Distributed photovoltaic engineering capacity optimization configuration method and system

Abstract

The application discloses a kind of distributed photovoltaic engineering capacity optimization configuration method and system, and the application relates to the technical field of distribution network distributed photovoltaic planning.It includes the following steps: obtaining the daily average temperature and environmental parameters of photovoltaic cell components connected to the distribution network, and generating a sample data set.Using linear regression algorithm to determine the fitting relationship between environmental parameters and battery component temperature, and establishing a temperature-environmental characterization formula. Calculate the daily average temperature of the photovoltaic to be connected through the formula, and analyze the similarity with similar samples in the sample data set to select similar samples. Establish a power characteristic parameter prediction model, and based on the access location and capacity of similar samples, predict the power characteristic parameters. With the minimum total power loss of the distribution network, the minimum voltage deviation and the maximum photovoltaic consumption ratio as the optimization goal, the optimal photovoltaic engineering capacity is determined through the constraint condition and optimization algorithm, to improve the operation efficiency of the distribution network and effectively utilize renewable energy.

Description

Technical Field

[0001] This invention relates to the field of distributed photovoltaic planning technology for power distribution networks, specifically a method and system for optimizing the capacity configuration of distributed photovoltaic projects. Background Technology

[0002] With the transformation of the global energy structure and the rapid development of renewable energy, distributed photovoltaic (PV) power generation, as an important form of renewable energy, has gradually attracted widespread attention. Due to its flexibility, reliability, and environmental friendliness, distributed PV systems have become an important component in achieving energy transition. However, the integration of distributed PV poses new challenges to the operation and management of distribution networks, especially in terms of capacity configuration and optimization. Existing integration technologies often face problems such as increased power losses in the distribution network, larger voltage deviations, and low PV integration rates, thus affecting the overall economic efficiency and stability of the system.

[0003] Traditional photovoltaic (PV) capacity configuration methods are mostly based on intuitive experience or simple calculation models, lacking in-depth analysis of the actual operating conditions of the distribution network. Particularly in node selection, they fail to fully consider the impact of environmental parameters on PV module performance, resulting in the PV system's power generation efficiency not being fully realized. Furthermore, most methods fail to effectively integrate data from different nodes, leading to significant discrepancies between the characteristics of PV power generation and the adaptability of the distribution network under different environmental conditions. Simultaneously, the size of the connected capacity can cause grid flow problems. If the connected PV capacity is too large, power flow will reverse, increasing line losses—even more than before the PV capacity was connected to the grid. Therefore, in practical applications, the integration of distributed PV often fails to improve the efficiency and economy of the distribution network as expected.

[0004] In the prior art, CN117132318A discloses a method for site selection and capacity determination of distributed photovoltaic (PV) in active distribution networks that considers the levelized cost of electricity (LCOE). This method can consider the static voltage stability index and loss sensitivity of nodes when selecting LCOE for distributed PV in active distribution networks, and consider the LCOE and IGBT junction temperature when determining capacity, which helps to reduce the operating cost and equipment investment cost of the grid. The method constructs an optimization model for distributed PV capacity in active distribution networks that considers the LCOE, with the peak LCOE of PV IGBT junction temperature, annual network loss, and LCOE as objective functions. However, this method does not consider the impact of environmental parameters on the performance of PV modules, nor does it consider grid power flow issues, thus reducing the accuracy and effectiveness of the optimization results.

[0005] The information disclosed in the background section is only intended to enhance the understanding of the background of this disclosure, and therefore may include information that does not constitute prior art known to those skilled in the art. Summary of the Invention

[0006] The purpose of this invention is to provide a method and system for optimizing the capacity configuration of distributed photovoltaic projects, so as to solve the problems mentioned in the background art.

[0007] To achieve the above objectives, the present invention provides the following technical solution:

[0008] A method for optimizing the capacity allocation of distributed photovoltaic projects, comprising the following steps:

[0009] When a distributed photovoltaic (PV) system connected to a power distribution network is in operation, the system obtains the average daily temperature of the distributed PV modules, environmental parameters of the location of the distributed PV system, and power characteristic parameters of the corresponding power distribution network. Based on the obtained data, a sample dataset is generated.

[0010] The fitting relationship between environmental parameters and the daily average temperature of battery modules at the location of distributed photovoltaics was determined by using sample datasets, and the temperature-environment characterization between each environmental parameter and the temperature of battery modules was determined by using linear regression algorithm.

[0011] Environmental parameters of the location of the distributed photovoltaic system to be connected are collected, and the daily average temperature of the distributed photovoltaic system to be connected is obtained through the temperature-environment characterization relationship. Based on the daily average temperature, the environmental similarity between the distributed photovoltaic system to be connected and each distributed photovoltaic cell module is calculated, and distributed photovoltaic systems with a similarity threshold less than the set threshold are regarded as similar samples.

[0012] A power characteristic parameter prediction model is established, with the access location and capacity of similar samples in the distribution network, the total power consumption of the distribution networks connected to the similar samples, and the distribution network topology as inputs, and the power characteristic parameters of the connected distribution networks as labels, to train the model;

[0013] The historical power characteristic parameters of the target distribution network to be connected to distributed photovoltaic (PV) are obtained before the connection of distributed PV. The predicted power characteristic parameters after the connection of distributed PV with different capacities are predicted by the power characteristic parameter prediction model. The predicted values ​​are then compared with the historical power characteristic parameters to analyze the power loss of the corresponding distribution network.

[0014] Based on the obtained power loss of the distribution network, an optimization target is set, with capacity as the optimization object. Based on the constraints of the distribution network operation, the optimal capacity for access is determined, and the capacity optimization configuration is completed.

[0015] Furthermore, the specific construction method of the sample dataset is as follows: the daily average temperature of the battery module is mapped one-to-one with the environmental parameters of the distributed photovoltaic location to generate the sample dataset; the power characteristic parameters of the distribution network specifically refer to: the active power, reactive power and terminal voltage values ​​of each branch and node in the distribution network topology; wherein the distribution network topology is specifically: the distribution network line is divided into multiple branches, and each branch has several nodes.

[0016] Further, environmental parameters of the location of the distributed photovoltaic are collected, where the environmental parameters of the location of the distributed photovoltaic include solar irradiance, ambient temperature, and ambient wind speed. The specific expression based on which the temperature - environment characterization formula is established is as follows:

[0017] T S = T air + 0.0138G s *(1 + 0.031T air )*(1 - 0.042V W )

[0018] In the formula, T S is the daily average temperature of the battery module, T air is the daily average temperature at the location of the distributed photovoltaic, G s is the daily average solar irradiance at the location of the distributed photovoltaic, V w is the daily average wind speed at the surface of the distributed photovoltaic;

[0019] Based on the daily average temperature, the environmental similarity between the to - be - connected distributed photovoltaic and each distributed photovoltaic battery module is calculated. The formula based on which the environmental similarity is calculated is as follows:

[0020] XS = ω1*|T1 - T2|+ω2*|R1 - R2|

[0021] In the formula, XS is the environmental similarity, T1 is the daily average temperature of the battery module of the to - be - connected distributed photovoltaic, T2 is the daily average temperature of the battery module of the distributed photovoltaic sample, R1 is the total resistance of the target distribution network, R2 is the total resistance of the distribution network to which the distributed photovoltaic sample is connected, ω1 and ω2 are the weight coefficients of the temperature difference and the total resistance difference of the distribution network respectively, where ω1 < ω2, ω1+ω2 = 1 and both ω1 and ω2 are greater than 0. The distributed photovoltaic sample is the distributed photovoltaic connected to the distribution network.

[0022] Further, similar samples with an environmental similarity less than the similarity threshold to the distribution network of the to - be - connected distributed photovoltaic are screened out. The logic based on which the samples are screened is as follows:

[0023] When XS ≥ yz, it indicates that the difference between the working environment of this distributed photovoltaic sample and the working environment of the to - be - connected distributed photovoltaic does not meet the requirements, so this distributed photovoltaic sample is not taken as a similar sample;

[0024] When XS < yz, it indicates that the difference between the working environment of this distributed photovoltaic sample and the working environment of the to - be - connected distributed photovoltaic meets the requirements, so this distributed photovoltaic sample is taken as a similar sample, where yz is the similarity threshold.

[0025] Furthermore, the predicted values ​​of power characteristic parameters are compared with historical power characteristic parameters to analyze the power loss of the distribution network under different photovoltaic capacities. The historical power characteristic parameters specifically refer to the power characteristic parameters of the target distribution network before the integration of distributed photovoltaics. The formula used to calculate the power loss of the distribution network under different photovoltaic capacities is as follows:

[0026]

[0027] In the formula, P′ loss To reduce the power loss in the target distribution network after photovoltaic integration, K i To monitor the switching operation of the i-th branch line of the target distribution network after photovoltaic integration, P mi Q represents the historical active power of the i-th branch line. mi U represents the historical reactive power of the i-th branch line. mi Let P be the historical terminal voltage value of the i-th branch line. pvi Q represents the predicted active power of the i-th branch line of the target distribution network after photovoltaic integration. pvi The predicted reactive power value, U, of the i-th branch line of the target distribution network after photovoltaic integration. Pi R represents the predicted terminal voltage of the i-th branch line of the target distribution network after photovoltaic integration. i Let be the resistance value of the i-th branch line, where i is the index of the branch line, i∈[1,n], and n is the total number of branch lines;

[0028] The switching operation status K of the i-th branch line i It is a numerical code, taking only two values: 0 and 1. When K i When K = 0, it means the circuit is not closed. i When the value is 1, it indicates that the circuit is closed.

[0029] Furthermore, the optimization objectives are specifically set as follows: minimizing the total power loss of the distribution network, minimizing the voltage deviation, and maximizing the photovoltaic absorption ratio, wherein the formula for calculating the voltage deviation is:

[0030]

[0031] In the formula, F2 is the voltage offset, and U i,ts U is the rated voltage of the t-th node in the i-th branch line. i,tN Let m be the predicted actual voltage value of the t-th node in the i-th branch line, where t is the index of the node in the branch line. i Let be the total number of nodes on the i-th branch line;

[0032] The formula used to calculate the photovoltaic power absorption ratio is as follows:

[0033]

[0034] In the formula, F3 is the photovoltaic absorption ratio, and G pv For distributed photovoltaic power generation to be connected, E C The target is the total electricity consumption of the distribution network, of which the output of distributed photovoltaic power to be connected is G. pv The formula used for the calculation is:

[0035]

[0036] In the formula, G pvx Let A be the per-unit value of the actual injected power of the x-th photovoltaic panel in the distributed photovoltaic system to be connected. x and δ x h refers to the area of ​​the x-th photovoltaic panel to be connected in the distributed photovoltaic system and its photoelectric conversion efficiency, respectively. max The maximum light intensity is given by x, where x is the index of the photovoltaic panel, and X is the total number of photovoltaic panels to be connected in the distributed photovoltaic system.

[0037] Furthermore, based on the obtained power loss of the distribution network, the comprehensive optimization objective is represented by a fitness function, with the optimization goals of minimizing the total power loss of the distribution network, minimizing the voltage deviation, and maximizing the photovoltaic absorption ratio. The specific formula for calculating the fitness function is as follows:

[0038]

[0039] In the formula, FZ is the fitness function, where ω1, ω2 and ω3 are the weighting coefficients of total power loss, voltage deviation and photovoltaic absorption ratio, respectively, where ω1≥ω2>ω3 and ω1, ω2 and ω3 are all greater than 0;

[0040] Since the total power loss, voltage deviation, and photovoltaic absorption ratio of the distribution network will vary depending on the photovoltaic capacity connected, the photovoltaic project capacity with the largest fitness function is selected from the range of available photovoltaic capacity.

[0041] Furthermore, based on the constraints of distribution network operation, specifically referring to the constraints of standard power flow, node voltage, and branch current, the specific expressions upon which these constraints are based are as follows:

[0042]

[0043] In the formula, ΔP and ΔQ represent the active power balance equation and the reactive power balance equation, respectively. i,Gt and Q i,Gt V represents the active and reactive power of the t-th node in the i-th branch line, respectively. i,j XC represents the maximum voltage at node t in the i-th branch line. i,tj and Bi,tj Let θ be the conductance and susceptance between nodes t and j in the i-th branch line. i,tj U represents the phase angle difference between nodes t and j in the i-th branch line. i,tmin and U i,tmax In the i-th branch line, let I be the minimum and maximum voltage amplitudes allowed by the t-th node. i Let I be the current in the i-th branch line. imax Let be the upper limit of the current for the i-th branch line.

[0044] This invention also provides a distributed photovoltaic (PV) project capacity optimization and configuration system, which is used to execute the above-described distributed PV project capacity optimization and configuration method, including:

[0045] The sample data acquisition module is used to obtain the average daily temperature of the distributed photovoltaic cell modules, the environmental parameters of the location of the distributed photovoltaic and the power characteristic parameters of the corresponding distribution network when the distributed photovoltaic connected to the distribution network is working, and to generate a sample dataset based on the acquired data.

[0046] The environmental impact characterization module is used to determine the fitting relationship between environmental parameters and the daily average temperature of battery modules at the location of distributed photovoltaics through sample datasets, and to determine the temperature-environment characterization between each environmental parameter and the temperature of battery modules using a linear regression algorithm.

[0047] The similar sample screening module is used to collect environmental parameters of the location of the distributed photovoltaic system to be connected, and obtain the daily average temperature of the distributed photovoltaic system to be connected through the temperature-environment characterization relationship. Based on the daily average temperature, the environmental similarity between the distributed photovoltaic system to be connected and each distributed photovoltaic cell module is calculated, and distributed photovoltaic systems with a similarity threshold less than the set threshold are regarded as similar samples.

[0048] The prediction model training module is used to establish a prediction model for power characteristic parameters. It takes the access location and capacity of similar samples in the distribution network, the total power consumption of the distribution networks connected to the similar samples, and the distribution network topology as inputs, and uses the power characteristic parameters of the connected distribution networks as labels to train the model.

[0049] The power loss characterization module is used to obtain the historical power characteristic parameters of the target distribution network to be connected to distributed photovoltaics when photovoltaics are not connected. It also uses a power characteristic parameter prediction model to predict the power characteristic parameters after connecting distributed photovoltaics of different capacities, and compares them with the historical power characteristic parameters to analyze the power loss of the corresponding distribution network.

[0050] The capacity optimization configuration module is used to set optimization targets based on the obtained power loss of the distribution network, take capacity as the optimization object, and determine the optimal capacity to be connected based on the constraints of the distribution network operation, thereby completing the capacity optimization configuration.

[0051] Compared with the prior art, the beneficial effects of the present invention are:

[0052] First, this method establishes a fitting relationship between environmental parameters and temperature by acquiring the daily average temperature and environmental parameters of distributed photovoltaic (PV) modules. This process not only lays the foundation for subsequent capacity optimization but also makes sensitivity analysis to environmental changes possible. The regression coefficients and temperature-environment characterization formula obtained through linear regression algorithms make PV module temperature prediction more accurate under different geographical and climatic conditions, thus providing reliable data support for the actual operation of PV systems.

[0053] Secondly, by calculating the similarity between the daily average temperature of the photovoltaic (PV) module to be connected and similar samples within the sample dataset, distributed PV modules with similar operating characteristics can be effectively screened. This improves the model's prediction accuracy and enhances its adaptability to different power characteristics of the distribution network. By establishing a power characteristic parameter prediction model and further analyzing the impact of connecting PV modules of different capacities on the power characteristic parameters of the distribution network based on the connection location and capacity of similar samples, a forward-looking analysis of the distribution network's operating status can be achieved.

[0054] Finally, through optimization analysis of power loss, node voltage, and branch current, this method can achieve optimal configuration of the photovoltaic system while ensuring the safe operation of the distribution network. With the optimization objectives of minimizing total power loss, minimizing voltage deviation, and maximizing photovoltaic absorption ratio, the optimal solution for the capacity of connected distributed photovoltaic projects is found. This not only improves the operational economy of the distribution network but also enhances the integration capacity of distributed photovoltaic systems, thereby promoting the widespread application of renewable energy. Attached Figure Description

[0055] Figure 1 This is a schematic diagram of the overall method flow of the present invention;

[0056] Figure 2 This is a typical topology of the node distribution network of the present invention;

[0057] Figure 3 A fitted curve of photovoltaic cell module temperature versus average daily solar irradiance;

[0058] Figure 4 A fitted curve of photovoltaic cell module temperature versus average daily wind speed at the photovoltaic surface;

[0059] Figure 5This is a fitted curve of photovoltaic cell module temperature versus the average daily temperature of the distribution network location;

[0060] Figure 6 This is a schematic diagram of the overall system structure of the present invention. Detailed Implementation

[0061] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to specific embodiments.

[0062] It should be noted that, unless otherwise defined, the technical or scientific terms used in this invention should have the ordinary meaning understood by one of ordinary skill in the art to which this invention pertains. The terms "first," "second," and similar terms used in this invention do not indicate any order, quantity, or importance, but are merely used to distinguish different components. Terms such as "comprising" or "including" mean that the element or object preceding the word encompasses the elements or objects listed following the word and their equivalents, without excluding other elements or objects. Terms such as "connected" or "linked" are not limited to physical or mechanical connections, but can include electrical connections, whether direct or indirect. Terms such as "upper," "lower," "left," and "right" are used only to indicate relative positional relationships; when the absolute position of the described object changes, the relative positional relationship may also change accordingly.

[0063] Example:

[0064] Please see Figures 1-5 The present invention provides a technical solution:

[0065] A method for optimizing the capacity allocation of distributed photovoltaic projects, comprising the following steps:

[0066] Step 1: Obtain the average daily temperature of the distributed photovoltaic (PV) modules, environmental parameters of the location of the distributed PV, and power characteristic parameters of the corresponding distribution network when the distributed PV is connected to the power distribution network. Generate a sample dataset based on the obtained data.

[0067] The sample dataset is constructed by mapping the daily average temperature of the battery modules to the environmental parameters of the distributed photovoltaic location to generate the sample dataset. The power characteristic parameters of the distribution network specifically refer to the active power, reactive power, and terminal voltage values ​​of each branch and node in the distribution network topology. The distribution network topology is specifically defined as dividing the distribution network lines into multiple branches, each with several nodes. Please refer to [link to details]. Figure 2 In the diagram, U0, U1, ..., U P ,…,U n Represents the voltage at each node, S0, S1, ..., S P ,…,Sn Indicates the load of each branch, l0, l1, ..., l P ,…,l n This indicates the distance between each branch road.

[0068] Based on a typical line segment in a distribution network topology, the voltage and network loss are analyzed, and formulas are derived using the selected line model. To analyze the changes in voltage and network loss after distributed photovoltaic (PV) power generation is connected to the distribution network, this analysis is used to examine the operation of the distribution network before the integration of PV power, specifically including historical power characteristic parameters.

[0069] The acquired power characteristic parameters are specifically the average values ​​of power characteristic parameters within a set collection period. The collection period can be a 24-hour day, but to ensure the accuracy of the power characteristic parameters, the collection period should be no less than 8 hours. Once the collection period is determined, all parameters collected below will be based on this collection period and will not be elaborated upon further in the following text.

[0070] Step 2: Determine the fitting relationship between environmental parameters and the daily average temperature of battery modules at the location of distributed photovoltaics through sample datasets, and use linear regression algorithm to determine the temperature-environment characterization between each environmental parameter and the battery module temperature.

[0071] Environmental parameters of the distributed photovoltaic (PV) location are collected, including solar irradiance, ambient temperature, and ambient wind speed. The specific expression upon which the temperature-environment characterization formula is based is as follows:

[0072] T S =T air +0.0138G s *(1+0.031T air )*(1-0.042V w )

[0073] In the formula, T S T is the daily average temperature of the battery module. air G represents the average daily temperature at the location of the distributed photovoltaic power generation facility. s V represents the average daily solar irradiance at the location of the distributed photovoltaic power generation system. W The average daily wind speed at the surface of the distributed photovoltaic system;

[0074] The average temperature of the battery module specifically refers to the daily average temperature of the battery module during the data collection period.

[0075] The same applies to the average daily solar irradiance and the average daily wind speed at the surface of distributed photovoltaic systems.

[0076] Ambient temperature has a significant impact on the operating efficiency of photovoltaic modules. The method for obtaining this information is as follows:

[0077] Temperature sensor: High-precision temperature sensors (such as thermocouples or digital thermometers) are installed near the photovoltaic modules to monitor the ambient temperature in real time.

[0078] Data logger: Connect the temperature sensor to the data logger to record temperature changes at regular intervals in order to analyze the temperature performance of the battery components under different conditions.

[0079] Weather station: Weather stations obtain the external ambient temperature and are usually equipped with temperature sensors, which can provide relatively accurate ambient temperature data.

[0080] Ambient wind speed is also an important factor affecting the heat dissipation and power generation efficiency of photovoltaic modules. Below are methods for obtaining wind speed:

[0081] Anemometer: A wind speed meter is used to measure the wind speed at a location. There are various types of anemometers, including rotary and ultrasonic types, and you can choose the one best suited to the site conditions.

[0082] Data logging system: Connect the anemometer to the data logging system to record wind speed data regularly, which facilitates environmental analysis and subsequent data processing.

[0083] Weather station: Modern weather stations are usually equipped with anemometers, which can provide real-time wind speed data.

[0084] Solar irradiance is one of the key parameters affecting the power generation efficiency of photovoltaic modules. It can be obtained through the following methods:

[0085] A pyranometer is a specialized radiometer used to measure the intensity of solar radiation received on the Earth's surface. Pyranometers provide real-time solar irradiance data, typically expressed in watts per square meter (W / m²). 2 (in units of )

[0086] Data acquisition system: Connect the radiometer to the data acquisition system to record irradiance data in real time. Appropriate time intervals can be selected for data recording for subsequent analysis.

[0087] Based on the collected data, the regression coefficients between various environmental parameters and battery component temperature were determined using a linear regression algorithm. Some of the collected data are shown in Table 1 below.

[0088] Table 1. Partial Environmental Data Collection Table

[0089]

[0090] Step 3: Collect environmental parameters of the location of the distributed photovoltaic system to be connected, and obtain the daily average temperature of the distributed photovoltaic system to be connected through the temperature-environment characterization relationship. Calculate the environmental similarity between the distributed photovoltaic system to be connected and each distributed photovoltaic cell module based on the daily average temperature, and take the distributed photovoltaic system with a similarity value less than the set similarity threshold as similar samples.

[0091] Based on the average daily temperature, the environmental similarity between the distributed photovoltaic system to be connected and each distributed photovoltaic cell module is calculated. The formula used for calculating the environmental similarity is as follows:

[0092] XS=ω1*|T1-T2|+ω2*|R1-R2|

[0093] In the formula, XS represents the environmental similarity, T1 represents the average daily temperature of the battery module to be connected to the distributed photovoltaic system, T2 represents the average daily temperature of the battery module of the distributed photovoltaic sample, R1 represents the total resistance of the target distribution network, R2 represents the total resistance of the distribution network to which the distributed photovoltaic sample is connected, and ω1 and ω2 are the weighting coefficients of the temperature difference and the total resistance difference of the distribution network, respectively, where ω1 < ω2, ω1 + ω2 = 1 and both ω1 and ω2 are greater than 0. The distributed photovoltaic sample is the distributed photovoltaic system connected to the distribution network.

[0094] It should be noted that the environmental similarity XS is characterized by the difference between the daily average temperature of the battery modules and the total resistance of the distribution network. The larger the environmental similarity XS value, the greater the difference in environment and structure between the sample and the distribution network to be connected to the distributed photovoltaic system.

[0095] Therefore, the environmental similarity XS is proportional to the absolute value of the average temperature difference of the battery module, |T1-T2|, and similarly proportional to the absolute value of the total resistance difference of the power distribution network, |R1-R2|.

[0096] The total resistance of a distribution network reflects the power transmission efficiency and system losses. High resistance leads to greater energy loss, affecting the overall performance and reliability of the power grid. When assessing the similarity of distribution networks, resistance is usually a key parameter affecting power transmission efficiency, stability, and load capacity. However, the effect of temperature is difficult to be completely similar due to constant fluctuations. Therefore, we set ω1 < ω2, ω1 + ω2 = 1, and both ω1 and ω2 are greater than 0.

[0097] Similar samples with a distribution network environment less than a similarity threshold to the distributed photovoltaic system to be connected are selected. The logic for selecting samples is as follows:

[0098] When XS≥yz, it means that the difference between the working environment of the distributed photovoltaic sample and the working environment of the distributed photovoltaic to be connected does not meet the requirements, and the distributed photovoltaic sample is not considered as a similar sample.

[0099] When XS < yz, it indicates that the difference in the working environment between the distributed PV sample and the distributed PV to be connected meets the requirements. Then, this distributed PV sample is regarded as a similar sample, where yz is the similarity threshold, which can be specifically set according to expert experience combined with the actual temperature and resistance difference. Generally, it is set between 1 and 100.

[0100] Step 4: Establish a power characteristic parameter prediction model. Using the connection position, capacity of the similar sample in the distribution network, the total power consumption of the distribution network connected by the similar sample, and the distribution network topology structure as inputs, and using the power characteristic parameters of the connected distribution network as labels, train the model.

[0101] The power characteristic parameter prediction model is established based on the LSTM model. For the long short-term memory network model LSTM model, select the activation function and optimization algorithm. Select the Tanh function as the activation function and select Adam as the optimization algorithm for the LSTM model; the formula of the Tanh function is:

[0102]

[0103] In the formula, f(r) represents the Tanh function, and the independent variable r represents the weighted sum of the neuron inputs, that is, the result after the inputs received by the neuron from the previous layer are weighted and summed.

[0104] At the same time, set the hyperparameters of the LSTM model. The hyperparameters of the LSTM model include: the number of network layers, the number of iterations, the learning rate, the batch size, the number of training times, the batch processing quantity, and the number of neurons in the hidden layer.

[0105] Among them, the number of network layers is set to a 3-layer network structure, the number of iterations is set to 200, the learning rate is set to 0.001, the batch size is set to 32, the number of training times is set to 100, the batch processing quantity is set to 256, and the number of neurons in the hidden layer is 32.

[0106] Step 5: Obtain the historical power characteristic parameters of the target distribution network of the distributed PV to be connected before the distributed PV is connected. Through the power characteristic parameter prediction model, predict the predicted values of the power characteristic parameters after connecting distributed PVs with different capacities, and compare them with the historical power characteristic parameters to analyze the power loss of the corresponding distribution network.

[0107] Regarding power loss, the historical power characteristic parameters specifically refer to the power characteristic parameters of the target distribution network before the distributed PV is connected. The formula based on which the power loss of the distribution network under different PV capacities is calculated is:

[0108]

[0109] In the formula, P′ lossTo reduce the power loss in the target distribution network after photovoltaic integration, K i To monitor the switching operation of the i-th branch line of the target distribution network after photovoltaic integration, P mi Q represents the historical active power of the i-th branch line. mi U represents the historical reactive power of the i-th branch line. mi Let P be the historical terminal voltage value of the i-th branch line. pvi Q represents the predicted active power of the i-th branch line of the target distribution network after photovoltaic integration. pvi The predicted reactive power value, U, of the i-th branch line of the target distribution network after photovoltaic integration. Pi R represents the predicted terminal voltage of the i-th branch line of the target distribution network after photovoltaic integration. i Let be the resistance value of the i-th branch line, where i is the index of the branch line, i∈[1,n], and n is the total number of branch lines;

[0110] The switching operation status K of the i-th branch line i It is a numerical code, taking only two values: 0 and 1. When K i When K = 0, it means the circuit is not closed. i When the value is 1, it indicates that the circuit is closed.

[0111] The power loss of the distribution network after photovoltaic (PV) grid connection is calculated using the above formula. The power loss will vary depending on the PV capacity connected. Therefore, the above method only calculates the power difference between the distribution network with and without PV grid connection. The above formula can be used to calculate the power loss of the distribution network with different PV capacities. The voltage offset calculation formula and the PV absorption ratio calculation formula are similar.

[0112] After photovoltaic (PV) power is connected, the active power loss of the line is proportional to the capacity of the PV power source. If the connected PV capacity is too large, the power flow will reverse, and the line loss will actually increase, even more than the loss before the PV capacity was connected to the grid.

[0113] Therefore, it can be seen that the line losses of each branch in the distribution network and the total network loss are related to the access capacity and access location of distributed power sources. If the power factor of all lines and the penetration rate of photovoltaic power sources do not change, then the grid connection location and capacity of photovoltaic power sources will become the main factors affecting the power parameters.

[0114] Step 6: Based on the obtained power loss of the distribution network, set the optimization target, take the capacity as the optimization object, and determine the optimal capacity for access based on the constraints of the distribution network operation, thereby completing the capacity optimization configuration.

[0115] The specific optimization objectives are: to minimize the total power loss of the distribution network, minimize the voltage deviation, and maximize the photovoltaic absorption ratio. The formula used to calculate the voltage deviation is:

[0116]

[0117] In the formula, F2 is the voltage offset, and U i,ts U is the rated voltage of the t-th node in the i-th branch line. i,tN Let m be the predicted actual voltage value of the t-th node in the i-th branch line, where t is the index of the node in the branch line. i Let be the total number of nodes on the i-th branch line;

[0118] The formula used to calculate the photovoltaic power absorption ratio is as follows:

[0119]

[0120] In the formula, F3 is the photovoltaic absorption ratio, and G pv For distributed photovoltaic power generation to be connected, E C The target is the total electricity consumption of the distribution network, of which the output of distributed photovoltaic power to be connected is G. pv The formula used for the calculation is:

[0121]

[0122] In the formula, G pvx Let A be the per-unit value of the actual injected power of the x-th photovoltaic panel in the distributed photovoltaic system to be connected. x and δ x h refers to the area of ​​the x-th photovoltaic panel to be connected in the distributed photovoltaic system and its photoelectric conversion efficiency, respectively. max The maximum light intensity is given by x, where x is the index of the photovoltaic panel, and X is the total number of photovoltaic panels to be connected in the distributed photovoltaic system.

[0123] It should be noted that the rated voltage of a node can be obtained in the following ways:

[0124] Monitoring systems: In modern power systems, monitoring systems (such as SCADA systems) can provide real-time voltage data and may include setting information about node rated voltages.

[0125] Power company database: Some power companies maintain a database that lists the relevant technical parameters of all nodes in their network, including rated voltage.

[0126] The area of ​​the xth photovoltaic panel, its photoelectric conversion efficiency, the per-unit value of the injected power, the maximum illuminance, and the total number of photovoltaic panels are all characterized by the actual values ​​of the photovoltaic distribution network connected to the grid. The maximum illuminance specifically refers to the maximum daily illuminance received by the photovoltaic panel.

[0127] Based on the obtained power loss of the distribution network, the optimization objectives of minimizing the total power loss of the distribution network, minimizing the voltage deviation, and maximizing the photovoltaic absorption ratio are expressed by a fitness function. The specific formula for calculating the fitness function is as follows:

[0128]

[0129] In the formula, FZ is the fitness function, where ω3, ω4 and ω5 are the weighting coefficients of total power loss, voltage deviation and photovoltaic absorption ratio, respectively, where ω3≥ω4>ω5 and ω3, ω4 and ω5 are all greater than 0;

[0130] The larger the fitness function FZ, the better the performance of the distribution network after the integration of distributed photovoltaic power and the smaller the loss. Since P′ loss The larger the value, the greater the power loss in the distribution network, therefore P′ loss It is inversely proportional to the fitness function FZ, through This indicates an inverse relationship; the larger the voltage offset F2, the greater the voltage deviation and the greater the distribution network error. Therefore, the voltage offset F2 is inversely proportional to the fitness function FZ. This indicates an inverse relationship; for the photovoltaic (PV) absorption ratio F3, a larger value indicates a larger distributed PV output and a smaller burden on the distribution network. Therefore, the PV absorption ratio F3 is directly proportional to the fitness function FZ. Voltage deviation affects the power supply quality of the distribution network; excessive voltage deviation may lead to equipment damage or performance degradation. Although voltage deviation is important, its impact is usually lower than that of power loss. The PV absorption ratio measures the effective utilization of PV power generation, and its importance is relatively low, especially in the basic operation of the distribution network. Therefore, we set ω3 ≥ ω4 > ω5, and ω3, ω4, and ω5 are all greater than 0.

[0131] Since the total power loss, voltage deviation, and photovoltaic absorption ratio of the distribution network will vary depending on the photovoltaic capacity connected, the photovoltaic project capacity with the largest fitness function is selected from the range of available photovoltaic capacity.

[0132] According to the constraints of distribution network operation, the constraints specifically refer to the constraints of standard power flow, node voltage, and branch current, and the specific expressions on which the constraints are based are as follows:

[0133]

[0134] In the formula, ΔP and ΔQ represent the active power balance equation and the reactive power balance equation, respectively. i,Gt and Q i,Gt V represents the active and reactive power of the t-th node in the i-th branch line, respectively. i,j XC represents the maximum voltage at node t in the i-th branch line. i,tj and B i,tj Let θ be the conductance and susceptance between nodes t and j in the i-th branch line. i,tj U represents the phase angle difference between nodes t and j in the i-th branch line. i,tmin and U i,tmax In the i-th branch line, let I be the minimum and maximum voltage amplitudes allowed by the t-th node. i Let I be the current in the i-th branch line. imax Let be the upper limit of the current for the i-th branch line.

[0135] Based on the established constraints, the optimal photovoltaic (PV) capacity for connection is determined through optimization algorithms, thus completing the optimized configuration of distributed PV capacity. These optimization algorithms include particle swarm optimization, ant colony optimization, and genetic algorithms, or optimization can be performed using the IEEE-69 node system. The IEEE-69 node system is one of the widely accepted standard test systems in the power system field, widely used in research on power system optimization and dispatching problems. It possesses a certain scale and complexity, encompassing components such as generators, loads, and transformers, while considering voltage and power balance constraints. It is a relatively medium-sized system, capable of reflecting the complexity of the problem without being overly large and complex, facilitating experimental verification and analysis of the algorithm. Its characteristics and topology represent some key features of actual power systems, such as different types of loads, generator limitations, and interconnected transmission lines. Furthermore, since the IEEE-69 node system is a standard test system, its topology and parameters are publicly available and widely shared. The specific optimization process uses common techniques and will not be elaborated upon here.

[0136] Please see Figure 6 The present invention also provides a distributed photovoltaic (PV) project capacity optimization configuration system, which is used to execute the above-described distributed PV project capacity optimization configuration method, including:

[0137] The sample data acquisition module is used to obtain the average daily temperature of the distributed photovoltaic cell modules, the environmental parameters of the location of the distributed photovoltaic and the power characteristic parameters of the corresponding distribution network when the distributed photovoltaic connected to the distribution network is working, and to generate a sample dataset based on the acquired data.

[0138] The environmental impact characterization module is used to determine the fitting relationship between environmental parameters and the daily average temperature of battery modules at the location of distributed photovoltaics through sample datasets, and to determine the temperature-environment characterization between each environmental parameter and the temperature of battery modules using a linear regression algorithm.

[0139] The similar sample screening module is used to collect environmental parameters of the location of the distributed photovoltaic system to be connected, and obtain the daily average temperature of the distributed photovoltaic system to be connected through the temperature-environment characterization relationship. Based on the daily average temperature, the environmental similarity between the distributed photovoltaic system to be connected and each distributed photovoltaic cell module is calculated, and distributed photovoltaic systems with a similarity threshold less than the set threshold are regarded as similar samples.

[0140] The prediction model training module is used to establish a prediction model for power characteristic parameters. It takes the access location and capacity of similar samples in the distribution network, the total power consumption of the distribution networks connected to the similar samples, and the distribution network topology as inputs, and uses the power characteristic parameters of the connected distribution networks as labels to train the model.

[0141] The power loss characterization module is used to obtain the historical power characteristic parameters of the target distribution network to be connected to distributed photovoltaics when photovoltaics are not connected. It also uses a power characteristic parameter prediction model to predict the power characteristic parameters after connecting distributed photovoltaics of different capacities, and compares them with the historical power characteristic parameters to analyze the power loss of the corresponding distribution network.

[0142] The capacity optimization configuration module is used to set optimization targets based on the obtained power loss of the distribution network, take capacity as the optimization object, and determine the optimal capacity to be connected based on the constraints of the distribution network operation, thereby completing the capacity optimization configuration.

[0143] The above formulas are all dimensionless calculations. The formulas are derived from software simulations based on a large amount of collected data to obtain the most recent real-world results. The preset parameters in the formulas are set by those skilled in the art according to the actual situation.

[0144] The above embodiments can be implemented, in whole or in part, by software, hardware, firmware, or any other combination thereof. When implemented in software, the above embodiments can be implemented, in whole or in part, as a computer program product. Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented by electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution.

[0145] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment, depending on actual needs.

[0146] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any changes or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application.

Claims

1. A distributed photovoltaic engineering capacity optimization configuration method, characterized in that, The specific steps include: Obtain the daily average temperature of the distributed photovoltaic cell assembly and the environmental parameters of the distributed photovoltaic location and the power characteristic parameters of the corresponding power distribution network when the distributed photovoltaic is working, and generate sample data sets according to the obtained data; Determine the fitting relationship between the environmental parameters of the distributed photovoltaic location and the daily average temperature of the cell assembly through the sample data sets, and determine the temperature-environmental characteristic between each environmental parameter and the cell assembly temperature using a linear regression algorithm; Collect the environmental parameters of the location where the to-be-attached distributed photovoltaic is located, and obtain the daily average temperature of the to-be-attached distributed photovoltaic through the temperature-environmental characteristic relationship, and calculate the environmental similarity of the to-be-attached distributed photovoltaic and each distributed photovoltaic cell assembly based on the daily average temperature, and take the distributed photovoltaic with a similarity less than a set similarity threshold as a similar sample; Establish a power characteristic parameter prediction model, take the access location, capacity of the similar sample in the power distribution network, total power consumption of the connected power distribution network and the topology structure of the power distribution network as inputs, and take the power characteristic parameters of the connected power distribution network as labels, and train the model; Obtain the historical power characteristic parameters of the target power distribution network of the to-be-attached distributed photovoltaic when the distributed photovoltaic is not connected, and predict the power characteristic parameter prediction value after connecting the distributed photovoltaic with different capacities through the power characteristic parameter prediction model, and compare it with the historical power characteristic parameters to analyze the power loss of the corresponding power distribution network; According to the obtained power loss of the power distribution network, set an optimization target, take the capacity as the optimization object, and determine the optimal capacity according to the constraint conditions of the operation of the power distribution network, and complete the capacity optimization configuration.

2. The method of claim 1, wherein: The specific construction method of the sample data set is to one-to-one map the daily average temperature of the cell assembly and the environmental parameters of the location of the distributed photovoltaic to generate the sample data set; the power characteristic parameters of the power distribution network specifically refer to the active power, reactive power and terminal voltage value of each branch and node in the topology structure of the power distribution network; wherein the topology structure of the power distribution network specifically refers to dividing the power distribution network line into multiple branches, and each branch is provided with a plurality of nodes.

3. The method of claim 2, wherein: Collect the environmental parameters of the location of the distributed photovoltaic, wherein the environmental parameters of the location of the distributed photovoltaic include solar irradiance, environmental temperature and environmental wind speed, wherein the specific expression for establishing the temperature-environmental characteristic formula is: T S = T air + 0.0138G s *(1+0.031T air )*(1-0.042V W ) where T S is the daily average temperature of the battery assembly, T air is the daily average temperature of the location of the distributed PV, G s is the daily average solar irradiance of the location of the distributed PV, V W is the daily average wind speed at the surface of the distributed PV; Based on the daily average temperature, calculate the environmental similarity of the to-be-attached distributed photovoltaic and each distributed photovoltaic cell assembly, wherein the formula for calculating the environmental similarity is: XS=ω1*|T1-T2|+ω2*|R1-R2| In the formula, XS is the environmental similarity, T1 is the daily average temperature of the cell assembly of the to-be-attached distributed photovoltaic, T2 is the daily average temperature of the cell assembly of the distributed photovoltaic sample, R1 is the total resistance value of the target power distribution network, R2 is the total resistance value of the power distribution network connected to the distributed photovoltaic sample, ω1 and ω2 are weight coefficients of the temperature difference and the total resistance value difference of the power distribution network respectively, wherein ω1<ω2, ω1+ω2=1 and ω1 and ω2 are greater than 0, and the distributed photovoltaic sample is the distributed photovoltaic connected to the power distribution network.

4. The method of claim 3, wherein: The similar samples with a similarity less than a similarity threshold value to the environment of the distribution network to which the distributed photovoltaic is to be connected are screened out, wherein the logic for screening the samples is as follows: When XS is greater than or equal to yz, it is indicated that the difference between the working environment of the distributed photovoltaic sample and the working environment of the distributed photovoltaic to be connected does not meet the requirements, and therefore the distributed photovoltaic sample is not used as a similar sample; When XS is less than yz, it is indicated that the difference between the working environment of the distributed photovoltaic sample and the working environment of the distributed photovoltaic to be connected meets the requirements, and therefore the distributed photovoltaic sample is used as a similar sample, wherein yz is the similarity threshold value.

5. The method of claim 4, wherein: The predicted value of the power characteristic parameter is compared with the historical power characteristic parameter, and the power loss of the distribution network under different photovoltaic capacities is analyzed, the historical power characteristic parameter specifically refers to the power characteristic parameter of the target distribution network before the distributed photovoltaic is connected, and the formula for calculating the power loss of the distribution network under different photovoltaic capacities is as follows: In the formula, P′ loss is the power loss of the target distribution network after accessing photovoltaic, K i is the switch operation of the ith branch line of the target distribution network after accessing photovoltaic, P mi represents the historical active power of the ith branch line, Q mi is the historical reactive power of the ith branch line, U mi is the historical terminal voltage value of the ith branch line, P pvi represents the active power prediction value of the ith branch line of the target distribution network after accessing photovoltaic, Q pvi is the reactive power prediction value of the ith branch line of the target distribution network after accessing photovoltaic, U Pi represents the terminal voltage prediction value of the ith branch line of the target distribution network after accessing photovoltaic, R i is the resistance value of the ith branch line, where i is the index of the branch line, i∈[1,n], where n is the total number of branch lines. K is the switch operating condition of the ith branch line i is a numerical code, taking 0 and 1 as two numerical values, when K i = 0, it represents not closed; when K i = 1, it represents closed.

6. The method of claim 1, wherein: The setting optimization target specifically is that the minimum total power loss, the minimum voltage deviation and the maximum photovoltaic consumption ratio of the distribution network are used as the optimization target, and the formula for calculating the voltage deviation is as follows: In the formula, F2 is a voltage offset, U i,ts is a rated voltage of the tth node in the ith branch line, U i,tN is a predicted actual voltage of the tth node in the ith branch line, t is an index of the node in the branch line, m i is a total number of nodes on the ith branch line; The formula for calculating the photovoltaic consumption ratio is as follows: In the formula, F3 is a photovoltaic consumption ratio, G pv is a distributed photovoltaic output to be connected, E C is a total power consumption of a target power distribution network, wherein the distributed photovoltaic output to be connected G pv The formula for calculation is: In the formula, G pvx is the actual injected power of the xth photovoltaic panel to be connected to the distributed photovoltaic system, A x and δ x respectively represent the area and the photoelectric conversion efficiency of the xth photovoltaic panel to be connected to the distributed photovoltaic system, h max is the maximum light intensity, x is the index of the photovoltaic panel, and X is the total number of photovoltaic panels to be connected to the distributed photovoltaic system.

7. The method of claim 6, wherein: According to the obtained power loss of the distribution network, the minimum total power loss, the minimum voltage deviation and the maximum photovoltaic consumption ratio of the distribution network are used as the optimization target, and the fitness function is used to represent the comprehensive optimization target, and the specific formula for calculating the fitness function is as follows: In the formula, FZ is the fitness function, ω3, ω4 and ω5 are weight coefficients of the total power loss, the voltage deviation and the photovoltaic consumption ratio respectively, ω3≥ω4>ω5 and ω3, ω4 and ω5 are greater than 0; Since the total power loss, the voltage deviation and the photovoltaic consumption ratio of the distribution network under different photovoltaic capacities are also different, in the range of the accessible photovoltaic capacity, the maximum fitness function is selected as the optimal photovoltaic engineering capacity.

8. The method of claim 5, wherein: According to the constraint conditions of the operation of the distribution network, the constraint conditions specifically refer to the constraint conditions of the standard power flow, the node voltage and the branch current, and the specific expression of the constraint conditions is as follows: where ΔP and ΔQ represent active power balance equation and reactive power balance equation, respectively, P i,Gt and Q i,Gt represent active and reactive power at the tth node in the ith branch line, V i,j is the maximum voltage at node t in the ith branch line, XC i,tj and B i,tj are conductance and susceptance between node t and j in the ith branch line, θ i,tj is the phase angle difference between node t and j in the ith branch line, U i,tmin and U i,tmax are the minimum and maximum voltage amplitude at the tth node in the ith branch line, I i is the current in the ith branch line, I imax is the current upper limit value in the ith branch line.

9. A distributed photovoltaic engineering capacity optimization configuration system, characterized in that: The distributed photovoltaic engineering capacity optimization configuration system is used to execute the distributed photovoltaic engineering capacity optimization configuration method of any one of claims 1-8, and comprises: A sample data acquisition module is configured to acquire the daily average temperature of the distributed photovoltaic cell assembly and the environmental parameters of the location of the distributed photovoltaic and the power characteristic parameters of the corresponding distribution network when the distributed photovoltaic connected to the distribution network is working, and to generate a sample data set according to the acquired data; An environmental influence characterization module is configured to determine the fitting relationship between the environmental parameters of the location of the distributed photovoltaic and the daily average temperature of the cell assembly by using the sample data set, and to determine the temperature-environment characterization between the environmental parameters and the cell assembly temperature by using a linear regression algorithm; A similar sample screening module is configured to acquire the environmental parameters of the location of the distributed photovoltaic to be connected, to acquire the daily average temperature of the distributed photovoltaic to be connected by using the temperature-environment characterization relationship, to calculate the environmental similarity of the distributed photovoltaic to be connected and each distributed photovoltaic cell assembly based on the daily average temperature, and to select the distributed photovoltaic with an environmental similarity less than a set similarity threshold value as a similar sample. The prediction model training module is configured to establish a power characteristic parameter prediction model, take the access position, capacity of the similar sample, total power consumption of the distribution network to which the similar sample is connected, and the topology structure of the distribution network as inputs, and take the power characteristic parameters of the distribution network as labels to train the model; The power loss representation module is configured to obtain historical power characteristic parameters of the target distribution network to which the distributed photovoltaic is to be connected when the target distribution network does not access the photovoltaic, predict the power characteristic parameter prediction values after the distributed photovoltaic of different capacities is accessed through the power characteristic parameter prediction model, compare the power characteristic parameter prediction values with the historical power characteristic parameters, and analyze the power loss of the corresponding distribution network; The capacity optimization configuration module is configured to set an optimization target according to the obtained power loss of the distribution network, take the capacity as an optimization object, determine the optimal capacity according to the constraint condition of the operation of the distribution network, and complete the capacity optimization configuration.

Images