Power distribution optimization method fusing source network simulation model and artificial intelligence algorithm

Abstract

The invention relates to the field of power systems, and discloses a power distribution optimization method fusing a source network simulation model and an artificial intelligence algorithm, comprising the following steps: collecting data in a power grid, the data comprising power grid load data, power generation data and meteorological data; based on collected data in the power grid, constructing a source-grid coupling simulation model for simulating power flow, node voltage and power grid dynamic response behaviors of the power grid under different operation conditions; and predicting the load data of the power grid by using an artificial intelligence algorithm, and generating load demand prediction and renewable energy output prediction according to a prediction result. According to the method, the source-network coupling simulation model and the artificial intelligence algorithm are fused, the power grid load and the renewable energy output are accurately predicted, the power generation plan and load distribution are dynamically adjusted, high efficiency, intelligence and real-time response of power grid dispatching are achieved, and the stability, economical efficiency and the renewable energy utilization rate of the power grid are improved.

Description

Technical Field

[0001] The present invention relates to the field of power systems, and in particular to a power distribution optimization method integrating a source-grid simulation model with an artificial intelligence algorithm. Background Art

[0002] With the rapid development of society and the continuous growth of electricity demand, the operation and management of power grids are facing tremendous pressure. Especially in the context of severe load fluctuations and the increasing proportion of renewable energy, traditional power grid dispatching methods can no longer meet the needs of modern power grids. In order to ensure the stable, efficient and economical operation of the power grid, it is necessary to rely on intelligent and efficient dispatching solutions. Existing technologies cannot cope well with complex load demands and fluctuations in renewable energy.

[0003] Existing power grid dispatching technologies mainly use simple mathematical models or rule-based empirical methods to perform load forecasting and power generation dispatching. These technologies can effectively maintain the stable operation of the power grid and ensure the basic stability of power supply when the power grid load changes little. Traditional methods can provide more intuitive operation decisions for the power grid by relying on historical data and certain load forecasting models, optimize the allocation of power grid resources to a certain extent, and avoid drastic fluctuations in power supply. However, these methods lack sufficient flexibility and accuracy when faced with complex power grid environments and real-time data changes.

[0004] Although the existing technology has a certain dispatch optimization effect, it has obvious shortcomings. First, the traditional load forecasting method cannot cope with the complexity of factors such as load fluctuations and weather changes, and the prediction accuracy is low. When the load of the power grid fluctuates sharply, the traditional method often lacks the ability to respond flexibly and cannot make real-time adjustments. Secondly, the existing dispatching system often relies on static rules and manual intervention, resulting in slow response speed of power grid dispatch and prone to decision-making errors, affecting the safety of the power grid. Finally, with the increase in the proportion of renewable energy, the existing technology is difficult to fully utilize the changing characteristics of renewable resources such as wind power and solar energy, and still over-relies on traditional power generation methods. Summary of the invention

[0005] In view of the shortcomings of the prior art, the present invention provides a distribution optimization method that integrates a source-grid simulation model and an artificial intelligence algorithm, which solves the problems in the prior art of insufficient grid load prediction accuracy, slow power generation scheduling response, and inability to fully utilize renewable energy.

[0006] To achieve the above objectives, the present invention is implemented through the following technical solutions: a power distribution optimization method integrating a source network simulation model and an artificial intelligence algorithm, comprising the following steps: Collecting data in the power grid, including power grid load data, power generation data and meteorological data; Based on the collected data from the power grid, a source-grid coupling simulation model is constructed to simulate the power flow, node voltage and dynamic response behavior of the power grid under different operating conditions; Use artificial intelligence algorithms to predict grid load data and generate load demand forecasts and renewable energy output forecasts based on the forecast results; Based on the load-source-grid coupling simulation model and prediction results, the dispatch optimization algorithm is used to dynamically adjust the power generation plan and load distribution to generate distribution optimization decisions; According to the distribution optimization decision, the corresponding dispatching instructions are executed through the power grid control system to adjust the operating status of the power grid.

[0007] Preferably, the source-network coupling simulation model includes: Power flow analysis, used to simulate power flow in the power grid and calculate power allocation and voltage distribution at each node; Grid dynamic response analysis is used to simulate the response of the grid under load fluctuations and external disturbances, including grid inertia and damping effects.

[0008] Preferably, the artificial intelligence algorithm comprises: Long short-term memory network models for processing time series data to predict grid load demand and renewable energy production; The LSTM network model is trained based on historical data to generate prediction results of future load and output.

[0009] Preferably, the input of the long short-term memory network model includes: Historical load data and historical renewable energy production data; The load forecast results and renewable energy output forecast results output by the LSTM model serve as the basis for scheduling optimization.

[0010] Preferably, the scheduling optimization algorithm includes: Feedback mechanism during training, used to adjust dispatch strategy according to real-time grid status; The power generation plan and load distribution are optimized through the reward function to maintain the stable operation of the power grid.

[0011] Preferably, the reward function includes: The operating cost of the power grid, the matching degree of load demand and the utilization rate of renewable energy; Through reward function optimization, the reinforcement learning model is able to adjust the dispatch strategy in real time to minimize the total cost of grid operation.

[0012] Preferably, the power distribution optimization decision includes: The economics of the grid, minimizing operating costs; The security of the power grid, ensuring that power flow and voltage stability indicators meet constraints; Environmental friendliness of the power grid and optimization of the use of renewable energy.

[0013] Preferably, the power distribution optimization decision further includes: By setting the weight of each objective function, multiple objective functions are weighted summed; The power flow, load limit and voltage constraints of the power grid are considered during the optimization process to ensure that the optimization results meet the actual operation requirements.

[0014] Preferably, the power grid dynamic response analysis includes: Grid stability analysis, used to evaluate the grid’s ability to respond to sudden load fluctuations or external disturbances; Dynamic dispatch is used to adjust the grid dispatch strategy in a timely manner according to the dynamic response results when a disturbance occurs in the grid.

[0015] Preferably, the artificial intelligence algorithm further comprises: An update mechanism for load and output forecast results, which is used to dynamically adjust the parameters of the forecast model based on historical data during grid operation; By adjusting the model parameters, the accuracy of the prediction results can be optimized.

[0016] The present invention provides a power distribution optimization method integrating a source network simulation model and an artificial intelligence algorithm. It has the following beneficial effects: 1. The present invention adopts a technical solution combining a source-grid coupling simulation model with an artificial intelligence algorithm to achieve the technical effect of accurately predicting grid load demand and renewable energy output. Compared with the prediction methods in the prior art that rely on simple linear models or empirical rules, the deep learning model of the present invention can better capture the complex time series characteristics of grid load and meteorological changes, solving the problems of low prediction accuracy and poor adaptability of traditional methods.

[0017] 2. The present invention realizes the dynamic adjustment of power grid generation plan and load distribution through the scheduling optimization algorithm, achieving the technical effect of optimizing power grid operation efficiency and improving renewable energy utilization. Different from the static scheduling strategy in the prior art, the dynamic adjustment algorithm of the present invention can flexibly adjust the power grid operation strategy according to real-time data and prediction results, solving the lag and inefficiency of the existing solutions in dealing with load fluctuations and renewable energy uncertainties.

[0018] 3. The present invention combines the close cooperation between power distribution optimization decision-making and power grid control system to achieve the technical effect of real-time optimization and safety assurance of power grid operation status. Compared with the control method with more manual intervention in traditional schemes, the present invention reduces the influence of human factors through automated dispatching instructions, improves the operation stability and response speed of the power grid, and effectively avoids the safety risks caused by operation delays or decision-making errors in the power grid.

[0019] 4. The present invention realizes the intelligent and continuous optimization of power grid dispatching through real-time data feedback and adaptive adjustment mechanism, and achieves the technical effect of improving the overall operation efficiency of the power grid. Compared with the relatively fixed dispatching mode in the prior art, the present invention can adjust the decision according to the real-time feedback of the power grid, so that the power grid can maintain the optimal performance under different operating conditions, solving the problem of lack of flexibility and inability to continuously optimize in the existing methods. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] Figure 1 The present invention is a flow chart of the method. DETAILED DESCRIPTION

[0021] The following will be combined with the drawings in the specification of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0022] Please see attached Figure 1 The embodiment of the present invention provides a power distribution optimization method integrating a source network simulation model and an artificial intelligence algorithm, comprising the following steps: S1. Collecting data from the power grid, including power grid load data, power generation data and meteorological data; Data collection is a basic component of the entire distribution optimization system. The accuracy and timeliness of data directly affect the subsequent source network simulation model construction, artificial intelligence prediction and dispatch optimization. The load data, power generation data and meteorological data of the power grid are the core information in the optimization process. Correctly collecting and processing this data provides important support for subsequent power grid operation analysis and decision-making.

[0023] In general, grid load data includes the load demand of each node in the grid and the time-varying characteristics of these demands. The change of grid load is affected by many factors, such as peak and off-peak hours, industrial production, and residents' lives. In order to accurately reflect the load demand of the grid, data collection is required from multiple levels. Specifically, load data should include but is not limited to: Total load data in each time period, that is, the total power demand of the power grid at that moment; Distribution data of each load node, that is, the power consumption of each node (such as substation, distribution room, etc.); Historical load data is used to analyze load fluctuation patterns and develop future load forecasts.

[0024] In one possible implementation, load data is collected in real time through power sensors, smart meters and other devices. During the data collection process, the device transmits the data to the control center through a wireless communication network. In some embodiments, the power grid load data includes short-term and long-term load fluctuation information. By integrating this data into the power grid simulation model, future load changes can be effectively predicted to support dynamic scheduling optimization.

[0025] Power generation data refers to the power generation sources of the power grid and its actual output. The power generation units in the power grid usually include traditional thermal power, nuclear power and renewable energy (such as wind power, solar power, etc.). The output of these power generation units is affected by many factors, such as equipment operating status, energy resource availability, meteorological conditions, etc.

[0026] Power generation data mainly includes the following aspects: Output power data of various types of power generation units, including conventional thermal power generation and renewable energy power generation; Equipment status data of the generator set, such as whether it is in the starting, stopping or running state; Renewable energy production data, especially wind power and solar power generation, are greatly affected by meteorological factors.

[0027] Specifically, in one implementation, the power generation data is collected in real time by power monitoring equipment, and the output power of the power generation equipment is continuously tracked using sensors and an automated monitoring system. For renewable energy, the data collection system can conduct a comprehensive analysis in combination with meteorological data to predict its future power generation capacity.

[0028] In the operation of the power grid, meteorological data has a direct impact on the power generation of renewable energy, especially wind and solar power. The meteorological data of the power grid includes not only conventional meteorological factors such as temperature and humidity, but also important parameters that affect the operation of the power grid, such as wind speed, wind direction, and radiation intensity. By collecting real-time meteorological data, the output of renewable energy can be predicted more accurately, thereby optimizing the scheduling decision of the power grid.

[0029] Meteorological data mainly include: Wind speed, direction and temperature data, which directly affect the efficiency of wind turbines and solar panels; Weather forecast data, used to predict meteorological conditions for a period of time in the future, especially those closely related to wind and light conditions; Real-time meteorological monitoring data, by deploying meteorological sensors, continuously tracks information such as wind speed and temperature changes, and transmits this data to the power grid data center.

[0030] In some embodiments, the meteorological data is obtained from real-time data released by local meteorological stations, or collected on-site by deployed meteorological stations. These data can be transmitted through the power network communication system and uniformly analyzed and processed in the power grid control center. The accuracy of meteorological data directly affects the accuracy of renewable energy output forecasts, thereby affecting the load distribution and power generation scheduling of the power grid.

[0031] To ensure the accuracy and timeliness of the data, data collection uses a combination of various equipment and technologies, including: Sensor network: Deploy sensors at key nodes of the power grid and in each power generation unit to monitor load and power generation data in real time; Meteorological monitoring equipment: collect meteorological data through meteorological equipment such as anemometers, temperature and humidity sensors to ensure the accuracy of forecasts; Communication network: The collected power grid data and meteorological data are transmitted to the data processing system in real time through wireless or wired communication.

[0032] In some implementations, a distributed data acquisition architecture is used to ensure the stability and anti-interference capability of data transmission. By integrating this data, sufficient information support can be provided for subsequent source-network coupling simulation, artificial intelligence prediction, and scheduling optimization.

[0033] Data collection is the basis for power grid operation analysis and decision-making. By collecting load, power generation and meteorological data in real time, accurate input can be provided for the construction of source-grid coupling simulation models, training and prediction of artificial intelligence algorithms. The accuracy and completeness of these data directly affect the optimization effect of power grid dispatching and the quality of decision-making.

[0034] Through this data collection mechanism, the system can effectively respond to factors such as grid load fluctuations, power generation uncertainty, and meteorological changes, providing strong support for distribution optimization.

[0035] S2. Based on the collected data in the power grid, a source-grid coupling simulation model is constructed to simulate the power flow, node voltage and dynamic response behavior of the power grid under different operating conditions; In the power grid optimization method, building a source-grid coupling simulation model is a very critical step, which can effectively simulate the power flow, node voltage and dynamic response behavior of the power grid under different operating conditions. Through this model, the operating state of the power grid can be predicted and provide strong support for subsequent dispatch optimization. In general, the accuracy of the simulation model directly affects the quality of the power grid dispatch decision.

[0036] In this embodiment, based on the aforementioned data collection module, the collected grid load data, power generation data and meteorological data will be used as inputs of the source-grid coupling simulation model. Through the real-time input of these data, the source-grid coupling simulation model can dynamically reflect the power flow, voltage distribution and dynamic response of the grid under different operating conditions.

[0037] In one possible implementation, the source-grid coupling simulation model first performs power flow analysis. This analysis module is mainly used to calculate the power distribution and voltage distribution of each node in the power grid. In the power flow calculation, the main parameters involved include node voltage, power flow, load demand, etc.

[0038] Specifically, power flow analysis is based on the grid topology and known generator output and load demand. Each node in the grid is connected by a power balance equation. The mathematical form of the power flow equation is as follows: ; ; in, represents the total number of nodes in the power grid, is the summation variable, and Node The active power and reactive power of and is the admittance matrix element of the power grid, and are the node voltage and phase angle respectively. Through this power flow equation, the power flow and voltage distribution of each node in the power grid can be calculated.

[0039] As an option, the power flow calculation can be simulated by power flow calculation software, solving the above equations using the most common Newton-Raphson method or Gauss-Seidel method.

[0040] In the dynamic response analysis of power grid, the main issue that needs to be paid attention to is the dynamic behavior of the power grid under load fluctuation, generation fluctuation or external disturbance. The dynamic response analysis of power grid aims to simulate the response process of the power grid when sudden load fluctuation, fault or other external disturbance occurs.

[0041] Specifically, the dynamic response of the power grid can be simulated by establishing a dynamic equation that includes the inertia and damping effects of the power grid. The dynamic behavior of the power grid can be expressed by a differential equation of the following general form: ; in, represents the second-order derivative of the node phase angle, represents the derivative of the node phase angle, is the inertia matrix of the power grid, is the damping matrix, is the external disturbance power (such as power fluctuation caused by load change or fault), is the phase angle of the node. By solving this equation, the transient response of the power grid under external disturbance can be simulated and the stability of the power grid can be predicted.

[0042] In one possible implementation, the dynamic response analysis of the power grid not only takes into account load fluctuations and power generation fluctuations, but also the automatic recovery capability of the power grid in the event of a fault. Through simulation, the stability of the power grid can be evaluated in real time to ensure that it can maintain stable operation when encountering external disturbances.

[0043] As the load and power generation of the power grid change, the source-grid coupling simulation model also needs to be continuously updated. In this embodiment, by continuously collecting real-time data from the power grid, the source-grid coupling simulation model will be dynamically adjusted according to the new data. For changes in the load, power generation and meteorological data of the power grid, the simulation model will update the calculation results in real time to ensure that the simulation of the power grid operation status is always accurate.

[0044] In some embodiments, an adaptive learning algorithm can be introduced to adjust the parameters of the simulation model by combining historical data and real-time data, so that the source-grid coupling simulation model is more consistent with the actual grid operation status. This model adaptive capability can effectively cope with external factors such as grid load fluctuations and meteorological changes, and improve the stability and reliability of the system.

[0045] The output of the source-grid coupling simulation model mainly includes the following results: Power flow distribution: Calculate the power flow of each node in the power grid and evaluate the operating status of the power grid.

[0046] Node voltage distribution: Outputs the voltage level of each node to detect the stability of the power grid.

[0047] Dynamic response behavior: simulate the dynamic performance of the power grid under load changes or external disturbances, and evaluate the resilience and stability of the power grid.

[0048] These results will serve as input for subsequent scheduling optimization and provide a basis for load distribution and power generation scheduling of the power grid.

[0049] S3. Use artificial intelligence algorithms to predict grid load data and generate load demand forecasts and renewable energy output forecasts based on the forecast results; In order to achieve dispatch optimization and load forecasting of the power grid, artificial intelligence algorithms are used to predict the load demand of the power grid and the output of renewable energy. Based on the data previously collected from the power grid, including power grid load data, power generation data and meteorological data, artificial intelligence models, especially deep learning algorithms, can efficiently process these data and predict the changing trend of future power grid load demand and the output of renewable energy. Accurate load forecasting results are the basis of dispatch optimization and directly affect the economy, safety and reliability of power grid operation.

[0050] Specifically, this embodiment uses a long short-term memory network (LSTM) model to predict the power grid load. The LSTM model is a deep learning model widely used in time series data processing, which can effectively capture the temporal relationship in load data, especially in the presence of long-term dependence and seasonal changes.

[0051] In general, load data has obvious time series characteristics, including daily load fluctuations, seasonal changes, and holiday effects. The LSTM model can generate accurate load demand forecasts by learning the time series characteristics of historical load data. The input data of the LSTM model includes historical power grid load data, which contains the changes in power grid load in different time periods, such as hourly, daily, or monthly load data.

[0052] In one possible implementation, the LSTM model extracts long-term dependent features by training historical load data, and then predicts future load demand. Specifically, the basic structure of the LSTM model can be expressed in the following form: ; ; in, represents the activation function, represents the hidden layer output at the current moment, is the input load data, is the hidden layer output at the previous moment, , , is the weight matrix, , is the bias term, is the output of the model and represents the predicted load demand.

[0053] By training on a large amount of historical data, the LSTM model can capture the time dependency in the load data and thus effectively predict future load demand.

[0054] In addition to load forecasting, the output of renewable energy in the power grid is also strongly affected by meteorological factors, especially wind power and solar power. In this embodiment, the LSTM model is used to predict the output of wind and solar power. By combining meteorological data (such as wind speed, wind direction, sunshine intensity, etc.) with historical power generation data, the LSTM model can predict future renewable energy output.

[0055] Specifically, when using LSTM to predict renewable energy production, the input data includes: Wind speed and direction data: used to predict wind power output; Sunshine intensity and temperature data: used to predict solar power output; Historical power generation data: Combine meteorological conditions and historical production data to enhance forecast accuracy.

[0056] In some embodiments, the LSTM model may need to be trained and adjusted in multiple stages, first by jointly training the meteorological data and the historical power generation data, and then optimizing and adjusting the power generation of renewable energy based on the prediction results. The output of the LSTM model is the predicted output of renewable energy in the future, and these predicted values ​​will play a key role in the subsequent scheduling optimization.

[0057] The prediction of load demand and renewable energy output are not just two independent processes, but closely linked. The predicted load demand and renewable energy output jointly determine the dispatch strategy of the power grid. Therefore, in practical applications, load demand forecast and renewable energy output forecast will serve as input data to provide a basis for the operation, dispatch and optimization of the power grid.

[0058] In one possible implementation, the forecast results of load demand and renewable energy production will be combined into an optimization algorithm input. The optimization algorithm will dynamically adjust the power generation plan of the power grid according to the predicted load demand and renewable energy production, ensuring the stable operation of the power grid and maximizing the utilization efficiency of renewable energy.

[0059] In order to ensure the accuracy of load forecasting and renewable energy production forecasting, the model needs to be updated regularly. In some embodiments, the LSTM model can be retrained regularly based on new historical data and can be adaptively adjusted based on real-time collected grid data. For example, when the grid load fluctuation changes significantly, the LSTM model will automatically adjust its prediction algorithm parameters to improve the accuracy of the prediction.

[0060] Through this adaptive adjustment, the model can effectively respond to sudden changes in load demand and renewable energy output, and improve the accuracy and stability of power grid scheduling.

[0061] S4. Based on the load-source-grid coupling simulation model and prediction results, the dispatch optimization algorithm is used to dynamically adjust the power generation plan and load distribution to generate distribution optimization decisions; Based on the aforementioned collected grid data, the constructed source-grid coupling simulation model and artificial intelligence prediction results, the dispatch optimization algorithm is used to dynamically adjust the power grid's power generation plan and load distribution, thereby generating distribution optimization decisions. Through this decision-making process, the utilization rate of renewable energy can be maximized while ensuring the stability of the grid, and an optimal balance can be achieved in the economy, environmental protection and safety of the grid.

[0062] The main task of the dispatch optimization algorithm is to optimize the power generation plan and load distribution of the power grid based on the output of the source-grid coupling simulation model and the prediction results of load demand and renewable energy output. To achieve this goal, the dispatch optimization algorithm needs to consider multiple factors, including the load demand of the power grid, the characteristics of the generator set, the output of renewable energy, and the safety constraints of the power grid.

[0063] In general, dispatch optimization algorithms can use methods such as integer linear programming (ILP), nonlinear programming (NLP) or mixed integer programming (MIP). These algorithms use mathematical models to represent the operating constraints of the power grid and optimize the generation plan and load distribution based on this.

[0064] Specifically, the scheduling optimization algorithm in this embodiment will dynamically adjust the power generation plan of the power grid according to the output results of the source-grid coupling simulation model. By analyzing the real-time load demand and renewable energy output of the power grid, the scheduling optimization algorithm will perform power generation scheduling according to the following objectives: Economic goal: Minimize the operating cost of the power grid. Specifically, select the appropriate combination of generators to ensure that the load demand is met at the lowest cost. Under this goal, the dispatch optimization algorithm will optimize based on the fuel cost of the generators and the utilization efficiency of renewable energy.

[0065] Environmental protection goal: Improve the utilization rate of renewable energy. For renewable energy generation such as wind and solar power, the dispatch optimization algorithm will prioritize the dispatch of renewable energy generation based on weather forecast results, reduce dependence on traditional thermal power generation, and reduce carbon emissions from the power grid.

[0066] In one possible implementation, the adjustment of the power generation plan can be solved by the following optimization model: ; in, The total number of generator sets, which indicates the number of all generator sets participating in the dispatch in the power grid. Indicates The operating cost of a generator set, is the power generation of the generator set. By solving the objective function, the dispatch optimization algorithm can determine the optimal output of each generator set to minimize the operating cost of the power grid.

[0067] While adjusting the power generation plan, the dispatch optimization algorithm also needs to optimize the load distribution of the power grid. In this embodiment, the optimization of load distribution not only takes into account the total load demand of the power grid, but also needs to balance the load distribution between different regions to avoid unstable operation caused by excessive load in some areas of the power grid.

[0068] Specifically, the load dispatch optimization problem can be expressed by the following objective function: ; in, For the The load weight factor of the area, For the Regional load demand. By optimizing the load distribution in each region, it is possible to ensure the reasonable distribution of load in the power grid and avoid load imbalance.

[0069] As an option, the dispatch optimization algorithm can be dynamically adjusted through the reinforcement learning algorithm. By providing real-time feedback on the load changes and power generation plans of the power grid, the reinforcement learning algorithm can optimize the load distribution strategy in continuous interaction and ultimately achieve the optimal state of power grid operation.

[0070] When optimizing dispatch, multiple constraints of the power grid need to be considered. These constraints mainly include: Power balance constraint: The total load demand in the grid must be equal to the total generated power of all generators and the contribution of renewable energy sources. For example, the power balance equation can be expressed as: ; in, For the The power generation capacity of each generator set, For the The output of renewable energy generation units, The load demand of the power grid.

[0071] Voltage constraint: The voltage of each node of the power grid must be kept within the allowed range to ensure the stability of the power grid. In this embodiment, the voltage constraint can be expressed by the following formula: ; in, and are the upper and lower limits of the node voltage, For the The voltage of a node.

[0072] Generator set operation constraints: The output power of each generator set must be within its minimum and maximum power output range. The specific constraints can be expressed as: ; in, and The minimum and maximum power generation capacity of each generator set.

[0073] By solving the dispatch optimization algorithm, the power distribution optimization decision of the power grid can be generated. These decisions include not only the dispatch plan of each generator set, but also the load distribution plan of each load node. Ultimately, the optimization decision will be transmitted to the power grid control system, which will perform actual operations based on the optimization decision, adjust the power generation plan and load distribution to ensure the stable and efficient operation of the power grid.

[0074] S5. According to the power distribution optimization decision, the corresponding dispatching instructions are executed through the power grid control system to adjust the operating status of the power grid.

[0075] The generation of distribution optimization decisions is a key step in the grid dispatching process. Through the above steps, the source-grid coupling simulation model and artificial intelligence prediction results are used to generate distribution optimization decisions through the dispatching optimization algorithm. This decision reflects the dispatching plan of each generator set in the grid and the load distribution plan of each load node. In order to make the grid operate in the optimal way, the corresponding dispatching instructions must be executed through the grid control system to dynamically adjust the operation status of the grid.

[0076] In general, the main function of the power grid control system is to adjust the power grid's power generation plan, load distribution, and operating parameters in real time based on the power distribution optimization decision. Specifically, the power grid control system will issue dispatch instructions based on the optimization decision, including but not limited to generator set start and stop, load transfer, load dispatch, backup power start, and other operations. The control system needs to ensure the safe, stable, and economical operation of the power grid, and make dynamic adjustments based on real-time data.

[0077] In one possible implementation, the workflow of the power grid control system includes the following main steps: Receiving distribution optimization decisions: The power grid control system receives distribution optimization decisions from the dispatch optimization algorithm and obtains information such as the dispatch plan and load distribution plan of each generator set; Execute dispatch instructions: Based on the optimization decision, the automated dispatch system executes specific operation instructions, such as scheduling the start or shutdown of generator sets, adjusting load distribution, etc.; Monitoring and feedback: The power grid control system monitors the power grid operation status in real time, feeds back actual operation data and compares it with the optimization decision to ensure that the power grid operates stably according to the predetermined plan.

[0078] Specifically, in the distribution optimization decision, the dispatch plan and load distribution plan of each generator set need to be executed by the power grid control system to adjust the operation status of the power grid. The dispatch instruction includes multiple aspects, which are mainly reflected in the following aspects: Generator group dispatching: According to the power generation plan in the distribution optimization decision, the power grid control system issues instructions to adjust the output level of each generator group. If a generator group needs to be started or shut down, the system will issue instructions based on the start and stop time of the generator group, the dispatching strategy and the changes in the power grid load to ensure that the generator group operates at the most appropriate time.

[0079] For example, if the grid load demand increases, the control system can instruct some units to increase power generation or start standby generators. The corresponding dispatch equation can be: ; in, For the The new output of the generator sets, is the original output of the unit. The power increment adjusted for the dispatch instruction.

[0080] Load distribution adjustment: According to the load distribution plan in the power distribution optimization decision, the power grid control system will dynamically adjust the load of each node in the power grid. The control system reasonably distributes the load demand in the power grid to various areas by adjusting load transfer and optimizing power flow, avoiding power grid instability caused by excessive load at a certain node.

[0081] Specifically, the power grid control system will adjust the load distribution in a timely manner according to the difference between the actual load demand and the predicted load. For example, by switching loads or enabling load adjustment devices, the load balance between different areas is maintained. The adjustment of load distribution can be calculated by the following formula: ; in, For the The new load distribution value for the zone, is the original load of the area, The load increase or decrease issued by the dispatch instruction.

[0082] Dispatching of renewable energy: In the distribution optimization decision, it is also necessary to ensure the use of renewable energy as much as possible and reduce the dependence on traditional energy. Therefore, the grid control system will prioritize the dispatch of renewable energy generation units such as wind power and solar power according to the forecast results of renewable energy production.

[0083] For example, when the wind power generation forecast is high, the grid control system will give priority to increasing the output of wind turbines while reducing the output of other generators. The control system will dynamically adjust the generators based on the forecast results to maximize the utilization of renewable energy.

[0084] The dispatch of renewable energy may need to meet the following constraints: ; in, For the The new output of renewable energy generation units, is the original output of the unit, The increase in renewable energy generation adjusted for dispatch instructions.

[0085] During the operation of the power grid, the power grid control system also needs to monitor the operation status of the power grid in real time, especially key parameters such as power grid load, power generation output, node voltage, etc. When there is a difference between the actual operation status of the power grid and the expected optimization decision, the control system will adjust the dispatching instructions in time through the feedback mechanism to ensure the stable operation of the power grid.

[0086] For example, if a voltage anomaly occurs at a node in the power grid, the power grid control system will immediately adjust the power generation plan or load distribution to ensure that the voltage at the node returns to a safe range.

[0087] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A power distribution optimization method integrating a source network simulation model and an artificial intelligence algorithm, characterized in that: The following steps are involved: Collecting data in the power grid, including power grid load data, power generation data and meteorological data; Based on the collected data from the power grid, a source-grid coupling simulation model is constructed to simulate the power flow, node voltage and dynamic response behavior of the power grid under different operating conditions; Use artificial intelligence algorithms to predict grid load data and generate load demand forecasts and renewable energy output forecasts based on the forecast results; Based on the load-source-grid coupling simulation model and prediction results, the dispatch optimization algorithm is used to dynamically adjust the power generation plan and load distribution to generate distribution optimization decisions; According to the distribution optimization decision, the corresponding dispatching instructions are executed through the power grid control system to adjust the operating status of the power grid.

2. The power distribution optimization method integrating the source network simulation model and the artificial intelligence algorithm according to claim 1 is characterized in that: The source-network coupling simulation model includes: Power flow analysis, used to simulate power flow in the power grid and calculate power allocation and voltage distribution at each node; Grid dynamic response analysis is used to simulate the response of the grid under load fluctuations and external disturbances, including grid inertia and damping effects.

3. The power distribution optimization method integrating source network simulation model and artificial intelligence algorithm according to claim 1 is characterized in that: The artificial intelligence algorithm includes: Long short-term memory network models for processing time series data to predict grid load demand and renewable energy production; The LSTM network model is trained based on historical data to generate prediction results of future load and output.

4. The power distribution optimization method integrating the source network simulation model and the artificial intelligence algorithm according to claim 3 is characterized in that: The input of the long short-term memory network model includes: Historical load data and historical renewable energy production data; The load forecast results and renewable energy output forecast results output by the long short-term memory network model serve as the basis for scheduling optimization.

5. The power distribution optimization method integrating source network simulation model and artificial intelligence algorithm according to claim 1 is characterized in that: The scheduling optimization algorithm includes: Feedback mechanism during training, used to adjust dispatch strategy according to real-time grid status; The power generation plan and load distribution are optimized through the reward function to maintain the stable operation of the power grid.

6. The power distribution optimization method integrating the source network simulation model and the artificial intelligence algorithm according to claim 5 is characterized in that: The reward function includes: The operating cost of the power grid, the matching degree of load demand and the utilization rate of renewable energy; Through reward function optimization, the reinforcement learning model is able to adjust the dispatch strategy in real time to minimize the total cost of grid operation.

7. The power distribution optimization method integrating source network simulation model and artificial intelligence algorithm according to claim 1 is characterized in that: The distribution optimization decision includes: The economics of the grid, minimizing operating costs; The security of the power grid, ensuring that power flow and voltage stability indicators meet constraints; Environmental friendliness of the power grid and optimization of the use of renewable energy.

8. The power distribution optimization method integrating source network simulation model and artificial intelligence algorithm according to claim 7 is characterized in that: The power distribution optimization decision further includes: By setting the weight of each objective function, multiple objective functions are weighted summed; The power flow, load limit and voltage constraints of the power grid are considered during the optimization process to ensure that the optimization results meet the actual operation requirements.

9. The power distribution optimization method integrating source network simulation model and artificial intelligence algorithm according to claim 1 is characterized in that: The power grid dynamic response analysis includes: Grid stability analysis, used to evaluate the grid’s ability to respond to sudden load fluctuations or external disturbances; Dynamic dispatch is used to adjust the grid dispatch strategy in a timely manner according to the dynamic response results when a disturbance occurs in the grid.

10. The power distribution optimization method integrating source network simulation model and artificial intelligence algorithm according to claim 1, characterized in that: The artificial intelligence algorithm further comprises: An update mechanism for load and output forecast results, which is used to dynamically adjust the parameters of the forecast model based on historical data during grid operation; By adjusting the model parameters, the accuracy of the prediction results can be optimized.

Images