Reactive power optimization scheduling method, device and equipment for converter equipment cluster

Abstract

The embodiment of the invention provides a reactive power optimization scheduling method, device and equipment for a converter equipment cluster. The method comprises the following steps: acquiring output prediction data and real-time operation data of a target converter equipment cluster; based on the day-ahead reactive power optimization scheduling model, processing the output prediction data of the target converter equipment cluster to obtain a day-ahead scheduling plan of the target converter equipment cluster in the next day; based on a rolling update mechanism, according to the real-time operation data, correcting the output prediction data to obtain corrected output prediction data of the target converter equipment cluster; and performing calculation processing on the intra-day scheduling plan of the target converter equipment cluster and the corrected output prediction data by adopting an improved pollen algorithm to obtain a reactive power optimization scheduling strategy of the target converter equipment cluster. According to the method, the adaptability of the power distribution network to the distributed energy output randomness and the load dynamic demand can be remarkably improved.

Description

Technical Field

[0001] This application relates to the field of power grid technology, and in particular to a reactive power optimization scheduling method, device and equipment for a converter equipment cluster. Background Technology

[0002] With the accelerated construction of new power systems, source-load side converter clusters, centered on distributed energy sources (such as photovoltaics and wind power) and flexible loads (such as electric vehicles and virtual power plants), are being connected to the power grid on a large scale. Among these, the output of distributed energy sources is highly random and volatile, while flexible loads, although flexible in response, suffer from scheduling uncertainties. This leads to the power grid facing challenges such as voltage fluctuations and harmonic distortion, which exacerbate power quality problems.

[0003] In existing technologies, taking wind power clusters as an example, an upper-level optimization model is constructed with the goal of minimizing network losses and voltage deviations, while the lower-level model coordinates the output of internal equipment through refined analysis of the reactive power potential of the wind farm.

[0004] However, the above methods do not fully integrate the randomness and flexibility of converter equipment in the model construction, resulting in insufficient fine-grained modeling of equipment collaborative scheduling and difficulty in covering the collaborative scheduling needs of multiple equipment. Summary of the Invention

[0005] This application provides a reactive power optimization scheduling method, apparatus, and equipment for converter equipment clusters, which can significantly improve the adaptability of the distribution network to the randomness of distributed energy output and dynamic load demand, and meet the requirements of real-time scheduling for rapid response.

[0006] In a first aspect, embodiments of this application provide a reactive power optimization scheduling method for a converter equipment cluster, including:

[0007] Obtain the output prediction data and real-time operation data of the target converter equipment cluster;

[0008] Based on the day-ahead reactive power optimization scheduling model, the output prediction data of the target converter equipment cluster is processed to obtain the day-ahead scheduling plan of the target converter equipment cluster for the next day; wherein, the optimization objectives of the day-ahead reactive power optimization scheduling model include power quality indicators and operating economic indicators of the distribution network.

[0009] Based on the rolling update mechanism, the output prediction data of the target converter equipment cluster is corrected according to the real-time operating data of the target converter equipment cluster, so as to obtain the corrected output prediction data of the target converter equipment cluster.

[0010] An improved pollen algorithm is used to calculate and process the intraday scheduling plan and the corrected output prediction data of the target converter equipment cluster, so as to obtain the reactive power optimization scheduling strategy of the target converter equipment cluster.

[0011] In one possible implementation, the step of employing an improved pollen algorithm to calculate and process the intraday scheduling plan and corrected output prediction data of the target converter equipment cluster to obtain a reactive power optimization scheduling strategy for the target converter equipment cluster includes:

[0012] Based on the intraday scheduling plan and corrected output forecast data of the target converter equipment cluster, an intraday reactive power optimization scheduling model is constructed; wherein, the intraday reactive power optimization scheduling model includes an intraday scheduling objective function and intraday scheduling constraints; the optimization objective of the intraday scheduling objective function includes the distribution network operating cost;

[0013] By using the improved pollen algorithm and based on the intraday scheduling constraints, the intraday reactive power optimization scheduling model is optimized to obtain the reactive power optimization scheduling strategy for the target converter equipment cluster.

[0014] In one possible implementation, the step of optimizing the intraday reactive power optimization scheduling model using the improved pollen algorithm based on the intraday scheduling constraints to obtain the reactive power optimization scheduling strategy for the target converter equipment cluster includes:

[0015] The intraday reactive power optimization scheduling model is calculated and initialized using a chaotic mapping algorithm to obtain a chaotic population; wherein the chaotic population includes at least one initial reactive power optimization scheduling strategy.

[0016] Let i=1, and repeat the following process until the i-th search result includes the reactive power optimization scheduling strategy that meets the intraday scheduling constraints: perform iterative search processing on the i-th chaotic population to obtain the i-th search result; update the i-th chaotic population according to the i-th search result to obtain the (i+1)-th chaotic population; let i=i+1.

[0017] In one possible implementation, the iterative search process for the i-th chaotic population to obtain the i-th search result includes:

[0018] If it is determined that the random number of the i-th chaotic population is less than the switching probability, then a local search is performed on the i-th chaotic population to obtain the i-th search result; wherein, the switching probability is obtained based on the number of iterations;

[0019] If it is determined that the random number of the i-th chaotic population is greater than the switching probability, then a global search is performed on the i-th chaotic population to obtain the i-th search result.

[0020] In one possible implementation, the step of performing a local search on the i-th chaotic population to obtain the i-th search result includes:

[0021] Based on a multi-strategy mutation mechanism, a local search is performed on the i-th chaotic population to obtain the i-th local search result; wherein, the multi-strategy mutation mechanism includes at least two of the following: basic algorithm mutation strategy, directed mutation strategy, and differential evolution mutation strategy; the i-th local search result includes at least two initial local search results;

[0022] The i-th search result is determined from the i-th local search result based on the fitness values ​​of the at least two initial local search results.

[0023] In one possible implementation, the intraday scheduling constraints include, but are not limited to: power balance constraints, node voltage constraints, branch power constraints, distributed energy output constraints, and intraday scheduling flexible load constraints.

[0024] In one possible implementation, the step of processing the output prediction data based on the day-ahead reactive power optimization scheduling model to obtain the day-ahead scheduling plan for the target converter equipment cluster in the next day includes:

[0025] Based on the day-ahead scheduling objective function and day-ahead scheduling constraints in the day-ahead reactive power optimization scheduling model, the output prediction data, flexible load demand data, and grid parameters of the target converter equipment cluster are calculated to obtain the day-ahead scheduling plan of the target converter equipment cluster for the next day.

[0026] In one possible implementation, the day-ahead scheduling constraints include, but are not limited to: power balance constraints, node voltage constraints, branch power constraints, distributed energy output constraints, and day-ahead scheduling flexible load constraints.

[0027] Secondly, embodiments of this application provide a reactive power optimization scheduling device for a converter equipment cluster, comprising:

[0028] The acquisition module is used to acquire the output prediction data and real-time operation data of the target converter equipment cluster;

[0029] The processing module is used to process the output prediction data of the target converter equipment cluster based on the day-ahead reactive power optimization scheduling model to obtain the day-ahead scheduling plan of the target converter equipment cluster for the next day; wherein, the optimization objectives of the day-ahead reactive power optimization scheduling model include power quality indicators and operating economic indicators of the distribution network.

[0030] The correction module is used to correct the output prediction data of the target converter equipment cluster based on the real-time operating data of the target converter equipment cluster using a rolling update mechanism, so as to obtain the corrected output prediction data of the target converter equipment cluster.

[0031] The optimization module is used to calculate and process the intraday scheduling plan and the corrected output prediction data of the target converter equipment cluster using an improved pollen algorithm, so as to obtain the reactive power optimization scheduling strategy of the target converter equipment cluster.

[0032] In one possible implementation, the optimization module is specifically used to: construct an intraday reactive power optimization scheduling model based on the intraday scheduling plan and corrected output prediction data of the target converter equipment cluster; wherein the intraday reactive power optimization scheduling model includes an intraday scheduling objective function and intraday scheduling constraints; the optimization objective of the intraday scheduling objective function includes the distribution network operating cost; and through the improved pollen algorithm, optimize the intraday reactive power optimization scheduling model according to the intraday scheduling constraints to obtain the reactive power optimization scheduling strategy of the target converter equipment cluster.

[0033] In one possible implementation, the optimization module is specifically used to: calculate and initialize the intraday reactive power optimization scheduling model using a chaotic mapping algorithm to obtain a chaotic population; wherein the chaotic population includes at least one initial reactive power optimization scheduling strategy; let i=1, and repeat the following process until the i-th search result includes the reactive power optimization scheduling strategy that meets the intraday scheduling constraints: perform iterative search processing on the i-th chaotic population to obtain the i-th search result; update the i-th chaotic population according to the i-th search result to obtain the (i+1)-th chaotic population; let i=i+1.

[0034] In one possible implementation, the optimization module is specifically configured to: if it is determined that the random number of the i-th chaotic population is less than the switching probability, then perform a local search on the i-th chaotic population to obtain the i-th search result; wherein the switching probability is obtained based on the number of iterations; if it is determined that the random number of the i-th chaotic population is greater than the switching probability, then perform a global search on the i-th chaotic population to obtain the i-th search result.

[0035] In one possible implementation, the optimization module is specifically used to: perform local search processing on the i-th chaotic population based on a multi-strategy mutation mechanism to obtain the i-th local search result; wherein the multi-strategy mutation mechanism includes at least two of the following: basic algorithm mutation strategy, directed mutation strategy, and differential evolution mutation strategy; the i-th local search result includes at least two initial local search results; and determine the i-th search result from the i-th local search results based on the fitness values ​​of the at least two initial local search results.

[0036] In one possible implementation, the intraday scheduling constraints include, but are not limited to: power balance constraints, node voltage constraints, branch power constraints, distributed energy output constraints, and intraday scheduling flexible load constraints.

[0037] In one possible implementation, the processing module is specifically used to: calculate the output forecast data, flexible load demand data, and grid parameters of the target converter equipment cluster based on the day-ahead scheduling objective function and day-ahead scheduling constraints in the day-ahead reactive power optimization scheduling model, so as to obtain the day-ahead scheduling plan of the target converter equipment cluster for the next day.

[0038] In one possible implementation, the day-ahead scheduling constraints include, but are not limited to: power balance constraints, node voltage constraints, branch power constraints, distributed energy output constraints, and day-ahead scheduling flexible load constraints.

[0039] Thirdly, embodiments of this application provide an electronic device, including: a memory and a processor;

[0040] The memory stores computer-executed instructions;

[0041] The processor executes computer execution instructions stored in the memory, causing the processor to perform the first aspect and / or various possible implementations of the first aspect as described above.

[0042] Fourthly, embodiments of this application provide a computer-readable storage medium storing computer-executable instructions, which, when executed by a processor, are used to implement the first aspect and / or various possible implementations of the first aspect.

[0043] Fifthly, embodiments of this application provide a computer program product, including a computer program that, when executed by a processor, implements the first aspect and / or various possible implementations of the first aspect.

[0044] The reactive power optimization scheduling method, apparatus, and equipment for converter equipment clusters provided in this application embodiment, through a day-ahead reactive power optimization scheduling model, and combined with two optimization objectives—power quality indicators and operational economic indicators of the distribution network—processes the output prediction data of the target converter equipment cluster to predict the day-ahead scheduling plan for the target converter equipment cluster in the next day. Based on the real-time operating data of the target converter equipment cluster, the output prediction data is corrected to obtain the corrected output prediction data. Using an improved pollen algorithm, the intraday scheduling plan and the corrected output prediction data are calculated and processed to obtain the reactive power optimization scheduling strategy for the target converter equipment cluster. Furthermore, by constructing a multi-time-scale scheduling framework (day-ahead to intraday) and combining it with the improved pollen algorithm for solving, dynamic collaborative optimization of the converter equipment cluster's grid connection is achieved. This can significantly improve the distribution network's adaptability to the randomness of distributed energy output and dynamic load demand, and meet the requirements of real-time scheduling for rapid response. Attached Figure Description

[0045] The accompanying drawings, which are incorporated in and form part of this specification, illustrate embodiments consistent with this application and, together with the description, serve to explain the principles of this application.

[0046] Figure 1 A flowchart illustrating a reactive power optimization scheduling method for a converter equipment cluster provided in this application embodiment;

[0047] Figure 2 A flowchart illustrating another reactive power optimization scheduling method for a converter equipment cluster provided in this application embodiment;

[0048] Figure 3 A schematic diagram of a reactive power optimization scheduling process for a source-load side converter cluster connected to the power grid, provided as an embodiment of this application;

[0049] Figure 4 A schematic diagram of the structure of a reactive power optimization scheduling device for a converter equipment cluster provided in this application embodiment;

[0050] Figure 5 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this application.

[0051] The accompanying drawings illustrate specific embodiments of this application, which will be described in more detail below. These drawings and descriptions are not intended to limit the scope of the concept in any way, but rather to illustrate the concept of this application to those skilled in the art through reference to particular embodiments. Detailed Implementation

[0052] Exemplary embodiments will now be described in detail, examples of which are illustrated in the accompanying drawings. When the following description relates to the drawings, unless otherwise indicated, the same numbers in different drawings denote the same or similar elements. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with this application. Rather, they are merely examples of apparatuses and methods consistent with some aspects of this application as detailed in the appended claims.

[0053] This application applies to distribution network scenarios where large-scale access of source-load (power supply-load) side converter equipment clusters is achieved in new power systems.

[0054] Based on the above scenarios, it can be seen that traditional methods use a single time scale (such as only the day before or only within the day) for static optimization, which cannot adapt to the real-time fluctuations in distributed energy output (such as sudden changes in wind power output) and the dynamic changes in load demand (such as the concentrated charging time of electric vehicles).

[0055] The reactive power optimization scheduling method for converter equipment clusters provided in this application achieves dynamic collaborative optimization of the grid connection of source-load side converter equipment clusters by constructing a reactive power optimization model that includes a multi-time-scale scheduling framework (day-ahead to intraday) and solving it using an improved pollen algorithm.

[0056] The technical solution of this application and how the technical solution of this application solves the above-mentioned technical problems are described in detail below with specific embodiments. These specific embodiments can be combined with each other, and the same or similar concepts or processes may not be described again in some embodiments. The embodiments of this application will now be described with reference to the accompanying drawings.

[0057] Figure 1 A flowchart illustrating a reactive power optimization scheduling method for a converter equipment cluster provided in this application embodiment is shown below. Figure 1 As shown, the method includes:

[0058] 201. Obtain the output prediction data and real-time operation data of the target converter equipment cluster.

[0059] For example, the execution entity in this embodiment can be an electronic device. The day-ahead reactive power optimization scheduling requires a dispatch center to generate output forecast data 24 hours in advance based on historical data and weather forecasts, including photovoltaic and wind power output forecasts for the current scheduling day. The device then obtains the output forecast data of the target converter equipment cluster from this day-ahead reactive power optimization scheduling requirement dispatch center. Simultaneously, the device obtains real-time operating data of the target converter equipment cluster, including real-time photovoltaic output data, through a real-time monitoring center.

[0060] The target converter equipment cluster can be a source-load side converter equipment cluster. This source-load side converter equipment cluster is a collection of distributed energy sources (such as photovoltaic and wind power) and flexible loads (such as loads that can be shifted or reduced) in the power grid, such as a distribution network area that includes multiple photovoltaic power plants, wind farms and electric vehicle charging stations.

[0061] 202. Based on the day-ahead reactive power optimization scheduling model, the output prediction data of the target converter equipment cluster is processed to obtain the day-ahead scheduling plan of the target converter equipment cluster for the next day; wherein, the optimization objectives of the day-ahead reactive power optimization scheduling model include the power quality index and the operation economic index of the distribution network.

[0062] For example, the equipment constructs a day-ahead reactive power optimization scheduling model based on historical data and weather forecasts (such as weather forecasts and load curves), which includes optimization objectives for power quality indicators and operational economic indicators of the distribution network. Based on this day-ahead reactive power optimization scheduling model, with an hourly scheduling step, the output prediction data of the target converter equipment cluster is processed to predict the day-ahead scheduling plan of the target converter equipment cluster for the next day, including the power generation resource output plan for the next 24 hours and the adjustment plan of each scheduling resource, so as to cover the macro-planning of distributed energy output and flexible load adjustment.

[0063] The established reactive power optimization scheduling model for the distribution network can fully leverage the synergistic effect of controllable scheduling resources such as flexible loads in the distribution network to achieve peak filling and valley shaving, and ensure the economic operation of the distribution network.

[0064] 203. Based on the rolling update mechanism, the output prediction data of the target converter equipment cluster is corrected according to the real-time operation data of the target converter equipment cluster, so as to obtain the corrected output prediction data of the target converter equipment cluster.

[0065] For example, the equipment is pre-set with a rolling update mechanism, and based on this rolling update mechanism (such as a 15-minute cycle), the output prediction data of the target converter equipment cluster is corrected according to the real-time operating data of the target converter equipment cluster. For example, during the day, the wind power and photovoltaic output prediction data for the next 4 hours are corrected through the rolling update mechanism to obtain the corrected output prediction data. That is, the prediction error is dynamically compensated through a periodic feedback correction mechanism, thereby effectively addressing the problem of prediction accuracy decaying over time.

[0066] The revised power output forecast data can also dynamically adjust the optimization objectives (such as power quality weights) and constraints (such as node voltage limits) of the day-ahead reactive power optimization scheduling model.

[0067] The rolling update mechanism model can be represented as follows:

[0068] ,

[0069] The target converter equipment cluster includes multiple photovoltaic and multiple wind turbines, P n,t+s P is the output of the target converter equipment cluster in the output prediction data, which is predicted from time t to time t+s based on the reactive power optimization scheduling resource n. n,t It is the initial value of the reactive power optimized scheduling resource n at time t, ΔP n,t+s The reactive power optimization scheduling resource n predicts the change in time t+s from time t to time t+s, ΔP. PVi,t+s ΔP WTj,t+s ΔP t+s These are the power output of the i-th photovoltaic unit, the power output of the j-th wind turbine, and the load change increment predicted from time t to time t+s based on real-time operating data.

[0070] 204. An improved pollen algorithm is used to calculate and process the intraday scheduling plan and the corrected output prediction data of the target converter equipment cluster to obtain the reactive power optimization scheduling strategy of the target converter equipment cluster.

[0071] For example, the device optimizes the traditional pollen algorithm to obtain an improved pollen algorithm. Based on this improved pollen algorithm, the intraday scheduling plan of the target converter equipment cluster and the corrected output prediction data are calculated and processed to obtain the reactive power optimization scheduling strategy for the target converter equipment cluster. This includes the scheduling strategies for multiple reactive power optimization scheduling resources at each time point in the future time period. The scheduling strategies include distributed energy output adjustment, flexible load response, and grid interaction power allocation. The entire process combines multi-timescale modeling and algorithm optimization to achieve efficient reactive power optimization scheduling of the distribution network under the scenarios of random distributed energy output and dynamic load response.

[0072] Among them, the pollen algorithm imitates two mechanisms of flower pollination in nature: cross-pollination relies on pollinators to spread pollen over long distances, which corresponds to a global search process; self-pollination involves pollination at relatively close physical locations, which corresponds to a local search process.

[0073] In one example, the improved pollen algorithm may include, but is not limited to, a mixture of the traditional pollen algorithm and the genetic algorithm, a mixture of the traditional pollen algorithm and the simulated annealing algorithm, a mixture of the traditional pollen algorithm and the gray wolf optimization algorithm, and a mixture of the traditional pollen algorithm and the simplex method.

[0074] For example, the intraday scheduling plan and corrected output forecast data of the target converter equipment cluster are calculated and processed, and a population of size R is randomly generated. The i-th pollen in the population is represented by the vector Xi = (Xi1, Xi2, ..., Xiθ, ..., XiD) to represent each possible scheduling strategy, where Xiθ (θ = 1, ..., D) is the position of the i-th pollen in the θ-th dimension, and D is the dimension of the problem being solved. A global search or local search is performed on X = (Xi, Xi, ..., Xi, ..., Xi) to obtain the optimal solution Xi, which is the reactive power optimization scheduling strategy for the target converter equipment cluster.

[0075] This embodiment provides a reactive power optimization scheduling method for converter equipment clusters. The day-ahead reactive power optimization scheduling model generates a macro-schedule plan based on long-term forecasts, achieving a balance between power quality (voltage stability, harmonic suppression) and economy (operating costs, network loss minimization) to ensure maximum economic efficiency and resource utilization. A rolling update mechanism (e.g., 15-minute cycle) dynamically corrects the forecast data, enabling rapid response to real-time fluctuations (e.g., sudden drops in wind power, concentrated charging of electric vehicles). The improved pollen algorithm, by combining traditional algorithms with other algorithms, significantly enhances the convergence speed and global search capability. Furthermore, this method significantly improves the distribution network's adaptability to the randomness of distributed energy output and dynamic load demands, meeting the rapid response requirements of real-time scheduling while reducing computational costs and ensuring the stability and economy of grid operation.

[0076] Figure 2 A flowchart illustrating another reactive power optimization scheduling method for a converter equipment cluster provided in this application embodiment is shown below. Figure 2 As shown, the method includes:

[0077] 301. Obtain the output prediction data and real-time operation data of the target converter equipment cluster.

[0078] For example, this step can be referred to as step 201, which will not be repeated here.

[0079] 302. Based on the day-ahead scheduling objective function and day-ahead scheduling constraints in the day-ahead reactive power optimization scheduling model, the output prediction data, flexible load demand data, and grid parameters of the target converter equipment cluster are calculated to obtain the day-ahead scheduling plan of the target converter equipment cluster for the next day.

[0080] For example, day-ahead reactive power optimization scheduling requires the dispatch center to generate output forecast data 24 hours in advance based on historical data and weather forecasts. This data includes photovoltaic and wind power output forecasts, flexible load demand data, and grid parameters for the current scheduling day. The equipment then obtains the output forecast data, flexible load demand data, and grid parameters for the target converter equipment cluster from this day-ahead reactive power optimization scheduling requirement dispatch center. The day-ahead reactive power optimization scheduling model is then invoked, which includes a day-ahead scheduling objective function and day-ahead scheduling constraints. Using the day-ahead scheduling objective function, the output forecast data, flexible load demand data, and grid parameters of the target converter equipment cluster are used to calculate the function value. If this function value meets the day-ahead scheduling constraints and minimizes the optimization objectives of the distribution network power quality indicators and operational economic indicators, then the function value is determined to be the optimal solution, which is the day-ahead scheduling plan for the target converter equipment cluster in the coming day. The day-ahead scheduling constraints include, but are not limited to: power balance constraints, node voltage constraints, branch power constraints, distributed energy output constraints, and day-ahead scheduling flexible load constraints.

[0081] In one example, the day-ahead scheduling objective function aims to optimize the power quality and operational economy of the distribution network within the target converter cluster. The power quality of the distribution network (i.e., the distribution network within the target converter cluster) is measured by the overall net load fluctuation, including the net load fluctuation rate and the net load fluctuation amplitude. This is based on the following formula:

[0082] ,

[0083] Where f1 represents the overall net load fluctuation situation, and P represents the minimum and maximum values ​​of the overall net load of the distribution network. net,t It is the power of the overall net load of the distribution network at time t. This represents the average power of the distribution network's comprehensive net load throughout the day on the dispatching day. ω1 and ω2 are the weights of the comprehensive net load fluctuation amplitude and fluctuation rate, respectively. The comprehensive net load fluctuation amplitude has a smaller impact on the distribution network's fluctuations than the comprehensive net load fluctuation rate; therefore, ω1 is set to 0.25 and ω2 to 0.75. Operating costs include distributed energy dispatch costs, flexible load dispatch costs, interaction costs with the upstream power grid, and power loss costs of the distribution network. This is based on the following formula:

[0084] ,

[0085] Where f2 is the operating cost of day-ahead reactive power optimization scheduling, and C DG It is the cost of distributed energy dispatch, C FL It is the cost of flexible load dispatching, C ESIt is the interaction cost between the distribution network and the upstream power grid, C loss This refers to the cost of electricity loss. The specific calculation formulas for each cost are as follows:

[0086] ,

[0087] ,

[0088] ,

[0089] ,

[0090] Among them, c PV,i c is the unit dispatch price of the i-th photovoltaic power generation. WT,j P is the unit dispatch price for the j-th wind power generation. PV,i,t P is the active power output of the i-th photovoltaic cell at time t. WT,j,t It is the active power output of the j-th wind turbine at time t, c py It is the unit dispatch price for loads that can be shifted, c xj It is the unit compensation cost for load reduction, c zd It is the unit compensation cost for interruptible loads, P py,t P xj,t P zd,t These represent the active power of the load that can be shifted, the load that can be reduced, and the load that can be interrupted at time t, respectively. pure It is the cost of purchasing electricity from the upper-level power grid, C sell It is the cost of selling electricity to the upstream power grid, c pure This is the unit price of electricity purchased, c sell It is the unit price of electricity sold, P. pure,t P is the amount of electricity purchased from the upper-level power grid at time t. sell,t It refers to the electricity sold to the higher-level power grid, c loss,net,t P is the unit loss cost of the distribution network at time t. loss,net,t It is the network loss value of the distribution network at time t.

[0091] In one example, power balance constraints can be represented as follows:

[0092] ,

[0093] Among them, P ES,t P is the interaction power between the distribution network and the upstream power grid at time t. un,t P is the active power required by the uncontrollable load at time t. FL,t P is the total active power of flexible load adjustments (including loads that can be shifted, loads that can be reduced, and loads that can be interrupted) at time t. loss,t It is the amount of power loss in the distribution network at time t.

[0094] Furthermore, the node voltage constraint can be represented as follows: , among which, U i It is the voltage amplitude at node i in the distribution network at time t. , These are the lower and upper limits of the voltage amplitude at node i, respectively.

[0095] Furthermore, the branch power constraint can be expressed as follows: , where P l It is the transmission power of distribution network line l. and These are the minimum and maximum values ​​of the line power, respectively.

[0096] Furthermore, the output constraint of distributed energy resources can be expressed as follows:

[0097] ,

[0098] ,

[0099] in, and These are the minimum and maximum actual active power outputs of the i-th photovoltaic cell in the distribution network, respectively. and These are the minimum and maximum actual active power outputs of the j-th wind power unit in the distribution network, respectively.

[0100] Furthermore, loads in the distribution network are divided into uncontrollable loads and flexible loads. The biggest difference between the two is that day-ahead dispatching flexible loads can self-adjust during distribution network operation and participate in reactive power optimization dispatch. Flexible load constraints can be divided into three types based on their response mode: transferable, reduceable, and shiftable. Shiftable loads mainly include commercial and residential loads, typically manifested as lighting and temperature control equipment. These load constraints have strong controllability, fast response speed, and short dispatch cycle, thus they can participate in day-ahead and intraday reactive power optimization dispatch. Shiftable load constraints can be represented as follows:

[0101] ,

[0102] ,

[0103] ,

[0104] in, and These represent the increased power and decreased power of the load that can be moved at time t, respectively. and These are the maximum increase and maximum decrease in power for the load that can be moved, respectively.

[0105] Reduceable loads mainly include industrial loads, which only participate in day-ahead reactive power optimization scheduling. Reduceable load constraints can be expressed as follows: ,in, and These are the lower and upper limits of the load reduction power that can be reduced at time t, respectively.

[0106] Interruptible loads refer to the portion of loads whose power supply can be suspended during peak electricity demand or emergency situations. This includes the duration, frequency, and amount of interruption for interruptible loads. Therefore, this type of load can only participate in day-ahead reactive power optimization dispatch. Interruptible load constraints can be expressed as follows:

[0107] ,

[0108] ,

[0109] ,

[0110] Among them, v n,t This indicates the call status of the interruptible load; 1 indicates a call, and 0 indicates no call. This is the upper limit of the interruption time. and These are the minimum and maximum single-interruption power of the interruptible load, respectively. This is the maximum number of interruptions.

[0111] By constructing a day-ahead reactive power optimization scheduling model that incorporates both power quality and economic objectives, and combining a multi-dimensional constraint system (power balance, node voltage, branch power, etc.), the optimization results achieve a balance between power quality (voltage stability, harmonic suppression) and economic efficiency (operating costs, network loss minimization). Specifically, power balance constraints ensure the feasibility of the scheduling strategy, node voltage constraints prevent voltage exceedance risks, branch power constraints prevent line overload, and flexible load response rules ensure that the scheduling strategy conforms to user agreements and equipment operating boundaries. This technique significantly improves the refined modeling capability of the day-ahead reactive power optimization scheduling model, reduces scheduling infeasibility caused by constraint conflicts, and provides a more reliable benchmark for intraday dynamic optimization.

[0112] 303. Based on the rolling update mechanism, the output prediction data of the target converter equipment cluster is corrected according to the real-time operation data of the target converter equipment cluster, so as to obtain the corrected output prediction data of the target converter equipment cluster.

[0113] For example, this step can be referred to as step 203, which will not be repeated here.

[0114] 304. Based on the intraday scheduling plan of the target converter equipment cluster and the corrected output forecast data, construct an intraday reactive power optimization scheduling model; wherein, the intraday reactive power optimization scheduling model includes the intraday scheduling objective function and the intraday scheduling constraints; the optimization objective of the intraday scheduling objective function includes the distribution network operating cost.

[0115] For example, cost calculations are performed on the intraday scheduling plan and corrected output forecast data of the target converter equipment cluster to obtain cost information for the target converter equipment cluster, including operating costs, distributed energy dispatch costs, flexible load dispatch costs (excluding loads that can be reduced or interrupted), interaction costs between the distribution network and the upstream grid, and energy loss costs. Based on the user-preset base model, the base model is adjusted using this cost information to obtain the intraday dispatch objective function included in the intraday reactive power optimization dispatch model. The optimization objective of this intraday dispatch objective function includes the distribution network operating cost. Simultaneously, the user configures the system according to preset conditions to enable the equipment to determine the intraday dispatch constraints included in the intraday reactive power optimization dispatch model. These intraday dispatch constraints include, but are not limited to: power balance constraints, node voltage constraints, branch power constraints, distributed energy output constraints, and intraday flexible load constraints.

[0116] For example, during intraday reactive power optimization scheduling, it is necessary to dynamically predict wind power output, photovoltaic power generation, and load demand for the next four hours based on real-time scheduling resource status, and update the prediction information dynamically every 15 minutes. When the operating conditions of the distribution network change, the equipment will implement rolling optimization of the intraday scheduling objective function based on the latest prediction data, with the goal of minimizing operating costs, and achieve economic operation by continuously correcting the output of distributed power sources and scheduling strategies.

[0117] In one example, the intraday scheduling objective function can be expressed by the following formula:

[0118] ,

[0119] Where f3 is the operating cost of intraday reactive power optimization scheduling, and C DG It is the cost of distributed energy dispatch, C FL The cost of flexible load dispatching (excluding loads that can be reduced or interrupted), C ES It is the interaction cost between the distribution network and the upstream power grid, C loss It is the cost of electricity loss.

[0120] In one example, the intraday scheduling flexible load constraint in the intraday scheduling constraints can be expressed by the following formula: , ,in, and These are the upper and lower limits of the output variable of the reactive power scheduling resource n at time t+s. and These are the upper and lower limits of the incremental change of reactive power dispatch resource n at time t+s within the day. In the intraday dispatch constraints, except for the intraday dispatch flexible load, the other constraints are the same as those in the day-ahead dispatch constraints.

[0121] 305. By improving the pollen algorithm and based on the intraday scheduling constraints, the intraday reactive power optimization scheduling model is optimized and calculated to obtain the reactive power optimization scheduling strategy for the target converter equipment cluster.

[0122] For example, the device invokes the improved pollen algorithm and uses it to optimize the objective function in the intraday reactive power optimization scheduling model to obtain the optimal solution. The optimal solution meets the conditions of minimizing operating costs and intraday scheduling constraints. Therefore, the optimal solution is the reactive power optimization scheduling strategy of the target converter equipment cluster, which may include distributed energy output correction for the next 4 hours, real-time adjustment of movable loads, and grid interaction power allocation.

[0123] In one example, the improved pollen algorithm can also be transformed into a binary pollen algorithm capable of handling discrete optimization problems such as feature selection and the knapsack problem by introducing transformation functions such as Sigmoid.

[0124] By using an intraday reactive power optimization scheduling model to quickly respond to real-time fluctuations and correct deviations in the day-ahead plan, and by combining rolling forecasting with an improved pollen algorithm, high-precision, low-computation-cost real-time scheduling can be achieved. In turn, the dynamic response capability of the distribution network to sudden changes in distributed energy output (such as a sudden drop in wind power) and fluctuations in load demand (such as concentrated charging of electric vehicles) can be significantly improved.

[0125] In one possible implementation, step 305 includes the following steps:

[0126] The first step is to use a chaotic mapping algorithm to calculate and initialize the intraday reactive power optimization scheduling model to obtain a chaotic population; wherein, the chaotic population includes at least one initial reactive power optimization scheduling strategy.

[0127] The second step is to set i=1 and repeat the following process until the i-th search result includes a reactive power optimization scheduling strategy that meets the intraday scheduling constraints: perform iterative search processing on the i-th chaotic population to obtain the i-th search result; update the i-th chaotic population based on the i-th search result to obtain the (i+1)-th chaotic population; set i=i+1.

[0128] Specifically, a chaotic mapping algorithm is used to calculate the intraday reactive power optimization scheduling model, resulting in multiple possible scheduling strategies. These strategies are then initialized using Tent chaotic mapping to obtain a chaotic population, including at least one initial reactive power optimization scheduling strategy. This chaotic population is designated as the first chaotic population, with i=1. An iterative search is performed on the i-th chaotic population to obtain the i-th search result. Based on this i-th search result, the i-th chaotic population is updated; for example, it is replaced to obtain the (i+1)-th chaotic population. If the i-th search result includes a reactive power optimization scheduling strategy that meets the intraday scheduling constraints, meaning the i-th search result is the optimal solution for the intraday reactive power optimization scheduling model, the process ends. If the i-th search result is not the optimal solution for the intraday reactive power optimization scheduling model, i=i+1, and the above process is repeated, i.e., a new search is performed on the (i+1)-th chaotic population to obtain the (i+1)-th search result.

[0129] In one example, the Tent mappings are relatively uniformly distributed. Using Tent mappings to generate chaotic sequences and initialize pollen individuals yields at least one initial reactive power optimization scheduling strategy, which can be calculated using the following expression:

[0130] ,

[0131] Among them, z i It generates a random sequence within the interval [0,1]. It is a chaotic sequence, with a set value of 0.5. Combining the chaotic sequence... The initial position sequence of pollen individuals within the search area is further generated, which represents the specific scheduling content in each initial reactive power optimization scheduling strategy. This can be calculated using the following expression:

[0132] ,

[0133] in, , They are The maximum and minimum values ​​in the sequence.

[0134] By initializing the population using chaotic mapping (such as the Tent mapping), the number of invalid iterations is reduced, while the initial solution covers a wider search space, avoiding local search bias. Thus, high-precision solutions can be achieved with low computational cost, meeting the needs of real-time dispatching of power distribution networks for rapid response.

[0135] In a possible implementation, the iterative search process for the $i$-th chaotic population in the second step to obtain the $i$-th search result includes: if it is determined that the random number of the $i$-th chaotic population is less than the switching probability, perform a local search process on the $i$-th chaotic population to obtain the $i$-th search result; where the switching probability is obtained based on the number of iterations. Or, if it is determined that the random number of the $i$-th chaotic population is greater than the switching probability, perform a global search process on the $i$-th chaotic population to obtain the $i$-th search result.

[0136] Specifically, the device sets the switching probability as an exponential function related to the number of iterations. In the iterative search process based on the pollen algorithm, for the $i$-th iterative search, according to the $i$-th number of iterations, determine the switching probability of the current $i$-th iteration. Compare the random number (rand) of the $i$-th chaotic population with this switching probability. If it is determined that the random number of the $i$-th chaotic population is less than the switching probability, perform self-pollination, that is, perform a local search process on the $i$-th chaotic population to obtain the $i$-th search result. If it is determined that the random number of the $i$-th chaotic population is greater than this switching probability, perform cross-pollination, that is, perform a global search process on the $i$-th chaotic population to obtain the $i$-th search result.

[0137] In one example, the switching probability is set as an exponential function related to the number of iterations, which can be expressed by the following formula: , where $t$ and $T$ respectively represent the current number of iterations and the maximum number of iterations.

[0138] In one example, introduce the switching probability $p (p\in[0,1])$. When the random number $rand < p$, perform self-pollination, which can be calculated according to the following expression: , where 、 [[$P_{t + 1}$]] and [[$P_{t}$]] are the pollen individuals of the $(t + 1)$-th and $t$-th generations respectively (i.e., the initial reactive power optimization scheduling strategy), 、 [[$r_{1}$]] and [[$r_{2}$]] are two random individuals different from the pollen in the population, and $\epsilon$ is a random number uniformly distributed on $[0,1]$. When $rand > p$, perform cross-pollination, which can be calculated according to the following expression: , where [[ID=2A]] [[$P_{best}$]] is the optimal pollen in the population, the control parameter $L$ is a $D$-dimensional pollination intensity vector, $D$ is the dimension of the problem to be solved. For example, in the reactive power optimization scheduling strategy, it includes 3 dimensions: the correction of distributed energy output in the next 4 hours, the real-time adjustment of shiftable load, and the distribution of grid interaction power. Among them, each dimension includes random numbers subject to the levy distribution. The calculation formula of the control parameter $L$ is as follows:

[0139] [[$L = \frac{\alpha}{\beta^{1 / D}}$]] [[$\alpha = \frac{2\gamma\Gamma(1 + D)\sin(\frac{\pi D}{2})}{\pi D\Gamma(\frac{1 + D}{2})2^{(D - 1) / 2}}$]] [[$\beta = 1$]]

[0140] [[$\gamma = 0.5$]] where [[$\Gamma$]] is the gamma functionIt is the standard gamma function, λ is 1.5, and s is the random step size.

[0141] By setting the switching probability as an exponential function related to the number of iterations, the improved pollen algorithm focuses on cross-pollination in the early stage and self-pollination in the later stage, which improves the convergence speed in the later stage and is conducive to balancing local and global searches. In this way, it can meet the needs of real-time dispatching of distribution networks for rapid response.

[0142] In one possible implementation, a local search is performed on the i-th chaotic population to obtain the i-th search result, including:

[0143] Step 1: Based on the multi-strategy mutation mechanism, perform local search processing on the i-th chaotic population to obtain the i-th local search result; wherein, the multi-strategy mutation mechanism includes at least two of the following: basic algorithm mutation strategy, directed mutation strategy and differential evolution mutation strategy; the i-th local search result includes at least two initial local search results.

[0144] Step 2: Based on the fitness values ​​of at least two initial local search results, determine the i-th search result from the i-th local search result.

[0145] Specifically, when the random number of the i-th chaotic population is determined to be less than the switching probability, self-pollination is performed. That is, when performing local search processing on the i-th chaotic population, the device invokes a preset multi-strategy mutation mechanism, including at least two of the following: basic algorithm mutation strategy, directed mutation strategy, and differential evolution mutation strategy. Based on this multi-strategy mutation mechanism, a local search is performed on the i-th chaotic population to obtain the i-th local search result, including at least two initial local search results, such as initial local search results obtained based on the basic algorithm mutation strategy, initial local search results obtained based on the directed mutation strategy, and initial local search results obtained based on the differential evolution mutation strategy. According to a preset adaptive algorithm, the fitness value of each initial local search result in the i-th local search result is calculated. These fitness values ​​are compared to obtain the optimal initial local search result in the i-th local search result, which is the i-th search result.

[0146] In one example, during the self-pollination stage, multiple mutation strategies are added and performed simultaneously to increase the mutation range. By calculating and comparing fitness values ​​and selecting the best option, the algorithm can be effectively prevented from getting trapped in local optima. The added targeted mutation strategies are represented as follows: The added differential evolution mutation strategy is represented as follows: , where ε` and ε`` are scaling factors.

[0147] In one example, Figure 3It is a schematic diagram of a reactive power optimal scheduling process for a source-load side converter equipment cluster connected to the power grid provided by an embodiment of the present application. As Figure 3 shown, a reactive power optimal scheduling model for the power grid connected with the source-load side converter equipment cluster is constructed, including constructing a day-ahead reactive power optimal scheduling model, then updating the prediction information and correcting the deviation. Based on the scheduling plan obtained from the day-ahead reactive power optimal scheduling model, an intra-day reactive power optimal scheduling model is constructed and solved by using an improved pollen algorithm, including: population initialization based on Tent chaotic mapping, dynamic switching probability update based on the number of iterations, and multi-strategy mutation mechanism based on self-pollination. Specifically, parameters such as the population number and the maximum number of iterations of the algorithm are initialized, calculations are performed for the intra-day reactive power optimal scheduling model, and the population positions are initialized based on Tent chaotic mapping to obtain a chaotic population; where the chaotic population includes at least one initial reactive power optimal scheduling strategy; and a dynamic switching probability p is set; if the random number rand of the population < p, self-pollination is performed, and if rand > p, cross-pollination is performed; if it is self-pollination, mutation is respectively performed by using a basic algorithm mutation strategy, a directional mutation strategy, and a differential evolution mutation strategy, the corresponding fitness values are calculated, and the optimal value among the 3 mutation strategies is selected as the result of self-pollination; the new generation of pollen is evaluated, and the optimal pollen is updated; if the termination condition is reached, the optimal solution is output, that is, the reactive power optimal scheduling strategy is output, including the scheduling strategies for flexible loads, photovoltaic and wind power systems.

[0148] By synchronously performing directional mutation (enhancing the search directionality) and differential evolution mutation (enhancing the diversity of solutions) in the self-pollination stage, the optimal result is selected through comparison, avoiding falling into local optima, enhancing the diversity of solutions, so as to achieve high-precision solution under low computational cost, adapt to the demand of fast response for the real-time scheduling of the distribution network, and at the same time reduce the situation of scheduling lag caused by slow convergence speed.

[0149] In this embodiment, on the basis of the above embodiment, on the one hand, by constructing a multi-dimensional constraint system including power quality indexes such as power balance, node voltage, branch power, and distributed energy output, the refined modeling of the collaborative control of source-load side equipment is realized, and the dynamic response ability of the distribution network scheduling strategy is improved; on the other hand, by using the improved pollen algorithm to solve the multi-objective optimization model, the problem of slow convergence speed of the traditional pollen algorithm in the later stage is effectively solved, and this method does not require a large amount of data sets for training, has low computational cost, and improves the real-time performance of the system.

[0150] Figure 4 It is a schematic diagram of the structure of a reactive power optimal scheduling device for a converter equipment cluster provided by an embodiment of the present application. As Figure 4 shown, the device includes:

[0151] An acquisition module 401, configured to acquire the output prediction data and real-time operation data of the target converter equipment cluster;

[0152] The processing module 402 is used to process the output prediction data of the target converter equipment cluster based on the day-ahead reactive power optimization scheduling model to obtain the day-ahead scheduling plan of the target converter equipment cluster for the next day; wherein, the optimization objectives of the day-ahead reactive power optimization scheduling model include the power quality indicators and the operation economic indicators of the distribution network.

[0153] The correction module 403 is used to correct the output prediction data of the target converter equipment cluster based on the real-time operating data of the target converter equipment cluster using a rolling update mechanism, so as to obtain the corrected output prediction data of the target converter equipment cluster.

[0154] The optimization module 404 is used to calculate and process the intraday scheduling plan and the corrected output prediction data of the target converter equipment cluster using the improved pollen algorithm, so as to obtain the reactive power optimization scheduling strategy of the target converter equipment cluster.

[0155] In one possible implementation, the optimization module 404 is specifically used to: construct an intraday reactive power optimization scheduling model based on the intraday scheduling plan of the target converter equipment cluster and the corrected output prediction data; wherein, the intraday reactive power optimization scheduling model includes an intraday scheduling objective function and intraday scheduling constraints; the optimization objective of the intraday scheduling objective function includes the distribution network operating cost; and by improving the pollen algorithm, the intraday reactive power optimization scheduling model is optimized according to the intraday scheduling constraints to obtain the reactive power optimization scheduling strategy of the target converter equipment cluster.

[0156] In one possible implementation, the optimization module 404 is specifically used to: calculate and initialize the intraday reactive power optimization scheduling model using a chaotic mapping algorithm to obtain a chaotic population; wherein the chaotic population includes at least one initial reactive power optimization scheduling strategy; let i=1, and repeat the following process until the i-th search result includes a reactive power optimization scheduling strategy that meets the intraday scheduling constraints: perform iterative search processing on the i-th chaotic population to obtain the i-th search result; update the i-th chaotic population according to the i-th search result to obtain the (i+1)-th chaotic population; let i=i+1.

[0157] In one possible implementation, the optimization module 404 is specifically used for: if it is determined that the random number of the i-th chaotic population is less than the switching probability, then performing a local search on the i-th chaotic population to obtain the i-th search result; wherein the switching probability is obtained based on the number of iterations; if it is determined that the random number of the i-th chaotic population is greater than the switching probability, then performing a global search on the i-th chaotic population to obtain the i-th search result.

[0158] In one possible implementation, the optimization module 404 is specifically used to: perform local search processing on the i-th chaotic population based on a multi-strategy mutation mechanism to obtain the i-th local search result; wherein, the multi-strategy mutation mechanism includes at least two of the following: basic algorithm mutation strategy, directed mutation strategy, and differential evolution mutation strategy; the i-th local search result includes at least two initial local search results; and determine the i-th search result from the i-th local search results based on the fitness values ​​of the at least two initial local search results.

[0159] In one possible implementation, intraday scheduling constraints include, but are not limited to: power balance constraints, node voltage constraints, branch power constraints, distributed energy output constraints, and intraday scheduling flexible load constraints.

[0160] In one possible implementation, the processing module 402 is specifically used to: calculate the output forecast data, flexible load demand data and grid parameters of the target converter equipment cluster based on the day-ahead scheduling objective function and day-ahead scheduling constraints in the day-ahead reactive power optimization scheduling model, and obtain the day-ahead scheduling plan of the target converter equipment cluster for the next day.

[0161] In one possible implementation, day-ahead scheduling constraints include, but are not limited to: power balance constraints, node voltage constraints, branch power constraints, distributed energy output constraints, and day-ahead scheduling flexible load constraints.

[0162] The apparatus in this embodiment can execute the technical solutions in the above method. Its specific implementation process and technical principles are the same, and will not be repeated here.

[0163] Figure 5 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this application, such as... Figure 5 As shown, the electronic device includes: a memory 501 and a processor 502; the memory 501 is a memory used to store instructions executable by the processor 502.

[0164] The processor 502 is configured to perform the method provided in the above embodiments.

[0165] The electronic device also includes a receiver 503 and a transmitter 504. The receiver 503 is used to receive instructions and data sent by other devices, and the transmitter 504 is used to send instructions and data to external devices.

[0166] The specific implementation process of the processor can be found in the above method embodiments, and its implementation principle and technical effect are similar, so it will not be repeated here.

[0167] In the above embodiments, it should be understood that the processor can be a Central Processing Unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), etc. The general-purpose processor can be a microprocessor or any conventional processor. The steps of the method disclosed in this invention can be directly implemented by a hardware processor, or implemented by a combination of hardware and software modules within the processor.

[0168] This application also provides a computer-readable storage medium storing computer-executable instructions, which, when executed on a computer, cause the computer to perform the technical solutions described above.

[0169] The aforementioned readable storage medium can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory, electrically erasable programmable read-only memory, erasable programmable read-only memory, programmable read-only memory, read-only memory, magnetic storage, flash memory, magnetic disk, or optical disk. The readable storage medium can be any available medium accessible to a general-purpose or special-purpose computer.

[0170] An exemplary readable storage medium is coupled to a processor, enabling the processor to read information from and write information to the readable storage medium. The readable storage medium can also be a component of the processor. The processor and the readable storage medium can reside in an application-specific integrated circuit (ASIC). Alternatively, the processor and the readable storage medium can exist as discrete components in a device.

[0171] This application also provides a computer program product, which includes a computer program stored in a computer-readable storage medium. At least one processor can read the computer program from the computer-readable storage medium, and when the at least one processor executes the computer program, it can implement the technical solutions in the above embodiments.

[0172] Those skilled in the art will understand that all or part of the steps of the above-described method embodiments can be implemented by hardware related to program instructions. The aforementioned program can be stored in a computer-readable storage medium. When executed, the program performs the steps of the above-described method embodiments; and the aforementioned storage medium includes various media capable of storing program code, such as magnetic disks or optical disks.

[0173] Finally, it should be noted that other embodiments of the invention will readily occur to those skilled in the art upon consideration of the specification and practice of the invention disclosed herein. This invention is intended to cover any variations, uses, or adaptations of the invention that follow the general principles of the invention and include common knowledge or customary techniques in the art not disclosed herein, and is not limited to the precise structures described above and shown in the accompanying drawings, and various modifications and changes can be made without departing from its scope. The scope of the invention is limited only by the appended claims.

Claims

1. A reactive power optimization scheduling method for a cluster of power conversion devices, characterized in that, The application relates to a power flow prediction data and real-time operation data of a target converter cluster are acquired. A day-ahead reactive power optimization scheduling model is used to process the power flow prediction data of the target converter cluster, so that a day-ahead scheduling plan of the target converter cluster in the future one day is obtained; wherein the optimization target of the day-ahead reactive power optimization scheduling model comprises power grid power quality indexes and operation economy indexes. A rolling update mechanism is used to correct the power flow prediction data of the target converter cluster according to the real-time operation data of the target converter cluster, so that corrected power flow prediction data of the target converter cluster is obtained. An improved pollen algorithm is used to calculate and process the day-ahead scheduling plan and the corrected power flow prediction data of the target converter cluster, so that a reactive power optimization scheduling strategy of the target converter cluster is obtained. The improved pollen algorithm is used to calculate and process the day-ahead scheduling plan and the corrected power flow prediction data of the target converter cluster, so that a reactive power optimization scheduling strategy of the target converter cluster is obtained.

2. The method of claim 1, wherein, The improved pollen algorithm is used to calculate and process the day-ahead scheduling plan and the corrected power flow prediction data of the target converter cluster, so that a reactive power optimization scheduling strategy of the target converter cluster is obtained. The improved pollen algorithm is used to calculate and process the day-ahead scheduling plan and the corrected power flow prediction data of the target converter cluster, so that a reactive power optimization scheduling strategy of the target converter cluster is obtained. The improved pollen algorithm is used to calculate and process the day-ahead scheduling plan and the corrected power flow prediction data of the target converter cluster, so that a reactive power optimization scheduling strategy of the target converter cluster is obtained.

3. The method of claim 2, wherein, The improved pollen algorithm is used to calculate and process the day-ahead scheduling plan and the corrected power flow prediction data of the target converter cluster, so that a reactive power optimization scheduling strategy of the target converter cluster is obtained. The improved pollen algorithm is used to calculate and process the day-ahead scheduling plan and the corrected power flow prediction data of the target converter cluster, so that a reactive power optimization scheduling strategy of the target converter cluster is obtained. The improved pollen algorithm is used to calculate and process the day-ahead scheduling plan and the corrected power flow prediction data of the target converter cluster, so that a reactive power optimization scheduling strategy of the target converter cluster is obtained.

4. The method of claim 3, wherein, The improved pollen algorithm is used to calculate and process the day-ahead scheduling plan and the corrected power flow prediction data of the target converter cluster, so that a reactive power optimization scheduling strategy of the target converter cluster is obtained. The improved pollen algorithm is used to calculate and process the day-ahead scheduling plan and the corrected power flow prediction data of the target converter cluster, so that a reactive power optimization scheduling strategy of the target converter cluster is obtained. The improved pollen algorithm is used to calculate and process the day-ahead scheduling plan and the corrected power flow prediction data of the target converter cluster, so that a reactive power optimization scheduling strategy of the target converter cluster is obtained.

5. The method of claim 4, wherein, The improved pollen algorithm is used to calculate and process the day-ahead scheduling plan and the corrected power flow prediction data of the target converter cluster, so that a reactive power optimization scheduling strategy of the target converter cluster is obtained. The improved pollen algorithm is used to calculate and process the day-ahead scheduling plan and the corrected power flow prediction data of the target converter cluster, so that a reactive power optimization scheduling strategy of the target converter cluster is obtained. The improved pollen algorithm is used to calculate and process the day-ahead scheduling plan and the corrected power flow prediction data of the target converter cluster, so that a reactive power optimization scheduling strategy of the target converter cluster is obtained. The improved pollen algorithm is used to calculate and process the day-ahead scheduling plan and the corrected power flow prediction data of the target converter cluster, so that a reactive power optimization scheduling strategy of the target converter cluster is obtained. The improved pollen algorithm is used to calculate and process the day-ahead scheduling plan and the corrected power flow prediction data of the target converter cluster, so that a reactive power optimization scheduling strategy of the target converter cluster is obtained. The improved pollen algorithm is used to calculate and process the day-ahead scheduling plan and the corrected power flow prediction data of the target converter cluster, so that a reactive power optimization scheduling strategy of the target converter cluster is obtained. The improved pollen algorithm is used to calculate and process the day-ahead scheduling plan and the corrected power flow prediction data of the target converter cluster, so that a reactive power optimization scheduling strategy of the target converter cluster is obtained. The improved pollen algorithm is used to calculate and process the day-ahead scheduling plan and the corrected power flow prediction data of the target converter cluster, so that a reactive power optimization scheduling strategy of the target converter cluster is obtained. The improved pollen algorithm is used to calculate and process the day-ahead scheduling plan and the corrected power flow prediction data of the target converter cluster, so that a reactive power optimization scheduling strategy of the target converter cluster is obtained. The improved pollen algorithm is used to calculate and process the day-ahead scheduling plan and the corrected power flow prediction data of the target converter cluster, so that a reactive power optimization scheduling strategy of the target converter cluster is obtained. The improved pollen algorithm is used to calculate and process the day-ahead scheduling plan and the corrected power flow prediction data of the target converter cluster, so that a reactive power optimization scheduling strategy of the target converter cluster is obtained. The improved pollen algorithm is used to calculate and process the day-ahead scheduling plan and the corrected power flow prediction data of the target converter cluster, so that a reactive power optimization scheduling strategy of the target converter cluster is obtained. The improved pollen algorithm is used to calculate and process the day-ahead scheduling plan and the corrected power flow prediction data of the target converter cluster, so that a reactive power optimization scheduling strategy of the target converter cluster is obtained. The improved pollen algorithm is used to calculate and process the day-ahead scheduling plan and the corrected power flow prediction data of the target converter cluster, so that a reactive power optimization scheduling strategy of the target converter cluster is obtained. The improved pollen algorithm is used to calculate and process the day-ahead scheduling plan and the corrected power flow prediction data of the target converter cluster, so that a reactive power optimization scheduling strategy of the target converter cluster is obtained. The improved pollen algorithm is used to calculate and process the day-ahead scheduling plan and the corrected power flow prediction data of the target converter cluster, so that a reactive power optimization scheduling strategy of the target converter cluster is obtained. The improved pollen algorithm is used to calculate and process the day-ahead scheduling plan and the corrected power flow prediction data of the target converter cluster, so that a reactive power optimization scheduling strategy of the target converter cluster is obtained. The improved pollen algorithm is used to calculate and process the day-ahead scheduling plan and the corrected power flow prediction data of the target converter cluster, so that a reactive power optimization scheduling strategy of the target converter cluster is obtained. The improved pollen algorithm is used to calculate and process the day-ahead scheduling plan and the corrected power flow prediction data of the target converter cluster, so that a reactive power optimization scheduling strategy of the target converter cluster is obtained. The improved pollen algorithm is used to calculate and process the day-ahead scheduling plan and the corrected power flow prediction data of the target converter cluster, so that a reactive power optimization scheduling strategy of the target converter cluster is obtained. The improved pollen algorithm is used to calculate and process the day-ahead scheduling plan and the corrected power flow prediction data of the target converter cluster, so that a reactive power optimization scheduling strategy of the target converter cluster is obtained. The improved pollen algorithm is used to calculate and process the day-ahead scheduling plan and the corrected power flow prediction data of the target converter cluster, so that a reactive power optimization scheduling strategy of the target converter cluster is obtained. The improved pollen algorithm is used to calculate and process the day-ahead scheduling plan and the corrected power flow prediction data of the target converter cluster, so that a reactive power optimization scheduling strategy of the target converter cluster is obtained. The improved pollen algorithm is used to calculate and process the day-ahead scheduling plan and the corrected power flow prediction data of the target converter cluster, so that a reactive power optimization scheduling strategy of the target converter cluster is obtained. The improved pollen algorithm is used to calculate and process the day-ahead scheduling plan and the corrected power flow prediction data of the target converter cluster, so that a reactive power optimization scheduling strategy of the target converter cluster is obtained. The improved pollen algorithm is used to calculate and process the day-ahead scheduling plan and the corrected power flow prediction data of the target converter cluster, so that a reactive power optimization scheduling strategy of the target converter cluster is obtained. The improved pollen algorithm is used to calculate and process the day-ahead scheduling plan and the corrected power flow prediction data of the target converter cluster, so that a reactive power optimization scheduling strategy of the target converter cluster is obtained. The improved pollen algorithm is used to calculate and process the day-ahead scheduling plan and the corrected power flow prediction data of the target converter cluster, so that a reactive power optimization scheduling strategy of the target converter cluster is obtained. The improved pollen algorithm is used to calculate and process the day-ahead scheduling plan and the corrected power flow prediction data of the target converter cluster, so that a reactive power optimization scheduling strategy of the target converter cluster is obtained. The improved pollen algorithm is used to calculate and process the day-ahead scheduling plan and the corrected power flow prediction data of the target converter cluster, so that a reactive power optimization scheduling strategy of the target converter cluster is obtained. The improved pollen algorithm is used to calculate and process the day-ahead scheduling plan and the corrected power flow prediction data of the target converter cluster, so that a reactive power optimization scheduling strategy of the target converter cluster is obtained. The improved pollen algorithm is used to calculate and process the day-ahead scheduling plan and the corrected power flow prediction data of the target converter cluster, so that a reactive power optimization scheduling strategy of the target converter cluster is obtained. The improved pollen algorithm is used to calculate and process the day-ahead scheduling plan and the corrected power flow prediction The i th chaotic population is subjected to local search processing based on a multi-strategy mutation mechanism to obtain an i th local search result; the multi-strategy mutation mechanism includes at least two of the following: a basic algorithm mutation strategy, a directional mutation strategy, and a differential evolution mutation strategy; the i th local search result includes at least two initial local search results; An i th search result is determined from the i th local search result according to fitness values of the at least two initial local search results.

6. The method of claim 2, wherein, The intraday scheduling constraint conditions include, but are not limited to, power balance constraints, node voltage constraints, branch power constraints, distributed energy output constraints, and intraday scheduling flexible load constraints.

7. The method according to any one of claims 1 to 6, characterized in that, The output prediction data of the target converter cluster is processed based on the day-ahead reactive power optimization scheduling model to obtain a day-ahead scheduling plan of the target converter cluster for the next day, including: The output prediction data, flexible load demand data, and grid parameters of the target converter cluster are calculated based on a day-ahead scheduling objective function and day-ahead scheduling constraint conditions in the day-ahead reactive power optimization scheduling model to obtain a day-ahead scheduling plan of the target converter cluster for the next day.

8. The method of claim 7, wherein, The day-ahead scheduling constraint conditions include, but are not limited to, power balance constraints, node voltage constraints, branch power constraints, distributed energy output constraints, and day-ahead scheduling flexible load constraints.

9. A reactive power optimization scheduling device of a cluster of power conversion devices, characterized in that, The method comprises: An acquisition module is configured to acquire output prediction data and real-time operation data of a target converter cluster; A processing module is configured to process the output prediction data of the target converter cluster based on a day-ahead reactive power optimization scheduling model to obtain a day-ahead scheduling plan of the target converter cluster for the next day; wherein an optimization objective of the day-ahead reactive power optimization scheduling model includes power quality indicators and operation economic indicators of a distribution network; A correction module is configured to correct the output prediction data of the target converter cluster based on a rolling update mechanism according to real-time operation data of the target converter cluster to obtain corrected output prediction data of the target converter cluster; An optimization module is configured to calculate and process the intraday scheduling plan and the corrected output prediction data of the target converter cluster using an improved pollen algorithm to obtain a reactive power optimization scheduling strategy of the target converter cluster.

10. An electronic device, comprising: The method comprises: A memory and a processor; The memory stores computer execution instructions; The processor executes the computer execution instructions stored in the memory, so that the processor executes the method according to any one of claims 1-8. The method comprises: A memory and a processor; The memory stores computer execution instructions; The processor executes the computer execution instructions stored in the memory, so that the processor executes the method according to any one of claims 1-8.

Images