, input the real-time energy dataset and the preset time-series scale into the joint energy distribution model, and output the complementarity probabilities between multiple energy sources based on the preset time-series scale to obtain a set of complementarity probabilities; a fifth module configured to identify complementary paths in the first energy node topology network according to the size of the set of complementarity probabilities, and output a second energy node topology network with complementary path identification, wherein the complementary path is a path composed of energy type pairs whose complementary probabilities are greater than a preset complementary probability; and a sixth module configured to perform power coordination control based on the second energy node topology network.

[0015] This application also provides an electronic device, including: a processor; and a memory configured to store executable instructions of the processor; wherein the processor is configured to execute the instructions to implement the above-described spatiotemporal scale coordinated control method for multi-energy power systems.

[0016] This application also provides a computer-readable storage medium, which, when the instructions in the computer-readable storage medium are executed by the processor of an electronic device, enables the electronic device to perform the above-described spatiotemporal scale coordinated control method for multi-energy power systems. Attached Figure Description

[0017] Figure 1 is a flowchart illustrating a spatiotemporal scale coordinated control method for a multi-energy power system provided in an embodiment of this application;

[0018] Figure 2 is a flowchart illustrating the construction of a multi-energy joint distribution model in a spatiotemporal scale coordinated control method for a multi-energy power system provided in an embodiment of this application.

[0019] Figure 3 is a schematic diagram of the structure of a spatiotemporal scale coordinated control device for a multi-energy power system provided in an embodiment of this application;

[0020] Figure 4 is a schematic diagram of the hardware structure of an electronic device provided in an embodiment of this application. Detailed Implementation

[0021] This application provides a spatiotemporal coordinated control method for multi-energy power systems, solving the technical problem in traditional multi-energy power dispatching methods where the inability to accurately quantify the complementary characteristics of multiple energy sources at the spatiotemporal scale prevents efficient energy coordinated dispatching. This method effectively quantifies the correlation, complementarity, and fluctuation characteristics of various energy types at the spatiotemporal scale, thereby achieving scientific and flexible multi-energy coordinated dispatching, significantly improving the utilization rate of renewable energy, reducing power system fluctuations, and ensuring the efficient and stable operation of the power system.

[0022] Referring to Figure 1, this application provides a spatiotemporal scale coordinated control method for multi-energy power systems, comprising the following steps:

[0023] S100: Identify the first energy node topology network of a multi-energy power system, which is obtained based on the spatial location of energy nodes within the target area.

[0024] First, identify the various energy types within the target area's multi-energy power system, such as renewable energy (e.g., wind power, solar power), traditional energy (e.g., thermal power, nuclear power), energy storage facilities, and load nodes. Next, based on the geographical locations of the energy types (e.g., wind farms, photovoltaic power plants, thermal power plants, energy storage systems) within the target area, obtain the spatial coordinates of these energy nodes. Node location data can be obtained through a Geographic Information System (GIS). Then, based on the geographical location and power grid structure, establish the connections between energy nodes. Connect multiple nodes using a power grid topology map to construct the first energy node topology network. For example, use power system modeling tools to visualize the energy nodes and their connections, generating a topology map. Each node includes a node identifier (ID) (a unique identifier for the node), node type (e.g., wind power, photovoltaic, thermal power, energy storage), and node location (coordinates of the node). In the topology network, connections between nodes are typically made through power transmission lines or the power grid. Identify these connections to form the "edges" in the network, i.e., the transmission paths of power flow.

[0025] Next, the topology network of the first energy node is identified, and the structure of the topology network is further improved and optimized by deeply analyzing the connection relationship between nodes, energy flow and other characteristics.

[0026] S200: Perform real-time monitoring of the first energy node topology network to obtain a real-time energy dataset.

[0027] To achieve efficient multi-energy power dispatch and coordinated control, real-time monitoring of the first energy node topology network is necessary to obtain timely real-time data from the energy nodes. Firstly, sensors deployed at energy nodes and key locations in the power grid are used to monitor the first energy node topology network in real time, acquiring a real-time energy dataset. This dataset includes energy node power generation / load data, environmental data, and energy storage device data. Energy node power generation / load data refers to the real-time power generation (e.g., wind power, photovoltaic, thermal power) or power consumption (e.g., load nodes) of each energy node. Environmental data refers to environmental parameters for renewable energy nodes, such as wind speed, solar radiation intensity, temperature, humidity, and other meteorological data. Energy storage device data includes the charging / discharging status, battery voltage, remaining capacity, and charging / discharging efficiency of the energy storage devices. Obtaining this real-time energy dataset provides accurate data support for subsequent multi-energy coordinated dispatch analysis.

[0028] S300: Train a multi-energy joint distribution model based on historical energy datasets.

[0029] As shown in Figure 2, step S300 of this application further includes: S310: obtaining multiple time-series scale samples, the multiple time-series scale samples including short-term time-series scale, medium-term time-series scale, and long-term time-series scale; S320: performing feature convolution on the energy historical dataset based on the multiple time-series scale samples, and outputting the energy feature dataset corresponding to each time-series scale sample.

[0030] First, the power system operation log is queried to obtain historical energy datasets, including energy type data, meteorological data, and time ranges (short-term, medium-term, and long-term data). Next, multiple time-series scale samples are obtained, including short-term, medium-term, and long-term time-series scales. Short-term time-series scales refer to time ranges from minutes to hours (e.g., 1 minute to 24 hours). During this period, the load demand of the energy system and the fluctuations in renewable energy generation (e.g., wind and solar power) are relatively significant but short-lived, without large-scale systemic changes. Medium-term time-series scales typically refer to time ranges from days to weeks (e.g., 1 day to 1 month). During this period, the load demand of the energy system fluctuates significantly, and the impact of climate and seasonal factors on renewable energy gradually increases. Long-term time-series scales typically refer to time ranges from months to years (e.g., 1 month to 1 year). During this period, the dispatch and regulation capabilities of the energy system are mainly determined by large-scale grid infrastructure, energy allocation, and energy storage capacity. Climate change and long-term load trends are important factors affecting system operation.

[0031] Then, feature convolution is performed on the energy historical dataset based on the multiple time-series scale samples. The feature convolution process is to extract useful feature information from the historical dataset through a convolutional neural network (CNN) or convolution operation to realize the spatiotemporal scale analysis of the energy system. In the scheduling of multi-energy power systems, energy data at different time scales (short-term time-series scale, medium-term time-series scale, and long-term time-series scale) have different characteristics. By performing feature convolution on these data, valuable features can be extracted, thereby providing a basis for the coordinated scheduling of the power system. This involves performing convolution operations on the energy historical dataset based on samples from multiple time-series scales. The goal of these convolution operations is to extract local features (such as periodic changes, trends, and volatility) from the historical data for subsequent analysis and scheduling. Convolution operations include short-term, medium-term, and long-term operations. Short-term convolution operations involve convolving short-term time-series data (such as power output per minute or hour, load changes, etc.) to capture instantaneous fluctuations. Medium-term convolution operations involve convolving medium-term time-series data (such as daily or weekly power generation data, load data) to identify periodic changes and trends. Long-term convolution operations... The operation refers to convolving long-term time-series data (such as monthly and annual load, power generation, etc.) to extract features such as long-term trends and seasonal changes; resulting in an energy feature dataset corresponding to each time-series scale sample, including short-term energy feature dataset (features of energy fluctuations in the short term, such as load fluctuations, instantaneous fluctuations of wind power and photovoltaic power generation, etc.), medium-term energy feature dataset (periodic fluctuation features in the medium term, such as changes in daily load demand, seasonal fluctuations of photovoltaic and wind power, etc.), and long-term energy feature dataset (containing long-term trend features, such as the impact of climate change on renewable energy power generation, annual load growth trends, etc.).

[0032] S330: Construct a multi-energy joint distribution sub-model for each time-scale sample using the energy feature dataset corresponding to each time-scale sample.

[0033] Step S330 of this application further includes: S331: constructing the marginal distribution of each energy type using the energy feature dataset corresponding to each time-series scale sample, and outputting multiple marginal distributions; S332: identifying the correlation between each energy type and obtaining the correlation identification result; S333: connecting the multiple marginal distributions according to the correlation identification result and outputting a multi-energy joint distribution sub-model, wherein the model parameters of the multi-energy joint distribution sub-model are obtained through gradient descent optimization.

[0034] First, using the energy feature dataset corresponding to each time-series sample, we construct the marginal distribution for each energy source. That is, based on short-term, medium-term, and long-term time-series data, we calculate the marginal probability distribution for each energy source. The marginal distribution for each energy source is a probability distribution that describes the fluctuation characteristics and change patterns of that energy source at different time scales. The goal is to identify the characteristics of energy sources through statistical methods, thereby providing a reliable basis for the coordinated scheduling of multi-energy power systems. For example, marginal probabilities can be calculated using histograms, kernel density estimation, autoregressive models, etc. For instance, by dividing the data into multiple intervals using histograms, we can calculate the frequency in each interval and then estimate its probability density; and output multiple marginal distributions.

[0035] Next, the correlation between each type of energy source is identified. This correlation can be measured by calculating the correlation coefficient between different energy sources. For example, the Pearson correlation coefficient measures the strength of the linear correlation between two variables. In addition to quantifying the correlation through the correlation coefficient, cluster analysis or graph theory methods can also be used to identify the correlation patterns between energy sources. For example, cluster analysis divides energy sources into different groups based on their correlation to obtain the correlation identification results.

[0036] Based on the correlation identification results, the multiple marginal distributions are connected to construct a multi-energy joint distribution sub-model. The joint distribution describes the probability of multiple random variables (energy types) occurring together. In a multi-energy power system, the joint distribution represents the common fluctuation characteristics and mutual influences of different energy types at different time scales. Connecting the marginal distributions of each energy type, the joint distribution represents the synergistic effect between different energy sources. The marginal distributions of each energy type are combined with the correlations between energy sources using probabilistic models (such as Gaussian distributions, Gaussian mixture models, etc.) to construct a multidimensional joint probability distribution. Furthermore, optimization methods such as gradient descent are used to adjust the model parameters, making the model more... To effectively fit historical data and accurately reflect the correlation and temporal variation patterns among energy sources, gradient descent is a common optimization method used to update model parameters through backpropagation to minimize the error or loss function. In multi-energy joint distribution models, gradient descent can be used to minimize the model's prediction error, enabling the model to more accurately reflect the spatiotemporal correlation between energy sources. For example, the parameters of the multi-energy joint distribution model are initialized based on preliminary estimates of the marginal distributions. Then, the parameters are continuously optimized using the gradient descent algorithm on a training dataset (such as historical load data and power generation data) until convergence. Through continuous iteration, an optimized multi-energy joint distribution sub-model is finally obtained. The final output multi-energy joint distribution sub-model is an efficient probabilistic model that reflects the spatiotemporal correlation between different energy types and their joint distribution characteristics at different time scales. It not only quantifies the relationships between different energy sources but also enables efficient energy dispatch optimization based on the needs of different time scales, supporting the stability, economy, and maximization of renewable energy utilization in multi-energy power systems.

[0037] S340: Integrate the multi-energy joint distribution sub-models corresponding to each time-scale sample to construct a multi-energy joint distribution model.

[0038] Through the above training steps, multiple joint distribution sub-models corresponding to samples at various time scales are obtained. Each time-scale sub-model models energy data at different time scales, capturing the energy fluctuation characteristics at different scales. Then, the multi-energy joint distribution sub-models corresponding to each time-scale sample are integrated. For example, for each time-scale sub-model, appropriate weights are assigned based on its contribution to power system dispatch. Generally, the short-term sub-model is used for immediate dispatch decisions, the medium-term sub-model for daily operations, and the long-term sub-model for seasonal dispatch and resource planning. Finally, a global multi-energy joint distribution model is constructed. By integrating multi-energy joint distribution sub-models at different time scales, the characteristics of energy fluctuations and their interrelationships can be accurately captured, thereby providing a more precise basis for power system dispatch decisions.

[0039] S400: Obtain a preset time scale, input the real-time energy dataset and the preset time scale into the joint energy distribution model, output the complementary probabilities between multiple energy sources based on the preset time scale, and obtain a set of complementary probabilities.

[0040] First, a preset time-series scale is obtained, which may include a short-term, medium-term, or long-term scale, set according to the regional power status. Next, the real-time energy dataset and the preset time-series scale are input into the joint energy distribution model for analysis, outputting the complementarity probability between multiple energy sources based on the preset time-series scale. The complementarity probability refers to the probability that the power generation or supply capacity of two or more energy sources complement each other at a given time-series scale. For example, when wind power is insufficient, solar energy may provide supplementation; when wind power is sufficient, thermal power generation may need to be adjusted to maintain system stability. In the joint distribution model, the complementarity probability is calculated by considering the marginal distribution of each energy source and their correlations. For instance, the joint occurrence probability between multiple energy sources is calculated using a joint distribution function, thereby obtaining their complementary relationships and a complementarity probability set. The complementarity probability set is a data structure containing information on the complementarity between multiple energy sources, representing the complementarity probability between each pair of energy sources. For example, at a short-term time-series scale, the complementarity probability of wind power and photovoltaic power is 0.8, indicating that these two energy sources have an 80% probability of complementing each other in the short term. For medium- and long-term scales, there may be different probability values ​​because the fluctuation characteristics and complementarity of energy may change at different time scales.

[0041] To obtain the preset timing scale, step S400 of this application further includes:

[0042] S410: Obtain the power load index of the target area based on the multi-energy power system; S420: Determine the power status of the target area based on the power load index. If the power status of the target area is in a peak power state, adjust the preset time series scale to a short-term scale; S430: If the power status of the target area is in a mid-peak power state, adjust the preset time series scale to a medium-term scale; S440: If the power status of the target area is in a low-peak power state, adjust the preset time series scale to a long-term scale.

[0043] First, based on the multi-energy power system, the power load indicators of the target area are obtained. These indicators include load demand, power supply capacity, and power fluctuations. Load demand refers to the power consumption demand of the target area at a given moment or time period (e.g., total electricity consumption per unit time). Power supply capacity refers to the power supply capacity of the target area, including the power generation from various energy sources (e.g., wind, solar, thermal power). Power fluctuations refer to the range of fluctuations in the power system over a time period, reflecting the system's stability and load balance. These power load indicators can be obtained through real-time monitoring equipment, smart meters, or predictive models to help the power system assess its current power status.

[0044] Next, the power status of the target area is determined based on the power load index. The power status includes peak power status, mid-peak power status, and off-peak power status. Peak power status refers to the period when the load demand is close to or exceeds the power system's supply capacity. At this time, it may be necessary to enhance the system's response capability and balance the load by dispatching efficient and stable energy sources. Mid-peak power status refers to the period when the power load is close to the system's medium supply capacity. At this time, the power system's load is relatively balanced, and medium-term dispatch strategies can be used to deal with power fluctuations. Off-peak power status refers to the period when the power demand is lower than the system's supply capacity and the system is in a relatively stable state. At this time, the system load can be easily handled, and resource allocation can usually be optimized through long-term dispatch strategies.

[0045] Then, the preset time-series scale is adjusted according to the power system's power status to make energy dispatch more flexible and efficient. If the target area's power status is at its peak, the preset time-series scale is adjusted to a short-term scale. The short-term scale can quickly react to energy fluctuations and adjust energy output in real time to ensure that the system meets demand in a short period of time. If the target area's power status is at its mid-peak, the preset time-series scale is adjusted to a medium-term scale. The medium-term scale can optimize energy utilization and dispatch efficiency within a certain time range and balance the power generation of different energy sources. If the target area's power status is at its low-peak, the preset time-series scale is adjusted to a long-term scale. The long-term scale helps to plan and optimize the allocation of energy resources, reduce unnecessary energy waste, and make long-term load adjustments and plans in advance. By dynamically adjusting the preset time-series scale according to power load indicators, it is possible to flexibly dispatch multiple types of energy under different power demand and supply conditions, thereby improving overall energy utilization and reducing operational risks.

[0046] The formula for calculating the complementarity probability is: ;in, For energy based on the preset time scale and energy The probability of complementarity between them To preset the timing scale, For energy and energy The joint edge distribution, where a and b are energy sources respectively. The upper and lower limits of the electricity index, c and d are energy... The upper limit and lower limit of electricity indicators.

[0047] The formula for calculating the complementarity probability is used to quantify the complementarity of different energy types at a specific time scale. If the complementarity probability is close to 1, it means that the two energy types have strong complementarity at that time scale, that is, they can complement each other and provide a stable power supply. If the complementarity probability is close to 0, it means that the complementarity between the two energy types is poor, their power generation capacity cannot be effectively coordinated, and more dispatching means or backup energy may need to be introduced to ensure a stable power supply.

[0048] S500: Based on the size of the complementary probability set, the first energy node topology network is marked with complementary paths, and a second energy node topology network with complementary path markings is output. The complementary path is a path composed of energy type pairs with complementary probabilities greater than a preset complementary probability.

[0049] First, complementary paths are defined. A complementary path refers to a path in an energy node topology network where the complementarity probability between two energy nodes is greater than a preset complementarity probability, and the connection paths between these energy nodes can effectively cooperate (e.g., one-to-one or one-to-many). That is, under a specific time scale, they can jointly meet power demand and complement each other. The preset complementarity probability can be set according to the power system operation scenario, for example, setting the preset complementarity probability to 0.8. Next, complementary paths are identified in the first energy node topology network according to the size of the complementary probability set. That is, the paths formed by energy types with complementary probabilities greater than the preset complementarity probability that can cooperate are identified as complementary paths, and a second energy node topology network with complementary path identification is output.

[0050] Step S500 of this application includes: S510: The second energy node topology network includes the energy type of each node, the transmission path between multiple nodes, and control parameters, wherein the transmission path includes the transmission direction. The second energy node topology network includes the energy type of each node (such as solar energy, wind energy, thermal power, hydropower, etc.); in the second energy node topology network, the connection between multiple nodes is represented by the transmission path, which describes the power flow relationship between different nodes. The transmission path includes the transmission direction. For example, the transmission path means that the two energy nodes connected by the path can transmit power to each other. For two complementary energy nodes, if one energy type has insufficient output, the other energy type can supplement the power to ensure that the load demand is met; the transmission direction refers to the direction of power flow between energy nodes. The transmission direction is crucial for the scheduling and control system because it determines the direction of power flow and how each node responds to changes in system demand.

[0051] Meanwhile, the second energy node topology network also includes control parameters between multiple nodes. These parameters are used to optimize the operation and dispatch of the power system, including power limiting, transmission efficiency, and energy storage strategies. By identifying complementary paths based on complementarity probability and combining the energy type, transmission path, and control parameters of each node, a flexible and efficient power dispatch network is constructed, providing technical support for realizing intelligent and coordinated energy dispatch and ensuring the efficient and stable operation of the power system.

[0052] After outputting the second energy node topology network with complementary path identifiers, the following is also included:

[0053] S520: Identify the number of path identifiers for multiple nodes in the second energy node topology network, and obtain power output data and power input data for multiple nodes; S530: Perform power output and power input load analysis on multiple nodes based on the power output and power input data, and obtain power output and power input load indicators; S540: If the power output and power input load indicators are greater than preset power output and power input load indicators, update the complementary path identifiers of the second energy node topology network based on the preset power output and power input load indicators, and output the third energy node topology network.

[0054] After outputting the second energy node topology network with complementary path identifiers, the number of path identifiers for multiple nodes in the second energy node topology network is identified. First, the number of path identifiers for each node in the second energy node topology network is identified. A path identifier refers to the power transmission path between each node and other nodes. The number of path identifiers indicates the number of paths that the node, as an energy source or load node, can transmit power to other nodes. The number of path identifiers reflects the diversity and flexibility of power transmission between nodes. The more path identifiers there are, the more complex the interconnection relationship between nodes and the more power flow options there are. Further, for each node, its power output data and power input data are obtained. Power output data refers to the power output from the node to other nodes; power input data refers to the power input from the node to other nodes. These data reflect the power flow between nodes. Output data represents the node's power supply capacity within a specific time period, while input data represents the node's power demand.

[0055] Next, based on the power output data and power input data, the power output load analysis is performed on multiple nodes. That is, the dynamic balance between power supply and demand of each node is analyzed, and the power output load index is calculated. For example, the power output load index is set as the ratio of power input to power output. This index helps to determine the role of the node in power supply and demand. Based on preset input / output load indicators, the load status of each node is determined. If the input / output load indicator is greater than the preset input / output load indicator, it indicates that the node is overloaded or has insufficient power supply. It may be necessary to adjust power dispatch or strengthen complementarity with other nodes to balance the load. At this time, the complementary path identifiers of the second energy node topology network are updated according to the preset input / output load indicators. That is, for nodes with excessive load, the system may need to add more complementary paths to obtain power from other energy nodes or transfer excess power to other nodes. For example, increase the path for power to flow to nodes with excessive load, or enhance the power transmission capacity between nodes; adjust the transmission direction or power of the power flow path to ensure the supply and demand balance of the system, etc. After the complementary path identifiers are updated, the updated third energy node topology network is obtained. This topology network reflects the connection paths and complementary relationships between multiple nodes under the new load state.

[0056] S600: The multi-energy power system performs power coordination control based on the second energy node topology network.

[0057] The multi-energy power system performs power coordination control based on the second energy node topology network. This involves identifying the input and output loads, complementary paths, and power transmission status of multiple nodes in the second energy node topology network to generate power coordination control strategies. For example, through precise load analysis, it ensures a balance between power demand and supply at multiple nodes within the system, avoiding power shortages or surpluses. Based on real-time energy conditions and the complementarity between nodes, the system dynamically adjusts the energy usage ratio to maximize the utilization of various energy sources. Through effective power coordination control, efficient collaboration between different energy sources within the system is ensured, thereby achieving global load balance, maximized energy utilization, and stable system operation.

[0058] The method in this application embodiment further includes: S610: performing anomaly detection on the real-time energy dataset according to the anomaly detection module, and returning anomaly detection result; if the anomaly detection result is empty, the multi-energy power system performs power coordination control according to the second energy node topology network; S620: if the anomaly detection result is not empty, the multi-energy power system updates the second energy node topology network in real time according to the anomaly detection result.

[0059] In multi-energy power systems, real-time monitoring and anomaly detection are essential to ensure efficient power coordination control and system stability. This function is implemented by an anomaly detection module, which identifies problems that may affect the stable operation of the system. The anomaly detection module monitors and analyzes real-time data to promptly detect abnormal conditions in the power system (such as equipment failure, abnormal load, and power supply-demand imbalance). When an anomaly is detected, the system needs to process it and update the power coordination control strategy in real time based on the detection results.

[0060] The multi-energy power system also includes an anomaly detection module. First, the anomaly detection module performs anomaly detection on the real-time energy dataset, detecting anomalies such as equipment failure, load anomalies, and power supply-demand imbalances, obtaining anomaly detection results. If the anomaly detection result returns empty, it indicates that no anomaly was detected, and the power system can continue to perform power coordination control according to the current strategy. In this case, the multi-energy power system performs power coordination control based on the second energy node topology network. If the anomaly detection result returns non-empty, it indicates that an anomaly or potential risk (such as equipment failure, supply-demand imbalance, or abnormal power flow) has been detected, and the system needs to be adjusted. In this case, the multi-energy power system updates the second energy node topology network in real time based on the anomaly detection result. That is, when an anomaly occurs, the system can respond promptly, restoring normal system operation by adjusting control strategies and optimizing power dispatch, minimizing the impact of anomalies on the power system. For example, if a node fails, the system will reconfigure the power flow direction to bypass the failed node, ensuring that the power supply to other nodes is not affected.

[0061] By introducing an anomaly detection module and a real-time update mechanism, multi-energy power systems can effectively respond to emergencies or system anomalies, ensuring the efficient operation and stability of the power system. In the absence of anomalies, the system continues to perform planned power coordination and control to ensure load balance and maximize energy utilization. When anomalies are detected, the system quickly identifies the anomaly type and updates in real time to avoid power supply interruptions or imbalances, thereby improving the robustness and response speed of the entire power system.

[0062] By identifying the first energy node topology network of a multi-energy power system, which is obtained based on the spatial location of energy nodes within a target area; real-time monitoring of the first energy node topology network to obtain a real-time energy dataset; training based on historical energy datasets to construct a multi-energy joint distribution model; obtaining a preset time-series scale, inputting the real-time energy dataset and the preset time-series scale into the joint energy distribution model, and outputting the complementarity probabilities between multiple energy sources based on the preset time-series scale to obtain a complementarity probability set; identifying complementary paths in the first energy node topology network according to the magnitude of the complementarity probability set, and outputting a second energy node topology network with complementary path identifiers, wherein complementary paths are paths composed of energy type pairs with complementary probabilities greater than preset complementary probabilities; and performing power coordination control based on the second energy node topology network. This effectively quantifies the correlation, complementarity, and fluctuation characteristics of multiple energy types in a spatiotemporal scale, thereby achieving scientific and flexible multi-energy coordinated scheduling, improving the utilization rate of renewable energy, reducing power system fluctuations, and ensuring the efficient and stable operation of the power system.

[0063] This application embodiment also provides a spatiotemporal scale coordinated control device for a multi-energy power system. This device is installed in an electronic device executing the method embodiment of this application. As shown in Figure 3, the device includes: a first module 1, configured to identify a first energy node topology network of the multi-energy power system, the first energy node topology network being obtained based on the spatial location of energy nodes within a target area; a second module 2, configured to monitor the first energy node topology network in real time and obtain a real-time energy dataset; a third module 3, configured to train a multi-energy joint distribution model based on a historical energy dataset; a fourth module 4, configured to obtain a preset time-series scale, input the real-time energy dataset and the preset time-series scale into the joint energy distribution model, and output the complementarity probabilities between multiple energy sources based on the preset time-series scale to obtain a complementarity probability set; a fifth module 5, configured to identify complementary paths in the first energy node topology network according to the size of the complementarity probability set, and output a second energy node topology network with complementary path identifiers, wherein the complementary path is a path composed of energy type pairs whose complementary probabilities are greater than a preset complementary probability; and a sixth module 6, configured to perform power coordination control based on the second energy node topology network.

[0064] In one embodiment, the third module 3 is configured to: acquire multiple time-series scale samples, including short-term, medium-term, and long-term time-series scales; perform feature convolution on the energy history dataset based on the multiple time-series scale samples to output an energy feature dataset corresponding to each time-series scale sample; construct a multi-energy joint distribution sub-model corresponding to each time-series scale sample using the energy feature dataset corresponding to each time-series scale sample; and integrate the multi-energy joint distribution sub-models corresponding to each time-series scale sample to construct the multi-energy joint distribution model.

[0065] In one embodiment, the third module 3 is configured to construct a multi-energy joint distribution sub-model corresponding to each time-scale sample using the energy feature dataset corresponding to each time-scale sample in the following manner: constructing a marginal distribution for each energy type using the energy feature dataset corresponding to each time-scale sample, and outputting multiple marginal distributions; identifying the correlation between each energy type, and obtaining the correlation identification result; connecting the multiple marginal distributions according to the correlation identification result, and outputting the multi-energy joint distribution sub-model, wherein the model parameters of the multi-energy joint distribution sub-model are obtained through gradient descent optimization.

[0066] In one embodiment, the formula for calculating the complementarity probability is:

[0067] ;

[0068] in, For energy based on the preset time scale and energy The probability of complementarity between them The preset timing scale. For the energy and the energy The joint edge distribution, where a and b are the energy sources respectively. The upper and lower limits of the power index, where c and d are the energy... The upper limit and lower limit of electricity indicators.

[0069] In one embodiment, the device further includes a seventh module, configured to, after outputting a second energy node topology network with complementary path identifiers, identify the number of path identifiers for multiple nodes in the second energy node topology network, obtain power output data and power input data for the multiple nodes; perform input / output load analysis on the multiple nodes based on the power output data and the power input data, and obtain input / output load indicators; in response to the input / output load indicators being greater than a preset input / output load indicator, update the complementary path identifiers of the second energy node topology network based on the preset input / output load indicator, and output a third energy node topology network.

[0070] In one embodiment, the device further includes an anomaly detection module, configured to perform anomaly detection on the real-time energy dataset, return anomaly detection results, and, in response to the anomaly detection results being empty, perform power coordination control based on the second energy node topology network; and, in response to the anomaly detection results being non-empty, update the second energy node topology network in real time based on the anomaly detection results.

[0071] In one embodiment, the fourth module 4 is configured to obtain a preset time-series scale in the following manner: obtaining the power load index of the target area based on the multi-energy power system; determining the power status of the target area based on the power load index; adjusting the preset time-series scale to a short-term scale in response to the power status of the target area being in a peak power state; adjusting the preset time-series scale to a medium-term scale in response to the power status of the target area being in a mid-peak power state; and adjusting the preset time-series scale to a long-term scale in response to the power status of the target area being in a low-peak power state.

[0072] In one embodiment, the second energy node topology network includes the energy type of each node, the transmission path between multiple nodes, and control parameters, wherein the transmission path includes the transmission direction.

[0073] The multi-energy power system spatiotemporal scale coordination control device in this embodiment is used to execute the above-described multi-energy power system spatiotemporal scale coordination control method, and can produce the same effect as the above method.

[0074] Figure 4 is a schematic diagram of the hardware structure of an electronic device provided in an embodiment of this application. The electronic device is located in a multi-energy power system. As shown in Figure 4, the electronic device provided in this application includes a processor 310 and a memory 320. The processor 310 in the electronic device can be one or more, and one processor 310 is used as an example in Figure 4. The memory 320 is configured to store one or more programs. The one or more programs are executed by the one or more processors 310, so that the one or more processors 310 implement the spatiotemporal scale coordinated control method for multi-energy power systems as described in the embodiment of this application.

[0075] The electronic device also includes a communication device 330, an input device 340, and an output device 350. The processor 310, memory 320, communication device 330, input device 340, and output device 350 in the electronic device can be connected via a bus or other means; Figure 4 shows an example of connection via a bus. The input device 340 can receive input digital or character information and generate key signal inputs related to user settings and function control of the electronic device. The output device 350 may include a display screen or other display device. The communication device 330 may include a receiver and a transmitter. The communication device 330 is configured to perform information transmission and reception communication according to the control of the processor 310. The memory 320, as a computer-readable storage medium, can be configured to store software programs, computer-executable programs, and modules, such as the program instructions / modules corresponding to the spatiotemporal scale coordinated control method for multi-energy power systems described in the embodiments of this application. The memory 320 may include a program storage area and a data storage area, wherein the program storage area may store the operating system and at least one application program required for a function; the data storage area may store data created based on the use of the electronic device, etc. Furthermore, memory 320 may include high-speed random access memory, and may also include non-volatile memory, such as at least one disk storage device, flash memory device, or other non-volatile solid-state storage device. In some instances, memory 320 may include memory remotely located relative to processor 310, and such remote memory may be connected to the electronic device via a network. Examples of such networks include, but are not limited to, the Internet, intranets, local area networks, mobile communication networks, and combinations thereof.

[0076] This application also provides an electronic device, including: a processor; and a memory configured to store executable instructions of the processor; wherein the processor is configured to execute the instructions to implement the above-described spatiotemporal coordinated control method for multi-energy power systems. The storage medium may be a non-transitory storage medium.

Claims

1. A spatiotemporal coordinated control method for multi-energy power systems, comprising: Identify the first energy node topology network of a multi-energy power system, which is obtained based on the spatial location of energy nodes within a target area; Real-time monitoring of the first energy node topology network is performed to obtain real-time energy datasets. A multi-energy joint distribution model is constructed by training on historical energy datasets. Obtain a preset time scale, input the real-time energy dataset and the preset time scale into the joint energy distribution model, and output the complementarity probability between multiple energy sources based on the preset time scale to obtain a set of complementarity probabilities; The first energy node topology network is identified by complementary paths based on the size of the complementary probability set, and a second energy node topology network with complementary path identification is output. The complementary path is a path composed of energy type pairs whose complementary probability is greater than a preset complementary probability. Power coordination control is performed based on the topology network of the second energy node.

2. The spatiotemporal scale coordinated control method for multi-energy power systems as described in claim 1, wherein, The process of training a multi-energy joint distribution model based on historical energy datasets includes: Multiple time-scale samples are obtained, including short-term time-scale, medium-term time-scale, and long-term time-scale samples; Based on the multiple time-scale samples, the energy history dataset is convolved with features to output the energy feature dataset corresponding to each time-scale sample. Using the energy feature dataset corresponding to each time-scale sample, construct a multi-energy joint distribution sub-model for each time-scale sample; The multi-energy joint distribution sub-model corresponding to each time-scale sample is integrated to construct the multi-energy joint distribution model.

3. The spatiotemporal scale coordinated control method for multi-energy power systems as described in claim 2, wherein, The construction of a multi-energy joint distribution sub-model for each time-scale sample, based on the energy feature dataset corresponding to each time-scale sample, includes: Using the energy feature dataset corresponding to each time-scale sample, construct the marginal distribution for each energy source and output multiple marginal distributions; Identify the correlations between each type of energy source and obtain the correlation identification results; The multiple edge distributions are connected according to the correlation identification results to output the multi-energy joint distribution sub-model, wherein the model parameters of the multi-energy joint distribution sub-model are obtained through gradient descent optimization.

4. The spatiotemporal scale coordinated control method for multi-energy power systems as described in claim 1, wherein, The formula for calculating the complementarity probability is: ； in, For energy based on the preset time scale Japanese Energy The probability of complementarity between them The preset timing scale. For the energy and the energy The joint edge distribution, where a and b are the energy sources respectively. The upper and lower limits of the power index, where c and d are the energy... The upper limit and lower limit of electricity indicators.

5. The spatiotemporal scale coordinated control method for multi-energy power systems as described in claim 1, further comprising, after outputting the second energy node topology network with complementary path identifiers: The number of path identifiers of multiple nodes in the second energy node topology network is identified, and the power output data and power input data of the multiple nodes are obtained. Based on the power output data and the power input data, perform input and output load analysis on the multiple nodes to obtain input and output load indicators. In response to the inflow / outflow load index being greater than a preset inflow / outflow load index, the complementary path identifier of the second energy node topology network is updated based on the preset inflow / outflow load index, and the third energy node topology network is output.

6. The spatiotemporal scale coordinated control method for multi-energy power systems as described in claim 1 further includes: Anomaly detection is performed on the real-time energy dataset, and anomaly detection results are returned. If the anomaly detection results return empty, power coordination control is performed according to the second energy node topology network. In response to the anomaly detection result being non-empty, the topology network of the second energy node is updated in real time based on the anomaly detection result.

7. The spatiotemporal scale coordinated control method for multi-energy power systems as described in claim 1, wherein, The acquisition of the preset time scale includes: Based on the multi-energy power system, obtain the power load index of the target area; The power status of the target area is determined based on the power load index. In response to the power status of the target area being in a peak power state, the preset time scale is adjusted to a short-term scale. In response to the target area's power status being at its mid-peak, the preset time scale is adjusted to a mid-term scale; In response to the target area being in a low-power state, the preset time scale is adjusted to a long-term scale.

8. The spatiotemporal scale coordinated control method for multi-energy power systems as described in claim 1, wherein, The second energy node topology network includes the energy type of each node, the transmission path between multiple nodes, and control parameters, wherein the transmission path includes the transmission direction.

9. A spatiotemporal scale coordinated control device for a multi-energy power system, comprising: The first module is configured to identify the first energy node topology network of a multi-energy power system, which is obtained based on the spatial location of energy nodes within the target area. The second module is configured to monitor the topology network of the first energy node in real time and obtain real-time energy datasets. The third module is designed to train a multi-energy joint distribution model based on historical energy datasets. The fourth module is configured to obtain a preset time scale, input the real-time energy dataset and the preset time scale into the joint energy distribution model, and output the complementarity probability between multiple energy sources based on the preset time scale to obtain a set of complementarity probabilities. The fifth module is configured to identify complementary paths in the first energy node topology network based on the size of the complementary probability set, and output a second energy node topology network with complementary path identification, wherein the complementary path is a path composed of energy type pairs whose complementary probability is greater than a preset complementary probability; The sixth module is configured to perform power coordination control based on the topology network of the second energy node.

10. An electronic device, comprising: processor; The memory is configured to store the processor-executable instructions; The processor is configured to execute the instructions to implement the spatiotemporal scale coordinated control method for multi-energy power systems as described in any one of claims 1 to 8.

11. A computer-readable storage medium, wherein when instructions in the computer-readable storage medium are executed by a processor of an electronic device, the electronic device is enabled to perform the spatiotemporal scale coordinated control method for a multi-energy power system as described in any one of claims 1 to 8.

Images