Power dispatching task generation method and device, equipment and storage medium

Abstract

The invention discloses a power dispatching task generation method and device, equipment and a storage medium, and the method comprises the steps: obtaining the current operation data of a target power grid, extracting global operation characteristics representing the global operation information of the target power grid from the operation data, and for each power equipment, generating a power dispatching task according to the global operation characteristics of the target power grid; the local features of the neighborhood with the power equipment as the topology center are extracted from the operation data, the global operation features and the local features of the power equipment are fused to obtain fusion features, a power dispatching task is generated based on the fusion features, and the power dispatching task comprises dispatching operation of the power equipment. Compared with an existing modeling method, the power dispatching task generation efficiency is improved, and the cost is reduced. In addition, the power dispatching task is generated by fusing the global operation characteristics of the global operation information of the target power grid and the local characteristics of the neighborhood with the power equipment as the topology center, and the accuracy of the power dispatching task is improved.

Description

Technical Field

[0001] The present invention relates to electric power dispatching technology, and in particular to an electric power dispatching task generation method, device, equipment and storage medium. Background Art

[0002] The specific work content of power dispatching is to judge the safety and economic operation status of the power grid based on the data information fed back by various information collection devices or the information provided by monitoring personnel, combined with the actual operating parameters of the power grid, such as voltage, current, frequency, load, etc., and comprehensively consider the progress of various production work, issue operating instructions, and adjust the control system, such as adjusting the generator output, adjusting the load distribution, switching capacitors and reactors, etc., so as to ensure the continuous safe and stable operation of the power grid.

[0003] Traditional power system dispatch can be defined as modeling each basic unit in the target power grid from the system level, determining the optimization variables, constraints and objectives in the dispatch optimization model, and then finding the optimal solution that meets the constraints under the guidance of the objectives. This is a dispatch optimization method guided by this type of theoretical framework.

[0004] However, with the rapid development of the power system, the physical characteristics of the power generation side and the power consumption side have gradually diversified. The large-scale application of new energy such as wind and solar power and the massive number of dispatchable units have significantly increased the modeling cost, the dispatch optimization problem has become increasingly complex, and the efficiency of traditional model-based dispatch optimization methods has significantly decreased. Summary of the invention

[0005] The present invention provides a method, device, equipment and storage medium for generating electric power dispatching tasks, so as to improve the generation efficiency and accuracy of electric power dispatching tasks and reduce the cost.

[0006] In a first aspect, the present invention provides a method for generating a power dispatching task, comprising:

[0007] Acquire current operating data of the target power grid, wherein the operating data includes operating parameters of each power device in the target power grid, wherein the operating parameters include current, voltage, power, and operating status;

[0008] Extracting global operation characteristics representing global operation information of the target power grid from the operation data;

[0009] For each of the electric power equipment, extracting local features of a neighborhood with the electric power equipment as a topological center from the operation data;

[0010] Fusion of the global operation feature and the local feature of each of the electric power devices to obtain a fusion feature;

[0011] A power dispatching task is generated based on the fusion features, and the power dispatching task includes a dispatching operation of each of the power devices.

[0012] Optionally, extracting a global operation feature representing global operation information of the target power grid from the operation data includes:

[0013] Performing embedding processing on the operation data to obtain a first embedding feature represented by a vector of the operation data;

[0014] Performing linearization, batch normalization, and nonlinear activation on the first embedded features to obtain a first regularized feature;

[0015] Encoding the first regularized feature based on a self-attention mechanism to obtain an encoded feature;

[0016] The encoding features are globally pooled to obtain global operation features representing global operation information of the target power grid.

[0017] Optionally, encoding the first regularized feature based on a self-attention mechanism to obtain an encoded feature includes:

[0018] Inputting the first regularized feature into a self-attention module for processing, the self-attention module includes a plurality of cascaded self-attention units, the input feature of the first self-attention unit is the first regularized feature, the input feature of the i-th self-attention unit is the output feature of the i-1-th self-attention unit, each of the self-attention units is used to output attention features of different feature scales, and i is a positive integer greater than 1;

[0019] Concatenate multiple attention features of different feature scales in the spatial dimension to obtain concatenated features;

[0020] The concatenated features are linearized, batch normalized, and nonlinearly activated to obtain encoding features.

[0021] Optionally, for each of the electric power devices, extracting local features of a neighborhood with the electric power device as a topological center from the operating data includes:

[0022] For each of the electric power devices, determining a neighboring device with the electric power device as a topological center;

[0023] Filtering the operation data corresponding to the neighboring device from the operation data as the neighboring operation data;

[0024] Embedding the neighborhood operation data to obtain a second embedding feature represented by a vector of the neighborhood operation data;

[0025] Performing linearization, batch normalization, and nonlinear activation on the second embedded feature to obtain a second regularized feature;

[0026] The second regularized feature is subjected to convolution processing to obtain a local feature of a neighborhood with the power device as the topological center.

[0027] Optionally, generating a power dispatch task based on the fusion feature includes:

[0028] Performing linearization processing on the fused features to obtain linearized features;

[0029] The linearization feature is mapped to the power dispatching task in the instruction library of the power dispatching task to generate the power dispatching task.

[0030] Optionally, after generating the power dispatch task based on the fusion feature, the method further includes:

[0031] The dispatching operation of each device in the generated power dispatching task is verified based on the pre-built rule engine.

[0032] Optionally, after generating the power dispatch task based on the fusion feature, the method further includes:

[0033] After executing the power dispatching task, collecting feedback data fed back by each of the power devices;

[0034] The feedback data is used as training data to optimize model parameters of the power dispatch task generation model.

[0035] In a second aspect, the present invention further provides a power dispatching task generating device, comprising:

[0036] An operation data acquisition module is used to acquire the current operation data of the target power grid, wherein the operation data includes the operation parameters of each power device in the target power grid, and the operation parameters include current, voltage, power, and operation status;

[0037] A global feature extraction module, used to extract global operation features representing global operation information of the target power grid from the operation data;

[0038] A local feature extraction module, for extracting, for each of the electric power equipment, local features of a neighborhood with the electric power equipment as a topological center from the operation data;

[0039] A feature fusion module, used for fusing the global operation feature and the local feature of each of the electric power equipment to obtain a fusion feature;

[0040] A scheduling task generation module is used to generate a power scheduling task based on the fusion features, and the power scheduling task includes a scheduling operation of each of the power devices.

[0041] In a third aspect, the present invention further provides an electronic device, comprising:

[0042] one or more processors;

[0043] A storage device for storing one or more programs;

[0044] When the one or more programs are executed by the one or more processors, the one or more processors implement the power scheduling task generation method provided in the first aspect of the present invention.

[0045] In a fourth aspect, the present invention further provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the power dispatching task generation method provided in the first aspect of the present invention.

[0046] The method for generating a power dispatching task provided by the present invention obtains the current operating data of the target power grid, the operating data includes the operating parameters of each power device in the target power grid, the operating parameters include current, voltage, power, and operating status, extracts global operating features that characterize the global operating information of the target power grid from the operating data, extracts local features of the neighborhood with the power device as the topological center from the operating data for each power device, fuses the global operating features and the local features of each power device to obtain the fused features, and generates a power dispatching task based on the fused features, the power dispatching task includes the dispatching operation of each power device. The present invention adopts a large model based on deep learning, and automatically generates power dispatching tasks using the current operating data of the target power grid. Compared with the existing modeling method, the generation efficiency of power dispatching task generation is improved and the cost is reduced. In addition, the present invention fuses the global operating features of the global operating information of the target power grid and the local features of the neighborhood with the power device as the topological center to generate power dispatching tasks, thereby improving the accuracy of power dispatching tasks.

[0047] 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

[0048] 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.

[0049] Figure 1 A flow chart of a method for generating a power dispatching task provided by the present invention;

[0050] Figure 2 A schematic diagram of the structure of a power dispatching task generation model provided by the present invention;

[0051] Figure 3 A schematic diagram of the structure of a power dispatching task generating device provided by the present invention;

[0052] Figure 4 The present invention provides a schematic structural diagram of an electronic device.

[0053] The above drawings have shown clear embodiments of the present application, which will be described in more detail later. These drawings and text descriptions are not intended to limit the scope of the present application in any way, but to illustrate the concept of the present application to those skilled in the art by referring to specific embodiments. DETAILED DESCRIPTION

[0054] 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.

[0055] 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.

[0056] Figure 1 This is a flow chart of a method for generating a power dispatching task provided by the present invention. This embodiment is applicable to the case where a power dispatching task is automatically generated based on the current operating data of the target power grid. The method can be executed by a power dispatching task generating device provided by an embodiment of the present invention. The device can be implemented by software and / or hardware and is usually configured in an electronic device, such as Figure 1 As shown, the power dispatching task generation method specifically includes the following steps:

[0057] S101. Obtain current operation data of a target power grid.

[0058] In an embodiment of the present invention, the power dispatching task generation method can be executed by a power dispatching platform, and the power dispatching platform obtains the current operating data of the target power grid in real time. The target power grid can be a power network responsible for a certain transformer substation or multiple transformer substations, and the operating data can include the operating parameters of each power device in the target power grid, and the operating parameters include current, voltage, power, and operating status (such as normal, overload, shutdown, etc.). Power equipment can include substations, power stations, load points, etc., which are not limited in the embodiment of the present invention.

[0059] Exemplarily, corresponding sensors or data acquisition devices can be set at the power equipment to collect the operating parameters of each power equipment, and upload them to the power dispatching platform in real time through the Internet of Things device. In a specific embodiment of the present invention, a stable real-time interactive link is established between the power dispatching platform and the Internet of Things device through a standard communication protocol (such as HTTP or gRPC). The operating data of the target power grid is transmitted to the power dispatching platform through a standardized interface (for example, an interface based on JSON, XML, etc.) to ensure that the data can be transmitted in real time and quickly parsed for the next task generation. In this link, the efficient transmission of data is guaranteed and the delay is reduced to ensure that the power dispatching task generation process can reflect the dynamic changes of the target power grid in real time.

[0060] S102: extracting global operation features representing global operation information of the target power grid from the operation data.

[0061] In an embodiment of the present invention, the operation data is vectorized (matrixed) and then input into the power dispatching task generation model provided in the embodiment of the present invention for processing, and the power dispatching task generation model performs reasoning to finally generate a power dispatching task. The power dispatching task generation model uses a large amount of historical dispatching data and power grid operation data for learning and training, and learns the complex relationship between the operating status of power equipment and the dispatching command. Exemplarily, the vectorized operation data is input into the feature extraction network of the power dispatching task generation model for feature extraction, and the global operation features representing the global operation information of the target power grid are extracted from the operation data. The global operation features represent the overall attributes of the entire operation data.

[0062] It should be noted that the above-mentioned global operation features can be extracted using a deep convolutional network or a Transformer architecture, which is not limited in the embodiments of the present invention.

[0063] In some embodiments of the present invention, extracting global operation features representing global operation information of the target power grid from the operation data includes:

[0064] S1021. Perform embedding processing on the operating data to obtain a first embedding feature represented by a vector of the operating data.

[0065] Exemplarily, embedding processing is performed on the operation data to obtain a first embedding feature represented by a vector of the operation data. Exemplarily, the vector representation of the first embedding feature is as follows:

[0066] X={x 1, x 2, ...x n}

[0067] Among them, X is the first embedding feature, x i is a vector composed of the operating parameters of the i-th power equipment, where 1≤i≤n, and i is a positive integer.

[0068] S1022. Linearize, batch normalize, and nonlinearly activate the first embedded feature to obtain a first regularized feature.

[0069] Figure 2 A schematic diagram of a power dispatching task generation model provided by the present invention is shown in FIG. Figure 2 As shown, the first embedded feature X is input into an LBR unit for processing. The LBR unit includes a linearization layer (Linear), a batch normalization layer (Batch Normalization) and a Rule function activation layer connected in sequence. The linearization layer linearizes the first embedded feature X, the batch normalization layer performs batch normalization processing on the features output by the linearization layer, and the Rule function activation layer performs nonlinear activation on the features output by the batch normalization layer, and finally obtains the first regularized feature.

[0070] S1023. Encode the first regularized feature based on the self-attention mechanism to obtain an encoded feature.

[0071] In an embodiment of the present invention, the first regularized feature is encoded based on the self-attention mechanism to capture the correlation between the elements in the first regularized feature, which is beneficial to improving the accuracy of the generated power scheduling task.

[0072] For example, Figure 2As shown, the first regularized feature is input into the self-attention module for processing. The self-attention module includes multiple cascaded self-attention units Attention, the input feature of the first self-attention unit is the first regularized feature, the input feature of the i-th self-attention unit is the output feature of the i-1-th self-attention unit, and each self-attention unit is used to output attention features of different feature scales, and i is a positive integer greater than 1. Each self-attention unit can process the input features in a common self-attention processing method, that is: linearly transform the input features respectively to obtain a query matrix, a key matrix and a value matrix, calculate the dot product of the query matrix and the key matrix and normalize them to obtain an attention weight matrix, calculate the dot product of the attention weight matrix and the value matrix to obtain a self-attention matrix, and finally calculate the sum of the self-attention matrix and the input feature to obtain the output feature of the self-attention unit. The embodiments of the present invention will not be repeated here.

[0073] like Figure 2 As shown, multiple attention features of different feature scales (i.e., the output features of each attention unit) are concatenated (Concatenate) in the spatial dimension (represented by C in the figure) to obtain concatenated features.

[0074] like Figure 2 As shown, the concatenated features are linearized, batch normalized, and nonlinearly activated to obtain the encoding features. Specifically, the concatenated features are input into an LBR unit for linearization, batch normalization, and nonlinear activation to obtain the encoding features. The structure of the LBR unit is the same as that of the LBR unit in the aforementioned embodiment, and the embodiments of the present invention are not repeated here.

[0075] S1024: Perform global pooling processing on the coded features to obtain global operation features that characterize global operation information of the target power grid.

[0076] like Figure 2 As shown, the coded features are globally pooled to obtain global operation features that characterize the global operation information of the target power grid. Exemplarily, in an embodiment of the present invention, the coded features are first input into a global maximum pooling layer (Global Max-Pooling, GMP), and the output result of the global maximum pooling layer is input into a global average pooling layer (Global Average-Pooling, GAP), and finally the global operation features that characterize the global operation information of the target power grid are obtained.

[0077] S103 . For each electric power device, extract local features of a neighborhood with the electric power device as the topological center from the operating data.

[0078] In an embodiment of the present invention, for each power device, local features of the neighborhood with the power device as the topological center are extracted from the operating data. Exemplarily, with each power device as the topological center, the neighborhood devices with the power device as the topological center and the operating data corresponding to the neighborhood devices are determined, and the local features representing the association relationship between the devices in the neighborhood devices are extracted from the operating data corresponding to the neighborhood devices using a feature extraction network.

[0079] In some embodiments of the present invention, for each power device, extracting local features of a neighborhood with the power device as the topological center from the operation data includes:

[0080] S1031. For each electric power device, determine neighboring devices with the electric power device as the topological center.

[0081] In the embodiment of the present invention, based on the topological structure of the target power grid, for each power device, a neighboring device with the power device as the topological center is determined, and the neighboring device includes the power device at the topological center and other power devices connected to the topological center.

[0082] S1032: Filter out operation data corresponding to the neighboring device from the operation data as the neighboring operation data.

[0083] After determining the neighboring device with the power device as the topological center, the operation data corresponding to the neighboring device is filtered out from the operation data obtained in the above step as the neighborhood operation data of the electronic device. Exemplarily, the operation parameters of each electronic device in the operation data have a unique identifier of the electronic device, and the operation data corresponding to the neighboring device can be filtered out from the operation data as the neighborhood operation data based on the identifier.

[0084] S1033. Embedding the neighborhood operation data to obtain a second embedding feature represented by a vector of the neighborhood operation data.

[0085] In an embodiment of the present invention, the neighborhood operation data may be embedded to obtain a second embedded feature represented by a vector of the neighborhood operation data. Exemplarily, the vector representation of the second embedded feature is as follows:

[0086] Y j ={y1,y2,...y n}

[0087] Among them, Y j is the second embedded feature corresponding to the jth power device, y i is a vector composed of the operating parameters of the i-th neighboring device of the j-th power device, where 1≤i≤n, and i is a positive integer.

[0088] S1034. Linearize, batch normalize, and nonlinearly activate the second embedded feature to obtain a second regularized feature.

[0089] In an embodiment of the present invention, the second embedded features corresponding to all power devices can be combined into a matrix Y, and then the matrix is ​​linearized, batch normalized and nonlinearly activated to obtain the second regularized features. For example, Figure 2 As shown, the matrix Y is input into an LBR unit for processing. The LBR unit includes a linearization layer (Linear), a batch normalization layer (Batch Normalization) and a Rule function activation layer connected in sequence. The linearization layer linearizes the matrix Y, the batch normalization layer performs batch normalization processing on the features output by the linearization layer, and the Rule function activation layer performs nonlinear activation on the features output by the batch normalization layer, and finally obtains the second regularized features.

[0090] S1035. Perform convolution processing on the second regularized feature to obtain local features of a neighborhood with the power equipment as the topological center.

[0091] In the embodiment of the present invention, Figure 2 As shown, the second regularized feature is input into a convolution unit (Conv) for convolution processing to obtain local features of the neighborhood with each power device as the topological center. In the embodiment of the present invention, the structure of the convolution unit is not limited, and may include one or more convolution layers.

[0092] S104: Fuse the global operation characteristics with the local characteristics of each power device to obtain a fused characteristic.

[0093] like Figure 2 As shown, in an embodiment of the present invention, the global operation characteristics and the local characteristics of each power device are concatenated (represented by C in the figure) in the characteristic dimension to obtain a fusion characteristic.

[0094] S105. Generate power dispatching tasks based on the fused features.

[0095] In the embodiment of the present invention, a power dispatching task is generated based on the fusion feature. The power dispatching task includes the dispatching operation of each power device. The dispatching operation includes starting or shutting down the device, adjusting the load distribution, switching the backup power supply, etc. The embodiment of the present invention is not limited here. The generated power dispatching task can be expressed as a sequence:

[0096] T={t1,t2,...,t k}

[0097] Among them, T is the entire power dispatching task, t i is the specific instruction of the i-th scheduling operation.

[0098] For example, Figure 2 As shown, the fused features are input into a linear layer (Linear), the fused features are linearized to obtain linearized features, and then the linearized features are input into an activation function layer (Softmax), the linearized features are mapped to the instruction library of the power dispatching task, and the power dispatching task is generated.

[0099] The method for generating a power dispatching task provided by the present invention obtains the current operating data of the target power grid, the operating data includes the operating parameters of each power device in the target power grid, the operating parameters include current, voltage, power, and operating status, extracts global operating features that characterize the global operating information of the target power grid from the operating data, extracts local features of the neighborhood with the power device as the topological center from the operating data for each power device, fuses the global operating features and the local features of each power device to obtain the fused features, and generates a power dispatching task based on the fused features, the power dispatching task includes the dispatching operation of each power device. The present invention adopts a large model based on deep learning, and automatically generates power dispatching tasks using the current operating data of the target power grid. Compared with the existing modeling method, the generation efficiency of power dispatching task generation is improved and the cost is reduced. In addition, the present invention fuses the global operating features of the global operating information of the target power grid and the local features of the neighborhood with the power device as the topological center to generate power dispatching tasks, thereby improving the accuracy of power dispatching tasks.

[0100] In some embodiments of the present invention, in order to improve the safety of power dispatching tasks and avoid unsafe dispatching operations in the generated power dispatching tasks, which may cause safety problems after execution, after generating the power dispatching tasks based on the fusion features, the following may also be included:

[0101] The dispatching operation of each device in the generated power dispatching task is verified based on the pre-built rule engine.

[0102] Exemplarily, the rule engine is a predefined set of business rule systems used to ensure that the generated dispatching tasks meet the dispatching rules such as the safety of the power grid, the carrying capacity of the equipment, and the priority. The rule engine will verify each dispatching operation in the generated power dispatching task. The verification content may include: the compliance of the task (whether it complies with the dispatching strategy), whether the equipment can withstand the task operation (such as whether the load exceeds the limit), the priority of task execution, etc., which are not limited in the present invention. Through the verification of the rule engine, it can be ensured that each dispatching operation in the power dispatching task is safe and executable after it is generated.

[0103] In some embodiments of the present invention, the verified power dispatching task is sent to the dispatcher, who executes the power dispatching task, or the verified power dispatching task is directly sent to the corresponding execution device to automatically execute the power dispatching task. After executing the power dispatching task, the feedback data fed back by each power device is collected. The feedback data may include the execution status of the task (success / failure), the specific time of execution (start time, end time), and the response status of the device (whether it meets expectations). These feedback data are transmitted to the power dispatching platform through the interface as the basis for optimizing the model. The feedback data is used as training data to optimize the model parameters of the power dispatching task generation model.

[0104] For example, R = {r1, r2, ..., r k}, r i Represents the feedback data of the i-th power device. The feedback data is used as training data to optimize the model parameters of the power dispatch task generation model, which can further improve the accuracy of the generated power dispatch task. The specific optimization process is based on the following loss function:

[0105]

[0106] Among them, f θ (x i ) is the specific instruction of the ith dispatch operation in the generated power dispatch task (i.e., t i ), by minimizing the loss function, the model parameters of the power dispatching task generation model can be adjusted to make the generated task more in line with the actual situation.

[0107] The model learns by receiving feedback data from task execution in real time. After each task is completed, the model uses the feedback data as new training data and updates the model parameters. The self-learning algorithm is based on an incremental learning strategy, so that the model can process new data without retraining. As the feedback data continues to increase, the model can better adapt to changes in the power grid state. The model dynamically optimizes the generated scheduling tasks by adjusting the task generation strategy. The dynamic optimization process of the model updates the parameters through the gradient descent algorithm to minimize the loss function and achieve continuous improvement in the generation of scheduling tasks.

[0108] Figure 3 A schematic diagram of a power dispatching task generating device provided by the present invention is shown in FIG. Figure 3 As shown, the power dispatching task generating device includes:

[0109] An operation data acquisition module 201 is used to acquire the current operation data of the target power grid, wherein the operation data includes the operation parameters of each power device in the target power grid, and the operation parameters include current, voltage, power, and operation status;

[0110] A global feature extraction module 202, configured to extract global operation features representing global operation information of the target power grid from the operation data;

[0111] A local feature extraction module 203 is used to extract, for each of the electric power equipment, local features of a neighborhood with the electric power equipment as a topological center from the operation data;

[0112] A feature fusion module 204, configured to fuse the global operation feature and the local feature of each of the electric power devices to obtain a fused feature;

[0113] The scheduling task generation module 205 is used to generate a power scheduling task based on the fusion features, and the power scheduling task includes the scheduling operation of each of the power devices.

[0114] In some embodiments of the present invention, the global feature extraction module 202 includes:

[0115] A first embedding submodule, configured to perform embedding processing on the operation data to obtain a first embedding feature represented by a vector of the operation data;

[0116] A first regularization submodule, used for performing linearization, batch normalization and nonlinear activation on the first embedding feature to obtain a first regularized feature;

[0117] A self-attention encoding submodule, used for encoding the first regularized feature based on a self-attention mechanism to obtain an encoded feature;

[0118] The global pooling processing module is used to perform global pooling processing on the coding features to obtain global operation features that characterize the global operation information of the target power grid.

[0119] In some embodiments of the invention, the self-attention encoding submodule includes:

[0120] A self-attention processing unit, configured to input the first regularized feature into a self-attention module for processing, wherein the self-attention module includes a plurality of cascaded self-attention units, wherein the input feature of the first self-attention unit is the first regularized feature, the input feature of the i-th self-attention unit is the output feature of the i-1-th self-attention unit, and each of the self-attention units is configured to output attention features of different feature scales, where i is a positive integer greater than 1;

[0121] A feature concatenation unit is used to concatenate multiple attention features of different feature scales in the spatial dimension to obtain a concatenated feature;

[0122] The coding feature calculation unit is used to perform linearization, batch normalization and nonlinear activation on the concatenated features to obtain coding features.

[0123] In some embodiments of the present invention, the local feature extraction module 203 includes:

[0124] A neighbor device determination submodule, used for determining, for each of the electric power devices, a neighbor device with the electric power device as a topological center;

[0125] A data screening submodule, configured to screen out the operating data corresponding to the neighborhood device from the operating data as the neighborhood operating data;

[0126] A second embedding submodule is used to perform embedding processing on the neighborhood operation data to obtain a second embedding feature represented by a vector of the neighborhood operation data;

[0127] A second regularization submodule, used for performing linearization, batch normalization and nonlinear activation on the second embedding feature to obtain a second regularized feature;

[0128] The local feature extraction submodule is used to perform convolution processing on the second regularized feature to obtain the local features of the neighborhood with the power equipment as the topological center.

[0129] In some embodiments of the present invention, the scheduling task generation module 205 includes:

[0130] A linearization processing submodule, used for performing linearization processing on the fusion features to obtain linearized features;

[0131] The task generation submodule is used to map the linearization characteristics to the power dispatching task and the instruction library of the power dispatching task to generate the power dispatching task.

[0132] In some embodiments of the present invention, the power dispatching task generating device further includes:

[0133] The verification module is used to verify the dispatching operation of each device in the generated power dispatching task based on the pre-built rule engine after the power dispatching task is generated based on the fusion feature.

[0134] In some embodiments of the present invention, the power dispatching task generating device further includes:

[0135] A feedback data collection module, used for collecting feedback data fed back by each of the power devices after generating a power dispatch task based on the fusion feature and executing the power dispatch task;

[0136] The model optimization module is used to optimize the model parameters of the power dispatching task generation model by using the feedback data as training data.

[0137] The above-mentioned power dispatching task generation device can execute the power dispatching task generation method provided by the above-mentioned embodiment of the present invention, and has the corresponding functional modules and beneficial effects for executing the power dispatching task generation method.

[0138] Figure 4 A schematic diagram of the structure of an electronic device provided by the present invention. 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 may 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.

[0139] like Figure 4 As shown, the electronic device includes at least one processor 11, and a memory connected to the at least one processor 11, such as a read-only memory (ROM) 12, a random access memory (RAM) 13, etc., wherein the memory stores a computer program that can be executed by at least one processor, and the processor 11 can perform various appropriate actions and processes according to the computer program stored in the read-only memory (ROM) 12 or the computer program loaded from the storage unit 18 to the random access memory (RAM) 13. In the RAM 13, various programs and data required for the operation of the electronic device can also be stored. The processor 11, the ROM 12, and the RAM 13 are connected to each other through a bus 14. An input / output (I / O) interface 15 is also connected to the bus 14.

[0140] A number of components in the electronic device are connected to the I / O interface 15, including: an input unit 16, such as a keyboard, a mouse, etc.; an output unit 17, such as various types of displays, speakers, etc.; a storage unit 18, such as a disk, an optical disk, etc.; and a communication unit 19, such as a network card, a modem, a wireless communication transceiver, etc. The communication unit 19 allows the electronic device to exchange information / data with other devices through a computer network such as the Internet and / or various telecommunication networks.

[0141] The processor 11 may be a variety of general and / or special processing components with processing and computing capabilities. Some examples of the processor 11 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various special artificial intelligence (AI) computing chips, various processors running machine learning model algorithms, a digital signal processor (DSP), and any appropriate processor, controller, microcontroller, etc. The processor 11 executes the various methods and processes described above, such as a method for generating a power scheduling task.

[0142] In some embodiments, the power dispatching task generation method may be implemented as a computer program, which is tangibly contained in a computer-readable storage medium, such as a storage unit 18. In some embodiments, part or all of the computer program may be loaded and / or installed on the electronic device via the ROM 12 and / or the communication unit 19. When the computer program is loaded into the RAM 13 and executed by the processor 11, one or more steps of the power dispatching task generation method described above may be performed. Alternatively, in other embodiments, the processor 11 may be configured to perform the power dispatching task generation method in any other appropriate manner (e.g., by means of firmware).

[0143] 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.

[0144] 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.

[0145] 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.

[0146] 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).

[0147] 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.

[0148] 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.

[0149] An embodiment of the present invention further provides a computer program product, including a computer program, which, when executed by a processor, implements the power scheduling task generation method provided in any embodiment of the present application.

[0150] In the process of implementation, the computer program product can be written in one or more programming languages ​​or a combination thereof to perform the computer program code of the present invention, including object-oriented programming languages, such as Java, Smalltalk, C++, and conventional procedural programming languages, such as "C" language or similar programming languages. The program code can be executed entirely on the user's computer, partially on the user's computer, as a separate software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In the case of a remote computer, the remote computer can be connected to the user's computer through any type of network, including a local area network (LAN) or a wide area network (WAN), or can be connected to an external computer (for example, using an Internet service provider to connect through the Internet).

[0151] 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.

[0152] 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 generating a power dispatching task, characterized in that: include: Acquire current operating data of the target power grid, wherein the operating data includes operating parameters of each power device in the target power grid, wherein the operating parameters include current, voltage, power, and operating status; Extracting global operation characteristics representing global operation information of the target power grid from the operation data; For each of the electric power equipment, extracting local features of a neighborhood with the electric power equipment as a topological center from the operation data; Fusion of the global operation feature and the local feature of each of the electric power devices to obtain a fusion feature; A power dispatching task is generated based on the fusion features, and the power dispatching task includes a dispatching operation of each of the power devices.

2. The method for generating power dispatching tasks according to claim 1, characterized in that: Extracting a global operation feature representing global operation information of the target power grid from the operation data includes: Performing embedding processing on the operation data to obtain a first embedding feature represented by a vector of the operation data; Performing linearization, batch normalization, and nonlinear activation on the first embedded features to obtain a first regularized feature; Encoding the first regularized feature based on a self-attention mechanism to obtain an encoded feature; The encoding features are globally pooled to obtain global operation features representing global operation information of the target power grid.

3. The method for generating power dispatching tasks according to claim 2, characterized in that: The first regularized feature is encoded based on the self-attention mechanism to obtain an encoded feature, including: Inputting the first regularized feature into a self-attention module for processing, the self-attention module includes a plurality of cascaded self-attention units, the input feature of the first self-attention unit is the first regularized feature, the input feature of the i-th self-attention unit is the output feature of the i-1-th self-attention unit, each of the self-attention units is used to output attention features of different feature scales, and i is a positive integer greater than 1; Concatenate multiple attention features of different feature scales in the spatial dimension to obtain concatenated features; The concatenated features are linearized, batch normalized, and nonlinearly activated to obtain encoding features.

4. The method for generating a power dispatching task according to any one of claims 1 to 3, characterized in that: For each of the electric power equipment, extracting local features of a neighborhood with the electric power equipment as a topological center from the operation data includes: For each of the electric power devices, determining a neighboring device with the electric power device as a topological center; Filtering the operation data corresponding to the neighboring device from the operation data as the neighboring operation data; Embedding the neighborhood operation data to obtain a second embedding feature represented by a vector of the neighborhood operation data; Performing linearization, batch normalization, and nonlinear activation on the second embedded feature to obtain a second regularized feature; The second regularized feature is subjected to convolution processing to obtain a local feature of a neighborhood with the power device as the topological center.

5. The method for generating a power dispatching task according to any one of claims 1 to 3, characterized in that: Generating a power dispatching task based on the fusion feature includes: Performing linearization processing on the fused features to obtain linearized features; The linearization feature is mapped to the power dispatching task in the instruction library of the power dispatching task to generate the power dispatching task.

6. The method for generating a power dispatching task according to any one of claims 1 to 3, characterized in that: After the power dispatching task is generated based on the fusion feature, the method further includes: The dispatching operation of each device in the generated power dispatching task is verified based on the pre-built rule engine.

7. The method for generating a power dispatching task according to any one of claims 1 to 3, characterized in that: After the power dispatching task is generated based on the fusion feature, the method further includes: After executing the power dispatching task, collecting feedback data fed back by each of the power devices; The feedback data is used as training data to optimize model parameters of the power dispatch task generation model.

8. A power dispatching task generating device, characterized in that: include: An operation data acquisition module is used to acquire the current operation data of the target power grid, wherein the operation data includes the operation parameters of each power device in the target power grid, and the operation parameters include current, voltage, power, and operation status; A global feature extraction module, used to extract global operation features representing global operation information of the target power grid from the operation data; A local feature extraction module, for extracting, for each of the electric power equipment, local features of a neighborhood with the electric power equipment as a topological center from the operation data; A feature fusion module, used for fusing the global operation feature and the local feature of each of the electric power equipment to obtain a fusion feature; A scheduling task generation module is used to generate a power scheduling task based on the fusion features, and the power scheduling task includes a scheduling operation of each of the power devices.

9. An electronic device, characterized in that: include: one or more processors; A storage device for storing one or more programs; When the one or more programs are executed by the one or more processors, the one or more processors implement the power dispatching task generation method as described in any one of claims 1-7.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the program is executed by a processor, the power dispatching task generation method as described in any one of claims 1 to 7 is implemented.

Images