Microgrid Reactive Power Prediction Method, Device, Electronic Equipment and Storage Medium

Abstract

The microgrid reactive power prediction method, device, electronic device and storage medium provided by the embodiments of the present disclosure. The method includes collecting target grid data; the target grid data includes the output voltage data of each node of the microgrid, shunt susceptance, the output voltage data of the power supply end and the corresponding time tags; calculating the instantaneous reactive power values of each node of the microgrid according to the target grid data; generating a target change curve corresponding to each node of the microgrid based on the instantaneous reactive power values and the corresponding time tags, and using the target change curve corresponding to each node of the microgrid as a sample to be tested; inputting the sample to be tested into a pre-trained microgrid reactive power prediction model, outputting a corresponding reactive power prediction identification result, and adjusting the microgrid operation control strategy based on the output reactive power prediction identification result. In this way, the accuracy and speed of the microgrid reactive power prediction result are improved.

Description

Technical Field

[0001] The present disclosure relates to the technical field of digital twins in power systems, and particularly to the technical fields of microgrid reactive power prediction methods, devices, electronic devices, and storage media. Background Art

[0002] In the power grid, there are two types of electric power supplied by the power source to the load: one is active power, and the other is reactive power. Active power is the electric power required to keep the electrical equipment running normally, that is, the electric power that converts electrical energy into other forms of energy (mechanical energy, light energy, heat energy). For example, a 5.5 kW motor converts 5.5 kW of electric power into mechanical energy to drive a water pump to pump water or a threshing machine to thresh; various lighting devices convert electrical energy into light energy for people's living and working lighting. Reactive power is relatively abstract. It is used for the electric and magnetic fields in the circuit and is used to establish and maintain the magnetic field in electrical equipment. Any electrical equipment with an electromagnetic coil needs to consume reactive power to establish a magnetic field. For example, in addition to the 40 W of active power required for a 40 W fluorescent lamp to emit light (the ballast also consumes a part of the active power), about 80 var of reactive power is also required for the coil of the ballast to establish an alternating magnetic field. Since it does not do work externally, it is called "reactive".

[0003] Reactive power is an important operating parameter of the microgrid and has a certain impact on the operating stability of the microgrid, mainly manifested as: (1) reducing the output of the generator's active power; (2) when the apparent power is constant, increasing the reactive power will reduce the power supply capacity of the transmission and transformation equipment; (3) the flow of reactive power in the power grid will cause an increase in line voltage loss and power loss; (4) when the system lacks reactive power, it will cause low power factor operation and voltage drop, so that the capacity of electrical equipment cannot be fully utilized. However, the microgrid is divided into four parts: distributed power sources, power electronic conversion equipment, distributed energy storage, and loads, and the causes of its reactive power generation are relatively complex. Moreover, the microgrid is divided into two operating modes: grid-connected and islanded, and the magnitude of reactive power has a great influence on the selection of the microgrid operating mode. The traditional prediction algorithms for reactive power in the microgrid are difficult to implement. Summary of the Invention

[0004] The present disclosure provides a microgrid reactive power prediction method, device, equipment, and storage medium.

[0005] According to a first aspect of the present disclosure, a microgrid reactive power prediction method is provided. The method includes:

[0006] Collecting target grid data; the target grid data includes the output voltage data of each node of the microgrid, the shunt susceptance, the output voltage data of the power supply end, and the corresponding time tags;

[0007] Calculate the instantaneous reactive power values of each node of the microgrid based on the target power grid data;

[0008] Generate the target change curves corresponding to each node of the microgrid based on the instantaneous reactive power values and the corresponding time tags, and use the target change curves corresponding to each node of the microgrid as the samples to be measured;

[0009] Input the samples to be measured into the pre-trained microgrid reactive power prediction model, output the corresponding reactive power prediction identification results, and adjust the microgrid operation control strategy based on the output reactive power prediction identification results.

[0010] Furthermore, the generation process of the training samples of the microgrid reactive power prediction model includes:

[0011] Collect historical power grid data; the historical power grid data includes the output voltage data of each node of the microgrid, shunt susceptance, the output voltage data of the power supply end, and the corresponding time tags;

[0012] Calculate the historical reactive power values of each node of the microgrid based on the historical power grid data;

[0013] Generate the historical change curves corresponding to each node of the microgrid based on the historical reactive power values and the corresponding time tags;

[0014] Use the historical change curves corresponding to each node of the microgrid as training samples.

[0015] Furthermore, the identification process of the samples of the microgrid reactive power prediction model includes:

[0016] Obtain economic cost data; the economic cost data includes local average electricity price, total power generation, and microgrid cost data;

[0017] Calculate the ratio of the microgrid cost data to the local average electricity price;

[0018] Calculate the difference between the total power generation and the ratio to obtain the critical reactive power;

[0019] Identify the training samples based on the ratio relationship between the historical reactive power values and the critical reactive power.

[0020] Furthermore, the training process of the microgrid reactive power prediction model includes:

[0021] Perform data segmentation on the identified training samples to obtain a training set, a validation set, and a test set;

[0022] Train the self-attention time convolutional neural network model based on the training set, validation set, and test set, and output the corresponding reactive power prediction results. When the difference between the output result and the identification result is greater than the first preset threshold, correct the parameters of the self-attention time convolutional neural network model;

[0023] Repeat the above process until the difference between the output result and the identification result is less than the first preset threshold, output the trained self-attention time convolutional neural network model, and obtain the microgrid reactive power prediction model.

[0024] Further, before inputting the sample to be measured into the pre-trained microgrid reactive power prediction model, it includes: performing phase space transformation processing on the sample to be measured using chaos theory.

[0025] Further, the calculation formula for obtaining the instantaneous reactive power value of each node of the microgrid according to the target grid data is as follows:

[0026]

[0027] Where: W i represents the shunt susceptance on node i, U i represents the voltage on node i, U ter represents the real-time output voltage of the power supply end in the microgrid, and the calculation result P i represents the instantaneous reactive power value.

[0028] Further, the adjustment of the microgrid operation control strategy based on the output reactive power prediction identification result includes:

[0029] If the output reactive power prediction identification result is the first preset identification, issue a normal operation instruction;

[0030] If the output reactive power prediction identification result is the second preset identification, issue an instruction to adjust the microgrid operation control strategy, so that the negative feedback control system performs autonomous regulation of the microgrid according to the instruction;

[0031] If the output reactive power prediction identification result is the third preset identification, issue an island operation instruction.

[0032] According to the second aspect of the present disclosure, a microgrid reactive power prediction device is provided. The device includes:

[0033] A data acquisition module for acquiring target grid data; the target grid data includes output voltage data of each node of the microgrid, shunt susceptance, output voltage data of the power supply end, and corresponding time tags;

[0034] A data calculation module, configured to calculate the instantaneous reactive power values of each node of the microgrid according to the target grid data;

[0035] A to-be-tested sample generation module, configured to generate a target change curve corresponding to each node of the microgrid based on the instantaneous reactive power value and the corresponding time tag, and use the target change curve corresponding to each node of the microgrid as a to-be-tested sample;

[0036] A prediction result output module, configured to input the to-be-tested sample into a pre-trained microgrid reactive power prediction model, output a corresponding reactive power prediction identification result, and adjust the microgrid operation control strategy based on the output reactive power prediction identification result.

[0037] According to a third aspect of the present disclosure, an electronic device is provided. The electronic device includes: a memory and a processor, where a computer program is stored on the memory, and when the processor executes the program, the method as described above is implemented.

[0038] According to a fourth aspect of the present disclosure, a computer-readable storage medium is provided, on which a computer program is stored, and when the program is executed by a processor, the method as described in the first aspect of the present disclosure is implemented.

[0039] The microgrid reactive power prediction method, device, electronic device and storage medium provided by the embodiments of the present disclosure. Using chaos theory to perform phase space conversion on relevant data, highlighting the potential features in the data, and then inputting the preprocessed data into a self-attention time convolutional neural network to generate a microgrid reactive power prediction model, so as to realize the prediction of the reactive power of the microgrid. It can solve the problem that it is difficult to predict reactive power by traditional prediction algorithms, and improve the accuracy and speed of the microgrid reactive power prediction results.

[0040] It should be understood that the content described in the summary of the invention section is not intended to limit the key or important features of the embodiments of the present disclosure, nor is it used to limit the scope of the present disclosure. Other features of the present disclosure will become easily understood through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] Combined with the drawings and referring to the following detailed description, the above and other features, advantages and aspects of the embodiments of the present disclosure will become more obvious. The drawings are used to better understand the solution and do not constitute a limitation to the present disclosure. In the drawings, the same or similar reference numerals represent the same or similar elements, where:

[0042] Figure 1 The flowchart of the microgrid reactive power prediction method according to the embodiment of the present disclosure is shown;

[0043] Figure 2The flowchart of a reactive power prediction method for a microgrid according to another embodiment of the present disclosure is shown;

[0044] Figure 3 The specific structural diagram of a self-attention time convolutional neural network according to another embodiment of the present disclosure is shown;

[0045] Figure 4 The unit structural diagram of a self-attention time convolutional neural network according to another embodiment of the present disclosure is shown;

[0046] Figure 5 The block diagram of a reactive power prediction device for a microgrid according to an embodiment of the present disclosure is shown;

[0047] Figure 6 The block diagram of an exemplary electronic device capable of implementing the embodiments of the present disclosure is shown. Detailed implementation manners

[0048] To make the objectives, technical solutions, and advantages of the embodiments of the present disclosure clearer, the technical solutions in the embodiments of the present disclosure will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present disclosure. Apparently, the described embodiments are some but not all of the embodiments of the present disclosure. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present disclosure without creative efforts shall fall within the protection scope of the present disclosure.

[0049] In addition, the term "and / or" herein is only a relationship describing associated objects, indicating that three relationships may exist. For example, A and / or B may represent: A exists alone, A and B exist simultaneously, and B exists alone. In addition, the character " / " herein generally represents an "or" relationship between the associated objects before and after.

[0050] Figure 1 The flowchart of a reactive power prediction method 100 for a microgrid according to an embodiment of the present disclosure is shown. The method 100 includes:

[0051] Step 110, collecting target power grid data; the target power grid data includes the output voltage data of each node of the microgrid, shunt susceptance, the output voltage data of the power supply end, and the corresponding time tags.

[0052] In some embodiments, the output voltage of each node of the microgrid to be predicted, various parameters, and the data of the power supply end are collected according to a preset time interval, sampling duration, and sampling frequency. For example, for predicting the current day, six samplings can be performed, at 0:00, 4:00, 8:00, 12:00, 16:00, and 20:00 of the current day respectively. The sampling duration is 1 minute, and the sampling frequency is 1024 Hz.

[0053] Step 120: Calculate the instantaneous reactive power values of each node in the microgrid based on the target grid data.

[0054] In some embodiments, based on the data collected in step 110, calculate the instantaneous reactive power values of each node in the microgrid. Among them, the calculation formula for the transient reactive power value of any node in the microgrid is as follows:

[0055]

[0056] Where: W i represents the shunt susceptance at node i, U i represents the voltage at node i, U ter represents the real-time output voltage of the power supply end in the microgrid, and the calculation result P i represents the instantaneous reactive power value.

[0057] Step 130: Generate the target change curves corresponding to each node in the microgrid based on the instantaneous reactive power values and the corresponding time tags, and use the target change curves corresponding to each node in the microgrid as the samples to be measured.

[0058] In some embodiments, based on the instantaneous reactive power values of each node in the microgrid calculated in step 120 and the corresponding time tag information, obtain the curves of the instantaneous reactive power changing with time corresponding to each node in the microgrid, and use the obtained curves as the samples to be measured.

[0059] Step 140: Input the samples to be measured into a pre-trained microgrid reactive power prediction model, output the corresponding reactive power prediction identification result, and adjust the microgrid operation control strategy based on the output reactive power prediction identification result.

[0060] In some embodiments, according to the samples to be measured obtained in step 130, input them into a pre-trained microgrid reactive power prediction model, output the corresponding reactive power prediction identification result, and then adjust the microgrid operation control strategy based on the output reactive power prediction identification result. For example, the reactive power prediction identification result can be label 0, label 1, or label 2.

[0061] In some embodiments, the adjustment of the microgrid operation control strategy based on the output reactive power prediction identification result specifically includes the following situations:

[0062] Situation 1: If the output reactive power prediction identification result is label 0, issue a normal operation instruction.

[0063] Situation 2: If the output reactive power prediction identification result is label 1, issue an instruction to adjust the microgrid operation control strategy, so that the negative feedback control system performs autonomous regulation of the microgrid according to the instruction.

[0064] Case 3: If the predicted reactive power identification result is Tag 2, an islanding operation instruction is issued.

[0065] In some embodiments, when the predicted reactive power identification result is Tag 0, it means that the reactive power is within the range of 0 to 0.2*w, and at this time, the microgrid can continue to operate normally; when the predicted reactive power identification result is Tag 1, it means that the reactive power is within the range of 0.2*w p to 0.8*w p ~0.8*w p In this range, the negative feedback control system needs to be started for autonomous regulation of the microgrid; when the predicted reactive power identification result is Tag 2, it means that the reactive power is greater than 0.8*w p , and at this time, the microgrid needs to perform islanding operation and conduct manual maintenance.

[0066] According to the embodiments of the present disclosure, the following technical effects are achieved:

[0067] The microgrid reactive power prediction method provided by the embodiments of the present disclosure uses chaos theory to perform phase space conversion on relevant data to highlight potential features in the data, and then inputs the preprocessed data into the self-attention time convolutional neural network to generate a microgrid reactive power prediction model, realizing the prediction of the reactive power of the microgrid. It can solve the problem of difficult reactive power prediction by traditional prediction algorithms, and improve the accuracy and speed of the microgrid reactive power prediction results.

[0068] Based on the above implementation manner, the flowchart of the training process 200 of the microgrid reactive power prediction model provided in another implementation manner of the present disclosure is as Figure 2 shown, and includes the following steps:

[0069] Step 210, perform data segmentation on the labeled training samples to obtain a training set, a validation set, and a test set.

[0070] Step 220, train the self-attention time convolutional neural network model based on the training set, the validation set, and the test set, output the corresponding reactive power prediction result, and when the difference degree between the output result and the labeled result is greater than the first preset threshold, correct the parameters of the self-attention time convolutional neural network model.

[0071] Step 230, repeat the above process until the difference degree between the output result and the labeled result is less than the first preset threshold, output the trained self-attention time convolutional neural network model, and obtain the microgrid reactive power prediction model.

[0072] In some embodiments, the labeled training samples are subjected to data segmentation to divide the training set, validation set, and test set of the model. The data of the first nine months of each year is used as the training set, the data of the tenth month is used as the validation set, and the data of the eleventh and twelfth months is used as the test set. The training set is input into the self-attention time convolutional neural network model to train the model. During the model iteration process, the validation set is input into the model every ten iterations to verify the iteration situation of the model. At the same time, when the number of iterations is greater than 200, on the basis of inputting the validation set every ten iterations, the validation set is randomly input into the model multiple times to verify the iteration situation. When the accuracy rate of the model is greater than a certain threshold B, the iteration is stopped, and the threshold B is selected as 0.99. The test set is input into the model to verify the reliability and robustness of the model. When the output accuracy rate is greater than the threshold 0.99 when the test set is input into the model, the model is feasible and is output as the power generation prediction model of the distributed power source.

[0073] In some embodiments, the attention mechanism is set to 3 layers. In the time convolutional neural network, each layer of the time convolutional layer is stacked with five layers of convolution. The dilation coefficient can be used to expand the field of view to 1024 data. The network depth is set to 8, indicating that in this model, eight layers of time convolutional neural network units are connected in series as the reactive power prediction model, and its output is the prediction result. The data division of the model makes full use of the data randomness, strengthens the robustness of the model, and enhances the reliability of the model under complex data and improves the model training rate.

[0074] Figure 3 The specific structure of the self-attention time convolutional neural network is shown. To fully utilize the spatio-temporal information in the data after phase space transformation, the self-attention time convolutional neural network model is used for prediction. There are 3 layers in total for the self-attention layer. The first layer is the input layer, which is directly input with the data after phase space transformation. The length of each layer is the same as the original data length, and the dimension of each layer is the same as the data dimension, which is (dMAX + 1) - dimensional.

[0075] Figure 4 The structure of the self-attention time convolutional neural network unit is shown. The time convolutional neural network relies on dilated causal convolution. The dilation of the convolutional layer can be regarded as the distance between the elements of the input sequence. As Figure 4 shown, the convolutional kernel adopted is 4, so the diffusion degree of this time convolutional network model is 4, and the visual range of each value at the output end is 1024. The depth of the time convolutional neural network is 8, and 8 layers of time convolutional neural network layers are stacked with each other to form a time convolutional neural network model for life prediction. The input data of its input layer is the data after the self-attention layer. The length of its input layer is the original data length, and the dimension is (dMAX + 1). The output after passing through the time convolutional neural network layer is the predicted value.

[0076] Based on the above embodiments, the generation process of the training samples of the microgrid reactive power prediction model provided in another embodiment of the present disclosure is as follows:

[0077] Collect historical power grid data; the historical power grid data includes the output voltage data of each node of the microgrid, shunt susceptance, the output voltage data of the power supply end, and the corresponding time tags;

[0078] Calculate the historical reactive power values of each node of the microgrid according to the historical power grid data;

[0079] Generate the historical change curves corresponding to each node of the microgrid based on the historical reactive power values and the corresponding time tags;

[0080] Use the historical change curves corresponding to each node of the microgrid as training samples.

[0081] In some embodiments, collect the output voltage and various parameters of each node of the microgrid and the data of the power supply end from 2019 to 2021 for a total of three years. Sample six times a day, at 0:00, 4:00, 8:00, 12:00, 16:00, and 20:00 every day. The sampling duration is 1 minute, and the sampling frequency is 1024 Hz. Calculate the instantaneous reactive power value of each node according to the collected data, and generate the curve of the instantaneous reactive power change with time corresponding to each node of the microgrid as the sample for model training.

[0082] In some embodiments, the calculation formula for the transient reactive power value of each node in the microgrid is as follows:

[0083]

[0084] Where: W i represents the shunt susceptance on node i, U i represents the voltage on node i, U ter represents the real-time output voltage of the power supply end in the microgrid, and the calculation result P i represents the instantaneous reactive power value.

[0085] Based on the above embodiments, the identification process of the samples of the microgrid reactive power prediction model provided in another embodiment of the present disclosure is as follows:

[0086] Obtain economic cost data; the economic cost data includes the local average electricity price, total power generation, and microgrid cost data;

[0087] Calculate the ratio of the microgrid cost data to the local average electricity price;

[0088] Calculate the difference between the total power generation and the ratio to obtain the critical reactive power;

[0089] Identify the training samples based on the ratio relationship between the historical reactive power value and the critical reactive power.

[0090] In some embodiments, a label is established by calculating the economic cost, and the samples are identified according to the label. The calculation formula is as follows:

[0091] p e (w0 - w p ) = p co

[0092] Where p e represents the local average electricity price, w0 is the total power generation, w p is the critical reactive power, and p co represents the microgrid cost.

[0093] In some embodiments, after calculating the critical reactive power, a label is established. When the reactive power is in the range of 0 to 0.2*w p , label 0 is given; when the reactive power is in the range of 0.2*w p to 0.8*w p , label 1 is given; when the reactive power is greater than 0.8*w p , label 2 is given. For example, if the critical reactive power is 1000 kw and the instantaneous reactive power value is 200 kwh, it indicates that when it is in the range of 0 to 0.2*w p , label 0 is given.

[0094] The microgrid reactive power prediction method provided by the embodiments of the present disclosure uses the chaos theory to perform phase space transformation on relevant data, highlighting the potential features in the data and improving the accuracy of the microgrid reactive power prediction results. Then, the preprocessed data is input into the self-attention time convolutional neural network to generate a microgrid reactive power prediction model. The sample to be measured is input into the trained model to output the identification result, and the accurate prediction of the microgrid reactive power can be quickly realized according to the identification result.

[0095] Based on the above implementation manner, before inputting the sample to be measured into the pre-trained microgrid reactive power prediction model provided by another implementation manner of the present disclosure, it includes: performing phase space transformation processing on the sample to be measured using the chaos theory.

[0096] In some embodiments, the autocorrelation function R(t) of each one-dimensional time-varying data is calculated, and the value approximately 1 / e times of R(0) is selected as the delay time. The overall false nearest neighbor value E(d) of the time series is calculated using the Cao's method, and then the appropriate embedding dimension d is selected according to the time series proximity value ratio E1(d). To ensure the dimensional unity of the instantaneous reactive power of each node input to the neural network, the largest embedding dimension dMAX in the data is selected for data padding, so that all data dimensions reach (dMAX + 1) dimensions.

[0097] The microgrid reactive power prediction method provided by the embodiments of the present disclosure, for the relatively complex operating conditions of the microgrid and the many components of reactive power, if the collected data is directly used to predict reactive power, due to the influence of relatively complex situations, it will have a greater impact on the results. To make full use of the feature information hidden in the data, the chaos theory is first used to perform phase space transformation on the data, and the data after phase space transformation is then input into the neural network, which enhances the accuracy and robustness of network prediction.

[0098] It should be noted that for the foregoing method embodiments, for the sake of simple description, they are all expressed as a series of action combinations. However, those skilled in the art should know that the present disclosure is not limited by the described action sequence, because according to the present disclosure, certain steps can be performed in other sequences or simultaneously. Secondly, those skilled in the art should also know that the embodiments described in the specification are all optional embodiments, and the actions and modules involved are not necessarily essential to the present disclosure.

[0099] The above is the introduction of the method embodiments. The following further illustrates the solution of the present disclosure through device embodiments.

[0100] Figure 5 The block diagram of the microgrid reactive power prediction device 500 according to the embodiments of the present disclosure is shown. As Figure 5 shown, the device 500 includes:

[0101] A data acquisition module 510, configured to acquire target power grid data; the target power grid data includes the output voltage data of each node of the microgrid, the shunt susceptance, the output voltage data of the power supply end, and the corresponding time tags;

[0102] A data calculation module 520, configured to calculate the instantaneous reactive power value of each node of the microgrid according to the target power grid data;

[0103] A to-be-tested sample generation module 530, configured to generate a target change curve corresponding to each node of the microgrid based on the instantaneous reactive power value and the corresponding time tag, and use the target change curve corresponding to each node of the microgrid as a to-be-tested sample;

[0104] A prediction result output module 540 is configured to input the to-be-tested sample into a pre-trained microgrid reactive power prediction model, output a corresponding reactive power prediction identification result, and adjust the microgrid operation control strategy based on the output reactive power prediction identification result.

[0105] Those skilled in the art can clearly understand that for the convenience and brevity of description, the specific working processes of the described modules can refer to the corresponding processes in the foregoing method embodiments and will not be elaborated herein.

[0106] According to an embodiment of the present disclosure, the present disclosure also provides an electronic device, a readable storage medium, and a computer program product.

[0107] Figure 6 FIG. shows a schematic block diagram of an electronic device 600 that can be used to implement embodiments of the present disclosure. The electronic device is intended to represent various forms of digital computers, such as, for example, a laptop computer, a desktop computer, a workbench, a personal digital assistant, a server, a blade server, a mainframe computer, and other suitable computers. The electronic device can also represent various forms of mobile devices, such as, for example, a personal digital processor, a cellular phone, a smart phone, a wearable device, and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely illustrative and are not intended to limit the implementation of the present disclosure described and / or claimed herein.

[0108] The device 600 includes a computing unit 601, which can perform various appropriate actions and processes according to a computer program stored in a read-only memory (ROM) 602 or a computer program loaded from a storage unit 608 into a random access memory (RAM) 603. In the RAM 603, various programs and data required for the operation of the device 600 can also be stored. The computing unit 601, the ROM 602, and the RAM 603 are connected to each other via a bus 604. An input / output (I / O) interface 605 is also connected to the bus 604.

[0109] A plurality of components in the device 600 are connected to the I / O interface 605, including: an input unit 606, such as a keyboard, a mouse, etc.; an output unit 607, such as various types of displays, speakers, etc.; a storage unit 608, such as a magnetic disk, an optical disk, etc.; and a communication unit 609, such as a network card, a modem, a wireless communication transceiver, etc. The communication unit 609 allows the device 600 to exchange information / data with other devices through a computer network such as the Internet and / or various telecommunication networks.

[0110] The computing unit 601 can be various general-purpose and / or special-purpose processing components with processing and computing capabilities. Some examples of the computing unit 601 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various dedicated artificial intelligence (AI) computing chips, various computing units running machine learning model algorithms, a digital signal processor (DSP), and any suitable processor, controller, microcontroller, etc. The computing unit 601 executes the various methods and processes described above, such as method 100. For example, in some embodiments, method 100 may be implemented as a computer software program tangibly embodied in a machine-readable medium, such as the storage unit 608. In some embodiments, part or all of the computer program may be loaded and / or installed onto the device 600 via the ROM 602 and / or the communication unit 609. When the computer program is loaded into the RAM 603 and executed by the computing unit 601, one or more steps of the method 100 described above can be executed. Alternatively, in other embodiments, the computing unit 601 can be configured to execute method 100 in any other suitable manner (e.g., by means of firmware).

[0111] Various embodiments of the systems and techniques described above in this document can be implemented in digital electronic circuitry, integrated circuit systems, field-programmable gate arrays (FPGA), application-specific integrated circuits (ASIC), application-specific standard products (ASSP), systems-on-chip (SOC), complex programmable logic devices (CPLD), computer hardware, firmware, software, and / or combinations thereof. These various embodiments 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 dedicated or general-purpose programmable processor that receives data and instructions from a storage system, at least one input device, and at least one output device, and transmits the data and instructions to the storage system, the at least one input device, and the at least one output device.

[0112] The program code for implementing the methods of the present disclosure can be written in any combination of one or more programming languages. These program codes can be provided to a processor or controller of a general-purpose computer, a special-purpose computer, or other programmable data processing device, such that when the program codes are executed by the processor or controller, the functions / operations specified in the flowcharts and / or block diagrams are implemented. The program codes can be executed entirely on the machine, partially on the machine, as a stand-alone software package partially on the machine and partially on a remote machine, or entirely on a remote machine or server.

[0113] In the context of this disclosure, a machine-readable medium can be a tangible medium that can contain or store a program for use by or in connection with an instruction execution system, apparatus, or device. A machine-readable medium can be a machine-readable signal medium or a machine-readable storage medium. A machine-readable medium can include, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatus, or devices, or any suitable combination of the foregoing. More specific examples of a machine-readable storage medium would include an electrical connection based on one or more wires, a portable computer diskette, 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 disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

[0114] To provide for interaction with a user, the systems and techniques described herein can be implemented on a computer 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 a trackball) by which the user can provide input to the computer. Other kinds of devices can also be used to provide for interaction with the user; for example, feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including acoustic, speech, or tactile input).

[0115] The systems and techniques described herein can be implemented in a computing system that includes back-end 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 front-end components (e.g., a user computer having a graphical user interface or a web browser through which the user can interact with an implementation of the systems and techniques described herein), or a computing system that includes any combination of such back-end components, middleware components, or front-end components. The components of the system can be interconnected by any form or medium of digital data communication (e.g., a communication network). Examples of a communication network include: a local area network (LAN), a wide area network (WAN), and the Internet.

[0116] A computer system can include a client and a server. The client and the server are generally remote from each other and typically interact through a communication network. The relationship of client and server arises by virtue of computer programs running on the respective computers and having a client-server relationship to each other. The server can be a cloud server, a server of a distributed system, or a server incorporating a blockchain.

[0117] It should be understood that the various forms of processes shown above can be used, with steps reordered, added or deleted. For example, the steps recited in this disclosure can be executed in parallel, sequentially or in a different order, as long as the desired results of the technical solutions disclosed in this disclosure can be achieved, and no limitation is imposed herein.

[0118] The above specific embodiments do not constitute a limitation on the protection scope of this disclosure. Those skilled in the art should understand that various modifications, combinations, sub - combinations and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions and improvements made within the spirit and principles of this disclosure shall be included within the protection scope of this disclosure.

Claims

1. A method for predicting the reactive power of a microgrid, characterized in that, Including: Collecting target power grid data; The target power grid data includes output voltage data of each node of the microgrid, shunt susceptance, output voltage data of the power supply end, and corresponding time tags; Calculating the instantaneous reactive power value of each node of the microgrid based on the target power grid data; Generating a corresponding target change curve for each node of the microgrid based on the instantaneous reactive power value and the corresponding time tag, and using the corresponding target change curve of each node of the microgrid as a sample to be tested; Inputting the sample to be tested into a pre-trained microgrid reactive power prediction model, outputting a corresponding reactive power prediction identification result, and adjusting the microgrid operation control strategy based on the output reactive power prediction identification result; wherein, The generation process of the training sample of the microgrid reactive power prediction model includes: Collecting historical power grid data; the historical power grid data includes output voltage data of each node of the microgrid, shunt susceptance, output voltage data of the power supply end, and corresponding time tags; Calculating the historical reactive power value of each node of the microgrid based on the historical power grid data; Generating a corresponding historical change curve for each node of the microgrid based on the historical reactive power value and the corresponding time tag; Using the corresponding historical change curve of each node of the microgrid as a training sample; The identification process of the sample of the microgrid reactive power prediction model includes: Obtaining economic cost data; the economic cost data includes local average electricity price, total power generation, and microgrid cost data; Calculating the ratio of the microgrid cost data to the local average electricity price; Calculating the difference between the total power generation and the ratio to obtain the critical reactive power; Identifying the training sample based on the ratio relationship between the historical reactive power value and the critical reactive power; The training process of the microgrid reactive power prediction model includes: Performing data segmentation on the identified training sample to obtain a training set, a validation set, and a test set; Training a self-attention time convolutional neural network model based on the training set, the validation set, and the test set, outputting a corresponding reactive power prediction result, and correcting the parameters of the self-attention time convolutional neural network model when the difference degree between the output result and the identification result is greater than a first preset threshold; Repeating the above process until the difference degree between the output result and the identification result is less than the first preset threshold, outputting the trained self-attention time convolutional neural network model, and obtaining the microgrid reactive power prediction model.

2. The method according to claim 1, characterized in that, Before inputting the sample to be tested into a pre-trained microgrid reactive power prediction model, it includes: performing phase space transformation processing on the sample to be tested using chaos theory.

3. The method according to claim 1, characterized in that, The calculation formula for calculating the instantaneous reactive power value of each node of the microgrid based on the target power grid data is as follows: Where: W i represents the shunt susceptance at node i, U i represents the voltage at node i, U ter represents the real-time output voltage at the power supply end in the microgrid, and the calculation result P i represents the instantaneous reactive power value.

4. The method according to claim 1, characterized in that, The adjusting of the microgrid operation control strategy based on the output reactive power prediction identification result includes: If the output reactive power prediction identification result is a first preset identification, sending a normal operation instruction; If the output reactive power prediction identification result is a second preset identification, sending an instruction to adjust the microgrid operation control strategy, so that the negative feedback control system performs autonomous regulation of the microgrid according to the instruction; If the predicted reactive power prediction identification result is the third preset identification, an islanding operation instruction is issued.

5. A device for predicting the reactive power of a microgrid, characterized in that, Including: A data acquisition module for acquiring target power grid data; The target power grid data includes output voltage data of each node of the microgrid, shunt susceptance, output voltage data of the power supply end, and corresponding time tags; A data calculation module for calculating the instantaneous reactive power value of each node of the microgrid based on the target power grid data; A to-be-tested sample generation module for generating a target change curve corresponding to each node of the microgrid based on the instantaneous reactive power value and the corresponding time tag, and using the target change curve corresponding to each node of the microgrid as a to-be-tested sample; A prediction result output module for inputting the to-be-tested sample into a pre-trained microgrid reactive power prediction model, outputting a corresponding reactive power prediction identification result, and adjusting the microgrid operation control strategy based on the output reactive power prediction identification result; wherein, The generation process of the training sample of the microgrid reactive power prediction model includes: Collecting historical power grid data; the historical power grid data includes output voltage data of each node of the microgrid, shunt susceptance, output voltage data of the power supply end, and corresponding time tags; Calculating the historical reactive power value of each node of the microgrid based on the historical power grid data; Generating a historical change curve corresponding to each node of the microgrid based on the historical reactive power value and the corresponding time tag; Using the historical change curve corresponding to each node of the microgrid as a training sample; The identification process of the sample of the microgrid reactive power prediction model includes: Obtaining economic cost data; the economic cost data includes local average electricity price, total power generation, and microgrid cost data; Calculating the ratio of the microgrid cost data to the local average electricity price; Calculating the difference between the total power generation and the ratio to obtain the critical reactive power; Identifying the training sample based on the ratio relationship between the historical reactive power value and the critical reactive power; The training process of the microgrid reactive power prediction model includes: Performing data segmentation on the identified training sample to obtain a training set, a validation set, and a test set; Training a self-attention time convolutional neural network model based on the training set, the validation set, and the test set, outputting a corresponding reactive power prediction result, and correcting the parameters of the self-attention time convolutional neural network model when the difference degree between the output result and the identification result is greater than a first preset threshold; Repeating the above process until the difference degree between the output result and the identification result is less than the first preset threshold, outputting the trained self-attention time convolutional neural network model, and obtaining the microgrid reactive power prediction model.

6. An electronic device, characterized in that, Including: At least one processor; And A memory communicatively connected to the at least one processor; wherein, The memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can execute the method according to any one of claims 1-4.

7. A non-transitory computer-readable storage medium storing computer instructions, characterized in that, The computer instructions are used to cause the computer to execute the method according to any one of claims 1-4.

Images