Power system reliability control method and system for distributed power supply collaborative optimization

Abstract

The invention provides a distributed power supply collaborative optimization power system reliability control method, which comprises the following steps of 1, obtaining real-time operation parameters of a power system, and obtaining an optimal output combination scheme of distributed power supply and conventional power supply collaborative optimization; 2, monitoring the running state of the distributed power supply in real time based on the Internet of Things technology; and step 3, scheduling and controlling active and reactive power output of the distributed power supply, and ensuring the power supply reliability of the important load. According to the technical scheme, important technical support is provided for constructing a power system optimization scheduling strategy with high robustness and high adaptability.

Description

Technical Field

[0001] The present invention relates to the technical field of power system reliability control, and in particular to a method and system for power system reliability control with coordinated optimization of distributed power sources. Background Art

[0002] In the existing technology, the reliability control of distributed power sources mainly adopts the following methods: First, in the distribution network planning stage, optimize the site selection and capacity configuration of distributed power sources to improve the static safety margin of the power grid; second, in the distribution network operation stage, improve the voltage quality of the distributed power access point through reactive power optimization, voltage control and other means; third, in the event of a power grid failure, reduce the scope of the accident through island operation and fault isolation. However, these methods are often limited to a single time scale and specific scenarios, lacking a coordinated interaction mechanism between distributed power sources and centralized power sources, and between distribution networks and active distribution networks, making it difficult to give full play to the flexible adjustment capabilities and system support role of distributed power sources. Summary of the invention

[0003] In view of this, the purpose of the present invention is to provide a power system reliability control method and system for coordinated optimization of distributed power sources, so as to provide important technical support for building a highly robust and adaptable power system optimization scheduling strategy.

[0004] To achieve the above object, the present invention adopts the following technical solution: a power system reliability control method for coordinated optimization of distributed power sources, comprising the following steps: Step 1: Obtain the real-time operating parameters of the power system and obtain the optimal output combination solution for the coordinated optimization of distributed power sources and conventional power sources; Step 2: Real-time monitoring of the operating status of distributed power sources based on Internet of Things technology; Step 3: Dispatch and control the active and reactive output of distributed power sources to ensure the reliability of power supply to important loads.

[0005] In a preferred embodiment, in the step 1, the real-time operating parameters of the power system are obtained, including data on load demand, output of conventional generator sets, transmission line flow, line voltage phase angle, and bus voltage amplitude, and the distributed power modeling parameters are combined, and the maximum entropy principle and adaptive dynamic programming algorithm are used to establish a multi-time scale power system optimization scheduling model that considers the spatiotemporal distribution characteristics and uncertainty factors of distributed power sources; the historical operating data of distributed power sources are analyzed using big data mining technology and deep learning algorithms to obtain the output probability distribution characteristics of different types of distributed power sources, and combined with the static safety constraints and stability constraints of the power system, the scheduling model is solved through the hybrid robust stochastic optimization theory to obtain the optimal output combination scheme for the coordinated optimization of distributed power sources and conventional power sources.

[0006] In a preferred embodiment, in step 2, when it is detected that a fault in a distributed power source causes a decrease in the power supply reliability of the power system, the conventional power source output and the switching status of the reactive compensation equipment are optimized and adjusted in real time by solving the power system optimization flow model considering the distributed power source failure scenario to ensure that the N-1 verification reliability of the power system meets the requirements; the multi-agent collaborative optimization theory is used, combined with the short-term load forecast and the distributed power source output forecast results, the output of the conventional power source and the distributed power source in each time period of the power system is dynamically optimized and dispatched to minimize the impact of the distributed power source access on the power grid, and a rolling optimization strategy is used to achieve real-time adaptive control of the power system.

[0007] In a preferred embodiment, in step 3, when an emergency fault occurs in the power system and causes power outage in some areas, graph theory and network topology analysis algorithms are used to identify and isolate the faulty area, and based on the distributed power supply configuration and load importance level in the power outage area, a regional microgrid recovery and reconstruction model is solved through multi-objective optimization, and the active and reactive output of distributed power supplies is reasonably dispatched and controlled. While ensuring the power supply reliability of important loads, the power supply of non-important loads is restored to the maximum extent, thereby improving the overall power supply reliability level and emergency resistance capability of the power system.

[0008] The present invention also provides a distributed power source collaboratively optimized power system reliability control system, and runs the distributed power source collaboratively optimized power system reliability control method.

[0009] Compared with the prior art, the present invention has the following beneficial effects: 1. Introduce the probability distribution constraints of distributed power output into the power system optimization dispatching model, consider the static safety constraints and stability constraints of the power grid, adopt the hybrid robust stochastic optimization theory to transform the optimization dispatching model into a robust optimization problem with probability constraints, and reduce the complexity of problem solving through scenario generation and simplification techniques. Use column generation algorithm, Benders decomposition algorithm and other algorithms to solve the hybrid robust stochastic optimization model, and obtain the optimal output combination scheme for the coordinated optimization of distributed power and conventional power under various constraints of the power system. Perform post-simulation evaluation, and make adaptive corrections and improvements to the optimization dispatching model based on the evaluation results.

[0010] 2. Analyze the historical operation data of distributed power sources, obtain the output probability distribution characteristics of different types of distributed power sources, and combine the static safety constraints and stability constraints of the power system to solve the dispatch model through hybrid robust stochastic optimization theory to obtain the optimal output combination plan for the coordinated optimization of distributed power sources and conventional power sources. By integrating data-driven and model-driven methods, we can fully explore the operating laws and uncertainty characteristics of distributed power sources, and provide important technical support for building a highly robust and adaptable power system optimization dispatch strategy. DETAILED DESCRIPTION

[0011] The present invention will be further described below in conjunction with the embodiments.

[0012] It should be noted that the following detailed descriptions are illustrative and are intended to provide further explanation of the present application. Unless otherwise specified, all technical and scientific terms used herein have the same meanings as those commonly understood by those skilled in the art to which the present application belongs.

[0013] It should be noted that the terms used herein are only for describing specific embodiments and are not intended to limit the exemplary embodiments according to the present application; as used herein, unless the context clearly indicates otherwise, the singular form is also intended to include the plural form. In addition, it should be understood that when the terms "comprise" and / or "include" are used in this specification, they indicate the presence of features, steps, operations, devices, components and / or their combinations.

[0014] The power system reliability control method for coordinated optimization of distributed generation includes the following steps: S101. Obtain the real-time operating parameters of the power system, including data on load demand, output of conventional generator sets, power flow of transmission lines, line voltage phase angle, and bus voltage amplitude, and combine the distributed power modeling parameters, apply the maximum entropy principle and adaptive dynamic programming algorithm, and establish a multi-time scale power system optimization scheduling model that considers the temporal and spatial distribution characteristics and uncertainty factors of distributed power sources; use big data mining technology and deep learning algorithms to analyze the historical operating data of distributed power sources, obtain the output probability distribution characteristics of different types of distributed power sources, and combine the static safety constraints and stability constraints of the power system, solve the scheduling model through hybrid robust stochastic optimization theory, and obtain the optimal output combination plan for the coordinated optimization of distributed power sources and conventional power sources; S102. Based on the Internet of Things technology, the operating status of distributed power sources is monitored in real time. When a distributed power source failure is detected, resulting in a decrease in the reliability of the power supply of the power system, the power system optimization flow model considering the distributed power source failure scenario is solved to optimize and adjust the output of conventional power sources and the switching status of reactive compensation equipment in real time to ensure that the N-1 verification reliability of the power system meets the requirements; the multi-agent collaborative optimization theory is used, combined with the short-term load forecast and the distributed power source output forecast results, to dynamically optimize the output of conventional power sources and distributed power sources in each period of the power system, to minimize the impact of distributed power source access on the power grid, and to adopt a rolling optimization strategy to achieve real-time adaptive control of the power system; S103. When an emergency fault occurs in the power system and causes power outage in some areas, graph theory and network topology analysis algorithms are used to identify and isolate the faulty area. Based on the configuration of distributed power sources and the importance level of loads in the power outage area, a regional microgrid recovery and reconstruction model is solved through multi-objective optimization to reasonably dispatch and control the active and reactive output of distributed power sources. While ensuring the power supply reliability of important loads, the power supply of non-important loads is restored to the maximum extent, thereby improving the overall power supply reliability level and emergency resistance capability of the power system.

[0015] Embodiment 1: Combined with the distributed power modeling parameters, the maximum entropy principle and adaptive dynamic programming algorithm are used to establish a multi-time scale power system optimization scheduling model that takes into account the temporal and spatial distribution characteristics and uncertainty factors of distributed power sources. The model includes: The key parameters of different types of distributed power sources are obtained through field research and equipment monitoring. The obtained distributed power source parameter data are preprocessed using data cleaning and feature engineering techniques. The mapping relationship between distributed power source parameters and their output characteristics is established using machine learning algorithms to obtain accurate distributed power source modeling parameters. The maximum entropy principle is introduced, and the geographical location distribution of distributed power sources in the power system is regarded as a random variable. The key attribute parameters of distributed power sources are used as constraints, and the maximum entropy model is established using the Lagrange multiplier method. The probability density function of the location distribution of distributed power sources is solved, and the Monte Carlo simulation method is used to randomly generate the location distribution of a large number of distributed power sources. The spatial distribution characteristics of distributed power sources in the power system and their impact on power grid flow calculation, voltage distribution, etc. are statistically analyzed. Considering the uncertainty of distributed power sources, scenario sets are introduced into the power system optimization dispatching model, and the complexity of model solution is reduced based on scenario simplification technology. The multi-time scale optimization dispatching problem is decomposed into multiple sub-problems by using adaptive dynamic programming algorithm. For each sub-problem, a power system optimization flow model considering the access of distributed power sources is established, and the constraints of the spatiotemporal distribution characteristics of distributed power sources are embedded to solve the optimal dispatching plan for each time period. In the dynamic planning process, the state variables of each sub-problem are updated in real time. The time domain and space granularity are optimized through adaptive adjustment, and the efficiency and accuracy of model solving are balanced to obtain the optimal dispatching strategy of the power system for the entire period. The multi-time scale optimization dispatching model is integrated into the power system energy management system, and data is interacted with functional modules such as load forecasting and electricity price forecasting to realize the coordinated optimization dispatching of distributed power sources and conventional power sources.

[0016] For example, In order to establish a multi-time scale power system optimization dispatch model that takes into account the temporal and spatial distribution characteristics of distributed generation and uncertainty factors, the following method steps need to be adopted: 1. Acquisition and processing of distributed power modeling parameters 1.1 Obtain key parameters of different types of distributed power sources through field research and equipment monitoring, including rated capacity, output power, efficiency curve, failure rate, etc.; 1.2 Use data cleaning and feature engineering techniques to pre-process the acquired distributed power parameter data, remove outliers, and achieve standardized representation of the data through normalization and standardization methods; 1.3 Use machine learning algorithms such as support vector machines and random forests to establish a mapping relationship between distributed power generation parameters and their output characteristics, and obtain accurate distributed power generation modeling parameters.

[0017] 2. Characterization of spatiotemporal distribution characteristics of distributed power sources based on the maximum entropy principle 2.1 The maximum entropy principle is introduced, the geographical distribution of distributed generation in the power system is regarded as a random variable, and its spatiotemporal distribution characteristics are characterized by constructing a probability distribution function; 2.2 Taking the key attribute parameters of distributed generation (such as capacity, type, distance from load, etc.) as constraints, the Lagrange multiplier method is used to establish the maximum entropy model and solve the probability density function of the location distribution of distributed generation; 2.3 The Monte Carlo simulation method is used to randomly generate the location distribution of a large number of distributed generation sources according to the probability density function, and their spatial distribution characteristics in the power system and their impact on power grid flow calculation, voltage distribution, etc. are statistically analyzed.

[0018] 3. Adaptive dynamic programming algorithm to solve multi-time scale optimization scheduling model 3.1 Considering the uncertainty of distributed generation, scenario sets are introduced into the power system optimization dispatch model, and the complexity of model solution is reduced based on scenario reduction technology; 3.2 Adopting an adaptive dynamic programming algorithm, the multi-time scale optimization scheduling problem is decomposed into multiple sub-problems. The sub-problems are coupled through state variables to form a nested recursive optimization structure. 3.3 For each sub-problem, an optimal power flow model of the power system considering the access of distributed generation is established, and the temporal and spatial distribution characteristic constraints of distributed generation are embedded to obtain the optimal dispatching plan for each time period; 3.4 In the dynamic programming process, the state variables of each sub-problem are updated in real time, and the time domain and spatial granularity are optimized through adaptive adjustment to balance the efficiency and accuracy of model solving, so as to obtain the optimal dispatching strategy of the power system for the entire period.

[0019] 4. Verification and application of multi-time scale optimization scheduling model 4.1 Build a power system simulation platform with distributed power sources, import typical power grid structure and load data, and test and verify the effectiveness of the optimization scheduling model; 4.2 In view of the problems that may arise in the process of solving the model, such as high computational complexity and slow convergence speed, we use parallel computing, heuristic search and other technical means to optimize and improve the practicality of the model; 4.3 Integrate the multi-time scale optimization scheduling model into the power system energy management system, interact with load forecasting, electricity price forecasting and other functional modules, realize the coordinated optimization scheduling of distributed power sources and conventional power sources, and improve the economy and reliability of the power system.

[0020] The above method steps elaborate on how to combine the distributed power modeling parameters, use the maximum entropy principle and adaptive dynamic programming algorithm, and establish a multi-time scale power system optimization dispatch model that takes into account the temporal and spatial distribution characteristics and uncertainty factors of distributed power. By using advanced technical means such as data mining and artificial intelligence, the output characteristics and distribution laws of distributed power sources are deeply characterized, and they are integrated with the power system optimization dispatch model. The impact of distributed power access on the power grid can be analyzed more accurately and efficiently, providing important theoretical basis and method support for the formulation of scientific and reasonable power system optimization dispatch strategies.

[0021] Embodiment 2: By using big data mining technology and deep learning algorithms, the historical operation data of distributed power sources are analyzed to obtain the output probability distribution characteristics of different types of distributed power sources. Combined with the static safety constraints and stability constraints of the power system, the scheduling model is solved through hybrid robust stochastic optimization theory to obtain the optimal output combination scheme for the coordinated optimization of distributed power sources and conventional power sources, including: The real-time operating parameters of distributed power sources are collected through smart meters, sensors and other IoT devices to form a massive historical operating data set. The collected historical operating data is cleaned and preprocessed, and the time series decomposition technology is used to divide the distributed power source historical operating data into trend items, periodic items and random items, respectively describing its long-term change trend, periodic fluctuation law and random disturbance characteristics; The association rule mining algorithm is used to discover the association pattern between the output of distributed power sources and environmental factors. The cluster analysis algorithm is used to divide the historical operation data of distributed power sources into several typical operation modes. The probabilistic graphical model is introduced to establish the spatiotemporal correlation model of distributed power output, which describes the dynamic evolution of its output in time and space. Adopting deep learning models such as long short-term memory neural network, learning the time series characteristics and nonlinear relationships contained in the historical operation data of distributed power sources, introducing attention mechanism and Transfer Learning technology, building a multi-time scale distributed power output probability prediction model with cross-scenario and cross-region generalization prediction capabilities, and integrating the prediction results of multiple deep learning sub-models through model integration strategy to improve the stability and robustness of distributed power output probability prediction; The probability distribution constraints of distributed power output are introduced into the power system optimization scheduling model. The static safety constraints and stability constraints of the power grid are considered. The hybrid robust stochastic optimization theory is used to transform the optimization scheduling model into a robust optimization problem with probability constraints. The complexity of problem solving is reduced through scenario generation and simplification techniques. The column generation algorithm, Benders decomposition algorithm and other algorithms are used to solve the hybrid robust stochastic optimization model. The optimal output combination scheme for the coordinated optimization of distributed power sources and conventional power sources under various constraints of the power system is obtained. A post-simulation evaluation is carried out, and the optimization scheduling model is adaptively corrected and improved according to the evaluation results.

[0022] For example, in order to use big data mining technology and deep learning algorithms to analyze the historical operation data of distributed power sources, obtain the output probability distribution characteristics of different types of distributed power sources, and solve the scheduling model through hybrid robust stochastic optimization theory in combination with power system constraints, and obtain the optimal output combination solution for the coordinated optimization of distributed power sources and conventional power sources, the following method steps need to be adopted: 1. Collection and preprocessing of historical operation data of distributed power generation 1.1 Through smart meters, sensors and other IoT devices, real-time operating parameters of distributed power sources are collected, including active power, reactive power, voltage, current, etc., to form a massive historical operating data set; 1.2 Clean and preprocess the collected historical operation data, remove missing values, outliers and other noise data, and achieve standardized representation of data through methods such as data normalization and feature scaling; 1.3 Using time series decomposition technology, the historical operation data of distributed power sources are divided into trend items, periodic items and random items, which respectively characterize their long-term change trends, periodic fluctuation laws and random disturbance characteristics.

[0023] 2. Analysis of the probability distribution characteristics of distributed power output based on big data mining technology 2.1 Use association rule mining algorithms to discover the association patterns between distributed power output and environmental factors (such as wind speed, light intensity, etc.), and characterize the output distribution rules of distributed power under different conditions; 2.2 Cluster analysis algorithms, such as K-means and DBSCAN, are used to divide the historical operation data of distributed generation into several typical operation modes, and statistically analyze the output probability distribution characteristics under each operation mode; 2.3 Introduce probabilistic graphical models, such as hidden Markov models and conditional random fields, to establish a spatiotemporal correlation model of distributed power output and characterize the dynamic evolution of its output in time and space.

[0024] 3. Deep learning algorithm to build a distributed power output probability prediction model 3.1 Use deep learning models such as long short-term memory neural network (LSTM) to learn the time series characteristics and nonlinear relationships contained in the historical operation data of distributed power sources, and build a multi-time scale output probability prediction model; 3.2 Introducing the attention mechanism and Transfer Learning technology to improve the generalization prediction ability of the deep learning model for different types of distributed power sources and realize the characterization of output probability distribution characteristics across scenarios and regions; 3.3 Through model integration strategies, such as Bagging and Boosting, the prediction results of multiple deep learning sub-models are integrated to improve the stability and robustness of the probability prediction of distributed power output.

[0025] 4. Hybrid robust stochastic optimization theory to solve the power system dispatch model 4.1 Introduce the probability distribution constraints of distributed generation output into the power system optimization dispatch model, and consider the static security constraints (such as power flow balance, line capacity limitation, etc.) and stability constraints (such as voltage stability, frequency stability, etc.) of the power grid; 4.2 Using hybrid robust stochastic optimization theory, the optimization scheduling model is transformed into a robust optimization problem with probability constraints, and the complexity of problem solving is reduced through scenario generation and simplification techniques; 4.3 Use the column generation algorithm, Benders decomposition algorithm, etc. to solve the hybrid robust stochastic optimization model and obtain the optimal output combination scheme of the coordinated optimization of distributed power sources and conventional power sources under various constraints of the power system; 4.4 Conduct post-simulation evaluation of the hybrid robust stochastic optimization dispatching strategy, analyze its impact on indicators such as the economy, reliability, and environmental protection of the power system, and adaptively modify and improve the optimization dispatching model based on the evaluation results.

[0026] The above method steps elaborate on how to use big data mining technology and deep learning algorithms to analyze the historical operation data of distributed power sources, obtain the output probability distribution characteristics of different types of distributed power sources, and combine the static safety constraints and stability constraints of the power system to solve the dispatch model through hybrid robust stochastic optimization theory to obtain the optimal output combination plan for the coordinated optimization of distributed power sources and conventional power sources. By integrating data-driven and model-driven methods, the operating laws and uncertainty characteristics of distributed power sources are fully explored, providing important technical support for building a highly robust and adaptable power system optimization dispatch strategy.

[0027] Embodiment 3: Based on the Internet of Things technology, the operating status of distributed power sources is monitored in real time. When a fault in the distributed power source is detected, resulting in a decrease in the reliability of the power supply of the power system, the power system optimization power flow model considering the distributed power source fault scenario is solved to optimize and adjust the output of conventional power sources and the switching status of reactive compensation equipment in real time to ensure that the N-1 verification reliability of the power system meets the requirements, including: Deploy smart sensors, smart meters and other IoT devices at the distributed power source site to collect distributed power source operating parameters in real time, and transmit the collected operating data to the cloud server through the wireless communication network to achieve centralized monitoring and management of the distributed power source operating status. Use edge computing technology to deploy smart gateways at the distributed power source site to pre-process and analyze the collected operating data locally, so as to achieve rapid detection and early warning of distributed power source faults. Based on historical operation data, a machine learning algorithm is used to establish a parameter model of distributed power under normal operating conditions, which is used to determine in real time whether the distributed power operating status is abnormal. When it is detected that the distributed power operating parameters deviate from the normal range, the fault diagnosis process is triggered, and the fault type and cause of the distributed power are inferred through methods such as fault tree analysis and expert knowledge base, and corresponding emergency response strategies are formulated for different types of distributed power faults. A set of distributed power generation fault scenarios is introduced on the basis of the conventional power system power flow model. Each scenario corresponds to a possible combination of distributed power generation faults and is assigned a corresponding probability of occurrence. A random optimization power flow model considering distributed power generation fault scenarios is established with the goal of minimizing the power system operation cost. The decision variables include conventional power output, reactive compensation equipment switching status, etc., and the power system N-1 verification reliability constraint is introduced into the optimization power flow model to ensure that the power system can meet the safety operation limit requirements under any distributed power generation fault scenario. When the IoT monitoring system detects a fault in a distributed power source, it triggers the real-time solution of the power system optimization power flow model, generates the optimal dispatching strategy to deal with the distributed power source fault, and sends the optimization dispatching strategy to each conventional power source and reactive power compensation equipment through the energy management system, so as to achieve real-time optimization and adjustment of the power system operation mode, and continuously monitor the changes in the distributed power source fault situation. When the fault is restored or expanded, the optimization power flow model of the power system is re-triggered to solve, and the optimization dispatching strategy of the power system is dynamically updated; Under the premise of considering the distributed power generation failure scenario, the Monte Carlo simulation and other methods are used to evaluate the power supply reliability indicators of the power system, quantify the impact of distributed power generation failure on the reliability of the power system, formulate power system reliability improvement measures for the weak links exposed in the reliability assessment, and continuously track the operation status and fault characteristics of distributed power sources, regularly update the power system optimization power flow model and reliability assessment model, and realize the adaptive optimization operation of the power system in the high penetration distributed power generation scenario.

[0028] For example, In order to monitor the operating status of distributed power sources in real time based on the Internet of Things technology, when a fault in the distributed power source is detected, resulting in a decrease in the power supply reliability of the power system, the power system optimization power flow model considering the distributed power source fault scenario is solved to optimize and adjust the conventional power output and the switching status of the reactive compensation equipment in real time to ensure that the N-1 verification reliability of the power system meets the requirements. The following methods and steps need to be adopted: 1. Build a distributed power supply IoT monitoring system 1.1 Deploy smart sensors, smart meters and other IoT devices at the distributed power source site to collect real-time operating parameters of the distributed power source, including active power, reactive power, voltage, current, temperature, etc.; 1.2 Transmit the collected distributed power supply operation data to the cloud server through wireless communication networks (such as 4G / 5G, NB-IoT, etc.) to achieve centralized monitoring and management of the distributed power supply operation status; 1.3 Using edge computing technology, deploy intelligent gateways at the distributed power generation site to pre-process and analyze the collected operating data locally to achieve rapid detection and early warning of distributed power generation faults.

[0029] 2. Distributed power supply fault detection and diagnosis method 2.1 Based on historical operation data, a machine learning algorithm (such as support vector machine, random forest, etc.) is used to establish a parameter model of the normal operation state of the distributed power supply, which is used to determine in real time whether the operation state of the distributed power supply is abnormal; 2.2 When it is detected that the operating parameters of the distributed power source deviate from the normal range, the fault diagnosis process is triggered, and the fault type and cause of the distributed power source are inferred through methods such as fault tree analysis and expert knowledge base; 2.3 Develop corresponding emergency response strategies for different types of distributed power generation failures, including fault isolation, power limitation, orderly shutdown, etc., to reduce the impact of distributed power generation failures on the power system.

[0030] 3. Power system optimization power flow model considering distributed generation failure scenario 3.1 Based on the conventional power system flow model, a set of distributed power generation fault scenarios is introduced. Each scenario corresponds to a possible combination of distributed power generation faults and is assigned a corresponding probability of occurrence. 3.2 With the goal of minimizing the operating cost of the power system, a stochastic optimization power flow model considering the distributed power supply failure scenario is established. The decision variables include the output of conventional power supply, the switching status of reactive power compensation equipment, etc.; 3.3 The N-1 verification reliability constraint of the power system is introduced into the optimization power flow model to ensure that the power system can meet the safe operation limit requirements such as line current and node voltage under any distributed power generation failure scenario.

[0031] 4. Real-time optimization and adjustment of power system operation mode 4.1 When the IoT monitoring system detects a fault in a distributed power source, it triggers the real-time solution of the power system optimization power flow model to generate the optimal dispatching strategy to deal with the distributed power source fault; 4.2 The optimized dispatching strategy is sent to each conventional power source and reactive power compensation equipment through the energy management system (EMS), so as to achieve real-time optimization and adjustment of the power system operation mode and maintain the power supply reliability of the power system; 4.3 Continuously monitor the changes in the fault conditions of distributed power sources. When the fault is restored or expanded, re-trigger the solution of the power system optimization power flow model and dynamically update the optimization scheduling strategy of the power system.

[0032] 5. Power system reliability assessment and improvement 5.1 Under the premise of considering the distributed power generation failure scenario, the power supply reliability indicators (such as SAIFI, SAIDI, etc.) of the power system are evaluated by using methods such as Monte Carlo simulation to quantify the impact of distributed power generation failure on the reliability of the power system; 5.2 Formulate measures to improve the reliability of the power system in response to the weak links exposed in the reliability assessment, such as optimizing the site selection of distributed power access points, adding distribution automation equipment, and introducing a demand-side response mechanism; 5.3 Continuously track the operating conditions and fault characteristics of distributed power sources, regularly update the power system optimization power flow model and reliability assessment model, and realize adaptive optimization operation of the power system in high penetration distributed power generation scenarios.

[0033] The above method steps elaborate on how to monitor the operating status of distributed power sources in real time based on the Internet of Things technology. When a distributed power source fault is detected, resulting in a decrease in the reliability of the power supply of the power system, the power system optimization flow model considering the distributed power source fault scenario is solved to optimize and adjust the conventional power output and the switching status of reactive compensation equipment in real time to ensure that the N-1 verification reliability of the power system meets the requirements. By deeply integrating the Internet of Things technology, intelligent fault diagnosis technology and power system optimization operation technology, a self-healing and robust smart grid operation control system is constructed to enhance the power system's ability to cope with the uncertainty risks of distributed power sources.

[0034] Embodiment 4: By using the theory of multi-agent collaborative optimization, combined with the short-term load forecast and distributed power output forecast results, the output of conventional power sources and distributed power sources in the power system in each period is dynamically optimized and dispatched to minimize the impact of distributed power access on the power grid, and a rolling optimization strategy is used to achieve real-time adaptive control of the power system, including: The power system is divided into multiple sub-areas. An intelligent dispatching agent is set up in each sub-area to be responsible for the optimal dispatch of conventional power sources and distributed power sources in the area. A coordination control agent is set up at the system level to coordinate the power balance and exchange between sub-areas and interact with the power market. A hierarchical and distributed information interaction and decision-making mechanism is established to achieve seamless integration of sub-area internal optimization and system-level collaborative optimization. For each sub-region, historical load data and distributed power output data are collected and pre-processed, and a regional load forecasting model integrating multiple models is constructed by comprehensively using statistical learning and machine learning methods. Output forecasting models that take into account the physical characteristics and uncertainty factors of different types of distributed power sources are established respectively. Taking intelligent dispatching agent as the basic unit, a dynamic optimization dispatching model of distributed power system is established. The objective functions include minimizing power generation cost, maximizing new energy consumption, and power balance. Taking into account the static safety constraints and dynamic stability constraints of the power system, a multi-period and multi-scenario stochastic optimization model is constructed. Game theory and collaborative learning mechanism are introduced to achieve the coordinated optimization dispatching of distributed power sources and conventional power sources through strategy iteration and information interaction between intelligent agents, and the market-oriented operation of the power system is achieved through the interaction between intelligent agents and the power market. The model predictive control concept is adopted to transform the dynamic optimization scheduling problem of the power system into a rolling optimization problem in a limited time domain. At the beginning of each scheduling cycle, the short-term load forecast and distributed power output forecast results are updated, and the multi-agent collaborative optimization algorithm is triggered to generate the optimal scheduling strategy for a period of time in the future in combination with the current power system operation status. The optimized scheduling strategy is decoded into control instructions and sent to each conventional power source and distributed power source through the energy management system to realize real-time adaptive control of the power system. In view of the uncertainties in short-term load forecasting and distributed generation output forecasting, the robust optimization theory is introduced to construct an optimal dispatching model for the power system with uncertainty immunity. Scenario generation and simplification techniques are used to improve the solution efficiency of the robust optimization model, and online self-optimization of the power system dispatch is achieved by adaptively and dynamically adjusting the scenario set and robustness parameters. Machine learning methods are introduced to continuously optimize the dispatching strategy of the intelligent agent through reinforcement learning, contrastive learning and other technologies, so as to enhance the resilience and recovery ability of the power system to uncertain disturbances.

[0035] For example, in order to apply the multi-agent collaborative optimization theory, combined with the short-term load forecast and distributed power output forecast results, dynamically optimize the output of conventional power sources and distributed power sources in each period of the power system, minimize the impact of distributed power access on the power grid, and adopt a rolling optimization strategy to achieve real-time adaptive control of the power system, the following methods and steps need to be adopted: 1. Constructing a hierarchical distributed power system optimization dispatch framework 1.1 Divide the power system into multiple sub-areas, and set up an intelligent dispatching agent in each sub-area to be responsible for the optimal dispatch of conventional power sources and distributed power sources in the area; 1.2 Set up a coordination control agent at the system level to coordinate power balance and exchange between sub-regions and interact with the power market; 1.3 Establish a hierarchical and distributed information interaction and decision-making mechanism to achieve seamless integration of sub-region internal optimization and system-level collaborative optimization.

[0036] 2. Develop high-precision short-term load forecasting and distributed power output forecasting models 2.1 For each sub-region, collect historical load data and distributed power output data, and perform data cleaning and preprocessing; 2.2 Comprehensively apply statistical learning and machine learning methods, such as multivariate linear regression, time series analysis, support vector machine, neural network, etc., to build a regional load forecasting model integrating multiple models and improve the accuracy of short-term load forecasting; 2.3 For different types of distributed power sources (such as wind power, photovoltaic power, small hydropower, etc.), output prediction models that take into account their physical characteristics and uncertainty factors are established, such as the wind power prediction model based on numerical weather forecast (NWP) and the photovoltaic power generation prediction model based on sky imaging.

[0037] 3. Power system dynamic dispatch model based on multi-agent collaborative optimization theory 3.1 Taking the intelligent dispatching agent as the basic unit, a dynamic optimization dispatching model of distributed power system is established, and the objective functions include minimizing the power generation cost, maximizing the consumption of new energy, and balancing the power supply. 3.2 Considering the static security constraints (such as line flow limits, node voltage limits, etc.) and dynamic stability constraints (such as N-1 verification, small disturbance angle stability, etc.) of the power system, a multi-period and multi-scenario stochastic optimization model is constructed; 3.3 Game theory and collaborative learning mechanism are introduced to achieve the coordinated optimization scheduling of distributed power sources and conventional power sources through strategy iteration and information interaction among intelligent agents, and the market-oriented operation of the power system is realized through the interaction between intelligent agents and the power market.

[0038] 4. Rolling optimization scheduling strategy 4.1 Using the model predictive control (MPC) concept, the dynamic optimization scheduling problem of the power system is transformed into a rolling optimization problem within a limited time domain; 4.2 At the beginning of each dispatch cycle, the short-term load forecast and distributed generation output forecast results are updated, and combined with the current power system operation status, the multi-agent collaborative optimization algorithm is triggered to generate the optimal dispatch strategy for the next period of time; 4.3 Decode the optimized dispatching strategy into control instructions and send them to each conventional power source and distributed power source through the energy management system to achieve real-time adaptive control of the power system.

[0039] 5. Enhanced robust optimization considering uncertainties 5.1 In view of the uncertainty in short-term load forecasting and distributed generation output forecasting, robust optimization theories such as stochastic optimization and distributed robust optimization are introduced to build an optimal dispatching model of power system with uncertainty immunity; 5.2 Use scenario generation and reduction techniques to improve the efficiency of solving robust optimization models, and achieve online self-optimization of power system dispatch by adaptively and dynamically adjusting scenario sets and robustness parameters; 5.3 Introduce machine learning methods, and continuously optimize the dispatching strategy of intelligent agents through reinforcement learning, contrastive learning and other technologies to enhance the resilience and recovery ability of the power system to cope with uncertain disturbances.

[0040] 6. Simulation verification and practical application 6.1 Build a power system simulation platform, including multiple sub-modules such as conventional power supply, distributed power supply, load, power grid, etc., and consider actual engineering constraints and restrictions; 6.2 Test and verify the power system dynamic dispatch method based on multi-agent collaborative optimization theory on the simulation platform, evaluate its optimization performance and computational efficiency, and compare and analyze it with the traditional centralized dispatch method; 6.3 Select a typical distribution network as a demonstration project, apply the proposed hierarchical distributed dynamic dispatching method in the actual power system, conduct long-term operation monitoring and evaluation, summarize the practical experience, and provide a basis for further improvement and promotion of the dispatching method.

[0041] The above method steps elaborate on how to use the theory of multi-agent collaborative optimization, combined with short-term load forecasting and distributed power output forecasting results, to dynamically optimize the output of conventional power sources and distributed power sources in each period of the power system, minimize the impact of distributed power access on the power grid, and use a rolling optimization strategy to achieve real-time adaptive control of the power system. By constructing a hierarchical distributed power system optimization and scheduling framework, developing a high-precision prediction model, and introducing robust optimization and machine learning techniques, the proposed method can significantly improve the power system's ability to cope with the uncertainty of distributed power sources and achieve efficient, economical, and safe operation in scenarios with a high proportion of new energy.

[0042] Embodiment 5: When an emergency fault occurs in the power system and causes power outage in some areas, graph theory and network topology analysis algorithms are used to identify and isolate the fault area. Based on the configuration of distributed power sources and the importance level of loads in the power outage area, the regional microgrid recovery and reconstruction model is solved through multi-objective optimization to reasonably dispatch and control the active and reactive output of distributed power sources, including: First, a physical topology model of the power system is constructed, and the physical devices of the power system are abstracted into nodes and edges of a graph. The nodes represent the busbars, the edges represent the lines or transformers, and the edge weights represent the electrical parameters of the lines or transformers. The information of intelligent electronic devices is embedded in the graph model to achieve integrated modeling of the physical layer and information layer of the power system. When an emergency fault occurs in the power system and causes power outages in some areas, the depth-first search or breadth-first search algorithm is used based on the constructed graph model to monitor the changes in the power system topology in real time, quickly locate the fault point, and use the minimum path tree algorithm to search all nodes directly or indirectly connected to the fault point with the fault point as the starting node to form the fault-affected area. At the boundary bus of the fault-affected area, the switch topology state is changed to isolate the fault area from the external power grid to form an island-operated regional microgrid; Then, taking the regional microgrid operating in an island as the research object, considering the distribution of distributed power sources and loads inside it, a multi-period and multi-scenario mixed integer nonlinear programming problem is established with the objective function of maximizing the restoration of power supply to important loads, minimizing the amount of load reduction and minimizing network losses, the active output, reactive output and switch state of distributed power sources as decision variables, and the power flow equation, voltage amplitude limit, line capacity limit and upper and lower limits of distributed power output as constraints; The established multi-objective optimization model for microgrid restoration and reconstruction is solved by using intelligent optimization algorithms such as genetic algorithm, particle swarm optimization algorithm or ant colony algorithm. The Pareto frontier is generated by constructing fitness function and non-dominated sorting, and the distributed optimization architecture is used to divide the problem into multiple sub-problems. The local intelligent agent achieves the convergence of the global optimal solution through iterative game. According to the solution results of multi-objective optimization, the optimal topological structure and distributed power output plan for microgrid restoration and reconstruction are determined. A hierarchical distributed control architecture is used to establish a coordinated control mechanism between the microgrid control center and each distributed power local controller to achieve automatic issuance and execution of dispatch instructions. Real-time adaptive control strategies such as rolling optimization and model predictive control are introduced to dynamically adjust the active and reactive output of distributed power sources to ensure the voltage and frequency stability of the microgrid. Finally, after the fault is repaired, the smooth grid connection between the isolated microgrid and the external grid is achieved through synchronous inspection and automatic reclosing technology. Power control and reactive power compensation are used to reduce the impact and disturbance of the grid connection. After the grid connection, the output dispatch and load management strategy of the distributed power source are dynamically optimized according to the operating status of the microgrid and the external grid to achieve flexible and autonomous operation of the microgrid.

[0043] For example, In order to identify and isolate the fault area by using graph theory and network topology analysis algorithms when an emergency fault occurs in the power system and causes power outage in some areas, and solve the regional microgrid recovery and reconstruction model through multi-objective optimization according to the configuration of distributed power sources and the importance level of loads in the power outage area, and reasonably dispatch and control the active and reactive output of distributed power sources, the following methods and steps need to be taken: 1. Constructing a physical topology model of the power system 1.1 Collect network topology data of the power system, including the connection relationship and electrical parameters of busbars, lines, transformers, switches and other equipment; 1.2 Using graph theory modeling methods, the physical equipment of the power system is abstracted into nodes and edges of a graph, and a directed weighted graph model is constructed. The nodes represent the busbars, the edges represent the lines or transformers, and the edge weights represent the electrical parameters of the lines or transformers. 1.3 Embed intelligent electronic device (IED) information, such as protection devices and measuring devices, into the graphical model to achieve integrated modeling of the physical layer and information layer of the power system.

[0044] 2. Fault area identification and isolation algorithm 2.1 Based on the depth-first search or breadth-first search algorithm of the graph model, the changes in the topology of the power system are monitored in real time. When an emergency fault occurs and causes power outages in some areas, the fault point can be quickly located; 2.2 Taking the fault point as the starting node, use the minimum path tree algorithm, such as the Dijkstra algorithm or the Bellman-Ford algorithm, to search for all nodes directly or indirectly connected to the fault point to form the fault impact area; 2.3 At the boundary bus of the fault-affected area, the fault area is isolated from the external power grid by changing the switch topology state, forming a regional microgrid operating in an islanded manner.

[0045] 3. Regional microgrid restoration and reconstruction model 3.1 Taking the regional microgrid operating in an isolated island as the research object, considering its internal distributed power sources (such as micro gas turbines, energy storage, renewable energy, etc.) and load distribution, a multi-objective optimization model for microgrid restoration and reconstruction is established; 3.2 The objective function includes maximizing the restoration of power supply to important loads, minimizing load reduction, minimizing network losses, etc. The decision variables include the active output, reactive output and switch status of distributed generation; 3.3 The constraints include power flow equation, voltage amplitude limit, line capacity limit, upper and lower limits of distributed generation output, etc., forming a multi-period and multi-scenario mixed integer nonlinear programming (MINLP) problem.

[0046] 4. Multi-objective optimization solution algorithm 4.1 In view of the multi-objective characteristics and non-convexity of the microgrid restoration and reconstruction model, intelligent optimization algorithms, such as genetic algorithm (GA), particle swarm optimization (PSO) algorithm, ant colony optimization (ACO), etc., are used to perform multi-objective optimization; 4.2 The concept of Pareto optimal solution is introduced. By constructing fitness function and non-dominated sorting, the Pareto frontier of microgrid restoration and reconstruction problem is generated to achieve the trade-off and compromise between multiple objectives. 4.3 A distributed optimization architecture is adopted to divide the microgrid restoration and reconstruction problem into multiple sub-problems. Local intelligent agents are responsible for solving each sub-problem, and convergence of the global optimal solution is achieved through iterative game.

[0047] 5. Distributed power generation optimization dispatching control strategy 5.1 According to the multi-objective optimization solution results, determine the optimal topology structure and distributed power output plan for microgrid restoration and reconstruction, and generate corresponding control instructions; 5.2 Adopt a hierarchical distributed control architecture, establish a coordinated control mechanism between the microgrid control center and each distributed power local controller, and realize the automatic issuance and execution of dispatch instructions; 5.3 In view of the intermittent and fluctuating characteristics of distributed power sources, real-time adaptive control strategies such as rolling optimization and model predictive control are introduced to dynamically adjust the active and reactive output of distributed power sources to ensure the voltage and frequency stability of the microgrid.

[0048] 6. Synchronous connection of microgrid and external grid 6.1 After the fault is repaired, synchronous inspection and automatic reclosing technology are used to achieve smooth grid connection between the isolated microgrid and the external grid; 6.2 Through measures such as power control and reactive power compensation, the impact and disturbance of grid connection can be reduced to ensure the synchronization of voltage, frequency and phase angle between the microgrid and the external grid; 6.3 After grid connection, the output dispatch and load management strategies of distributed power sources are dynamically optimized according to the operating status of the microgrid and the external power grid to achieve flexible and autonomous operation of the microgrid.

[0049] 7. Simulation analysis and practical application 7.1 Based on the physical parameters and topological structure of the actual power system, a simulation platform for microgrid restoration and reconstruction is built to test and verify the proposed fault area identification and isolation algorithm, multi-objective optimization model and solution method; 7.2 Carry out a large number of simulation experiments for different fault scenarios and load demands to evaluate the feasibility, effectiveness and robustness of microgrid restoration and reconstruction strategies, and conduct comparative analysis with traditional centralized restoration methods; 7.3 Select typical distribution networks and industrial parks as demonstration projects, apply the proposed microgrid restoration and reconstruction technology to actual fault recovery and emergency management, and conduct long-term operation monitoring and effect evaluation to provide a practical basis for the promotion and application of the technology.

[0050] The above method steps elaborate on how to use graph theory and network topology analysis algorithms to identify and isolate fault areas when an emergency fault occurs in the power system and causes power outages in some areas, and solve the regional microgrid recovery and reconstruction model through multi-objective optimization according to the configuration of distributed power sources in the power outage area and the importance level of loads, and reasonably dispatch and control the active and reactive output of distributed power sources. Through a series of innovative methods such as constructing a physical topology model of the power system, developing a fault area identification and isolation algorithm, establishing a microgrid recovery and reconstruction optimization model, using an intelligent optimization algorithm to solve, realizing the optimal dispatching and control of distributed power sources, and synchronously connecting microgrids to external power grids, the proposed technical solution can significantly improve the self-healing ability and recovery speed of the power system in response to emergency faults, reduce the scope of accident impact, and ensure the power supply reliability of important loads, and has broad application prospects.

Claims

1. A power system reliability control method for coordinated optimization of distributed power sources, characterized in that: The following steps are involved: Step 1: Obtain the real-time operating parameters of the power system and obtain the optimal output combination solution for the coordinated optimization of distributed power sources and conventional power sources; Step 2: Real-time monitoring of the operating status of distributed power sources based on Internet of Things technology; Step 3: Dispatch and control the active and reactive output of distributed power sources to ensure the reliability of power supply to important loads.

2. The power system reliability control method for coordinated optimization of distributed power sources according to claim 1 is characterized in that: In the step 1, the real-time operating parameters of the power system are obtained, including data on load demand, output of conventional generator sets, transmission line flow, line voltage phase angle, and bus voltage amplitude, and the distributed power modeling parameters are combined, and the maximum entropy principle and adaptive dynamic programming algorithm are used to establish a multi-time scale power system optimization scheduling model that considers the spatiotemporal distribution characteristics and uncertainty factors of distributed power sources; the historical operating data of distributed power sources are analyzed by big data mining technology and deep learning algorithms to obtain the output probability distribution characteristics of different types of distributed power sources, and the scheduling model is solved by hybrid robust stochastic optimization theory in combination with the static safety constraints and stability constraints of the power system to obtain the optimal output combination scheme for the coordinated optimization of distributed power sources and conventional power sources.

3. The power system reliability control method for coordinated optimization of distributed power sources according to claim 1 is characterized in that: In step 2, when it is detected that a fault in a distributed power source causes a decrease in the power supply reliability of the power system, the power system optimization flow model considering the distributed power source fault scenario is solved to optimize the conventional power source output and the switching state of the reactive compensation equipment in real time to ensure that the N-1 verification reliability of the power system meets the requirements; the multi-agent collaborative optimization theory is used, combined with the short-term load forecast and the distributed power source output forecast results, to dynamically optimize the output of conventional power sources and distributed power sources in each time period of the power system, to minimize the impact of the distributed power source access on the power grid, and to adopt a rolling optimization strategy to achieve real-time adaptive control of the power system.

4. The power system reliability control method for coordinated optimization of distributed power sources according to claim 1 is characterized in that: In step 3, when an emergency fault occurs in the power system and causes power outage in some areas, graph theory and network topology analysis algorithms are used to identify and isolate the faulty area, and based on the configuration of distributed power sources and the importance level of loads in the power outage area, a regional microgrid recovery and reconstruction model is solved through multi-objective optimization, and the active and reactive output of distributed power sources is reasonably dispatched and controlled. While ensuring the power supply reliability of important loads, the power supply of non-important loads is restored to the maximum extent, thereby improving the overall power supply reliability level and emergency resistance capability of the power system.

5. The power system reliability control system based on the coordinated optimization of distributed power sources is characterized by Run the power system reliability control method for coordinated optimization of distributed power sources as described in any one of claims 1 to 4.