Power equipment load scheduling method and device, equipment, storage medium and product

Through real-time data acquisition and machine learning algorithms to predict power load trends, identify unbalanced areas and optimize power distribution, the problem of slow response in dealing with dynamic changes in power loads is solved, and the grid load balance and safe operation are achieved.

CN119030042APending Publication Date: 2024-11-26CHINA SOUTHERN POWER GRID COMPANY
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202411317289.X
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-09-20
Publication Date
2024-11-26

AI Technical Summary

Technical Problem

Traditional power scheduling management methods are slow to respond and lack of flexibility when dealing with dynamic changes in power loads and dispatching distributed power equipment, resulting in unbalanced loads of the power grid, which may cause equipment overload, system instability, and even power outages.

Method used

By collecting and analyzing the power consumption data of each node of the power network in real time, building the original data set, removing noise and outliers, identifying load changes patterns and patterns, using machine learning algorithms to predict the load change trend in the short future, and identifying potential loads Unbalanced areas and generate optimal load distribution schemes, adjust power distribution, optimize resource configuration, and avoid equipment overload.

Benefits of technology

It has achieved optimization of the distribution of power resources, ensured the safe operation of the power grid, avoided equipment overload and system instability, and improved the response speed and flexibility of the power grid.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119030042A_ABST
    Figure CN119030042A_ABST
Patent Text Reader

Abstract

The invention discloses a power equipment load scheduling method and device, equipment, a storage medium and a product. The method comprises the following steps: acquiring power consumption data of each equipment node in a power network, and determining real-time load data of each equipment node according to the power consumption data; clustering the equipment nodes according to the real-time load data of the equipment nodes to obtain at least one target node area; inputting the real-time load data of each equipment node in each target node area into a pre-trained load trend prediction model to obtain predicted load data corresponding to each equipment node in each target node area output by the model; determining an unbalanced node area according to the real-time load data and the predicted load data of each equipment node in each target node area; and carrying out equipment load scheduling on each equipment node in the unbalanced node area according to the predicted load data of each equipment node in the unbalanced node area.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the technical field of electric power Internet of Things, and in particular to a method, device, equipment, storage medium and product for load scheduling of electric power equipment. Background Art

[0002] In modern power systems, with the continuous growth and complexity of electricity demand, how to efficiently manage power resources and ensure the stable operation of the power grid has become an important technical challenge.

[0003] The traditional power dispatching management method mainly relies on manual experience and static rules. This method has the problems of slow response and insufficient flexibility when dealing with dynamic changes in power load and the dispatching of distributed power equipment. With the rapid development of Internet of Things technology, power Internet of Things terminal devices are gradually applied to various nodes of the power network, realizing real-time monitoring of power equipment status and power consumption. However, it is still difficult to solve the problem of unbalanced power grid load by relying solely on data collection and monitoring, especially during peak load periods. The unbalanced load of power equipment may lead to equipment overload, system instability, and even power outages. Summary of the invention

[0004] The present invention provides a method, device, equipment, storage medium and product for dispatching load of electric power equipment to solve the problem of unbalanced load of power grid.

[0005] According to one aspect of the present invention, a method for load scheduling of electric power equipment is provided, the method comprising:

[0006] Acquire power consumption data of each device node in the power network, and determine real-time load data of each device node based on the power consumption data;

[0007] Clustering each of the device nodes according to the real-time load data of each of the device nodes to obtain at least one target node area; the target node area includes at least one device node;

[0008] Input the real-time load data of each device node in each target node area into the pre-trained load trend prediction model to obtain the predicted load data corresponding to each device node in each target node area output by the model;

[0009] Determine an unbalanced node area according to the real-time load data and the predicted load data of each of the device nodes in each of the target node areas;

[0010] According to the predicted load data of each of the device nodes in the unbalanced node area, device load scheduling is performed on each of the device nodes in the unbalanced node area.

[0011] According to another aspect of the present invention, there is provided a load dispatching device for electric power equipment, the device comprising:

[0012] A real-time load determination module, used to obtain power consumption data of each device node in the power network, and determine the real-time load data of each device node according to the power consumption data;

[0013] A clustering module, used for clustering each of the device nodes according to the real-time load data of each of the device nodes to obtain at least one target node area; the target node area includes at least one device node;

[0014] A load prediction module, used for inputting the real-time load data of each device node in each target node area into a pre-trained load trend prediction model, and obtaining the predicted load data corresponding to each device node in each target node area output by the model;

[0015] An unbalanced area determination module, used to determine an unbalanced node area according to the real-time load data and predicted load data of each of the device nodes in each of the target node areas;

[0016] The load scheduling module is used to schedule the equipment load of each equipment node in the unbalanced node area according to the predicted load data of each equipment node in the unbalanced node area.

[0017] According to another aspect of the present invention, an electronic device is provided, the electronic device comprising:

[0018] at least one processor; and

[0019] a memory communicatively connected to the at least one processor; wherein,

[0020] The memory stores a computer program that can be executed by the at least one processor, and the computer program is executed by the at least one processor so that the at least one processor can execute the power equipment load scheduling method described in any embodiment of the present invention.

[0021] According to another aspect of the present invention, a computer-readable storage medium is provided, wherein the computer-readable storage medium stores computer instructions, and the computer instructions are used to enable a processor to implement the power equipment load scheduling method described in any embodiment of the present invention when executed.

[0022] According to another aspect of the present invention, a computer program product is provided. The computer program product includes a computer program. When the computer program is executed by a processor, the steps in the above-mentioned power equipment load scheduling method are implemented.

[0023] The present invention collects and analyzes the power consumption data and equipment status data of each node in the power network in real time, constructs the original data set, and removes noise and outliers through data preprocessing to ensure the accuracy of the data. Subsequently, the preprocessed data is analyzed in real time to obtain the current load status of each transformer and power node, reflecting the real-time load distribution of the system; based on the historical load data and real-time load status data, the system identifies the law and pattern of load change, and predicts the load change trend in the short term in the future through a machine learning algorithm. This prediction capability can identify potential load imbalance areas in advance; on this basis, the system accurately identifies load imbalance areas by comparing and analyzing real-time load data with predicted data, and generates the optimal load distribution plan. The plan adjusts the power distribution of each area according to the specific conditions of the load forecast and the imbalance area, optimizes resource allocation, and avoids equipment overload. Finally, by remotely issuing instructions to each transformer and control device, the efficient execution of the plan is ensured, the problem of power grid load imbalance is solved, the optimization of power resource distribution is achieved, and the safe operation of the power grid is guaranteed.

[0024] It should be understood that the contents described in this section are not intended to identify the key or important features of the embodiments of the present invention, nor are they intended to limit the scope of the present invention. Other features of the present invention will become easily understood through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0025] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0026] Figure 1 is a flow chart of a method for load scheduling of electric power equipment provided according to the first embodiment of the present invention;

[0027] Figure 2 is a structural schematic diagram of a load dispatching device for electric power equipment provided according to Embodiment 2 of the present invention;

[0028] Figure 3 It is a structural schematic diagram of an electronic device for implementing the power equipment load dispatching method according to an embodiment of the present invention. DETAILED DESCRIPTION

[0029] In order to enable those skilled in the art to better understand the scheme of the present invention, the technical scheme in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of the present invention.

[0030] It should be noted that the terms "first", "second", etc. in the specification and claims of the present invention and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchanged where appropriate, so that the embodiments of the present invention described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device that includes a series of steps or units is not necessarily limited to those steps or units that are clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0031] Embodiment 1

[0032] Figure 1 This is a flow chart of a method for load dispatching of electric power equipment provided in the first embodiment of the present invention. This embodiment can be applied to the case of load dispatching of unbalanced electric power equipment nodes in an electric power equipment network. The method can be executed by an electric power equipment load dispatching device. The electric power equipment load dispatching device can be implemented in the form of hardware and / or software. The electric power equipment load dispatching device can be configured in an electronic device. Figure 1 As shown, the method includes:

[0033] S110, acquiring power consumption data of each device node in the power network, and determining real-time load data of each device node based on the power consumption data.

[0034] S120. Clustering each device node according to the real-time load data of each device node to obtain at least one target node area; the target node area includes at least one device node.

[0035] S130, inputting the real-time load data of each device node in each target node area into a pre-trained load trend prediction model, and obtaining the predicted load data corresponding to each device node in each target node area output by the model.

[0036] S140. Determine the unbalanced node area according to the real-time load data and predicted load data of each device node in each target node area.

[0037] S150. Perform device load scheduling on each device node in the unbalanced node area according to the predicted load data of each device node in the unbalanced node area.

[0038] The device nodes in the power network may be, for example, transformers, circuit breakers, relays and other related equipment. The power consumption data may include the current, voltage and power of the device nodes. The real-time load data of the device nodes may be determined as follows:

[0039] P i (t) = V i (t)×I i (t)×cos(φ i );

[0040] Among them, P i (t) represents the real-time load data of device node i; V i (t) represents the voltage value of device node i at time t; I i (t) represents the current value of device node i at time t; cos(φ i ) represents the power factor of device node i, which can be preset by relevant technical personnel, for example, it can be set to 0.9.

[0041] Optionally, the device status data of each device node in the power network can be obtained in real time to facilitate real-time monitoring or supervision of the health status of the device node, so as to effectively identify the potential risks of the device node. The device status data may include device temperature, operating time, load rate, etc.

[0042] For example, the device temperature, operating time and load rate of the device node can be detected in real time to determine whether the device node is in an abnormal state; if so, an alarm is issued to remind relevant personnel that the device node is abnormal and request relevant personnel to handle the device. When it is found that the device abnormality of the device node is caused by overload, the device load scheduling can be performed based on this embodiment.

[0043] Optionally, before determining the real-time load data of the device node, the power consumption data and / or the device status data may be preprocessed. The preprocessing method may include denoising and outlier detection and elimination. Specifically, a moving average filter or a low-pass filter may be used to denoise the power consumption data and / or the device status data, filter out high-frequency noise in the signal, and obtain denoised power consumption data and / or device status data. The triple standard deviation method is used to detect and eliminate outliers in the denoised power consumption data and / or the device status data, and obtain power consumption data and / or device status data after eliminating outliers.

[0044] Optionally, the load ratio corresponding to each device node can be determined according to the power consumption data, and whether to perform device load scheduling on the device node can be determined according to the load ratio. Specifically, the load ratio of the device node is determined as follows:

[0045]

[0046] Among them, P i (t) represents the real-time load of device node i; P rated,i Indicates rated capacity; R i Indicates the load ratio of device node i.

[0047] If the load ratio is greater than the preset ratio threshold, an overload alarm signal is generated, and load scheduling is performed based on the device load scheduling method of this embodiment; if the load ratio is not greater than the preset ratio threshold, no load scheduling is required. The preset ratio threshold can be pre-set according to actual needs, for example, the preset ratio threshold can be set to 1.

[0048] Based on the real-time load data of each device node, each device node is clustered to obtain at least one target node region, wherein the number of target node regions is at least two, and the number of device nodes contained in the target node region is at least one.

[0049] In an optional embodiment, each device node is clustered according to the real-time load data of each device node to obtain at least one target node area, including:

[0050] Step a1: clustering each device node according to the real-time load data of each device node to obtain at least one initial node area and the regional center point of each initial node area.

[0051] Among them, the clustering method can be based on K-means (K-means clustering algorithm) to cluster device nodes. Specifically, based on the K-means clustering algorithm, device nodes with similar real-time load characteristics can be divided into the same area, thereby obtaining at least one initial node area. Each initial node area corresponds to its own regional center point, and for any initial node area, the area contains device nodes with similar load characteristics.

[0052] Step a2: determine the node weight value corresponding to each device node.

[0053] Among them, the node weight value of the device node can be pre-set by relevant technical personnel, and specifically, the weight value can be randomly assigned, or the weight value can be manually assigned according to the characteristics of the device node.

[0054] In order to further improve the accuracy of determining the node weight value of the device node and realize automatic determination of the node weight of the device node, optionally, determining the node weight value corresponding to each device node includes:

[0055] Step a21: Determine the node load mean corresponding to each device node according to the historical load data of each device node.

[0056] The historical load data of the device node may be load data monitored by the device node in a historical time period, and may be directly obtained.

[0057] For any device node, the load mean of the historical load data of the device node is used as the node load mean of the device node.

[0058] Step a22: Determine the node load standard deviation corresponding to each device node according to the historical load data and the node load mean of each device node.

[0059] Step a23: Determine the node weight value corresponding to each device node according to the node load mean value and node load standard deviation of each device node.

[0060] Exemplarily, the node weight value of the device node is determined as follows:

[0061]

[0062] Among them, P i represents the load mean of device node i; σ i represents the load standard deviation of device node i; w i Indicates the node weight value of device node i.

[0063] Optionally, the node weight value of the device node is normalized to obtain the normalized node weight value for subsequent calculations.

[0064] Step a3: According to the node weight value of each device node and based on the regional center point of each initial node area, determine the weighted distance between each device node and the regional center point of each initial node area.

[0065] Exemplarily, the method for determining the region center point of the initial node region is as follows:

[0066]

[0067] Among them, C k represents the center point of the kth initial node area; S k Represents all device nodes in the kth initial node area; xj Represents the real-time load data of the jth device node.

[0068] It should be noted that, in the process of performing the initial clustering of the device nodes, clustering can be based on different dimensions, such as based on the mean dimension or the standard deviation dimension of the device nodes, and clustering results under different dimensions can be obtained.

[0069] In the process of determining the weighted distance, the clustering results of different dimensions and the node weights of each device node are considered. The process of determining the weighted distance is as follows:

[0070]

[0071] Among them, d(x i ,C k ) represents the weighted Euclidean distance between node i and the initial node region, i.e., the weighted distance; w i ′ represents the normalized node weight value of device node i; represents the real-time load of device node i in the mth dimension; represents the center point of the kth initial node area in the m dimension; M represents the number of dimensions of the load data of the device node, for example, two dimensions including the mean dimension and the standard deviation dimension.

[0072] Step a4: Update the center point of each region according to the weighted distance to obtain the center point of the target region.

[0073] Specifically, according to the weighted distance, the device node is reallocated to the nearest regional center point, and the regional center point is updated. The process is repeated until at least the regional center point does not change, or the change amplitude is less than a preset threshold. The updated regional center point, i.e., the target regional center point, is determined as follows:

[0074]

[0075] in, Indicates the center point of the target area; S k Represents all device nodes in the kth initial node area; x j represents the real-time load data of the jth device node; w i ′ represents the normalized node weight value of device node i.

[0076] Step a5: Determine at least one target node area according to the target distance between the center point of the target area and each device node.

[0077] According to the target distance between the center point of the target area and each device node, re-clustering and division are performed to obtain at least one target node area.

[0078] The real-time load data of each device node in each target node area is input into the pre-trained load trend prediction model, and the load trend prediction model predicts the load trend of the device node in the future time period, and obtains the predicted load data corresponding to each device node in each target node area output by the model.

[0079] It is understandable that the model training accuracy of the load area prediction model affects the model's load prediction accuracy for device nodes in future areas. Therefore, to further ensure the model prediction accuracy, this embodiment also provides a model training method for the load area prediction model. Optionally, the model training method for the load trend prediction model is as follows:

[0080] Step b1: Obtain historical load data of each device node, and determine the load peak data and load change law data of each device node based on the historical load data of each device node.

[0081] Specifically, the autocorrelation analysis recognition algorithm can be used to identify the periodic change law in the historical load data of each device node, and the regression analysis recognition algorithm can be used to identify the long-term trend in the historical load data, and based on this, the load growth stage and the load reduction stage are divided to obtain the load change law data. Based on the results of the regression analysis, the local extreme value algorithm is used to identify the peak value in the historical load data, and based on the set threshold, the load peak data of each device node is determined.

[0082] Optionally, if the load peak exceeds a preset normal load range, an abnormality detection is performed on the device node.

[0083] Step b2: clustering each device node according to the historical load data of each device node to obtain at least one historical node area; the historical node area includes at least one device node.

[0084] The node clustering method based on historical load data may be the same as the method of clustering device nodes to determine the target node area in step 120 above, which will not be described in detail in this embodiment.

[0085] Step b3: Use the historical load data, load peak data and load change law data of each device node as training sample data, and use the area identifier of the historical node area of ​​each device node as the sample label of the training sample data.

[0086] The regional identifier of the historical node area is used as the sample label of the sample training data, so that the model can learn the corresponding feature data based on the load characteristics in different historical node areas during the learning process. In addition, during the model training process, based on the load peak data and load change law data, the load change trend is learned, so as to facilitate more accurate prediction of load data in the future time period.

[0087] Step b4: input the training sample data into the preset network model to obtain the model prediction result.

[0088] The network model may be pre-set by relevant technical personnel based on actual experience values ​​or experimental values. For example, the network model may be an LSTM (Long Short-Term Memory) model.

[0089] Step b5: According to the model prediction results, the network model is trained until the model training end conditions are met to obtain the load trend prediction model.

[0090] In order to improve the accuracy of model training, a loss function is constructed based on the model training scenario of this embodiment. Specifically, the loss function L in the model training process is constructed as follows:

[0091]

[0092] In the formula, represents Fourier transform, which is used to convert time domain data into frequency domain; β is the weight of the importance factor that balances the time domain error and the frequency domain error; N represents the number of samples of training sample data; P i (t+Δt) represents the historical load value of device node i, that is, the actual load at time t+Δt; It represents the predicted load value predicted by the LSTM model, that is, the load predicted by the model at time t+Δt.

[0093] Since the power load in time series data not only has characteristics in the time domain, but also reflects important trends in the frequency domain, during the model training process, the loss function is constructed by combining the time domain and the frequency domain, which not only considers the characteristics in the time domain but also the frequency domain. By balancing the differences between the two aspects, the model can better learn the time domain and frequency domain characteristics in the load data, thereby improving the overall prediction performance.

[0094] According to the real-time load data and predicted load data of each device node in each target node area, an unbalanced node area is determined. The unbalanced node area may be an area where the load of the device nodes in the area is unbalanced, for example, the device nodes in the area are in excess and overloaded, and the distribution is uneven.

[0095] In an optional embodiment, determining the unbalanced node area according to the real-time load data and the predicted load data of each device node in each target node area includes:

[0096] Step c1: Determine the load difference value corresponding to each device node according to the real-time load data and predicted load data of each device node in the target node area.

[0097] Specifically, the load difference value of the device node is determined as follows:

[0098]

[0099] Where ΔP i It represents the load difference between the predicted load data of device node i at the future time t+Δt and the real-time load data in the current time period, that is, the load difference value; represents the predicted load data of device node i at the future time t+Δt; P i (t) represents the real-time load data of device node i at the current moment.

[0100] Step c2: Determine the load difference mean of each device node according to the load difference value of each device node.

[0101] Specifically, the load difference mean μ of the device nodes ΔP is determined as follows:

[0102]

[0103] Where N represents the number of device nodes in the target node area.

[0104] Step c3: Determine the load difference standard deviation of each device node according to the load difference value and the load difference mean of each device node.

[0105] Specifically, the load difference standard deviation of the device node σ ΔP is determined as follows:

[0106]

[0107] Step c4: Determine the load imbalance degree of the target node area according to the load difference standard deviation and load difference mean of each device node in the target node area.

[0108] If the number of device nodes in the target node area whose load difference standard deviation exceeds the preset standard deviation threshold reaches a predetermined number threshold, it indicates that there is a significant load imbalance in the target node area. In order to further accurately determine the node area with load imbalance, the load imbalance degree U of each target node area can also be determined. imbalance , the specific determination method is as follows:

[0109]

[0110] Among them, σ ΔP represents the standard deviation of load difference in the target node area; μ ΔP Represents the mean load difference of the target node area.

[0111] Step c5: Determine the unbalanced node area according to the load imbalance degree of the target node area.

[0112] The target node area whose load imbalance degree is greater than the preset balance degree is determined as the unbalanced node area.

[0113] According to the predicted load data of each device node in the unbalanced node area, device load scheduling is performed on each device node in the unbalanced node area.

[0114] As an implementable manner of this embodiment, according to the predicted load data of each device node in the unbalanced node area, device load scheduling is performed on each device node in the unbalanced node area, including:

[0115] Step d1: Determine the remaining load capacity or excess load of each device node based on the predicted load data of each device node in the unbalanced node area.

[0116] Step d2: Determine the maximum transferable load between the device nodes based on the remaining load capacity or excess load of each device node.

[0117] Step d3: Generate an optimal load transfer path based on the maximum transferable load and unit transfer load cost between each device node.

[0118] Step d4: According to the optimal load transfer path, the device load of each device node in the unbalanced node area is scheduled.

[0119] Specifically, if the predicted load data of a device node If the capacity limit is exceeded, part of the load of the device node needs to be transferred to other adjacent nodes in the unbalanced node area, and the following conditions must be met:

[0120]

[0121] Among them, Pi,max Indicates the upper limit of the capacity of device node i.

[0122] The optimal load transfer path is determined by the minimum cost maximum flow algorithm. Specifically, the nodes in the unbalanced node area and the edges between the nodes are defined. The device nodes in the unbalanced node area are represented as a topological network node in the topological graph. Each topological network node has its own node attributes, namely, the remaining capacity. It represents the remaining load capacity that device node i can carry in the future time period t+Δt.

[0123] The definition of an edge in a topological graph is that each edge e in the graph ij represents the available load transfer path between device node i and device node j; the capacity of the edge u ij The maximum load that can be transferred from device node i to device node j is limited by the transmission line capacity and the node's remaining capacity. ij It is defined as the cost of transferring a unit load from device node i to device node j, which is related to transmission loss or economic cost.

[0124] All device nodes whose predicted load exceeds the node capacity are defined as excess load nodes, denoted as set S. For each i∈S, its excess load is All device nodes with positive remaining capacity are defined as remaining capacity nodes, denoted as set T. Add virtual source point s and sink point t to the topology graph model. Add an edge (s, i) with a capacity of excess load from source point s to each excess load node i, with a weight of 0; add an edge with a capacity of node remaining capacity C from each excess capacity node j to sink point t. j The edge (j,t) has a weight of 0.

[0125] Set the flow of all edges to 0; use Dijkstra algorithm to find the minimum cost path from source point s to sink point t; the available flow on the augmenting path is the minimum remaining capacity of all edges on the path; increase the flow along the augmenting path found, and update the flow f of each edge ij and the corresponding remaining capacity u ij , the updated flow and capacity are:

[0126] f ij ←f ij +flow

[0127] u ij ←u ij -flow

[0128] Among them, flow is dynamically calculated according to the maximum flow rate on the augmenting path during the algorithm operation, which represents the actual increase in flow on the augmenting path found.

[0129] Update Fees:

[0130]

[0131] Among them, Total cost It is the edge property, and its initial value is 0.

[0132] Repeat the search for augmenting paths to update costs until no new augmenting paths can be found; the traffic distribution obtained by the minimum cost maximum flow algorithm is the optimal load transfer solution. This solution will transfer excess load at the lowest transmission cost while meeting capacity constraints.

[0133] The generated load transfer plan is sent to each transformer and control device to ensure that the actual load is distributed and transferred according to the plan.

[0134] The present invention collects and analyzes the power consumption data and equipment status data of each node in the power network in real time, constructs the original data set, and removes noise and outliers through data preprocessing to ensure the accuracy of the data. Subsequently, the preprocessed data is analyzed in real time to obtain the current load status of each transformer and power node, reflecting the real-time load distribution of the system; based on the historical load data and real-time load status data, the system identifies the law and pattern of load change, and predicts the load change trend in the short term in the future through a machine learning algorithm. This prediction capability can identify potential load imbalance areas in advance; on this basis, the system accurately identifies load imbalance areas by comparing and analyzing real-time load data with predicted data, and generates the optimal load distribution plan. The plan adjusts the power distribution of each area according to the specific conditions of the load forecast and the imbalance area, optimizes resource allocation, and avoids equipment overload. Finally, by remotely issuing instructions to each transformer and control device, the efficient execution of the plan is ensured, the problem of power grid load imbalance is solved, the optimization of power resource distribution is achieved, and the safe operation of the power grid is guaranteed.

[0135] Embodiment 2

[0136] Figure 2 A schematic diagram of the structure of a power equipment load dispatching device provided in the second embodiment of the present invention. A power equipment load dispatching device provided in the embodiment of the present invention can be applied to the case of load dispatching of unbalanced power equipment nodes in a power equipment network. The power equipment load dispatching device can be implemented in the form of hardware and / or software, such as Figure 2 As shown, the power equipment load dispatching device can be configured in an electronic device, and specifically includes: a real-time load determination module 201, a clustering module 202, a load prediction module 203, an unbalanced area determination module 204 and a load dispatching module 205. Among them,

[0137] The real-time load determination module 201 is used to obtain power consumption data of each device node in the power network, and determine the real-time load data of each device node according to the power consumption data;

[0138] A clustering module 202 is used to cluster each of the device nodes according to the real-time load data of each of the device nodes to obtain at least one target node area; the target node area includes at least one device node;

[0139] The load prediction module 203 is used to input the real-time load data of each device node in each target node area into the pre-trained load trend prediction model to obtain the predicted load data corresponding to each device node in each target node area output by the model;

[0140] An unbalanced region determination module 204 is used to determine an unbalanced node region according to the real-time load data and the predicted load data of each of the device nodes in each of the target node regions;

[0141] The load scheduling module 205 is used to schedule the device load of each device node in the unbalanced node area according to the predicted load data of each device node in the unbalanced node area.

[0142] The present invention collects and analyzes the power consumption data and equipment status data of each node in the power network in real time, constructs the original data set, and removes noise and outliers through data preprocessing to ensure the accuracy of the data. Subsequently, the preprocessed data is analyzed in real time to obtain the current load status of each transformer and power node, reflecting the real-time load distribution of the system; based on the historical load data and real-time load status data, the system identifies the law and pattern of load change, and predicts the load change trend in the short term in the future through a machine learning algorithm. This prediction capability can identify potential load imbalance areas in advance; on this basis, the system accurately identifies load imbalance areas by comparing and analyzing real-time load data with predicted data, and generates the optimal load distribution plan. The plan adjusts the power distribution of each area according to the specific conditions of the load forecast and the imbalance area, optimizes resource allocation, and avoids equipment overload. Finally, by remotely issuing instructions to each transformer and control device, the efficient execution of the plan is ensured, the problem of power grid load imbalance is solved, the optimization of power resource distribution is achieved, and the safe operation of the power grid is guaranteed.

[0143] Optionally, the clustering module 202 includes:

[0144] An initial clustering unit, used for clustering each of the device nodes according to the real-time load data of each of the device nodes, to obtain at least one initial node area and a regional center point of each of the initial node areas;

[0145] A weight value determination unit, used to determine the node weight value corresponding to each of the device nodes;

[0146] A weighted distance determination unit, configured to determine, according to the node weight value of each device node and based on the regional center point of each initial node region, the weighted distance between each device node and the regional center point of each initial node region;

[0147] A center point updating unit, used for updating the center points of each region according to the weighted distance to obtain the center point of the target region;

[0148] The target area determination unit is used to determine at least one target node area according to the target distance between the center point of the target area and each of the device nodes.

[0149] Optionally, the weight value determining unit includes:

[0150] A load mean value determination subunit, used to determine the node load mean values ​​corresponding to each of the device nodes according to the historical load data of each of the device nodes;

[0151] A load standard deviation determination subunit, used to determine the node load standard deviation corresponding to each of the device nodes according to the historical load data of each of the device nodes and the node load mean;

[0152] The weight value determination subunit is used to determine the node weight value corresponding to each of the device nodes according to the node load mean value and the node load standard deviation of each of the device nodes.

[0153] Optionally, the model training method of the load trend prediction model is as follows:

[0154] Acquire historical load data of each of the device nodes, and determine load peak data and load variation law data of each of the device nodes according to the historical load data of each of the device nodes;

[0155] Clustering each of the device nodes according to the historical load data of each of the device nodes to obtain at least one historical node area; the historical node area includes at least one device node;

[0156] Using the historical load data, load peak data and load variation law data of each of the device nodes as training sample data, and using the region identifier of the historical node region of each of the device nodes as a sample label of the training sample data;

[0157] Input the training sample data into a preset network model to obtain a model prediction result;

[0158] According to the model prediction results, the network model is trained until the model training end condition is met to obtain a load trend prediction model.

[0159] Optionally, the imbalance area determination module 204 includes:

[0160] A difference value determining unit, configured to determine the load difference value corresponding to each of the device nodes according to the real-time load data and the predicted load data of each of the device nodes in the target node area;

[0161] A load difference mean value determining unit, configured to determine a load difference mean value of each of the device nodes according to the load difference value of each of the device nodes;

[0162] A load difference standard deviation determining unit, configured to determine the load difference standard deviation of each of the device nodes according to the load difference value and the load difference mean of each of the device nodes;

[0163] A load imbalance determination unit, configured to determine the load imbalance of the target node area according to the load difference standard deviation and the load difference mean of each of the device nodes in the target node area;

[0164] The unbalanced node area determining unit is used to determine the unbalanced node area according to the load unbalance degree of the target node area.

[0165] Optionally, the load scheduling module 205 includes:

[0166] A capacity determination unit, configured to determine the remaining load capacity or excess load of each of the device nodes according to the predicted load data of each of the device nodes in the unbalanced node area;

[0167] A maximum transfer amount determination unit, used to determine the maximum transferable load amount between each of the device nodes according to the remaining load capacity or excess load amount of each of the device nodes;

[0168] An optimal path generation unit, used for generating an optimal load transfer path according to the maximum transferable load amount and the unit transfer load cost between the device nodes;

[0169] The load scheduling unit is used to schedule the equipment load of each equipment node in the unbalanced node area according to the optimal load transfer path.

[0170] The power equipment load dispatching device provided in the embodiment of the present invention can execute the power equipment load dispatching method provided in any embodiment of the present invention, and has the corresponding functional modules and beneficial effects of the execution method.

[0171] Embodiment 3

[0172] Figure 3 A schematic diagram of the structure of an electronic device 30 that can be used to implement an embodiment of the present invention is shown. The electronic device is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device can also represent various forms of mobile devices, such as personal digital processing, cellular phones, smart phones, wearable devices (such as helmets, glasses, watches, etc.) and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely examples and are not intended to limit the implementation of the present invention described and / or required herein.

[0173] like Figure 3 As shown, the electronic device 30 includes at least one processor 31, and a memory connected to the at least one processor 31, such as a read-only memory (ROM) 32, a random access memory (RAM) 33, etc., wherein the memory stores a computer program that can be executed by at least one processor, and the processor 31 can perform various appropriate actions and processes according to the computer program stored in the read-only memory (ROM) 32 or the computer program loaded from the storage unit 38 to the random access memory (RAM) 33. In the RAM 33, various programs and data required for the operation of the electronic device 30 can also be stored. The processor 31, the ROM 32, and the RAM 33 are connected to each other through a bus 34. An input / output (I / O) interface 35 is also connected to the bus 34.

[0174] A number of components in the electronic device 30 are connected to the I / O interface 35, including: an input unit 36, such as a keyboard, a mouse, etc.; an output unit 37, such as various types of displays, speakers, etc.; a storage unit 38, such as a disk, an optical disk, etc.; and a communication unit 39, such as a network card, a modem, a wireless communication transceiver, etc. The communication unit 39 allows the electronic device 30 to exchange information / data with other devices through a computer network such as the Internet and / or various telecommunication networks.

[0175] The processor 31 may be a variety of general and / or special processing components with processing and computing capabilities. Some examples of the processor 31 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various dedicated artificial intelligence (AI) computing chips, various processors running machine learning model algorithms, digital signal processors (DSPs), and any appropriate processors, controllers, microcontrollers, etc. The processor 31 executes the various methods and processes described above, such as a power equipment load scheduling method.

[0176] In some embodiments, the power equipment load scheduling method may be implemented as a computer program, which is tangibly contained in a computer-readable storage medium, such as a storage unit 38. In some embodiments, part or all of the computer program may be loaded and / or installed on the electronic device 30 via the ROM 32 and / or the communication unit 39. When the computer program is loaded into the RAM 33 and executed by the processor 31, one or more steps of the power equipment load scheduling method described above may be performed. Alternatively, in other embodiments, the processor 31 may be configured to execute the power equipment load scheduling method in any other appropriate manner (e.g., by means of firmware).

[0177] Various implementations of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, field programmable gate arrays (FPGAs), application specific integrated circuits (ASICs), application specific standard products (ASSPs), systems on chips (SOCs), load programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various implementations can include: being implemented in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which can be a special purpose or general purpose programmable processor that can receive data and instructions from a storage system, at least one input device, and at least one output device, and transmit data and instructions to the storage system, the at least one input device, and the at least one output device.

[0178] Computer programs for implementing the methods of the present invention may be written in any combination of one or more programming languages. These computer programs may be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing device, so that when the computer program is executed by the processor, the functions / operations specified in the flow chart and / or block diagram are implemented. The computer program may be executed entirely on the machine, partially on the machine, partially on the machine and partially on a remote machine as a stand-alone software package, or entirely on a remote machine or server.

[0179] In the context of the present invention, a computer-readable storage medium may be a tangible medium that may contain or store a computer program for use by or in combination with an instruction execution system, device or equipment. A computer-readable storage medium may include, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, devices or equipment, or any suitable combination of the foregoing. Alternatively, a computer-readable storage medium may be a machine-readable signal medium. A more specific example of a machine-readable storage medium may include an electrical connection based on one or more lines, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

[0180] To provide interaction with a user, the systems and techniques described herein may be implemented on an electronic device having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user; and a keyboard and a pointing device (e.g., a mouse or trackball) through which the user can provide input to the electronic device. Other types of devices may also be used to provide interaction with the user; for example, the feedback provided to the user may be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user may be received in any form (including acoustic input, voice input, or tactile input).

[0181] The systems and techniques described herein may be implemented in a computing system that includes backend components (e.g., as a data server), or a computing system that includes middleware components (e.g., an application server), or a computing system that includes frontend components (e.g., a user computer with a graphical user interface or a web browser through which a user can interact with implementations of the systems and techniques described herein), or a computing system that includes any combination of such backend components, middleware components, or frontend components. The components of the system may be interconnected by any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include: a local area network (LAN), a wide area network (WAN), a blockchain network, and the Internet.

[0182] A computing system may include a client and a server. The client and the server are generally remote from each other and usually interact through a communication network. The client and server relationship is generated by computer programs running on the corresponding computers and having a client-server relationship with each other. The server may be a cloud server, also known as a cloud computing server or cloud host, which is a host product in the cloud computing service system to solve the defects of difficult management and weak business scalability in traditional physical hosts and VPS services.

[0183] It should be understood that the various forms of processes shown above can be used to reorder, add or delete steps. For example, the steps described in the present invention can be executed in parallel, sequentially or in different orders, as long as the desired results of the technical solution of the present invention can be achieved, and this document does not limit this.

[0184] The above specific implementations do not constitute a limitation on the protection scope of the present invention. It should be understood by those skilled in the art that various modifications, combinations, sub-combinations and substitutions can be made according to design requirements and other factors. Any modification, equivalent substitution and improvement made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. A method for dispatching load of electric power equipment, It is characterized in that include: Acquire power consumption data of each device node in the power network, and determine real-time load data of each device node based on the power consumption data; Clustering each of the device nodes according to the real-time load data of each of the device nodes to obtain at least one target node area; The target node area includes at least one device node; Input the real-time load data of each device node in each target node area into the pre-trained load trend prediction model to obtain the predicted load data corresponding to each device node in each target node area output by the model; Determine an unbalanced node area according to the real-time load data and the predicted load data of each of the device nodes in each of the target node areas; According to the predicted load data of each of the device nodes in the unbalanced node area, device load scheduling is performed on each of the device nodes in the unbalanced node area.

2. The method according to claim 1, It is characterized in that The step of clustering the device nodes according to the real-time load data of the device nodes to obtain at least one target node area includes: Clustering each of the device nodes according to the real-time load data of each of the device nodes to obtain at least one initial node area and a regional center point of each of the initial node areas; Determine the node weight value corresponding to each of the device nodes; According to the node weight value of each device node and based on the regional center point of each initial node area, determining the weighted distance between each device node and the regional center point of each initial node area; According to the weighted distance, the center point of each region is updated to obtain the center point of the target region; At least one target node area is determined according to the target distance between the center point of the target area and each of the device nodes.

3. The method according to claim 2, It is characterized in that The determining of the node weight values ​​corresponding to the device nodes respectively includes: Determine the node load mean value corresponding to each of the device nodes according to the historical load data of each of the device nodes; Determine the node load standard deviation corresponding to each of the device nodes according to the historical load data and the node load mean of each of the device nodes; The node weight values ​​corresponding to the device nodes are determined according to the node load mean and the node load standard deviation of the device nodes.

4. The method according to claim 1, It is characterized in that The model training method of the load trend prediction model is as follows: Acquire historical load data of each of the device nodes, and determine load peak data and load variation law data of each of the device nodes according to the historical load data of each of the device nodes; Clustering each of the device nodes according to the historical load data of each of the device nodes to obtain at least one historical node area; the historical node area includes at least one device node; Using the historical load data, load peak data and load variation law data of each of the device nodes as training sample data, and using the region identifier of the historical node region of each of the device nodes as a sample label of the training sample data; Input the training sample data into a preset network model to obtain a model prediction result; According to the model prediction results, the network model is trained until the model training end condition is met to obtain a load trend prediction model.

5. The method according to claim 1, It is characterized in that Determining the unbalanced node area according to the real-time load data and the predicted load data of each device node in each target node area includes: Determine the load difference value corresponding to each of the device nodes according to the real-time load data and the predicted load data of each of the device nodes in the target node area; Determining a load difference mean value of each of the device nodes according to the load difference value of each of the device nodes; Determining the load difference standard deviation of each of the device nodes according to the load difference value and the load difference mean of each of the device nodes; Determining the load imbalance degree of the target node area according to the load difference standard deviation and the load difference mean of each of the device nodes in the target node area; An unbalanced node area is determined according to the load imbalance degree of the target node area.

6. The method according to claim 1, It is characterized in that The step of performing device load scheduling on each device node in the unbalanced node area according to the predicted load data of each device node in the unbalanced node area includes: Determine the remaining load capacity or excess load of each of the device nodes according to the predicted load data of each of the device nodes in the unbalanced node area; Determining the maximum transferable load between the device nodes according to the remaining load capacity or excess load of each device node; Generate an optimal load transfer path according to the maximum transferable load amount and unit transfer load cost between the device nodes; According to the optimal load transfer path, device load scheduling is performed on each device node in the unbalanced node area.

7. A load dispatching device for electric power equipment, It is characterized in that include: A real-time load determination module, used to obtain power consumption data of each device node in the power network, and determine the real-time load data of each device node according to the power consumption data; A clustering module, used for clustering each of the device nodes according to the real-time load data of each of the device nodes to obtain at least one target node area; The target node area includes at least one device node; A load prediction module, used for inputting the real-time load data of each device node in each target node area into a pre-trained load trend prediction model, and obtaining the predicted load data corresponding to each device node in each target node area output by the model; An unbalanced area determination module, used to determine an unbalanced node area according to the real-time load data and predicted load data of each of the device nodes in each of the target node areas; The load scheduling module is used to schedule the equipment load of each equipment node in the unbalanced node area according to the predicted load data of each equipment node in the unbalanced node area.

8. An electronic device, It is characterized in that The electronic device comprises: at least one processor; and a memory communicatively connected to the at least one processor; wherein, The memory stores a computer program executable by the at least one processor, and the computer program is executed by the at least one processor so that the at least one processor can execute the power equipment load scheduling method according to any one of claims 1 to 6.

9. A computer-readable storage medium, It is characterized in that The computer-readable storage medium stores computer instructions, and the computer instructions are used to enable a processor to implement the power equipment load scheduling method according to any one of claims 1 to 6 when executed.

10. A computer program product, It is characterized in that The computer program product comprises a computer program, and when the computer program is executed by a processor, the computer program implements the method for load dispatching of electric power equipment according to any one of claims 1 to 6.