Microgrid Dynamic Disturbance Analysis System and Method

Abstract

This disclosure provides a microgrid dynamic disturbance analysis system and method. The microgrid dynamic disturbance analysis system includes: a detection and monitoring unit configured to collect operational data of the microgrid; a data platform configured to: calculate the dynamic stability parameters of the microgrid system based on a microgrid mathematical model using the collected operational data, and predict the dynamic disturbance parameters of the microgrid system based on a neural network model; and visualize the dynamic stability parameters and dynamic disturbance parameters through a microgrid digital twin model and return microgrid control commands.

Description

Technical Field

[0001] This disclosure relates to the field of microgrids, and more specifically, to a microgrid dynamic disturbance analysis system and method. Background Technology

[0002] Renewable energy sources, such as renewable energy, are beneficial to achieving sustainable development and carbon neutrality goals. Microgrids include micro-sources, loads, energy storage systems, and control devices. Microgrids can operate in grid-connected or islanded mode and can increase the renewable energy penetration rate of the power system; however, systems primarily powered by wind and solar power have relatively low system inertia. Dynamic disturbances in microgrids are instantaneous or short-lived. Therefore, the analysis and prediction of dynamic disturbances in microgrids are crucial for mitigating them through control measures. Microgrid dynamic disturbance control systems are applied to microgrid clusters at the edge of the power system to address dynamic disturbances and grid connection impacts in microgrid systems primarily powered by renewable energy.

[0003] Traditional power systems typically handle transient and dynamic disturbances by disconnecting generating equipment or loads, or by adding compensation devices to the power system. These methods are reactive, taking action only after the disturbance occurs, without prior prediction or analysis, and are costly. Currently, analyzing the development of dynamic disturbances in microgrids using mathematical models of the power system is still in the offline analysis stage. Existing big data platforms suffer from long development cycles, high costs, limited storage options for massive amounts of data, and poor data throughput and real-time performance. Summary of the Invention

[0004] This disclosure proposes a big data cloud platform for microgrids to analyze and predict dynamic disturbance behavior online. Faced with the increasing volume of operational data from diverse microgrids, it proposes using a big data cloud platform to address issues such as limited computing and storage capacity in existing microgrid management systems, lack of data sharing between different microgrid systems, and the inability to fully utilize real-time and historical data from microgrid operations.

[0005] According to one aspect of this disclosure, a microgrid dynamic disturbance analysis system is provided, the microgrid dynamic disturbance analysis system comprising: a detection and monitoring unit configured to collect operating data of the microgrid; a data platform configured to: calculate dynamic stability parameters of the microgrid system based on a microgrid mathematical model using the collected operating data, and predict dynamic disturbance parameters of the microgrid system based on a neural network model; and visualize the dynamic stability parameters and dynamic disturbance parameters through a microgrid digital twin model and return microgrid control commands.

[0006] The inputs to the neural network model include wind power generation parameters, photovoltaic power generation parameters, and energy storage system parameters. The dynamic disturbance parameters predicted by the neural network model include disturbance oscillation frequency, maximum new energy power generation boundary, and minimum energy storage power boundary.

[0007] When the microgrid is connected to the grid, the input to the neural network model also includes regional grid system parameters, and the dynamic disturbance parameters predicted by the neural network model also include regional grid capacity assessment; when the microgrid is disconnected from the grid, the input to the neural network model also includes microgrid load parameters, and the dynamic disturbance parameters predicted by the neural network model also include load upper limit power boundary.

[0008] Microgrid parameters can be displayed and control commands can be received via mobile client or remote web interface.

[0009] The data platform performs data storage through a relational database, performs searches through a real-time database, and performs offline analysis through a hot database.

[0010] According to another aspect of this disclosure, a microgrid system is provided, the microgrid system including at least one of a wind turbine, a photovoltaic power generation module, a diesel generator, and an energy storage system, and the microgrid system further includes the microgrid dynamic disturbance analysis system as described above.

[0011] According to another aspect of this disclosure, a method for dynamic disturbance analysis of a microgrid is provided, the microgrid including a wind power generation system, a photovoltaic power generation system, and an energy storage system. The method includes: collecting operational data of the microgrid; using the collected operational data, calculating dynamic stability parameters of the microgrid system based on a microgrid mathematical model, and predicting dynamic disturbance parameters of the microgrid system based on a neural network model; and visualizing the dynamic stability parameters and dynamic disturbance parameters through a microgrid digital twin model and returning microgrid control commands.

[0012] The method includes: preprocessing the collected operational data, wherein the preprocessing includes at least one of cleaning, aggregation, dimensionality reduction, sampling, and clustering.

[0013] The inputs to the neural network model include wind power generation parameters, photovoltaic power generation parameters, and energy storage system parameters. The dynamic disturbance parameters predicted by the neural network model include disturbance oscillation frequency, maximum new energy power generation boundary, and minimum energy storage power boundary.

[0014] When the microgrid is connected to the grid, the input to the neural network model also includes regional grid system parameters, and the dynamic disturbance parameters predicted by the neural network model also include regional grid capacity assessment; when the microgrid is disconnected from the grid, the input to the neural network model also includes microgrid load parameters, and the dynamic disturbance parameters predicted by the neural network model also include load upper limit power boundary.

[0015] When a microgrid includes multiple wind power generation systems and multiple energy storage systems, the inputs to the neural network model also include the number of wind power generation systems, the wake effect of each wind power generation system, and the number of energy storage systems.

[0016] The neural network model includes an input layer, an output layer, and multiple hidden layers.

[0017] According to another aspect of this disclosure, a computer-readable storage medium storing a computer program is provided, which, when executed by a processor, implements the microgrid dynamic disturbance analysis method as described above.

[0018] According to another aspect of this disclosure, a computer device is provided, the computer device comprising: a processor; and a memory storing a computer program, which, when executed by the processor, implements the microgrid dynamic disturbance analysis method as described above.

[0019] This disclosure utilizes big data and digital technologies to perform detailed modeling and digital analysis of the dynamic disturbance behavior of microgrid groups, predicting in advance the impact of various electrical and non-electrical quantities on dynamic disturbance behavior. A big data platform provides massive data throughput capabilities and ensures data real-time performance. Attached Figure Description

[0020] The above and other aspects, features, and advantages of this disclosure will become more clearly understood from the following description, taken in conjunction with the accompanying drawings and specific embodiments, in which:

[0021] Figure 1 This is a schematic diagram of a microgrid dynamic disturbance analysis system according to an embodiment.

[0022] Figure 2 This is a schematic diagram of a microgrid dynamic disturbance analysis method according to an embodiment.

[0023] Figure 3 This is a schematic diagram of a microgrid dynamic disturbance neural network training model according to an embodiment.

[0024] Figure 4 and Figure 5 These are schematic diagrams of the dynamic disturbance neural network training models for microgrids during grid-connected and off-grid operation, respectively, according to the embodiments.

[0025] Figure 6This is a schematic diagram of a dynamic perturbation neural network training model for the joint operation of a wind power generation system and an energy storage system in a microgrid according to an embodiment.

[0026] Figure 7 This is a flowchart of the front-end and back-end data request process according to an embodiment.

[0027] Figure 8 This is the PostgreSQL database table structure of the big data platform according to the embodiment.

[0028] Figure 9 This is the basic graphics library architecture according to the embodiment.

[0029] Figure 10 This is the load prediction interface of a big data platform according to an embodiment.

[0030] Figure 11 This is a flowchart of the remote control process of a big data platform according to an embodiment.

[0031] Figure 12 It is a real-time data structure in ES according to the embodiment.

[0032] Figure 13 This is a flowchart illustrating the real-time data storage to ES data flow according to an embodiment. Detailed Implementation

[0033] The following detailed embodiments are provided to aid the reader in gaining a comprehensive understanding of the methods, apparatus, and / or systems described herein. However, various variations, modifications, and equivalents of the methods, apparatus, and / or systems described herein will be apparent upon understanding the disclosure of this application. For example, the order of operations described herein is merely illustrative and is not limited to the order set forth herein; changes may be made that will be apparent upon understanding the disclosure of this application, except for operations that must occur in a specific order. Furthermore, descriptions of features known in the art may be omitted for clarity and brevity. To enable those skilled in the art to better understand this disclosure, specific embodiments of the disclosure are described in detail below with reference to the accompanying drawings.

[0034] Technical terminology definition

[0035] Python: A cross-platform computer programming language that combines interpreted, compiled, interactive, and object-oriented scripting features.

[0036] Django: An open-source web application framework written in Python that uses the MTV framework pattern (i.e., Model M, Template T, and View V).

[0037] PostgreSQL: A feature-rich, free software object-relational database management system (ORDBMS).

[0038] Elastic Search (ES): A search server based on Lucene, using a RESTful web interface, providing a distributed, multi-user full-text search engine for cloud computing and real-time searching. It is stable, reliable, fast, and easy to install and use.

[0039] Redis (Remote Dictionary Server) is an open-source, ANSI C-written, network-enabled, in-memory or persistent log-structured key-value database with advantages such as high performance, strong scalability, and high availability.

[0040] React: A JavaScript library for building user interfaces and UIs, designed to solve the problem of building large-scale applications with data that changes over time.

[0041] The microgrid dynamic disturbance control system according to embodiments of this disclosure can be applied in microgrid clusters at the edge of power systems. Distributed power sources in microgrids include new energy power generation systems such as wind power systems, photovoltaic power systems, and energy storage systems. Most microgrids operate at voltage levels of 35kV and below. Building microgrid clusters at 10kV voltage levels can accommodate more loads and cater to a wider variety of electricity demands. Islanded operation or weak connection to the grid are hallmarks of microgrids. However, microgrid clusters have a high penetration rate of new energy sources. New energy power generation is mainly interconnected via power electronic devices with almost no inertial characteristics. Furthermore, new energy power generation, such as wind power and photovoltaic power generation, is intermittent, volatile, and uncertain, leading to unstable dynamic disturbances in the system. When microgrid clusters or incremental distribution networks are dominated by new energy power generation, it will affect the dynamic stability of the power system in edge areas and the difficulty of grid connection. Multi-energy interconnected networks formed by multiple interconnected microgrids also experience system dynamic disturbance problems when operating independently or with weak connections to the grid.

[0042] Traditional dynamic stability parameter analysis can be performed offline or online using mathematical modeling of generator sets. However, microgrid power system models exhibit strong nonlinearity, with large time and spatial scales, leading to increasingly prominent dynamic disturbance problems. When modeling distributed power sources, power systems, control systems, and dynamic disturbance induction mechanisms in microgrids, wind turbine models, involving numerous disciplines such as electrical engineering, mechanics, structure, fluid mechanics, and meteorology, cannot meet the practical requirements for establishing detailed and accurate traditional mathematical models.

[0043] Figure 1This is a schematic diagram of a microgrid dynamic disturbance analysis system according to an embodiment.

[0044] Reference Figure 1 The microgrid dynamic disturbance analysis system according to an embodiment of the present disclosure includes a detection and monitoring unit 110 and a data platform 120.

[0045] The detection and monitoring unit 110 collects operational data from the microgrid. The collected operational data includes wind power generation parameters, photovoltaic power generation parameters, and energy storage system parameters.

[0046] Data platform 120 uses collected operational data to establish a mathematical model of the traditional power system and calculates the dynamic stability parameters of the microgrid system based on the microgrid mathematical model, giving full play to the accuracy, descriptivity and computability of the mathematical model.

[0047] Meanwhile, the data platform 120 combines big data algorithms to predict the dynamic disturbance parameters of the microgrid system, employing digital methods for implicit instability factors (which cannot be precisely described by mathematical models) caused by dynamic disturbances. For example, the data platform 120 analyzes the digital correlation of dynamic disturbances caused by implicit instability factors in microgrid operating parameters based on a neural network model. For parts where precise mathematical models cannot be established due to the integration of renewable energy and the strong nonlinearity and power decoupling of power electronic devices, the data model is trained using a neural network model. This combines traditional precise mathematical models with big data models to establish dynamic disturbance analysis of the microgrid, thereby achieving digital reproduction and early prediction of disturbance problems. Here, the neural network model includes an input layer, an output layer, and multiple hidden layers. The inputs to the neural network model may include wind power generation parameters, photovoltaic power generation parameters, and energy storage system parameters, while the dynamic disturbance parameters predicted by the neural network model may include the disturbance oscillation frequency, the maximum renewable energy power generation boundary, and the minimum energy storage power boundary.

[0048] Data platform 120 visualizes dynamic stability parameters and dynamic disturbance parameters through a microgrid digital twin model and returns microgrid control commands. It describes the dynamic relationship between input and output through the correlation between data, thereby realizing the predictability, controllability and observability of microgrid system parameters.

[0049] In addition, the data platform 120 can perform data storage through a relational database, perform searches through a real-time database, and perform offline analysis through a hot database.

[0050] Alternatively, in the microgrid dynamic disturbance analysis system according to the embodiment, microgrid parameters and control commands can be displayed and received via a mobile client or a remote web interface.

[0051] Figure 2This is a schematic diagram of a microgrid dynamic disturbance analysis method according to an embodiment.

[0052] Reference Figure 2 This paper illustrates a method for dynamic disturbance analysis of microgrids according to embodiments of the present disclosure. The microgrid includes wind power generation systems, photovoltaic power generation systems, and energy storage systems.

[0053] The microgrid dynamic disturbance analysis method includes: in step 210, collecting the microgrid's operating data.

[0054] In step 220, the dynamic stability parameters of the microgrid system are calculated based on the collected operational data and a microgrid mathematical model. In step 230, the dynamic disturbance parameters of the microgrid system are predicted based on a neural network model. In step 240, the dynamic stability parameters and dynamic disturbance parameters are visualized using a microgrid digital twin model, and microgrid control commands are returned. Here, the neural network model includes an input layer, an output layer, and multiple hidden layers.

[0055] As mentioned above, the inputs to the neural network model can include wind power generation parameters, photovoltaic power generation parameters, and energy storage system parameters. The dynamic disturbance parameters predicted by the neural network model can include the disturbance oscillation frequency, the maximum renewable energy generation power boundary, and the minimum energy storage power boundary. When the microgrid is connected to the grid, the inputs to the neural network model also include regional grid system parameters, and the dynamic disturbance parameters predicted by the neural network model also include regional grid capacity assessment. On the other hand, when the microgrid is off-grid, the inputs to the neural network model also include microgrid load parameters, and the dynamic disturbance parameters predicted by the neural network model also include the load ceiling power boundary. Optionally, when the microgrid includes multiple wind power generation systems and multiple energy storage systems, the inputs to the neural network model also include the number of wind power generation systems, the wake effect of each wind power generation system, and the number of energy storage systems.

[0056] According to embodiments of this disclosure, the collected operational data can also be preprocessed, including at least one of data cleaning, aggregation, dimensionality reduction, sampling, and clustering. The microgrid dynamic disturbance analysis method can predict the control parameters of the microgrid system, simultaneously track the system's control feedback, and optimize the continuity of control parameters, enabling the system to quickly recover stability in response to step changes. When the rate of change of environmental parameters in the microgrid is too large, the microgrid dynamic disturbance analysis method of this disclosure can predict dynamic disturbances in real time and optimize system control parameters.

[0057] According to embodiments of this disclosure, a microgrid system is also provided, the microgrid system including at least one of a wind turbine, a photovoltaic power generation module, a diesel generator, and an energy storage system, and the microgrid system also includes the microgrid dynamic disturbance analysis system described above.

[0058] This disclosure employs a neural network training model to train dynamic disturbance parameter data for which an accurate mathematical model cannot be established, and mines the digital relationship between input and output data to obtain the dynamic disturbance influence factor and disturbance boundary.

[0059] Figure 3 This is a schematic diagram of a microgrid dynamic disturbance neural network training model according to an embodiment.

[0060] Reference Figure 3 The input conditions for the neural network training model according to the embodiment may include, but are not limited to, relevant data from wind power generation systems, photovoltaic power generation systems, and energy storage systems. After training and processing a large amount of input data through the neural network, the required data is output, thereby displaying seemingly unrelated input and output data through digital training and correlation. This enables the analysis of oscillation frequencies, maximum renewable energy boundaries, and minimum energy storage boundaries that cause dynamic disturbances in the microgrid system, improving the system's predictive capabilities and economic efficiency.

[0061] Figure 4 and Figure 5 These are schematic diagrams of the dynamic disturbance neural network training models for microgrids during grid-connected and off-grid operation, respectively, according to the embodiments.

[0062] Reference Figure 4 and Figure 5 Considering the dynamic disturbances of microgrids under both grid-connected and off-grid operation scenarios, this disclosure provides a separately trained neural network model for the dynamic problems of grid-connected microgrids, addressing the issues of weak grid connectivity and the intermittency, uncertainty, and volatility of renewable energy. The input to the neural network model includes regional grid system parameters, enabling real-time assessment of the grid-connected capacity of the microgrid in areas with weak grid connectivity (i.e., the dynamic disturbance parameters predicted by the neural network model also include regional grid capacity assessment). Based on the assessment results, control of renewable energy generation and energy storage systems is implemented to ensure friendly microgrid integration and avoid disconnection due to dynamic stability issues.

[0063] Off-grid microgrids are autonomous systems. In a microgrid, stable power sources act as U / f sources, renewable energy sources as P / Q sources, and reactive power compensation and filtering equipment as PV sources. Both P / Q and U / f types of power sources need to provide stable power to the load. With the increase in load and P / Q sources, off-grid microgrids will also experience dynamic disturbance instability. To address load volatility, this disclosure provides a separately trained neural network model for the dynamic problems of off-grid microgrids. The input to the neural network model includes microgrid load parameters, allowing the neural network to obtain various dynamic disturbance instability boundaries under load and P / Q source variations, such as the upper limit power boundary (i.e., the dynamic disturbance parameters predicted by the neural network model also include the upper limit power boundary).

[0064] Figure 6 This is a schematic diagram of a dynamic perturbation neural network training model for the joint operation of a wind power generation system and an energy storage system in a microgrid according to an embodiment.

[0065] Reference Figure 6 In scenarios where multiple wind turbines and energy storage systems operate jointly on-grid or off-grid, a model can be developed using empirical values. Inputs include the number of wind turbines, wind turbine power variations with wind speed, energy storage system capacity, number of converters, and control strategies. A big data model can then be used to obtain the maximum boundary value for wind power generation and the minimum boundary value for energy storage in response to dynamic disturbances, serving as the outputs. When the microgrid includes multiple wind power systems and multiple energy storage systems, the neural network model's inputs also include the number of wind power systems, the wake effect of each wind power system, and the number of energy storage systems.

[0066] The neural network model according to embodiments of the present disclosure includes an input layer, an output layer, and multiple hidden layers. For example, three hidden layers are shown in the figures, but the present disclosure is not limited thereto and may include fewer or more hidden layers.

[0067] The big data platform module in this disclosure includes a backend and a frontend.

[0068] The backend is used to implement the neural network model according to embodiments of this disclosure. The backend is primarily compiled using Python, employs the Django framework as the backend web framework, PostgreSQL as the relational database, Elasticsearch as the real-time data storage engine, and Redis as a cold / hot data separation cache. The frontend includes a web client and an app client, primarily using React, with Redux for centralized state management. Data interaction between the frontend and backend uses RESTful API communication technology, exchanging data via JSON format data packets, achieving frontend / backend separation and thus solving problems such as insufficient manpower, unclear division of labor, and complex coupling between the frontend and backend.

[0069] This disclosure uses Python's Django framework to build the backend management configuration system web framework. Its main functions include permission distribution, scene registration, data configuration, data reporting, sending control, and alarm reporting. Table 1 is a detailed list of backend management configurations.

[0070] Table 1

[0071]

[0072]

[0073]

[0074] Figure 7 This is a flowchart of the front-end and back-end data request process according to an embodiment.

[0075] Reference Figure 7 The front-end and back-end data communication API can define front-end request parameters, receive and calculate them, and return corresponding results. It can also configure route mapping globally, so that the corresponding module responds when a client accesses the API. In step 710, the data request processing receives data to obtain the parameter scene; in step 720, the data request processing obtains a data list based on the parameters; in step 730, the data request processing determines whether the first node (pos) is 0; if the result is negative, the data request processing adds a data entry to the empty array in step 740; and in step 750, the data request processing returns the new array.

[0076] Building a connection with PostgreSQL involves configuring PostgreSQL database information globally, including specific parameters such as:

[0077] ENGINE: Database engine

[0078] NAME: Database name,

[0079] USER: Owner

[0080] PASSWORD: Password

[0081] HOST: Hostname

[0082] PORT: Port.

[0083] Then, define the table structure in the model.py file under the corresponding module, including:

[0084] name=models.CharField(max_length=128)

[0085] uid=models.CharField(max_length=128,null=True,blank=True)

[0086] address=models.CharField(max_length=256)

[0087] long=models.FloatField(default=True,null=True)

[0088] lat=models.FloatField(default=True,null=True)

[0089] avatar=models.ImageField(null=True,blank=True)

[0090] code = models.CharField(max_length = 128, null = True, blank = True, verbose_name = "region code")

[0091] owners=models.ManyToManyField(settings.AUTH_USER_MODEL, related_name='scenes')

[0092] is_used = models.BooleanField(default = False, verbose_name = "Whether it is enabled")

[0093] Finally, data tables are generated in the configured PostgreSQL database.

[0094] Figure 8 This is the PostgreSQL database table structure of the big data platform according to the embodiment.

[0095] The front-end can be developed using the React framework. The project is built using the Create React App scaffolding tool. The main dependencies to install include Echarts, Antd, and Bootstrap. Echarts is used to implement the chart files on the page, Antd is used to implement the basic UI components on the page, and Bootstrap is used to implement the responsive layout of the page.

[0096] The front-end can also use the React Native cross-platform framework to develop apps. Key plugins include native-echarts, react-native-baidu-map, antd-mobile, and react-native-clear-cache. native-echarts is used to implement chart components on the page, react-native-baidu-map is used to implement the map functionality on the homepage, antd-mobile is used to implement basic UI components on the page, and react-native-clear-cache is used to clear the cache.

[0097] The front-end module includes:

[0098] Components

[0099] Containers: Pages

[0100] Redux: State Management and Data Requests

[0101] Service: Request URL and method configuration related

[0102] Themes: Theme style configuration related

[0103] Use fetch to make HTTP data requests, define key-value pairs for the request URL, and store the results, entities, and status in Redux after a successful request for global use by the project.

[0104] Figure 9 This is the basic graphics library architecture according to the embodiment.

[0105] Reference Figure 9 The front-end visualization analysis can be based on basic graphics libraries and components such as SVG, PNG, JPG, and ICON (910), combined with JavaScript graphics processing libraries such as svg.js (920), to build a data visualization interface (930), aiming to use graphical means to clearly and effectively convey and communicate information.

[0106] This disclosure utilizes a big data platform for load forecasting, enabling the mining and prediction of historical and real-time data based on meteorological information and selected meters. Big data forecasting models may include Monte Carlo methods, decision trees, and support vector machines.

[0107] Figure 10 This is the load prediction interface of a big data platform according to an embodiment.

[0108] Reference Figure 10According to the big data platform analysis and prediction of the day's load data in the embodiment, the actual value and the predicted value can be displayed in the form of a curve graph.

[0109] The front end can send parameters through the RESTful API interface. The parameters can include meter ID, selected date, and prediction type (Monte Carlo: knc, decision tree: dtc, support vector machine: svc).

[0110] Table 2

[0111]

[0112] After receiving the parameters sent by the frontend, the backend performs calculations on the received data.

[0113] Monte Carlo mainly uses the GaussianNB method (Gaussian Naive Bayes) under sklearn.naive_bayes.

[0114] The decision tree mainly uses the KNeighborsClassifier method under sklearn.neighbors.

[0115] Support Vector Machines primarily employ the DecisionTreeClassifier method from sklearn.tree.

[0116] The embodiments of this disclosure can also achieve remote control via a web client or an app client. Remote control commands are sent to the backend for a specific device or a group of devices. The command information includes: scene ID, device ID, RTU ID, point number, and corresponding value. The backend then pushes the command to the corresponding monitoring system receiving program. The receiving program parses the JSON command into monitoring system control command data and then transmits it to the front-end process. Upon receiving the command, the front-end process calls the protocol parsing interface to distribute the actual command to each device.

[0117] The front end sends control commands through the RESTful API interface, which specifically includes the controlled device command ID, control type (remote control 0, remote adjustment 1), and setting value.

[0118] Table 3

[0119]

[0120] After receiving the control commands from the front end, the back end encapsulates the received data and sends it to the accessor via a socket.

[0121] Figure 11 This is a flowchart of the remote control process of a big data platform according to an embodiment.

[0122] Reference Figure 11Upon receiving control commands, the accessor parses the command words and sends them to the monitoring system via a socket. According to the embodiment, the big data platform's remote control can maintain a long socket connection, enabling real-time reception of control commands from the front end and their transmission to business modules. The issuance of commands is subject to access control, determined in conjunction with the user's permission system. This provides the remote control with excellent stability, real-time performance, and security.

[0123] The accessor serves as a data transmission and acquisition module, acting as a bridge between the monitoring system and the big data platform. It employs the Google Protobuf protocol for compressed transmission, directly uploading data from the monitoring system to the big data platform, thus avoiding drawbacks such as unstable communication and data complexity. It enables real-time data storage, database association, and the issuance of remote control commands.

[0124] Figure 12 It is a real-time data structure in ES according to the embodiment. Figure 13 This is a flowchart illustrating the real-time data storage to ES data flow according to an embodiment.

[0125] Reference Figure 12 and 13 Storing data in Elastic Search includes: establishing an ES connection and configuring the anlogInput telemetry and statusInput teleindication data structures.

[0126] The received data is parsed to determine if it is a data report. Then, it is unpacked to obtain the real-time data list (measurelist). The task is submitted to the task queue, and the task queue stores the data in Elasticsearch for use.

[0127] The data transmission and acquisition module meets the following requirements:

[0128] Horizontal scaling: Horizontal scaling can be achieved through storage decoupling and front-end load balancing;

[0129] High concurrency: Asynchronous I / O operations in the program improve concurrency;

[0130] Exception handling: When data I / O fails, it can be recorded through logs, which facilitates subsequent unified operations such as supplementing the abnormal data.

[0131] An exemplary embodiment of the present invention also provides a computer-readable storage medium storing a computer program. The computer-readable storage medium stores a computer program that, when executed by a processor, causes the processor to execute the microgrid dynamic disturbance analysis method according to the present invention. The computer-readable recording medium is any data storage device capable of storing data read from a computer system. Examples of computer-readable recording media include: read-only memory, random access memory, read-only optical disk, magnetic tape, floppy disk, optical data storage device, and carrier waves (such as data transmission via the Internet through wired or wireless transmission paths).

[0132] An exemplary embodiment of the present invention also provides a computer device. The computer device includes a processor and a memory. The memory stores a computer program. The computer program is executed by the processor, causing the processor to execute the computer program of the microgrid dynamic disturbance analysis method according to the present invention.

[0133] The embodiments disclosed herein integrate data twins with NewIT, providing users with application services such as load forecasting, real-time monitoring, and remote control through accurate, reliable, and high-fidelity virtual models, multi-source, massive, and trustworthy twin data, and real-time dynamic virtual-real interaction. Through intelligent matching of optimal algorithms, data preparation, analysis, and fusion are automatically performed to conduct deep knowledge mining of the twin data, thereby generating various types of services. This achieves accurate construction of multi-dimensional / multi-scale models, deep fusion of full-process / full-business data, intelligent / personalized on-demand use, and real-time / dynamic interaction.

[0134] To address the dynamic disturbance problem in microgrids, effective predictive measures are proposed, identifying boundary technical issues in energy storage and renewable energy generation operation, thereby reducing microgrid construction costs and facilitating economically optimized operation. An online analysis and prediction method for system dynamic disturbance behavior is proposed, deeply integrating big data and cloud computing. This method enables data collection, storage and retrieval, computation and analysis, and visualization of microgrid big data, improving the operation, maintenance, and optimization levels of microgrid applications. This publicly available big data platform adopts a front-end and back-end separation approach, reducing development cycles and personnel costs. A hybrid storage model enables efficient and stable access to massive amounts of data, improving exponentially growing high-concurrency capabilities and the parallelization capabilities of data mining in cloud computing. Multiple types and sources of business data are collected, and hierarchical extraction, cleaning, and filtering preprocessing are performed to improve data quality. Dynamic disturbance analysis helps improve system stability and renewable energy penetration, enhancing the economic benefits and stability of the energy internet.

[0135] For the massive amounts of data from microgrid systems, the big data platform's data storage, serving as the foundation for parallel computing in cloud computing, adopts a hybrid storage model of Redis + Elastic Search + PostgreSQL, improving its ability to store data concurrently. The data acquisition module utilizes a Celery asynchronous task queue, demonstrating excellent throughput and responsiveness under concurrent requests.

[0136] The specific embodiments of this disclosure have been described in detail above. Although some embodiments have been shown and described, those skilled in the art should understand that modifications and variations can be made to these embodiments without departing from the principles and spirit of this disclosure, which are defined by the claims and their equivalents. Such modifications and variations should also be within the protection scope of the claims of this disclosure.

Claims

1. A microgrid dynamic disturbance analysis system, characterized in that, The microgrid dynamic disturbance analysis system includes: The detection and monitoring unit is configured to collect operational data of the microgrid; The data platform is configured to: calculate the dynamic stability parameters of the microgrid system based on a microgrid mathematical model using collected operational data, and predict the dynamic disturbance parameters of the microgrid system based on a neural network model; and visualize the dynamic stability parameters and dynamic disturbance parameters through a microgrid digital twin model and return microgrid control commands. The inputs to the neural network model include wind power generation parameters, photovoltaic power generation parameters, and energy storage system parameters. The dynamic disturbance parameters predicted by the neural network model include disturbance oscillation frequency, maximum new energy power generation boundary, and minimum energy storage power boundary.

2. The system according to claim 1, wherein, When the microgrid is connected to the grid, the input of the neural network model also includes regional power grid system parameters, and the dynamic disturbance parameters predicted by the neural network model also include regional power grid capacity assessment. When the microgrid is off-grid, the input to the neural network model also includes microgrid load parameters, and the dynamic disturbance parameters predicted by the neural network model also include the upper limit power boundary of the load.

3. The system of claim 1, wherein, Microgrid parameters can be displayed and control commands can be received via mobile client or remote web interface.

4. The system of claim 1, wherein, The data platform performs data storage through a relational database, performs searches through a real-time database, and performs offline analysis through a hot database.

5. A microgrid system, the microgrid system comprising at least one of a wind turbine, a photovoltaic power generation module, a diesel generator, and an energy storage system, and the microgrid system further comprising a microgrid dynamic disturbance analysis system as described in any one of claims 1 to 4.

6. A microgrid dynamic disturbance analysis method, characterized in that, The microgrid includes a wind power generation system, a photovoltaic power generation system, and an energy storage system, and the method includes: Collect operational data of the microgrid; Using collected operational data, the dynamic stability parameters of the microgrid system are calculated based on a microgrid mathematical model, and the dynamic disturbance parameters of the microgrid system are predicted based on a neural network model; and The dynamic stability parameters and dynamic disturbance parameters are visualized using a microgrid digital twin model, and microgrid control commands are returned. The inputs to the neural network model include wind power generation parameters, photovoltaic power generation parameters, and energy storage system parameters. The dynamic disturbance parameters predicted by the neural network model include disturbance oscillation frequency, maximum new energy power generation boundary, and minimum energy storage power boundary.

7. The microgrid dynamic disturbance analysis method of claim 6, wherein, The method includes: preprocessing the collected operational data, wherein the preprocessing includes at least one of cleaning, aggregation, dimensionality reduction, sampling, and clustering.

8. The microgrid dynamic disturbance analysis method according to claim 6, wherein, When the microgrid is connected to the grid, the input of the neural network model also includes regional power grid system parameters, and the dynamic disturbance parameters predicted by the neural network model also include regional power grid capacity assessment. When the microgrid is off-grid, the input to the neural network model also includes microgrid load parameters, and the dynamic disturbance parameters predicted by the neural network model also include the upper limit power boundary of the load.

9. The microgrid dynamic disturbance analysis method of claim 6, wherein, When a microgrid includes multiple wind power generation systems and multiple energy storage systems, the inputs to the neural network model also include the number of wind power generation systems, the wake effect of each wind power generation system, and the number of energy storage systems.

10. The microgrid dynamic disturbance analysis method of claim 6, wherein, The neural network model includes an input layer, an output layer, and multiple hidden layers.

11. A computer readable storage medium storing a computer program, characterized in that, When the computer program is executed by a processor, it implements the microgrid dynamic disturbance analysis method as described in any one of claims 6 to 10.

12. A computer device, comprising: The computer device includes: processor; The memory stores a computer program that, when executed by a processor, implements the microgrid dynamic disturbance analysis method as described in any one of claims 6 to 10.

Images