A flexible direct current power transmission network loss optimization scheduling method and system and storage medium

Abstract

The present application belongs to the technical field of power transmission loss optimization, and particularly relates to a flexible DC power transmission network loss optimization scheduling method and system and a storage medium. The loss scheduling scheme of the flexible DC power transmission network is analyzed, and the power transmission region and the corresponding power transmission background in which the loss does not meet the scheduling loss standard under the loss scheduling scheme are determined. The power transmission region in which the loss does not meet the scheduling loss standard is taken as a scheduling optimization region, the regional monitoring data of the scheduling optimization region is obtained in real time, whether the scheduling optimization region is in the power transmission background is judged in real time based on the current regional monitoring data, the current probabilistic load flow analysis result of the scheduling optimization region is obtained if it is in the power transmission background, and the loss scheduling scheme is optimized according to the loss scheduling scheme, the target monitoring data and the current probabilistic load flow analysis result of the scheduling optimization region. The loss characteristics of the flexible DC power transmission network under different operating conditions can be accurately adapted, and the network loss is significantly reduced.

Description

Technical Field

[0001] This invention belongs to the field of power transmission loss optimization technology, specifically relating to a method, system, and storage medium for optimizing the scheduling of losses in flexible DC transmission networks. Background Technology

[0002] With the continuous expansion of flexible DC transmission networks and their widespread application in power systems, network losses have become increasingly prominent, becoming a key factor restricting their further development and efficient operation. For example, Chinese patent application CN115241885A discloses a method, system, computer equipment, and storage medium for real-time power grid dispatch optimization. This method acquires power grid model parameters and power grid operation data; based on the power grid model parameters and operation data, it obtains real-time power grid dispatch adjustment strategies through a pre-set real-time power grid dispatch reinforcement learning training model. The state space of the reinforcement learning training environment includes generator active power output, generator reactive power output, generator voltage amplitude, load active power, load reactive power, load voltage amplitude, energy storage battery charging and discharging power, line status, line load rate, and power grid losses.

[0003] Network losses in flexible DC transmission networks not only directly lead to energy waste and increased operating costs, but can also affect system voltage stability and power quality, reducing the overall efficiency and reliability of the power system. Therefore, reducing network losses, improving energy efficiency, and enhancing the economic efficiency and stability of power system operation are of paramount practical significance. For each region of a flexible DC transmission network, a fixed network loss scheduling scheme is typically formulated based on the power equipment in that region. However, in actual operation, flexible DC transmission networks face the complex influence of various uncertainties. These uncertainties are intertwined and coupled, posing a significant challenge to the accurate assessment and effective control of network losses, thus highlighting the urgent necessity to optimize loss scheduling methods for user applications.

[0004] For example, with the vigorous development and utilization of clean energy, the proportion of renewable energy sources such as wind and solar power integrated into flexible DC transmission networks is constantly increasing. However, the output of renewable energy is characterized by significant intermittency, volatility, and unpredictability. Wind power output is affected by meteorological conditions such as wind speed and direction, while photovoltaic output is closely related to factors such as sunlight intensity and temperature. These uncertainties lead to real-time fluctuations in renewable energy power generation, making it difficult for the relatively fixed loss scheduling schemes of flexible DC transmission networks to cope with the rapid changes in renewable energy output. This results in inconsistencies between the original flexible DC network losses and the actual operating losses. When the actual network losses are high, it reduces the power quality of the grid and may also cause grid voltage fluctuations, thereby affecting the operational stability of the entire power system. Summary of the Invention

[0005] The purpose of this invention is to provide a method, system, and storage medium for optimizing the scheduling of losses in flexible DC transmission networks, in order to solve the following problems.

[0006] To address the aforementioned technical problems, the first aspect of this invention provides a method for optimizing the scheduling of losses in a flexible DC transmission network, comprising: The loss scheduling scheme of flexible DC transmission network is analyzed to determine the transmission areas whose losses do not meet the scheduling loss standard and the corresponding transmission background under the loss scheduling scheme. The transmission areas whose losses do not meet the scheduling loss standard are taken as the scheduling optimization areas. Real-time acquisition of regional monitoring data for the scheduling optimization area; real-time determination of whether the scheduling optimization area is under power transmission background based on the current regional monitoring data; if it is, acquisition of the current probability power flow analysis results for the scheduling optimization area. Real-time target monitoring data of the target scheduling area of ​​the flexible DC transmission network is acquired, and the loss scheduling scheme is optimized based on the loss scheduling scheme, target monitoring data, and the current probabilistic power flow analysis results of the scheduling optimization area.

[0007] In one possible implementation, the process of determining the transmission area where losses do not meet the scheduling loss criteria and the corresponding transmission background includes: Obtain historical loss scheduling data for the application loss scheduling scheme of flexible DC transmission network. The historical loss scheduling data is the loss scheduling data from the current time to the time node corresponding to the last adjustment or change of flexible DC transmission network or loss scheduling scheme. Based on historical loss scheduling data, transmission network areas that do not meet the scheduling loss criteria are identified. These transmission network areas are then merged to obtain the transmission areas whose losses do not meet the scheduling loss criteria. Finally, the transmission background corresponding to each transmission network area is integrated to obtain the transmission background corresponding to that transmission area.

[0008] In one possible implementation, the process of determining whether the scheduling optimization area is under a power transmission background includes: performing feature identification on the current area monitoring data according to background acquisition items to obtain corresponding background feature combinations, and determining whether the scheduling optimization area is under a power transmission background based on the background feature combinations; the background acquisition items are parameter items of the power transmission background.

[0009] In one possible implementation, the current probabilistic power flow analysis result of the scheduling optimization area is obtained by analyzing the power flow input data using a power flow analysis model, and the power flow input data is generated based on the current area monitoring data.

[0010] In one possible implementation, the current probabilistic power flow analysis result of the scheduling optimization area is determined by combining the results obtained from analyzing the power flow input data using the power flow analysis model and the power flow prediction model respectively. The power flow input data is generated based on the current area monitoring data, and the power flow prediction model is obtained by training a deep learning algorithm or machine learning model using the power flow input data and the probabilistic power flow analysis result.

[0011] In one possible implementation, the probabilistic power flow analysis results include loss scheduling patterns in the scheduling optimization region under the current transmission context.

[0012] In one possible implementation, the regional monitoring data includes power transmission data, time period data, meteorological data, and output data of renewable energy power generation equipment.

[0013] In one possible implementation, the transmission areas whose losses do not meet the scheduling loss criteria and the corresponding transmission background are obtained by analyzing historical loss scheduling data using an intelligent evaluation model. The historical loss scheduling data is the loss scheduling data from the current time to the time node corresponding to the last adjustment or change of the flexible DC transmission network or the adjustment or change of the loss scheduling scheme. The intelligent evaluation model adopts deep learning algorithms or machine learning algorithms.

[0014] To address the aforementioned technical problems, a second aspect of the present invention provides a loss optimization scheduling system for flexible DC transmission networks, comprising a processor that executes a computer program to implement the steps of the above-described method.

[0015] To address the aforementioned technical problems, a third aspect of the present invention provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of the above-described method.

[0016] The beneficial effects of this invention are as follows: By analyzing the loss scheduling scheme of flexible DC transmission networks, this invention identifies transmission areas and transmission backgrounds where losses do not meet the scheduling loss standards under the scheme. Through real-time monitoring of data in the corresponding areas, when it is determined that the transmission area is in a transmission background that does not meet the scheduling loss standards, the scheduling scheme is re-determined through power flow analysis and target monitoring data to optimize the loss scheduling method. This allows for precise adaptation to the loss characteristics of the flexible DC transmission network under different operating conditions, significantly reducing network losses. It directly reduces energy waste during transmission, improves energy utilization efficiency, and has positive significance for alleviating energy shortages and achieving sustainable energy use. The optimized loss scheduling method can better cope with the complex influence of various uncertainties in flexible DC transmission networks, such as the intermittency, volatility, and unpredictability of renewable energy output, effectively stabilizing system voltage, improving power quality, and reducing voltage fluctuations and power quality degradation caused by network losses. This enhances the operational stability of the entire power system and ensures the reliability of power supply. Attached Figure Description

[0017] Figure 1 This is a schematic diagram illustrating the principle of loss optimization scheduling in flexible DC transmission networks according to the present invention; Figure 2 This is a schematic diagram of the composition of the flexible DC transmission network loss optimization scheduling system of the present invention. Detailed Implementation

[0018] To make the objectives, technical solutions, and advantages of the present invention clearer, the specific embodiments of the present invention will be further described below with reference to the accompanying drawings.

[0019] This invention analyzes the loss scheduling scheme of flexible DC transmission networks in real time, identifies the scheduling optimization area and transmission background where the loss does not meet the scheduling loss standard, and determines whether the scheduling optimization area is in the corresponding transmission background. If it is, the loss scheduling scheme is optimized based on the probability power flow analysis results and target monitoring data of the scheduling optimization area in the corresponding transmission background. This invention can accurately adapt to the loss characteristics of flexible DC transmission networks under different operating conditions.

[0020] Implementation of Loss Optimization Scheduling Method for Flexible DC Transmission Networks A loss optimization scheduling method for flexible DC transmission networks, the implementation principle of which is as follows: Figure 1 As shown, it includes the following steps: 1. Analyze the loss scheduling scheme of the flexible DC transmission network, identify the transmission areas whose losses do not meet the scheduling loss standard and the corresponding transmission background under the loss scheduling scheme, and take the transmission areas whose losses do not meet the scheduling loss standard as the scheduling optimization areas.

[0021] 2. Real-time acquisition of regional monitoring data for the scheduling optimization area, and real-time determination of whether the scheduling optimization area is under power transmission background based on the current regional monitoring data. If it is, the calculation of the current probability power flow analysis results for the scheduling optimization area.

[0022] 3. Acquire target monitoring data of the target scheduling area of ​​the flexible DC transmission network in real time, and optimize the loss scheduling scheme based on the loss scheduling scheme, target monitoring data and the current probability power flow analysis results of the scheduling optimization area.

[0023] For the same transmission area, due to the influence of changes in uncertain factors, the dispatch loss standard may be met in some cases. Therefore, it is necessary to mark the various combinations of factors that do not meet the dispatch loss standard as the transmission background, so as to facilitate subsequent targeted optimization and adjustment.

[0024] Therefore, in step 1, the current loss scheduling scheme for the flexible DC transmission network is analyzed in real time to determine the scheduling optimization area and the corresponding transmission background where the loss does not meet the scheduling loss standard. This indicates that the current loss scheduling scheme cannot meet the scheduling loss standard in the scheduling optimization area and transmission background, and the loss scheduling in the scheduling optimization area and transmission background needs to be optimized and adjusted so that it can meet the scheduling loss standard in the future. The scheduling optimization area is the transmission area that does not meet the scheduling loss standard.

[0025] By identifying in real-time the scheduling optimization areas and transmission backgrounds where the current loss scheduling scheme does not meet the loss scheduling criteria, the loss scheduling scheme can be optimized specifically for those areas and backgrounds. This can be achieved by directly adjusting the current loss scheduling scheme or by establishing a supplementary loss scheduling scheme for the optimized area and background. When the system is in the optimized area and background, this supplementary scheme is switched for loss scheduling. Therefore, by dynamically determining the optimization area and transmission background, the loss scheduling scheme is continuously updated. Preferably, the loss scheduling scheme is optimized using methods that minimize the scale of adjustments.

[0026] The dispatch loss standard can be set according to actual conditions. For example, it can be designed according to relevant industry standards, or it can be set according to the requirements of the power grid superior, user needs, etc.

[0027] The process of determining the transmission areas whose losses do not meet the dispatch loss criteria and the corresponding transmission background includes: Historical loss scheduling data for flexible DC transmission network application loss scheduling schemes are obtained; based on the historical loss scheduling data, transmission network areas that do not meet the scheduling loss criteria are identified; after merging the transmission network areas, the transmission areas whose losses do not meet the scheduling loss criteria are obtained; and after integrating the transmission background corresponding to the transmission network areas, the transmission background corresponding to the transmission areas is obtained.

[0028] Historical loss scheduling data refers to data that reflects the current loss scheduling status of the flexible DC transmission network, with the time point being the time from the current time to the time corresponding to the last adjustment or change in the flexible DC transmission network or loss scheduling scheme. Correspondingly, historical loss scheduling data is the loss scheduling data from the current time to the time point corresponding to the last adjustment or change in the flexible DC transmission network or loss scheduling scheme.

[0029] In one embodiment, the process of determining the scheduling optimization area and transmission background includes: Obtain the power grid network requirements applicable to probabilistic power flow calculations for analysis, and mark these requirements as power flow region standards. Power flow region standards include complete random variable data, high-quality power grid topology and operational data, etc. Set background data collection items based on factors influencing network losses; both the power flow region standards and background data collection items are pre-set.

[0030] Users set scheduling loss standards to obtain the target scheduling area of ​​the flexible DC transmission network. The target scheduling area is the overall transmission area corresponding to the flexible DC transmission network.

[0031] Real-time acquisition of historical loss scheduling data for the application of loss scheduling schemes in flexible DC transmission networks; identification of corresponding historical loss scheduling results based on historical loss scheduling data, such as loss scheduling results for converter losses, DC line losses, and transformer losses.

[0032] Based on the dispatch loss standard, historical loss dispatch results are evaluated to identify transmission network areas that do not meet the standard. The number of such areas can vary depending on the actual situation. If all the above criteria are met, and no transmission network areas fail to meet the standard, the corresponding number is 0, indicating that the loss dispatch scheme meets the standard and no further analysis is needed. If one or more transmission network areas fail to meet the standard, the historical loss dispatch data is processed using background data collection to determine the corresponding transmission background for these areas.

[0033] Based on the power flow area standard, transmission network areas that meet the power flow area standard but do not meet the dispatch loss standard are merged to obtain the corresponding transmission areas. If there is only one transmission network that does not meet the dispatch loss standard, no merging analysis is required. The corresponding transmission backgrounds are identified based on the transmission network areas corresponding to each transmission area, and these transmission backgrounds are integrated into the transmission background corresponding to that transmission area. In other words, this transmission background is a set of transmission backgrounds corresponding to each transmission network.

[0034] In another embodiment, the scheduling optimization area and transmission background can be determined by existing evaluation methods, such as analyzing historical loss scheduling data in real time based on preset scheduling loss standards and evaluation algorithms to determine the transmission areas that do not meet the scheduling loss standards and the corresponding transmission background.

[0035] For example, an intelligent evaluation model can be established based on machine learning or deep learning algorithms. After successful training, the intelligent evaluation model can be used to analyze historical loss scheduling data in real time to determine the transmission areas where the loss does not meet the scheduling loss standard and the corresponding transmission background.

[0036] In step 2, the regional monitoring data includes relevant data useful for loss analysis in the transmission area, such as transmission data, time-period data, meteorological data, and renewable energy power generation equipment output data. Based on the regional monitoring data, it is determined in real time whether the dispatch optimization area is in a transmission background that does not meet the dispatch loss standards.

[0037] The process of determining in real time whether the scheduling optimization area is under a power transmission background based on current regional monitoring data includes: identifying features of the current regional monitoring data according to background acquisition items to obtain corresponding background feature combinations, and determining whether the scheduling optimization area is under a power transmission background based on the background feature combinations. Background acquisition items are parameters related to the power transmission background.

[0038] When judging the combination of background features using a preset background calibration model, the obtained background calibration results include qualified and unqualified. When the background calibration result is qualified, the transmission area is judged to be in a transmission background; when the background calibration result is unqualified, the transmission area is judged not to be in a transmission background.

[0039] In one embodiment, the expression for the background judgment model is: Where s is the background feature combination, U is the power transmission background set of the scheduling optimization area; BJ(s,U) is the output background judgment value; when the background calibration value is 1, the background calibration result is qualified and the scheduling optimization area is in the power transmission background; when the background calibration value is 0, the background calibration result is unqualified and the scheduling optimization area is not in the power transmission background.

[0040] In one embodiment, the background judgment model can also be built based on existing algorithms, such as machine learning or deep learning algorithms.

[0041] In one embodiment, the determination of whether a power transmission area is under power transmission background can be made in real time based on regional monitoring data, which can be based on existing assessment methods.

[0042] In one embodiment, the probabilistic power flow analysis result is obtained by analyzing power flow input data using a power flow analysis model. The power flow input data is generated based on current area monitoring data and includes: The process involves identifying the scheduling optimization area and transmission background, and establishing a targeted power flow analysis model based on these factors. Specifically, the power flow analysis model is built using probabilistic power flow calculations tailored to the specific transmission area and background. This model is used to perform analysis within the defined transmission area and background to determine the corresponding probabilistic power flow analysis results. These results include the loss scheduling method within the scheduling optimization area under the current transmission background.

[0043] Real-time acquisition of regional monitoring data for the scheduling optimization area, and real-time determination of whether the power transmission area is under power transmission background based on the regional monitoring data.

[0044] If it is determined that the system is not in a power transmission background, no corresponding operation will be performed.

[0045] When it is determined that the system is in a power transmission background, power flow input data for the power flow analysis model is generated based on regional monitoring data. The power flow analysis model is then used to analyze the power flow input data to obtain the corresponding probabilistic power flow analysis results.

[0046] In one embodiment, issues such as timeliness, computing resources, and flexible adjustment in probabilistic power flow calculations lead to other methods for analyzing and optimizing transmission areas and transmission backgrounds based on probabilistic power flow calculations. In other words, the probabilistic power flow analysis results are determined by combining the results obtained from analyzing power flow input data using a power flow analysis model and a power flow prediction model, respectively. The power flow input data is generated based on current regional monitoring data, and the power flow prediction model is obtained by training deep learning algorithms or machine learning models using the power flow input data and probabilistic power flow analysis results.

[0047] For example, when it is determined that the system is in a power transmission background, power flow input data is generated based on regional monitoring data to form a power flow analysis model. This input data is then analyzed by the power flow analysis model to obtain corresponding probabilistic power flow analysis results. These results are then input into a power flow prediction model to obtain predicted probabilistic power flow analysis results. Finally, the probabilistic power flow analysis result is determined based on the two probabilistic power flow analysis results and their corresponding weights.

[0048] A power flow prediction model is established based on deep learning algorithms. This model is trained using prediction data, and once successfully trained, it can predict the probabilistic power flow analysis results based on the power flow input data. Analysis can be performed using historical data, or prediction data can be set based on validation data of the power flow analysis model. The prediction data includes power flow input data and probabilistic power flow analysis results. Specifically, by establishing targeted power flow analysis models for scheduling optimization areas and transmission backgrounds, and using these models to simulate and analyze the scheduling optimization areas and transmission backgrounds, corresponding prediction data is obtained. Historical data refers to the probabilistic power flow analysis results determined by statistically analyzing and matching a large amount of historical power flow input data.

[0049] Specifically, the regional monitoring data of the scheduling optimization area is acquired in real time, and the power transmission area is judged in real time whether it is under power transmission background based on the regional monitoring data.

[0050] No corresponding operation is performed when it is determined that the system is not in a power transmission background.

[0051] When it is determined that the system is in a power transmission background, power flow input data is generated based on regional monitoring data. The power flow input data is then analyzed using both a power flow prediction model and a power flow analysis model to obtain the corresponding probabilistic power flow analysis results. Based on the analysis results obtained from the two models, the current probabilistic power flow analysis results for the scheduling optimization area are determined.

[0052] To differentiate them, the probabilistic power flow analysis results obtained by the power flow analysis model are marked as standard analysis results. The standard analysis results and power flow input data can also be integrated into self-learning data, and the power flow prediction model can be learned and adjusted based on the self-learning data.

[0053] In one embodiment, the power flow analysis model is built based on existing methods.

[0054] For example, the establishment of a power flow analysis model includes: Define the system scope: Determine the power grid scope covered by the power flow analysis model, including power plants, substations, transmission lines, distribution networks, and various load nodes. For example, for a regional power grid, it is necessary to define the power generation facilities it includes, such as thermal power plants, hydropower plants, and wind farms, as well as substations of different voltage levels and the transmission lines connecting them.

[0055] Select model type: Node admittance matrix model: Applicable to small and medium-sized power grids, it describes the topology and electrical parameters of the power grid based on the node admittance matrix. The elements of the node admittance matrix reflect the electrical connections and admittance values ​​between nodes. The power flow distribution of the power grid can be obtained by solving the node voltage equations.

[0056] Network topology models focus more on describing the physical connections of the power grid, representing components (such as lines and transformers) as edges and nodes as connection points. This model facilitates the analysis of power grid connectivity and reliability, and has certain advantages when dealing with large-scale power grids.

[0057] Component modeling: Generator: Establish active and reactive power output models for the generator, considering its rated power, power factor, speed governor, and excitation system characteristics.

[0058] Transformer: Determine the transformer's turns ratio, impedance, and admittance parameters, considering its tap adjustment capability. The transformer's turns ratio affects the voltage level and power flow distribution of the power grid, while tap adjustment can be used to regulate voltage and reactive power.

[0059] Transmission lines: A line model is established based on the line's resistance, reactance, and susceptance parameters, taking into account the influence of factors such as line length, type, and operating temperature on the parameters.

[0060] Determine the initial running conditions: Reference voltage and power: Selecting appropriate reference voltage and power is used to normalize the actual parameters of the power grid, facilitating calculation and analysis. For example, for high-voltage transmission networks, the rated voltage is usually selected as the reference voltage, and the rated capacity as the reference power.

[0061] Initial power flow distribution: The power flow distribution of the power grid under given initial operating conditions is calculated using deterministic power flow calculation methods (such as the Newton-Raphson method, PQ decomposition method, etc.), which serves as the basis for probabilistic power flow calculation.

[0062] Identifying uncertain variables: Load uncertainty: Analyze the randomness and volatility of the load to identify the main factors affecting the load, such as time, season, weather, and economic activities.

[0063] Uncertainty in renewable energy output: Considering the intermittency and volatility of renewable energy sources such as wind and solar power, meteorological data (such as wind speed and solar intensity) and characteristic parameters of renewable energy power generation equipment are collected to analyze the probability distribution of their output.

[0064] Equipment parameter uncertainty: Identify the uncertainty of equipment parameters in the power grid, such as the parameters of components such as generators, transformers, and lines may deviate due to factors such as manufacturing process, operating environment, and aging degree.

[0065] Establish an uncertainty model: Probability distribution model: Based on the collected data, a suitable probability distribution function is selected for each uncertain variable, such as the normal distribution, Weibull distribution, or log-normal distribution. For example, the daily variation of load may follow a normal distribution, and wind speed may follow a Rayleigh distribution.

[0066] Correlation model: Considering the correlation between uncertain variables, an appropriate method (such as the Copula function) is used to describe their joint probability distribution.

[0067] Sampling method selection: Monte Carlo simulation: This method generates a large number of uncertain variable samples through random sampling, performs deterministic power flow calculations on each sample, and obtains the statistical characteristics of the output variables (such as node voltage, branch power flow, network loss, etc.). The Monte Carlo simulation method is simple in principle, but computationally intensive.

[0068] Latin hypercube sampling: Compared to Monte Carlo simulation, Latin hypercube sampling can obtain more accurate statistical results with a smaller sample size. It divides the range of values ​​for each uncertain variable into several equally probable intervals, and then samples within each interval.

[0069] Quasi-Monte Carlo simulation: This method uses low-difference sequences instead of random sequences for sampling, further improving sampling efficiency and computational accuracy.

[0070] Establish the probabilistic power flow equations: By combining deterministic power flow equations with the probability distributions of uncertain variables, probabilistic power flow equations are established.

[0071] Network loss calculation: In each deterministic power flow calculation, the network loss of the power grid is calculated based on the power flow distribution. The network loss mainly includes the active power loss of transmission lines and the active power loss of transformers.

[0072] Model validation: Comparison with actual data: Collect actual power grid operation data, including load data, renewable energy output data, and network loss data, and compare the probabilistic power flow calculation results with the actual data to verify the accuracy of the model.

[0073] Sensitivity analysis: This analyzes the impact of changes in uncertain variables on network loss. Sensitivity analysis can identify key factors that have a significant impact on network loss, providing a basis for model optimization and improvement.

[0074] In one embodiment, probabilistic power flow calculation can directly determine the loss scheduling of the optimized scheduling area under the corresponding transmission background; that is, the probabilistic power flow analysis result includes the loss scheduling of the optimized scheduling area under the corresponding transmission background. Based on the probabilistic power flow analysis result, an optimal scheduling strategy can be formulated to reduce system losses. For example, by adjusting generator output and connecting energy storage for peak shaving and valley filling, the power flow distribution can be changed, reducing line losses. Therefore, it can be combined with existing loss scheduling to form a probabilistic power flow analysis result that includes the corresponding loss scheduling methods.

[0075] In step 3, target monitoring data for the target scheduling area and the current probabilistic power flow analysis results for the scheduling optimization area are acquired in real time. Target monitoring data integrates the monitoring data from the target scheduling area, and scheduling processing is performed based on the loss scheduling scheme, target monitoring data, and probabilistic power flow analysis results. For example, the loss scheduling scheme is continuously adjusted and optimized based on actual operational data and probabilistic power flow analysis results to improve the scheme's adaptability and effectiveness.

[0076] Based on the power grid risk indicators and optimized dispatch schemes obtained from probabilistic power flow calculations, the power grid control system can automatically adjust generator output, transformer tap positions, and reactive power compensation equipment switching to reduce power grid operational risks and improve power grid security and economy. For example, when probabilistic power flow calculations show that a certain transmission line has a high probability of overload, the power grid control system can promptly adjust the output of relevant generators to reduce the power flow on that line and avoid line overload.

[0077] Implementation of Flexible DC Transmission Network Loss Optimization Scheduling System A flexible DC transmission network loss optimization scheduling system, such as Figure 2 As shown, the system includes a memory, a processor, a system bus, and a computer program stored in the memory. The processor and the memory communicate and exchange data via the system bus. The processor executes the computer program to implement the steps of the flexible DC transmission network loss optimization scheduling method of the present invention. The processor can be a microprocessor (MCU) or other processing device; the memory can be any type of memory that stores information using electrical energy, such as non-volatile storage media (including computer programs, databases), or other types of memory.

[0078] The system's functional modules include a scheduling analysis module, an analysis and optimization module, and a scheduling module.

[0079] The scheduling analysis module analyzes the loss scheduling scheme of the flexible DC transmission network, identifies transmission areas and transmission backgrounds where losses do not meet the scheduling loss criteria under the given scheme, and designates these areas as scheduling optimization areas. The analysis and optimization module acquires real-time monitoring data for the scheduling optimization areas and determines whether the area is under a transmission background based on this data. If so, it calculates the current probabilistic power flow analysis results for the optimization area. The scheduling module acquires real-time target monitoring data for the target scheduling area of ​​the flexible DC transmission network and optimizes the loss scheduling scheme based on the loss scheduling scheme, target monitoring data, and the current probabilistic power flow analysis results for the optimization area.

[0080] Specifically, the scheduling analysis module performs real-time analysis on the current loss scheduling scheme for the flexible DC transmission network, identifies the scheduling optimization area and corresponding transmission background where the loss does not meet the scheduling loss standard, indicating that the current loss scheduling scheme cannot meet the scheduling loss standard in the scheduling optimization area and transmission background, and needs to be optimized and adjusted in the scheduling optimization area and transmission background so that it can meet the scheduling loss standard in the future. The scheduling optimization area is the transmission area that does not meet the scheduling loss standard.

[0081] The dispatch loss standard can be set according to actual conditions. For example, it can be designed according to relevant industry standards, or it can be set according to the requirements of the power grid superior, user needs, etc.

[0082] The scheduling analysis module acquires historical loss scheduling data of the flexible DC transmission network application loss scheduling scheme; based on the historical loss scheduling data, it determines the transmission network areas that do not meet the scheduling loss standard, merges the transmission network areas to obtain the transmission areas whose losses do not meet the scheduling loss standard, and integrates the transmission background corresponding to the transmission network areas to obtain the transmission background corresponding to the transmission area.

[0083] Historical loss scheduling data refers to data that reflects the current loss scheduling status of the flexible DC transmission network, with the time point being the time from the current time to the time corresponding to the last adjustment or change in the flexible DC transmission network or loss scheduling scheme. Correspondingly, historical loss scheduling data is the loss scheduling data from the current time to the time point corresponding to the last adjustment or change in the flexible DC transmission network or loss scheduling scheme.

[0084] In one embodiment, the process by which the scheduling analysis module determines the scheduling optimization area and transmission background includes: Obtain the power grid network requirements applicable to probabilistic power flow calculations for analysis, and mark these requirements as power flow region standards. Power flow region standards include complete random variable data, high-quality power grid topology and operational data, etc. Set background data collection items based on factors influencing network losses; both the power flow region standards and background data collection items are pre-set.

[0085] Users set scheduling loss standards to obtain the target scheduling area of ​​the flexible DC transmission network. The target scheduling area is the overall transmission area corresponding to the flexible DC transmission network.

[0086] Real-time acquisition of historical loss scheduling data for the application of loss scheduling schemes in flexible DC transmission networks; identification of corresponding historical loss scheduling results based on historical loss scheduling data, such as loss scheduling results for converter losses, DC line losses, and transformer losses.

[0087] Based on the dispatch loss standard, historical loss dispatch results are evaluated to identify transmission network areas that do not meet the standard. The number of such areas can vary depending on the actual situation. If all the above criteria are met, and no transmission network areas fail to meet the standard, the corresponding number is 0, indicating that the loss dispatch scheme meets the standard and no further analysis is needed. If one or more transmission network areas fail to meet the standard, the historical loss dispatch data is processed using background data collection to determine the corresponding transmission background for these areas.

[0088] Based on the power flow area standard, transmission network areas that meet the power flow area standard but do not meet the dispatch loss standard are merged to obtain the corresponding transmission areas. If there is only one transmission network that does not meet the dispatch loss standard, no merging analysis is required. The corresponding transmission backgrounds are identified based on the transmission network areas corresponding to each transmission area, and these transmission backgrounds are integrated into the transmission background corresponding to that transmission area. In other words, this transmission background is a set of transmission backgrounds corresponding to each transmission network.

[0089] In another embodiment, the scheduling analysis module can evaluate and determine the scheduling optimization area and transmission background using existing evaluation methods, such as performing real-time analysis of historical loss scheduling data based on preset scheduling loss standards and evaluation algorithms to determine the transmission areas that do not meet the scheduling loss standards and the corresponding transmission background.

[0090] For example, an intelligent evaluation model can be established based on machine learning or deep learning algorithms. After successful training, the intelligent evaluation model can be used to analyze historical loss scheduling data in real time to determine the transmission areas and transmission backgrounds that do not meet the scheduling loss standards.

[0091] Regional monitoring data includes transmission data, time-period data, meteorological data, and renewable energy power generation equipment output data—all relevant data useful for loss analysis in transmission areas. Based on this regional monitoring data, it is determined in real time whether the dispatch optimization area is in a transmission background that does not meet dispatch loss standards.

[0092] The analysis and optimization module's process of determining whether the scheduling optimization area is under a power transmission background based on current regional monitoring data includes: performing feature identification on the current regional monitoring data according to background acquisition items to obtain corresponding background feature combinations, and determining whether the scheduling optimization area is under a power transmission background based on these background feature combinations. Background acquisition items are parameters related to the power transmission background.

[0093] When judging the combination of background features using a preset background calibration model, the obtained background calibration results include qualified and unqualified. When the background calibration result is qualified, the transmission area is judged to be in a transmission background; when the background calibration result is unqualified, the transmission area is judged not to be in a transmission background.

[0094] In one embodiment, the expression for the background judgment model is: Where s is the background feature combination, U is the power transmission background set of the scheduling optimization area; BJ(s,U) is the output background judgment value; when the background calibration value is 1, the background calibration result is qualified and the scheduling optimization area is in the power transmission background; when the background calibration value is 0, the background calibration result is unqualified and the scheduling optimization area is not in the power transmission background.

[0095] In one embodiment, the background judgment model can also be built based on existing algorithms, such as machine learning or deep learning algorithms.

[0096] In one embodiment, the determination of whether a power transmission area is under power transmission background can be made in real time based on regional monitoring data, which can be based on existing assessment methods.

[0097] In one embodiment, the probabilistic power flow analysis result is obtained by analyzing power flow input data using a power flow analysis model. The power flow input data is generated based on current area monitoring data and includes: The process involves identifying the scheduling optimization area and transmission background, and establishing a targeted power flow analysis model based on these factors. Specifically, the power flow analysis model is built using probabilistic power flow calculations tailored to the specific transmission area and background. This model is used to perform analysis within the defined transmission area and background to determine the corresponding probabilistic power flow analysis results. These results include the loss scheduling method within the scheduling optimization area under the current transmission background.

[0098] Real-time acquisition of regional monitoring data for the scheduling optimization area, and real-time determination of whether the power transmission area is under power transmission background based on the regional monitoring data.

[0099] If it is determined that the system is not in a power transmission background, no corresponding operation will be performed.

[0100] When it is determined that the system is in a power transmission background, power flow input data for the power flow analysis model is generated based on regional monitoring data. The power flow analysis model is then used to analyze the power flow input data to obtain the corresponding probabilistic power flow analysis results.

[0101] In one embodiment, issues such as timeliness, computing resources, and flexible adjustment in probabilistic power flow calculations lead to other methods for analyzing and optimizing transmission areas and transmission backgrounds based on probabilistic power flow calculations. In other words, the probabilistic power flow analysis results are determined by combining the results obtained from analyzing power flow input data using a power flow analysis model and a power flow prediction model, respectively. The power flow input data is generated based on current regional monitoring data, and the power flow prediction model is obtained by training deep learning algorithms or machine learning models using the power flow input data and probabilistic power flow analysis results.

[0102] A power flow prediction model is established based on deep learning algorithms. This model is trained using prediction data, and once successfully trained, it can predict the probabilistic power flow analysis results based on the power flow input data. Historical data can be used for analysis, or prediction data can be set based on validation data of the power flow analysis model. The prediction data includes power flow input data and probabilistic power flow analysis results. In other words, by establishing targeted power flow analysis models for scheduling optimization areas and transmission backgrounds, and using these models to simulate and analyze the scheduling optimization areas and transmission backgrounds, corresponding prediction data is obtained.

[0103] Specifically, the regional monitoring data of the scheduling optimization area is acquired in real time, and the power transmission area is judged in real time whether it is under power transmission background based on the regional monitoring data.

[0104] No corresponding operation is performed when it is determined that the system is not in a power transmission background.

[0105] When it is determined that the system is in a power transmission background, power flow input data is generated based on regional monitoring data. The power flow input data is then analyzed using both a power flow prediction model and a power flow analysis model to obtain the corresponding probabilistic power flow analysis results. Based on the analysis results obtained from the two models, the current probabilistic power flow analysis results for the scheduling optimization area are determined.

[0106] To differentiate them, the probabilistic power flow analysis results obtained by the power flow analysis model are marked as standard analysis results. The standard analysis results and power flow input data can also be integrated into self-learning data, and the power flow prediction model can be learned and adjusted based on the self-learning data.

[0107] The scheduling module acquires target monitoring data for the target scheduling area and the current probabilistic power flow analysis results for the scheduling optimization area in real time. Target monitoring data integrates monitoring data from the target scheduling area and performs scheduling processing based on the loss scheduling scheme, target monitoring data, and probabilistic power flow analysis results. For example, based on actual operational data and probabilistic power flow analysis results, the loss scheduling scheme is continuously adjusted and optimized to improve its adaptability and effectiveness.

[0108] Computer-readable storage medium implementation A computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of a loss optimization scheduling method for a flexible DC transmission network, including: The loss scheduling scheme of flexible DC transmission network is analyzed to determine the transmission area and transmission background where the loss does not meet the scheduling loss standard under the loss scheduling scheme. The transmission area where the loss does not meet the scheduling loss standard is taken as the scheduling optimization area. Real-time acquisition of regional monitoring data of the scheduling optimization area; real-time determination of whether the scheduling optimization area is under power transmission background based on the current regional monitoring data; if it is, calculation of the current probability power flow analysis result of the scheduling optimization area. Real-time target monitoring data of the target scheduling area of ​​the flexible DC transmission network is acquired, and the loss scheduling scheme is optimized based on the loss scheduling scheme, target monitoring data, and the current probabilistic power flow analysis results of the scheduling optimization area.

[0109] The specific implementation process of this method has been described in detail in the implementation method of loss optimization scheduling for flexible DC transmission networks, and will not be repeated here.

[0110] The method proposed in this invention optimizes the loss scheduling approach, accurately adapting to the loss characteristics of flexible DC transmission networks under different operating conditions and significantly reducing network losses. This directly reduces energy waste during transmission, improves energy utilization efficiency, and is of positive significance for alleviating energy shortages and achieving sustainable energy use. The optimized loss scheduling approach better addresses the complex impact of various uncertainties in flexible DC transmission networks, such as the intermittency, volatility, and unpredictability of renewable energy output. It effectively stabilizes system voltage, improves power quality, and reduces voltage fluctuations and power quality degradation caused by network losses, thereby enhancing the operational stability of the entire power system and ensuring the reliability of power supply.

Claims

1. A loss optimization scheduling method for flexible DC transmission networks, characterized in that, include: The loss scheduling scheme of flexible DC transmission network is analyzed to determine the transmission areas whose losses do not meet the scheduling loss standard and the corresponding transmission background under the loss scheduling scheme. The transmission areas whose losses do not meet the scheduling loss standard are taken as the scheduling optimization areas. Real-time acquisition of regional monitoring data for the scheduling optimization area; real-time determination of whether the scheduling optimization area is under power transmission background based on the current regional monitoring data; if it is, acquisition of the current probability power flow analysis results for the scheduling optimization area. Real-time target monitoring data of the target scheduling area of ​​the flexible DC transmission network is acquired, and the loss scheduling scheme is optimized based on the loss scheduling scheme, target monitoring data, and the current probabilistic power flow analysis results of the scheduling optimization area.

2. The loss optimization scheduling method for flexible DC transmission networks according to claim 1, characterized in that, The process of determining the transmission areas whose losses do not meet the dispatch loss criteria and the corresponding transmission background includes: Obtain historical loss scheduling data for the application loss scheduling scheme of flexible DC transmission network. The historical loss scheduling data is the loss scheduling data from the current time to the time node corresponding to the last adjustment or change of flexible DC transmission network or loss scheduling scheme. Based on historical loss scheduling data, transmission network areas that do not meet the scheduling loss criteria are identified. These transmission network areas are then merged to obtain the transmission areas whose losses do not meet the scheduling loss criteria. Finally, the transmission background corresponding to each transmission network area is integrated to obtain the transmission background corresponding to that transmission area.

3. The loss optimization scheduling method for flexible DC transmission networks according to claim 1, characterized in that, The process of determining whether the scheduling optimization area is under a power transmission background includes: performing feature identification on the current area monitoring data according to the background collection items, obtaining the corresponding background feature combination, and determining whether the scheduling optimization area is under a power transmission background based on the background feature combination; the background collection items are parameter items of the power transmission background.

4. The loss optimization scheduling method for flexible DC transmission networks according to claim 1, characterized in that, The current probabilistic power flow analysis results for the scheduling optimization area are obtained by analyzing the power flow input data using a power flow analysis model. The power flow input data is generated based on the current area monitoring data.

5. The loss optimization scheduling method for flexible DC transmission networks according to claim 1, characterized in that, The current probabilistic power flow analysis results for the scheduling optimization area are determined by combining the results obtained from analyzing the power flow input data using the power flow analysis model and the power flow prediction model. The power flow input data is generated based on the current area monitoring data, and the power flow prediction model is obtained by training a deep learning algorithm or machine learning model using the power flow input data and the probabilistic power flow analysis results.

6. The loss optimization scheduling method for flexible DC transmission networks according to claim 1, 4, or 5, characterized in that, The probabilistic power flow analysis results include loss scheduling methods in the current power transmission context within the scheduling optimization region.

7. The loss optimization scheduling method for flexible DC transmission networks according to claim 1, characterized in that, The regional monitoring data includes power transmission data, time period data, meteorological data, and output data of renewable energy power generation equipment.

8. The loss optimization scheduling method for flexible DC transmission networks according to claim 1, characterized in that, The transmission areas whose losses do not meet the scheduling loss standards and the corresponding transmission backgrounds are obtained by analyzing historical loss scheduling data using an intelligent evaluation model. The historical loss scheduling data is the loss scheduling data from the current time to the time node corresponding to the last adjustment or change of the flexible DC transmission network or the adjustment or change of the loss scheduling scheme. The intelligent evaluation model adopts deep learning algorithms or machine learning algorithms.

9. A loss optimization scheduling system for flexible DC transmission networks, comprising a processor, characterized in that, The processor executes a computer program to implement the steps of the method according to any one of claims 1-8.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When executed by a processor, the computer program implements the steps of the method according to any one of claims 1-8.

Images