A photovoltaic power station energy storage site selection and capacity planning method, system, device and medium

Abstract

This invention provides a method, system, equipment, and medium for photovoltaic (PV) power plant energy storage site selection and capacity planning, belonging to the field of PV energy storage optimization configuration in power systems. The method includes the following steps: acquiring annual historical data of PV radiation; clustering PV output scenarios based on daily average PV output, daily PV output volatility, and daily PV output distribution skewness to obtain multiple PV output scenarios and the proportion of each scenario; constructing a multi-objective PV-storage site selection and capacity planning model based on the economic, reliability, and environmental indicators of distributed PV and energy storage batteries in the distribution network; analyzing the weights of the economic, reliability, and environmental indicators, and solving the multi-objective PV-storage site selection and capacity planning model to obtain the site selection and capacity planning scheme for each PV output scenario. This invention can more accurately characterize the uncertainty of PV output.

Description

Technical Field

[0001] This invention belongs to the field of photovoltaic energy storage optimization configuration in power systems, and specifically relates to a method, system, equipment and medium for photovoltaic power plant energy storage site selection and capacity planning. Background Technology

[0002] High-penetration renewable energy grid integration will become a major characteristic and development trend of the power system. Given the current widespread use of photovoltaics, challenges may arise such as difficulties in peak-hour consumption and off-peak supply assurance. The bidirectional power transmission characteristics and flexible adjustment capabilities of energy storage technology can effectively address the mismatch between energy supply and demand in time and space. Therefore, research on photovoltaic-energy storage planning within a photovoltaic-integrated scenario is of great significance.

[0003] Due to the large amount of historical photovoltaic (PV) data, directly using it as the basis for PV planning can easily lead to problems such as high solution difficulty and long computation time. Aggregating a high proportion of distributed PV and energy storage into a small number of characteristic clusters based on typical peak-shaving characteristics significantly reduces the solution difficulty and the problem of exploding decision variables for participating in power system peak shaving. Furthermore, based on the principle of weighted allocation, the adjustable resource control priority of distributed PV, energy storage, and interruptible flexible loads is clustered to achieve power fluctuation smoothing based on resource aggregation. Therefore, it is feasible to simplify the computational complexity caused by historical PV data through PV scenario clustering. However, the above method mainly utilizes PV clustering for peak shaving and is rarely applied to the site selection and capacity planning of PV power plants and energy storage.

[0004] Regarding the site selection and capacity planning of photovoltaic power plant energy storage, some scholars have adopted a photovoltaic output clustering method based on the equal division of the four seasons, using a certain day in each season to represent the entire season. This method of clustering is uncertain, leading to inaccurate clustering results and a limited number of clustering scenarios, which in turn affects subsequent site selection and capacity planning. Summary of the Invention

[0005] To overcome the inaccuracy of clustering results due to uncertainty in photovoltaic output clustering, this invention provides a method for site selection and capacity planning of photovoltaic power plant energy storage, comprising the following steps: We obtain the annual historical data of photovoltaic radiation from photovoltaic power plants, and cluster the annual historical data of photovoltaic radiation from photovoltaic power plants into photovoltaic power output scenarios from three dimensions: daily average photovoltaic power output, daily photovoltaic power output volatility, and daily photovoltaic power output distribution skewness. This yields multiple photovoltaic power output scenarios and the proportion of each photovoltaic power output scenario. Based on the economic, reliability, and environmental indicators of distributed photovoltaic (PV) and energy storage batteries in the power distribution network, a multi-objective PV-storage site selection and capacity planning model is constructed to cluster PV scenarios. The weights of the economic, reliability, and environmental indicators in the multi-objective PV-storage site selection and capacity planning model are analyzed. Based on the weights and using the particle swarm optimization algorithm, the multi-objective PV-storage site selection and capacity planning model is solved to obtain the site selection and capacity planning schemes for each PV power output scenario. By combining the proportion of each photovoltaic scenario, the site selection and capacity determination schemes for each photovoltaic power output scenario are integrated to obtain the optimal site selection and capacity determination scheme for photovoltaic power station energy storage.

[0006] Preferably, the step of clustering photovoltaic output scenarios based on the annual historical data of photovoltaic radiation from three dimensions—daily average photovoltaic output, daily photovoltaic output volatility, and daily photovoltaic output distribution skewness—includes the following steps: The daily average photovoltaic power output, daily photovoltaic power output volatility, and daily photovoltaic power output distribution skewness are used as aggregation indicators and normalized. Using the elbow method, based on the relationship between the number of clusters and the sum of squared errors (SSE), observe the inflection point of the curve in the relationship graph, and define the number of clusters corresponding to the inflection point as the optimal number of clusters, that is, the optimal number of photovoltaic power output scenarios. Cluster centers were selected from the normalized values ​​of daily average photovoltaic power output, daily photovoltaic power output volatility, and daily photovoltaic power output distribution skewness using a probability distribution method. Calculate the Euclidean distance from each aggregation index to the cluster center, and update the cluster center; Repeat the above steps until the convergence condition is met, at which point the clustering process ends.

[0007] Preferably, a multi-objective photovoltaic-storage site selection and capacity planning model is constructed based on the economic, reliability, and environmental indicators of distributed photovoltaic and energy storage batteries in the power distribution network. This model includes the following steps: A photovoltaic-storage planning economic model is constructed based on economic indicators; a model characterizing the power quality stability level is constructed based on reliability indicators; and a carbon emission calculation model is constructed based on environmental indicators. Define the constraints of the multi-objective photovoltaic-storage site selection and capacity planning model; Based on the aforementioned economic model for photovoltaic and energy storage planning, power quality stability level model, carbon emission calculation model, and constraints, a multi-objective photovoltaic and energy storage site selection and capacity planning model is constructed.

[0008] The preferred multi-objective photovoltaic-storage site selection and capacity planning model is as follows: In the formula, , and The first The weights corresponding to the economic indicators, reliability indicators, and environmental indicators; , and The first The coefficients corresponding to the economic, reliability, and environmental indicators.

[0009] Preferably, the weights of the economic, reliability, and environmental indicators in the multi-objective photovoltaic-storage site selection and capacity planning model are obtained through the analytic hierarchy process (AHP), specifically including the following steps: Judgment matrices are constructed for economic indicators, reliability indicators, and environmental indicators, respectively; The judgment matrix is ​​subjected to a consistency test. The rationality of the judgment matrix is ​​determined by the consistency ratio threshold. If it is reasonable, the weight matrix of economic indicators, reliability indicators and environmental indicators is calculated by the judgment matrix.

[0010] Preferably, the economic indicators include construction investment cost, operation and maintenance cost, electricity purchase cost and energy storage battery revenue; the reliability indicators include voltage stability and load stability; and the environmental indicators include node emissions and branch emissions.

[0011] Preferably, the K-means++ algorithm is used to cluster photovoltaic power output scenarios based on three dimensions: daily average photovoltaic power output, daily photovoltaic power output volatility, and daily photovoltaic power output distribution skewness, using the annual historical data of photovoltaic radiation.

[0012] This invention also provides a photovoltaic power plant energy storage site selection and capacity planning system, comprising: The photovoltaic output scenario clustering module is used to obtain the annual historical data of photovoltaic radiation from photovoltaic power plants. It performs photovoltaic output scenario clustering on the annual historical data of photovoltaic radiation from three dimensions: daily average photovoltaic output, daily photovoltaic output volatility, and daily photovoltaic output distribution skewness, to obtain multiple photovoltaic output scenarios and the proportion of each photovoltaic output scenario. The model building module is used to construct a multi-objective photovoltaic-storage site selection and capacity planning model for photovoltaic scenarios based on the economic, reliability, and environmental indicators of distributed photovoltaic and energy storage batteries in the power distribution network; analyze the weights of the economic, reliability, and environmental indicators in the multi-objective photovoltaic-storage site selection and capacity planning model; and solve the multi-objective photovoltaic-storage site selection and capacity planning model according to the weights and using the particle swarm optimization algorithm to obtain the site selection and capacity planning scheme for each photovoltaic power output scenario. The site selection and capacity determination scheme acquisition module is used to integrate the site selection and capacity determination schemes of each photovoltaic scenario based on the proportion of each photovoltaic scenario, so as to obtain the optimal photovoltaic power station energy storage site selection and capacity determination scheme.

[0013] The present invention also provides a computer device, including a memory and a processor; the memory stores a computer program, and the processor is used to run the computer program in the memory to execute the photovoltaic power station energy storage site selection and capacity planning method.

[0014] The present invention also provides a computer-readable storage medium storing a computer program adapted for loading by a processor to execute the photovoltaic power plant energy storage site selection and capacity planning method.

[0015] The photovoltaic power plant energy storage site selection and capacity planning method, system, equipment, and medium provided by this invention have the following beneficial effects: This invention clusters photovoltaic (PV) output scenarios based on historical PV data across three dimensions: daily average PV output, daily PV output volatility, and daily PV output distribution skewness. This clustering method can generate multiple clustering scenarios; the more clustering scenarios generated, the more accurate the clustering results. Furthermore, this invention constructs a multi-objective PV-storage site selection and capacity planning model for PV scenario clustering based on the economic, reliability, and environmental indicators of distributed PV and energy storage batteries in the power distribution network. This model can more accurately characterize the uncertainty of PV output, reflect the actual PV output, and thus obtain a superior PV power plant energy storage site selection and capacity planning scheme. Attached Figure Description

[0016] To more clearly illustrate the embodiments and design schemes of the present invention, the accompanying drawings required for this embodiment will be briefly described below. The drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0017] Figure 1 This is a flowchart of the photovoltaic power plant energy storage site selection and capacity planning method according to an embodiment of the present invention; Figure 2 Here is a flowchart of the K-means++ algorithm; Figure 3 This is a photovoltaic-storage planning and evaluation model based on the analytic hierarchy process (AHP). Figure 4 For the IEEE-33 node distribution network; Figure 5 This is a typical daily load curve; Figure 6 The SSE descent curve; Figure 7 Clustering results for photovoltaic scenarios; Figure 8 The photovoltaic power output curve; Figure 9 The proportion of each scenario. Detailed Implementation

[0018] To enable those skilled in the art to better understand and implement the technical solutions of the present invention, the present invention will be described in detail below with reference to the accompanying drawings and specific embodiments. The following embodiments are only used to more clearly illustrate the technical solutions of the present invention and should not be construed as limiting the scope of protection of the present invention.

[0019] In the description of this invention, it should be understood that the terms "center," "longitudinal," "lateral," "length," "width," "thickness," "upper," "lower," "front," "rear," "left," "right," "vertical," "horizontal," "top," "bottom," "inner," "outer," "axial," "radial," and "circumferential" indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings. They are only for the convenience of describing the technical solution of this invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, they should not be construed as limitations on this invention.

[0020] Furthermore, the terms "first," "second," etc., are used for descriptive purposes only and should not be construed as indicating or implying relative importance. In the description of this invention, it should be noted that, unless otherwise explicitly specified or limited, the terms "connected" or "linked" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral connection; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium. Those skilled in the art can understand the specific meaning of the above terms in this invention according to the specific circumstances. In the description of this invention, unless otherwise stated, "a plurality of" means two or more, which will not be elaborated further here.

[0021] Example This invention provides a method for site selection and capacity planning of energy storage in photovoltaic power plants, specifically as follows: Figure 1 As shown, it includes the following steps: Step 1: Obtain the annual historical data of photovoltaic radiation. Use the K-means++ algorithm to cluster the annual historical data of photovoltaic radiation from three dimensions: daily average photovoltaic power output, daily photovoltaic power output volatility, and daily photovoltaic power output distribution skewness, to obtain the optimal number of photovoltaic power output scenarios and the proportion of each photovoltaic power output scenario.

[0022] Due to cloud cover during cloudy, rainy, or snowy weather, photovoltaic (PV) power output characteristics exhibit fluctuations and randomness. Scientifically describing the power output characteristics of PV power plants is beneficial for better understanding their fluctuation patterns and improving the accuracy of PV-storage site selection and capacity determination. This invention extracts three typical characteristics of PV power plant output characteristics: daily average PV power output, daily PV power output volatility, and daily PV power output distribution skewness, to form a PV power output aggregation index, specifically: Considering the characteristics of daily average photovoltaic power output, calculate the average photovoltaic power output over one cycle: In the formula, For the average daily photovoltaic output, for Solar power output at all times One cycle is 24 hours.

[0023] The daily photovoltaic output volatility rate is used to describe the level of fluctuation in photovoltaic output. The smaller this characteristic value, the smaller the fluctuation in photovoltaic output on that day, and the more stable the output.

[0024] In the formula, For the volatility of solar power output, For the average daily photovoltaic output, for Solar power output at all times One cycle is 24 hours.

[0025] The daily skewness of photovoltaic (PV) output distribution is used to describe the degree of daily skewness of PV output. When the skewness value is between [-0.5, 0.5], the PV output distribution is relatively symmetrical; when the skewness value is between [0.5, ..., ...], the PV output distribution is relatively symmetrical. When the skewness value is between [ ], the photovoltaic output distribution is positively skewed, with the photovoltaic output concentrated to the right of the mean; when ... , When the value is between 0.5 and 0.5, the photovoltaic power output distribution is negatively skewed, with the photovoltaic power output concentrated to the left of the mean.

[0026] In the formula, The skewness of solar power output distribution. For the volatility of solar power output, For the average daily photovoltaic output, For the first Photovoltaic output during the time period One cycle is 24 hours.

[0027] The 0-means algorithm exhibits excellent aggregation capabilities when dealing with massive amounts of photovoltaic data, effectively addressing the challenges posed by the large volume and high randomness of historical photovoltaic data. However, the traditional K-means algorithm randomly selects initial cluster centers, which can easily lead to local optima if poor selection is not made, and it also requires pre-specifying the number of clusters. Therefore, this invention employs the K-means++ algorithm to address the aggregation problem in photovoltaic scenarios, achieving better aggregation results. The algorithm flow is as follows: Figure 2 As shown.

[0028] 1) Indicator normalization Since the selected aggregate indicators have different meanings and magnitudes, they need to be normalized. According to the selected indicators, they can be divided into positive and negative indicators. The calculation of their normalized values ​​is shown in equation (4). The higher the value of the positive indicator or the lower the value of the negative indicator, the higher its normalized value. The daily average photovoltaic power output is a positive indicator, while the daily photovoltaic power output volatility and the daily photovoltaic power output distribution skewness are negative indicators.

[0029] In the formula, , These are the quantified positive and negative indicators, respectively. These are the original indicator values; , These are the maximum and minimum values ​​in the original index, respectively.

[0030] 2) Determine the number of clusters This invention introduces a metric for measuring clustering effectiveness: the sum of squared errors (SSE). A smaller SSE value indicates that data points are closer to cluster centers, resulting in better clustering. Then, using the elbow method, the inflection point of the curve is observed based on the relationship between the number of clusters and SSE; the number of clusters corresponding to this inflection point is the optimal number of clusters.

[0031] In the formula: This is the sum of squared errors for clustering. It is the sum of the number of clusters. For the divided first Clusters, for The index value after internal normalization The index value corresponding to the selected cluster center.

[0032] 3) Initialize cluster centers The K-means++ algorithm is an improvement on the K-means algorithm, mainly reflected in the original random selection. KBased on the initial cluster centers of 100 data points, cluster centers are selected using a probability distribution method to distribute the cluster center values ​​as evenly as possible. The specific process is as follows:

[0033] a. Construct a dataset of daily average photovoltaic power output, daily photovoltaic power output volatility, and daily photovoltaic power output distribution skewness, and randomly select a data point from the dataset as the first cluster center.

[0034] b. For each data point in the dataset, calculate its shortest distance to the cluster center.

[0035] c. For each data point, normalize its shortest distance to the selected cluster center to obtain the probability distribution of the distance.

[0036] In the formula, For the first The probability that a sample is selected as the next cluster center. For the first The distance between each sample and the cluster center For the first j The distance between each sample and the cluster center.

[0037] d. Based on the calculated distance probability distribution, select the next cluster center using a weighted probability method. Data points that are farther away have a higher weight in the probability distribution, thus ensuring that the initial cluster centers are more dispersed.

[0038] e. Repeat steps c and d until a result is selected. K Cluster centers.

[0039] 4) Update cluster centers Calculate the Euclidean distance from each aggregation index to the cluster center, i.e.: In the formula, For sample points and cluster center The Euclidean distance between them For sample points No. Feature values ​​in each dimension Cluster center No. Feature values ​​in each dimension The dimension of each sample point in the dataset is the number of aggregated metrics selected.

[0040] Each sample point is assigned to the cluster containing the cluster center with the smallest Euclidean distance. After clustering, the cluster centers within each cluster are updated, i.e.: In the formula, For the updated cluster centers, The number of sample points within the cluster; For clusters Sample points within.

[0041] Repeat the above steps until the cluster center positions no longer change, at which point the clustering process ends.

[0042] Step 2: Construct a multi-objective photovoltaic-storage site selection and capacity planning model for photovoltaic scenarios based on the economic, reliability, and environmental indicators of distributed photovoltaic and energy storage batteries in the power distribution network; analyze the weights of the economic, reliability, and environmental indicators in the multi-objective photovoltaic-storage site selection and capacity planning model, and solve the multi-objective photovoltaic-storage site selection and capacity planning model according to the weights and using the particle swarm optimization algorithm to obtain the site selection and capacity planning scheme for each photovoltaic power output scenario.

[0043] The outer-layer photovoltaic (PV) scenario aggregation obtains hourly solar irradiance data from the NASA POWER website, then uses the PV power generation model to obtain historical PV power generation data for the whole year. Based on the PV output aggregation index, the K-means++ algorithm is used to obtain the optimal number of PV scenarios. With the current aggregation scenario determined, the inner-layer planning model compares the importance of each pair of indicators based on parameters such as voltage stability using a nine-level scaling method, and uses the Analytic Hierarchy Process (AHP) to determine the weights of each indicator. The objective function of the Particle Swarm Optimization (PSO) algorithm is determined based on the weights. Finally, the PSO algorithm is used to solve the planning model and determine the optimal site selection and capacity allocation scheme.

[0044] 1) Economic indicators The economic viability of distributed photovoltaic (PV) and energy storage batteries in power distribution networks primarily considers construction investment costs, operation and maintenance costs, PV curtailment costs, and energy storage battery revenue. Therefore, economic indicators reflecting the PV-energy storage planning model are crucial. The formula is expressed as:

[0045] In the formula, The construction investment cost for distributed photovoltaic and energy storage, For the operation and maintenance costs of distributed photovoltaic and energy storage, For electricity purchase costs, For the revenue of energy storage batteries, For the first The weights corresponding to each economic indicator will be obtained by solving the analytic hierarchy process.

[0046] Construction investment costs: In the formula, , These are the construction investment costs for distributed photovoltaic power and energy storage, respectively. , These are the discount rates for distributed photovoltaic power and energy storage, respectively. , These refer to the service life of distributed photovoltaic power and energy storage, respectively. , These are the unit capacity construction investment costs for distributed photovoltaic and energy storage, respectively. Planned construction capacity for distributed photovoltaic power; For grid-connected energy storage capacity.

[0047] Operation and maintenance costs: In the formula, The construction investment cost for distributed photovoltaic and energy storage, The percentage used to calculate operation and maintenance costs is typically 10%, based on engineering experience. Cost of curtailment of distributed photovoltaic power; Cost per unit of curtailment for distributed photovoltaic power; The replacement rate of energy storage (considering the replacement of aging energy storage batteries); , Distributed photovoltaic in The predicted power generation and the actual power generation at any given time.

[0048] Electricity purchase cost: In the formula, The time-of-use electricity price per unit capacity purchased from the main grid; for The main online shopping volume at any given time.

[0049] Benefits of energy storage batteries: In the formula: , These represent the load before and after the energy storage charging and discharging process.

[0050] 2) Environmental protection indicators Considering the carbon emissions generated during the operation of the power distribution network The calculation formula is as follows: In the formula, , These are the carbon emissions at the node and the carbon emissions at the branch road, respectively. Carbon emission conversion factor; for t Time Node Electricity consumption; for t Power consumption at any given moment; For the first The weights corresponding to each environmental protection indicator.

[0051] 3) Reliability Indicators Large-scale grid-connected distributed photovoltaic systems (PV) exhibit inherent volatility and randomness, leading to fluctuations in node loads and consequently exacerbating node voltage fluctuations, thus reducing the system's power quality. Therefore, node voltage stability and load stability are used to characterize the system's power quality stability level. .

[0052] In the formula, , These are voltage stability and load stability, respectively. For nodes exist The voltage value at that moment; For nodes The average voltage over the entire time period; for The load power value at any given time; This represents the average load frequency over the entire time period. For the first The weights corresponding to each reliability metric.

[0053] The specific constraints of the multi-objective photovoltaic-storage site selection and capacity planning model are as follows: 1) Current constraints In the formula: , They are nodes Active power and reactive power; , They are nodes , The voltage; , They are nodes , The susceptance and conductivity between them; For nodes , The voltage phase angle between them.

[0054] 2) Voltage constraint In the formula, , They are nodes The upper and lower limits of voltage.

[0055] 3) Distributed photovoltaic installation capacity constraints In the formula, This represents the upper limit of the installed capacity of distributed photovoltaic systems.

[0056] 4) Constraints of energy storage batteries In the formula, To improve the charging and discharging efficiency of energy storage batteries; , For energy storage batteries , Battery capacity at any given time; For energy storage batteries The charging and discharging power at any given moment; positive values ​​represent discharging power, and negative values ​​represent charging power. , These are the lower and upper limits of the charging and discharging power of energy storage batteries, respectively. The capacity installed for energy storage batteries; This represents the upper limit of the installed capacity of energy storage batteries; , These are the lower and upper limits of the state of charge of the energy storage battery, respectively. , These represent the state of charge (SOC) of the energy storage battery at the beginning and end of the time period, respectively.

[0057] The Analytic Hierarchy Process (AHP) is a decision analysis method that combines qualitative and quantitative approaches to handle multiple objectives. This method introduces quantitative analysis into multi-objective decision-making and gives corresponding merit scores based on the results of pairwise comparisons between multiple objectives by the decision-maker. Then, it performs hierarchical analysis and uses the judgment results to comprehensively calculate the weight of each indicator.

[0058] The steps for determining the multi-objective weights of photovoltaic energy storage based on the analytic hierarchy process are as follows: 1) Establishing an evaluation model. To assess the rationality of photovoltaic-storage planning, this invention analyzes the interactions between various indicators during the planning process and establishes evaluation indicators from three perspectives: reliability, economy, and environmental friendliness. Furthermore, specific evaluation indicators are subdivided under these three main evaluation indicators. The photovoltaic-storage planning evaluation model is as follows: Figure 3 As shown.

[0059] 2) Constructing the Judgment Matrix. To address the subjectivity of qualitative analysis in the evaluation process, expert experience and a nine-level scaling method are used to quantify and correct the relative importance of each pair of elements, thereby effectively improving the reliability of the evaluation process. Then, by comparing each indicator pairwise, the corresponding scale is obtained, resulting in the judgment matrix. :

[0060] In the formula, As an indicator For indicators The scale values ​​between the indicators represent their relative importance. The specific values ​​are based on the nine-level scale method, as shown in Table 1.

[0061] Table 1 Nine-level scale method according to Figure 3 Based on the indicator system shown, establish judgment matrices for each level of indicators. First, establish the judgment matrix for the first-level indicators. :

[0062] Establish corresponding judgment matrices for the secondary indicators under the economic indicators. for: Establish corresponding judgment matrices for the secondary indicators under the reliability index. for: Establish corresponding judgment matrices for the secondary indicators under the environmental protection indicators. for: 3) Consistency Check. The obtained judgment matrix also needs to undergo a consistency check to determine its scientific validity and rationality. The method for performing a consistency check on the judgment matrix is ​​as follows:

[0063] In the formula, Let the consistency ratio be the ratio when the consistency ratio is... If the condition is met, the judgment matrix passes the consistency check; otherwise, a reasonable judgment matrix needs to be reconstructed. As a consistency indicator; This is a random consistency index, and its value is related to the order of the judgment matrix. related; To determine the matrix The largest eigenvalue; 4) Determine the weight of each indicator. Calculate the weight matrix of the indicators based on the judgment matrices of each level of indicator.

[0064] There are three calculation methods: arithmetic mean, geometric mean, and eigenvalue method. This invention uses the arithmetic mean method. The specific calculation formula is as follows:

[0065] The overall weight of a secondary indicator is the product of its weight and the weight of its parent primary indicator. The weights of each level of indicator are shown in Table 2.

[0066] Table 2 Weighting Table of Indicators at Each Level The final multi-objective photovoltaic-storage site selection and capacity planning model is as follows: In the formula, , and The first The weights corresponding to the economic indicators, reliability indicators, and environmental indicators; , and The first The coefficients corresponding to the economic, reliability, and environmental indicators.

[0067] This invention takes the classic IEEE 33-node distribution network system as an example, and its topology is as follows: Figure 4 As shown.

[0068] The relevant parameters for photovoltaic energy storage are shown in Table 3, and the time-of-use electricity price is shown in Table 4.

[0069] Table 3 Photovoltaic Energy Storage Parameter Table The total system load is 3715kW + j2300kvar, and a typical daily load curve is shown below. Figure 5 As shown, the reference voltage is 12.66kV, and the allowable range of node voltage is 0.9~1.1pu. Due to geographical location and other limitations, the nodes that can be selected for photovoltaic power stations and energy storage are 4, 7, 8, 14, 18, 19, 20, 21, 22, 23, 24, 25, 29, 30, 31, and 32.

[0070] Table 4 Time-of-use electricity prices Using annual photovoltaic (PV) data from a specific region as the example data for this invention, with a PV data sampling interval of 1 hour, PV scenario clustering was performed using three different indicators. An improved K-means++ algorithm was then used to establish a PV aggregation model. The SSE (Solution Scale Index) decline curve is shown below. Figure 6 As shown, the elbow method is used to determine the optimal cluster centers.K Value, by Figure 6 It can be seen that when the number of aggregation clusters is selected as 4, an inflection point appears in the process of SSE change, and the rate of decrease slows down. At this point, it can be regarded as the optimal aggregation effect.

[0071] Clustering results as follows Figure 7 As shown, based on the daily average photovoltaic (PV) output, daily PV output volatility, and daily PV output distribution skewness as aggregation features, the improved K-means++ algorithm aggregates the annual PV data into four categories. Yellow and blue colors represent different aggregation clusters, and red circles represent the aggregation centers corresponding to different clusters. The PV output corresponding to each aggregation center is shown in the figure. Figure 8 As shown, the proportion of each scene is as follows: Figure 9 As shown.

[0072] Figure 9 This represents the weight of each scenario after clustering the annual photovoltaic data. The weight of the cluster to which the data of each day belongs in each scenario is different, reflecting the uncertainty of the annual photovoltaic data. This method can describe the annual photovoltaic data more accurately than the traditional four-season equal division method.

[0073] Three schemes were set up for comparative analysis: Scheme 1, which does not consider photovoltaic scene clustering and adopts the traditional four-season division method, has uncertain results and lacks rationality; Scheme 2, which does not consider AHP index weights and only considers the economic cost of particle swarm algorithm to solve the problem; Scheme 3, a photovoltaic scene clustering-based photovoltaic-storage planning model.

[0074] Based on the photovoltaic output and typical daily load corresponding to the aggregation center, and combined with the content of Section 2, the multi-objective weights are obtained. The planning model is solved using the PSO algorithm, and the planning schemes under each mode are shown in Table 5. The planning results are shown in Table 6.

[0075] Table 5 Planning Scheme Table 6 Planning Results As shown in Table 6, compared with Scheme 3, Scheme 1 reduces the operation and maintenance costs by 86.48% and improves the revenue of energy storage batteries, proving the economic value of photovoltaic-storage synergy in the system. Conversely, the photovoltaic output obtained by the traditional seasonal equal division method deviates from the actual photovoltaic output, and the timing of photovoltaic output is mismatched, resulting in poor photovoltaic-storage synergy. The photovoltaic scene clustering method proposed in this invention better reflects the actual photovoltaic output and effectively suppresses the power and voltage fluctuations caused by photovoltaic power generation, thereby improving the system's reliability indicators. Scheme 2, considering only economic costs, significantly reduces the planned capacity of the photovoltaic power station, and the resulting economic costs are also reduced accordingly. However, the total economic cost is only reduced by 0.57% compared to Scheme 3, while reliability and environmental protection indicators are improved. In modern distribution networks, reducing the economic cost of the system by sacrificing system reliability and environmental protection is obviously unreasonable. Therefore, it is proved that the photovoltaic-storage planning model based on photovoltaic scene clustering can effectively improve the economy, reliability, and environmental protection of the power grid.

[0076] This invention uses cluster analysis of historical photovoltaic (PV) output data, employing the K-means++ algorithm to cluster characteristic indicators, thereby obtaining typical PV output scenarios. A multi-objective model for PV-storage site selection and capacity determination is established, and the analytic hierarchy process (AHP) is used to determine the weights of each objective, thus more accurately quantifying the relative importance of each indicator. The Proof-of-Stake (PSO) algorithm is then used to solve the planning scheme. The planning model of this invention simplifies the calculation of historical PV data and comprehensively considers the economy, reliability, and environmental friendliness of the distribution network. The main conclusions are as follows:

[0077] 1) Aggregating photovoltaic power generation output characteristic scenarios can effectively simplify the calculation of historical photovoltaic data. At the same time, clustering photovoltaic scenarios with multiple indicators can better reflect the uncertainty of photovoltaic output throughout the year, and the clustering effect is more reliable.

[0078] 2) The multi-objective photovoltaic-storage planning model based on AHP can effectively reduce the economic cost of the distribution network, while taking into account power supply reliability and environmental protection, which is more in line with the development trend of modern distribution network planning.

[0079] This invention currently only analyzes photovoltaic power output and does not yet consider the site selection and capacity planning of other distributed power sources. The next step is to study the integration of different types of distributed power sources and energy storage into the joint planning of the distribution network.

[0080] Step 3: Based on the proportion of each photovoltaic scenario, integrate the site selection and capacity determination schemes for each photovoltaic power output scenario to obtain the optimal photovoltaic power station energy storage site selection and capacity determination scheme.

[0081] This invention also provides a photovoltaic power plant energy storage site selection and capacity planning system, comprising: The photovoltaic output scenario clustering module is used to obtain the annual historical data of photovoltaic radiation from photovoltaic power plants. It performs photovoltaic output scenario clustering on the annual historical data of photovoltaic radiation from three dimensions: daily average photovoltaic output, daily photovoltaic output volatility, and daily photovoltaic output distribution skewness, to obtain multiple photovoltaic output scenarios and the proportion of each photovoltaic output scenario. The model building module is used to construct a multi-objective photovoltaic-storage site selection and capacity planning model for photovoltaic scenarios based on the economic, reliability, and environmental indicators of distributed photovoltaic and energy storage batteries in the power distribution network; analyze the weights of the economic, reliability, and environmental indicators in the multi-objective photovoltaic-storage site selection and capacity planning model; and solve the multi-objective photovoltaic-storage site selection and capacity planning model according to the weights and using the particle swarm optimization algorithm to obtain the site selection and capacity planning scheme for each photovoltaic power output scenario. The site selection and capacity determination scheme acquisition module is used to integrate the site selection and capacity determination schemes of each photovoltaic scenario based on the proportion of each photovoltaic scenario, so as to obtain the optimal photovoltaic power station energy storage site selection and capacity determination scheme.

[0082] The present invention also provides a computer device, including a memory and a processor; the memory stores a computer program, and the processor is used to run the computer program in the memory to execute the photovoltaic power station energy storage site selection and capacity planning method.

[0083] The present invention also provides a computer-readable storage medium storing a computer program adapted for loading by a processor to execute the photovoltaic power plant energy storage site selection and capacity planning method.

[0084] The above-described embodiments are merely preferred embodiments of the present invention, and the scope of protection of the present invention is not limited thereto. Any simple changes or equivalent substitutions of the technical solutions that can be obviously obtained by those skilled in the art within the scope of the technology disclosed in the present invention shall fall within the scope of protection of the present invention.

Claims

1. A method for site selection and capacity planning of energy storage in photovoltaic power plants, characterized in that, Includes the following steps: Historical annual data of photovoltaic (PV) radiation from PV power plants was obtained. The K-means++ algorithm was used to cluster PV power output scenarios based on three dimensions: daily average PV output, daily PV output volatility, and daily PV output distribution skewness. This determined the optimal number of PV power output scenarios and the proportion of each scenario. Specifically, the PV power output scenario clustering included: Step 1: Using the daily average PV output, daily PV output volatility, and daily PV output distribution skewness as aggregation indicators and normalizing them. The daily average PV output was used as the positive indicator, and the daily PV output volatility and daily PV output distribution skewness were used as the inverse indicators. Step 2: Introducing SSE (Sum of Squared Errors) as a metric for clustering effectiveness. A smaller SSE value indicates that the data points are closer to the cluster centers, indicating better clustering. Using the elbow method, the inflection point of the curve in the relationship graph between the number of clusters and the sum of squared errors (SSE) was observed, and the inflection point was determined. The corresponding number of clusters is defined as the optimal number of clusters, i.e., the optimal number of photovoltaic power output scenarios; Step 3: Construct the dataset of daily average photovoltaic power output, daily photovoltaic power output volatility, and daily photovoltaic power output distribution skewness. Randomly select a data point from the dataset as the first cluster center. For each data point, normalize its Euclidean distance with the first cluster center to obtain the probability distribution of the distance; Based on the probability distribution of the distance, select the next cluster center in a weighted probability manner. When the number of cluster centers reaches the optimal number of clusters, the cluster center update is completed. Among them, the data points with greater distances have higher weights in the probability distribution of distances; Step 4: Calculate the Euclidean distance from the daily average photovoltaic power output, daily photovoltaic power output volatility, and daily photovoltaic power output distribution skewness to the cluster centers. Assign each data point to the cluster containing the cluster center with the smallest Euclidean distance, and update the aggregation center within the cluster; Step 5: Repeat Step 4 until the position of the cluster center no longer changes, and the clustering ends; Based on the economic, reliability, and environmental indicators of distributed photovoltaic (PV) and energy storage batteries in the power distribution network, a multi-objective PV-storage site selection and capacity planning model is constructed to cluster PV scenarios. The weights of the economic, reliability, and environmental indicators in the multi-objective PV-storage site selection and capacity planning model are analyzed. Based on the weights and using the particle swarm optimization algorithm, the multi-objective PV-storage site selection and capacity planning model is solved to obtain the site selection and capacity planning schemes for each PV power output scenario. By combining the proportion of each photovoltaic scenario, the site selection and capacity determination schemes for each photovoltaic power output scenario are integrated to obtain the optimal site selection and capacity determination scheme for photovoltaic power station energy storage.

2. The photovoltaic power plant energy storage site selection and capacity planning method according to claim 1, characterized in that, Based on the economic, reliability, and environmental performance indicators of distributed photovoltaic (PV) and energy storage batteries in the power distribution network, a multi-objective PV-storage site selection and capacity planning model is constructed, which includes the following steps: A photovoltaic-storage planning economic model is constructed based on economic indicators; a model characterizing the power quality stability level is constructed based on reliability indicators; and a carbon emission calculation model is constructed based on environmental indicators. Define the constraints of the multi-objective photovoltaic-storage site selection and capacity planning model; Based on the aforementioned economic model for photovoltaic and energy storage planning, power quality stability level model, carbon emission calculation model, and constraints, a multi-objective photovoltaic and energy storage site selection and capacity planning model is constructed.

3. The photovoltaic power plant energy storage site selection and capacity planning method according to claim 2, characterized in that, The multi-objective photovoltaic-storage site selection and capacity planning model is as follows: In the formula, , and The first The weights corresponding to the economic indicators, reliability indicators, and environmental indicators; , and The first The coefficients corresponding to the economic, reliability, and environmental indicators.

4. The photovoltaic power plant energy storage site selection and capacity planning method according to claim 1, characterized in that, The weights of the economic, reliability, and environmental indicators in the multi-objective photovoltaic-storage site selection and capacity planning model are obtained through the analytic hierarchy process (AHP), specifically including the following steps: Judgment matrices are constructed for economic indicators, reliability indicators, and environmental indicators, respectively; The judgment matrix is ​​subjected to a consistency test. The rationality of the judgment matrix is ​​determined by the consistency ratio threshold. If it is reasonable, the weight matrix of economic indicators, reliability indicators and environmental indicators is calculated by the judgment matrix.

5. The photovoltaic power station energy storage site selection and capacity planning method according to claim 1, characterized in that, The economic indicators include construction investment cost, operation and maintenance cost, electricity purchase cost and energy storage battery revenue; the reliability indicators include voltage stability and load stability; and the environmental indicators include node emissions and branch emissions.

6. A photovoltaic power station energy storage site selection and capacity planning system, characterized in that, include: The photovoltaic (PV) output scenario clustering module is used to acquire the annual historical data of PV power plant radiation. The K-means++ algorithm is used to cluster the annual historical PV power output data from three dimensions: daily average PV output, daily PV output volatility, and daily PV output distribution skewness. This yields the optimal number of PV output scenarios and the proportion of each scenario. Specifically, the PV output scenario clustering includes step 1: using the daily average PV output, daily PV output volatility, and daily PV output distribution skewness as aggregation indicators and normalizing them. The daily average PV output is used as the positive indicator in the aggregation indicators, and the daily average PV output is used as the positive indicator. The power output volatility and daily photovoltaic power output distribution skewness are used as inverse indicators in the aggregation index; Step 2: Introduce SSE as an indicator to measure the clustering effect, where the smaller the SSE value, the closer the data point is to the cluster center, and the better the clustering effect; Using the elbow method, based on the relationship graph between the number of clusters and the sum of squared errors (SSE), observe the inflection point of the curve in the relationship graph, and define the number of clusters corresponding to the inflection point as the optimal number of clusters, i.e., the optimal number of photovoltaic power output scenarios; Step 3: Construct the dataset of daily average photovoltaic power output, daily photovoltaic power output volatility, and daily photovoltaic power output distribution skewness, and randomly select a data point from the dataset as the first cluster center. For each data point, normalize its Euclidean distance to the first cluster center to obtain a probability distribution of the distance. Based on the probability distribution of the distance, select the next cluster center in a weighted probability manner. When the number of cluster centers reaches the optimal number of clusters, the cluster center update is completed. Among them, the data points with greater distances have higher weights in the probability distribution of the distance. Step 4: Calculate the Euclidean distance from the daily average photovoltaic power output, daily photovoltaic power output fluctuation rate, and daily photovoltaic power output distribution skewness to the cluster centers. Assign each data point to the cluster containing the cluster center with the smallest Euclidean distance, and update the cluster centers within the cluster. Step 5: Repeat step 4 until... The clustering ends when the cluster center location no longer changes. A multi-objective photovoltaic (PV) and energy storage (ESS) site selection and capacity planning model is constructed based on the economic, reliability, and environmental indicators of distributed PV and energy storage batteries in the power distribution network. The weights of the economic, reliability, and environmental indicators in the multi-objective PV site selection and capacity planning model are analyzed. The model is then solved using the particle swarm optimization algorithm based on these weights to obtain the site selection and capacity planning schemes for each PV power output scenario. Finally, the site selection and capacity planning schemes for each PV power output scenario are integrated based on the proportion of each PV scenario to obtain the optimal PV power plant energy storage site selection and capacity planning scheme. The model building module is used to construct a multi-objective photovoltaic-storage site selection and capacity planning model for photovoltaic scenarios based on the economic, reliability, and environmental indicators of distributed photovoltaic and energy storage batteries in the power distribution network; analyze the weights of the economic, reliability, and environmental indicators in the multi-objective photovoltaic-storage site selection and capacity planning model; and solve the multi-objective photovoltaic-storage site selection and capacity planning model according to the weights and using the particle swarm optimization algorithm to obtain the site selection and capacity planning scheme for each photovoltaic power output scenario. The site selection and capacity determination scheme acquisition module is used to integrate the site selection and capacity determination schemes of each photovoltaic scenario based on the proportion of each photovoltaic scenario, so as to obtain the optimal photovoltaic power station energy storage site selection and capacity determination scheme.

7. A computer device, characterized in that, It includes a memory and a processor; the memory stores a computer program, and the processor is used to run the computer program in the memory to execute the photovoltaic power plant energy storage site selection and capacity planning method according to any one of claims 1-5.

8. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program adapted for loading by a processor to execute the photovoltaic power plant energy storage site selection and capacity planning method according to any one of claims 1-5.

Images