Distributed power generation scene and cost planning method, system, equipment and medium

Abstract

The invention discloses a distributed power generation scene and cost planning method and system, equipment and a medium, and the method comprises the steps: obtaining historical wind power basic data per hour, and the wind power basic data comprises the power generation amount of a wind turbine generator, the photovoltaic power generation amount and the load demand; generating a representative scene matrix based on the wind power basic data; the scene matrix meets the expectation, the standard deviation, the skewness, the kurtosis and the correlation of the historical scene, and the representative scene of the generating capacity of the wind turbine generator, the photovoltaic generating capacity and the load demand is obtained; integrating the scene matrix into a power flow equation, and establishing a random DGIP optimization model; and constructing an objective function and a constraint condition, solving the random DGIP optimization model, and outputting an optimal DGIP solution. According to the invention, based on the uncertainty of wind power generation and photovoltaic power generation and the random model of load demand and electricity price fluctuation, various possible future scenes can be dynamically simulated, and the adaptability of the decision process to the future uncertainty is improved.

Description

Technical Field

[0001] This invention relates to the field of photovoltaic power generation technology, specifically to a distributed power generation scenario and cost planning method, system, equipment, and medium. Background Technology

[0002] Integrating distributed generation (DG) into the distribution system can significantly improve the system's operational performance, reduce power losses and voltage deviations, and ensure a reliable power supply. However, intermittent renewable energy generation also introduces various operational uncertainties. Failure to adequately consider these uncertainties in distributed generation planning (DGP) can lead to inappropriate DGP allocation, which not only exacerbates power losses and voltage deviations but also reduces power quality and results in inadequate coordination of protection devices.

[0003] Therefore, existing distributed generation planning methods do not take into account the various operational uncertainties brought about by these intermittent renewable energy generation, which may lead to improper allocation of distributed generation and failure to achieve optimal resource allocation. Summary of the Invention

[0004] The present invention aims to provide a distributed generation scenario and cost planning method, system, equipment, and medium. By pre-setting a scenario matrix and ensuring it meets the statistical characteristics of historical scenarios (such as expectation, standard deviation, skewness, kurtosis, and correlation), this scheme can capture historical trends and extreme situations in wind power, photovoltaic power generation, and load demand, making the generated scenarios highly representative. The scenario matrix can more comprehensively reflect the complexity and uncertainty of the system. Based on the uncertainties of wind power and photovoltaic power generation, as well as stochastic models of load demand and electricity price fluctuations, it can dynamically simulate multiple possible future scenarios, improving the decision-making process's adaptability to future uncertainties. Employing cost-benefit analysis, this model can quantify the economic benefits under different strategies, helping decision-makers identify the most cost-effective operation or investment plan and achieve optimal resource allocation.

[0005] This invention is achieved through the following technical solution: In a first aspect, the present invention provides a distributed generation scenario and cost planning method, the method comprising: Obtain historical hourly wind power basic data, which includes wind turbine power generation, photovoltaic power generation, and load demand; Based on wind power fundamental data, a representative scenario matrix is ​​generated. The scenario matrix satisfies the expectation, standard deviation, skewness, kurtosis and correlation of historical scenarios, and derives a representative scenario for wind turbine power generation, photovoltaic power generation and load demand. The scenario matrix is ​​integrated into the power flow equation to establish a stochastic DGIP optimization model; the stochastic DGIP optimization model is formed by integrating the scenario matrix with deterministic decision variables. Construct the objective function and constraints for the stochastic DGIP optimization model, solve the stochastic DGIP optimization model, and output the optimal DGIP solution.

[0006] Furthermore, based on wind power baseline data, a scenario matrix is ​​generated, including: Based on wind power fundamental data, the target moments and target correlation matrices for historical scenarios are calculated; the target moments and target correlation matrices for historical scenarios refer to the target moments and target correlation matrices for historical hourly wind turbine power generation, photovoltaic power generation, and load demand. The objective moments are normalized to obtain the normalized objective moments; Based on the normalized target moments, the first matrix is ​​established by randomly generating scenarios with uncertainties from a normal distribution. The first matrix is ​​transformed to obtain the second matrix; the second matrix satisfies the target relevance matrix of the historical scenario. The second matrix is ​​cubically transformed to obtain the third matrix; the third matrix is ​​the normalized target moment that satisfies the historical scenario. The third matrix is ​​then inverted to satisfy the target time, resulting in a representative scenario matrix.

[0007] Furthermore, based on wind power baseline data, the generation of a scenario matrix also includes: Perform error standard tests on the third and second matrices respectively; If the error standard verification meets the corresponding preset error, then continue to invert the third matrix; If at least one of the error standard checks fails to meet the corresponding preset error, then return to normalize the target moment.

[0008] Furthermore, the formula for the cubic transformation is: , In the formula, From column vectors that follow a normal distribution Obtain a univariate normal random column vector with four given moments; The transformation coefficients are those assumed in the target scenario. The moment equals the historical scenario The calculation is performed under the condition of the target moment; Corresponding to the first four moments: expectation, standard deviation, skewness, and kurtosis; These refer to wind turbine power generation, photovoltaic power generation, and load demand, respectively.

[0009] Furthermore, an objective function is constructed for the stochastic DGIP optimization model, including: Maximizing the net present value (NPV) of DNO over the planning period is the objective function of the stochastic DGIP optimization model; the NPV is calculated using cost-benefit analysis based on the difference between the present value of revenue and the present value of costs.

[0010] Furthermore, the constraints include: First constraint: representing the total active power supply and demand balance of each bus under all representative scenarios; Second constraint: representing the node power balance of total reactive power supply and demand under all representative scenarios; Third constraint: defining the bus voltage and branch current under all cases based on active power flow and reactive power flow; Fourth constraint: representing the lower and upper limits of bus voltage and branch current; Fifth constraint: limiting the total installed capacity of wind turbine units within the preset range of peak load demand; Sixth constraint: representing the upper and lower limits of the installed capacity of wind turbine units and photovoltaic units on each bus.

[0011] Secondly, the present invention provides a distributed generation scenario and cost planning system, the system comprising: The acquisition unit is used to acquire historical hourly wind power basic data, which includes wind turbine power generation, photovoltaic power generation, and load demand. The scenario matrix generation unit is used to generate a representative scenario matrix based on wind power basic data. The scenario matrix satisfies the expectation, standard deviation, skewness, kurtosis and correlation of historical scenarios to obtain representative scenarios of wind turbine power generation, photovoltaic power generation and load demand. The optimization model building unit is used to integrate the scenario matrix into the power flow equation to establish a stochastic DGIP optimization model; the stochastic DGIP optimization model is formed by integrating the scenario matrix with deterministic decision variables. The model solving unit is used to construct the objective function and constraints for the stochastic DGIP optimization model, solve the stochastic DGIP optimization model, and output the optimal DGIP solution.

[0012] Furthermore, the scenario matrix generation unit includes: The calculation subunit is used to calculate the target moments and target correlation matrices for historical scenarios based on wind power basic data; the target moments and target correlation matrices for historical scenarios refer to the target moments and target correlation matrices for historical hourly wind turbine power generation, photovoltaic power generation, and load demand; The normalization sub-unit is used to normalize the target moments to obtain the normalized target moments; The matrix establishes sub-units, which are used to establish the first matrix based on the normalized target moments and the scenario of randomly generating uncertain factors by sampling from the normal distribution; The matrix transformation subunit is used to transform the first matrix to obtain the second matrix; the second matrix satisfies the target relevance matrix of the historical scenario. The cubic transformation sub-unit is used to perform a cubic transformation on the second matrix to obtain the third matrix; the third matrix is ​​the normalized target moment that satisfies the historical scenario. The inversion subunit is used to invert the third matrix to meet the target time and obtain a representative scenario matrix.

[0013] Furthermore, the objective function and constraints for the stochastic DGIP optimization model are constructed, including: Maximizing the net present value (NPV) of DNO over the planning period is the objective function of the stochastic DGIP optimization model; the NPV is calculated using cost-benefit analysis based on the difference between the present value of revenue and the present value of costs. The constraints include: First constraint: representing the total active power supply and demand balance of each bus under all representative scenarios; Second constraint: representing the node power balance of total reactive power supply and demand under all representative scenarios; Third constraint: defining the bus voltage and branch current under all conditions based on active power flow and reactive power flow; Fourth constraint: representing the lower and upper limits of bus voltage and branch current; Fifth constraint: limiting the total installed capacity of wind turbine units within the preset range of peak load demand; Sixth constraint: representing the upper and lower limits of the installed capacity of wind turbine units and photovoltaic units on each bus.

[0014] Thirdly, the present invention also provides a computer device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the steps of the above-described distributed generation scenario and cost planning method.

[0015] Fourthly, the present invention also provides a computer-readable storage medium having a computer program stored thereon, wherein the computer program, when executed by a processor, implements the steps of the above-described distributed generation scenario and cost planning method.

[0016] Compared with the prior art, the present invention has the following advantages and beneficial effects: This invention discloses a distributed generation scenario and cost planning method, system, equipment, and medium. By pre-setting a scenario matrix and ensuring it meets the statistical characteristics of historical scenarios (such as expectation, standard deviation, skewness, kurtosis, and correlation), this scheme can capture historical trends and extreme situations in wind power, photovoltaic power generation, and load demand, making the generated scenarios highly representative. The scenario matrix can more comprehensively reflect the complexity and uncertainty of the system. Based on the uncertainties of wind power and photovoltaic power generation, as well as stochastic models of load demand and electricity price fluctuations, it can dynamically simulate multiple possible future scenarios, improving the decision-making process's adaptability to future uncertainties. Employing cost-benefit analysis, this model can quantify the economic benefits under different strategies, helping decision-makers identify the most cost-effective operation or investment options and achieve optimal resource allocation. This ensures the sustainable development of the energy system; it can also be used for short-term operational scheduling, flexibly responding to immediate changes and improving the overall efficiency and reliability of the system. Attached Figure Description

[0017] The accompanying drawings, which are included to provide a further understanding of embodiments of the invention and form part of this application, do not constitute a limitation thereof. In the drawings: Figure 1 This is a schematic diagram of a scenario matrix according to an embodiment of the present invention; Figure 2 This is a schematic diagram of scenario matrix modeling according to an embodiment of the present invention; Figure 3 This is a schematic diagram illustrating the development of a stochastic distributed generation planning model according to an embodiment of the present invention; Figure 4 This is a bus test feeder topology diagram of an embodiment 53 of the present invention; Figure 5 This is an embodiment of the IEEE 123 bus test feeder topology diagram of the present invention; Figure 6 This is a schematic diagram of the torque error between a generated scenario and a historical scenario according to an embodiment of the present invention; Figure 6 The horizontal axis in the graph represents the number of scenarios generated, and the vertical axis represents the error. Figure 7 This is a schematic diagram of the scenario selection for an embodiment of the present invention, where (a) is a 53 bus test feeder and (b) is an IEEE 123 bus test feeder. Figure 7 The horizontal axis, Number of scenarios, represents the number of scenarios; the vertical axis, Multi-objective optimization index, represents the multi-objective index; and the vertical axis, Computation time, represents the computation time. Figure 8 This is a schematic diagram of the normalized values ​​of a scenario matrix containing 40 scenarios according to an embodiment of the present invention; Figure 8 The horizontal axis in the matrix represents the scenario matrix, and the vertical axis represents the normalized value. Figure 9 This is a schematic diagram of a scenario-based stochastic DGIP model framework according to an embodiment of the present invention; Figure 10 This is a 138 bus test distribution network topology diagram according to an embodiment of the present invention; Figure 11 This is a schematic diagram of the generation of a scenario matrix according to an embodiment of the present invention; (a) is a historical scenario of wind power generation and load demand, (b) is a historical scenario of photovoltaic power generation, (c) is the percentage error between the historical scenario and the scenario obtained by the scenario matrix modeling method, and (d) is a normalized scenario of wind power, photovoltaic power generation and load demand obtained by the scenario matrix modeling method. Figure 11 In this context, "Historical scenarios" refers to historical scenarios, "Number of scenarios" refers to the number of scenarios, "Normalized value" refers to the normalized value, and "Moment error" refers to the error. Figure 12 This is a schematic diagram of the scenario selection standard according to an embodiment of the present invention; (a) is a 53 bus test feeder and (b) is a 138 bus test feeder. Figure 12 In this context, "Number of scenarios" represents the number of scenarios, i.e., scenario numbers; "Net present value" represents the net present value. Figure 13 This is a schematic diagram of the optimal configuration of a wind turbine and a photovoltaic generator according to an embodiment of the present invention; (a) is a 53 bus test feeder and (b) is a 138 bus test feeder. Figure 13 In this context, "Optimal location" represents the best location, and "Optimal size" represents the best capacity. Figure 14 This is a cost diagram ignoring uncertainties in an embodiment of the present invention; (a) is a 53 bus test feeder and (b) is a 138 bus test feeder. Figure 14 In this context, the scenario number represents the scenario matrix, CIU represents the cost of ignoring uncertainty, and Variability represents the value of variation. Figure 15 This is a schematic diagram of the relationship between net present value and ECIU and DG penetration level in an embodiment of the present invention; (a) is a 53 bus test feed line and (b) is a 138 bus test feed line. Detailed Implementation

[0018] To make the objectives, technical solutions, and advantages of the present invention clearer, the present invention will be further described in detail below with reference to the embodiments and accompanying drawings. The illustrative embodiments and descriptions of the present invention are only used to explain the present invention and are not intended to limit the present invention.

[0019] This invention includes a pre-constructed scenario matrix that satisfies the expected value, standard deviation, skewness, kurtosis, and correlation of historical scenarios, resulting in representative scenarios. Based on this scenario matrix, a stochastic DGIP optimization model is constructed. This model, considering the uncertainties of wind and solar power generation, as well as the impact of load demand and electricity price fluctuations, employs cost-benefit analysis to maximize benefits. By using a pre-constructed scenario matrix and ensuring it meets the statistical characteristics of historical scenarios, this invention can capture historical trends and extreme cases of wind and solar power generation and load demand, making the generated scenarios highly representative. The stochastic DGIP optimization model employs cost-benefit analysis, which can quantify the economic benefits under different strategies, helping decision-makers identify the most cost-effective operation or investment plan and achieve optimal resource allocation.

[0020] Example 1 like Figures 1 to 15 As shown, this invention provides a distributed generation scenario and cost planning method, the method comprising: Step 1: Obtain historical hourly wind power basic data, which includes wind turbine power generation, photovoltaic power generation, and load demand. Step 2: Based on the basic wind power data, generate a representative scenario matrix. The scenario matrix satisfies the expected value, standard deviation, skewness, kurtosis and correlation of historical scenarios, and obtains representative scenarios for wind turbine power generation, photovoltaic power generation and load demand. Step 3: Integrate the scenario matrix into the power flow equation to establish a stochastic DGIP optimization model; the stochastic DGIP optimization model is formed by integrating the scenario matrix with deterministic decision variables. Step 4: Construct the objective function and constraints for the stochastic DGIP optimization model, solve the stochastic DGIP optimization model, and output the optimal DGIP solution.

[0021] This invention, through a pre-defined scenario matrix and ensuring it meets the statistical characteristics of historical scenarios (such as expectation, standard deviation, skewness, kurtosis, and correlation), captures historical trends and extreme situations in wind power, photovoltaic power generation, and load demand, resulting in highly representative scenarios. The scenario matrix comprehensively reflects the complexity and uncertainty of the system. Based on the uncertainties of wind and photovoltaic power generation, as well as stochastic models of load demand and electricity price fluctuations, it can dynamically simulate various possible future scenarios, improving the decision-making process's adaptability to future uncertainties. Employing cost-benefit analysis, the model quantifies the economic benefits of different strategies, helping decision-makers identify the most cost-effective operating or investment options and achieve optimal resource allocation. This ensures the sustainable development of the energy system and can also be used for short-term operational scheduling, flexibly responding to immediate changes and improving the overall efficiency and reliability of the system.

[0022] This invention proposes a scenario-based stochastic model for distributed generation planning that explicitly considers the uncertainties of wind power generation, photovoltaic power generation, and electricity demand. The model provides optimal capacity and location for wind turbines and distributed photovoltaic generators, aiming to minimize active and reactive power losses as well as voltage deviations in the active power distribution network.

[0023] First, a scenario matrix is ​​generated using a heuristic moment matching method, which considers stochastic characteristics (i.e., expected value, standard deviation, and kurtosis) and the correlation between historical wind power, photovoltaic power generation, and electricity demand. The scenario matrix effectively transforms a large number of historical scenarios into significantly reduced scenarios. Subsequently, a stochastic programming model is developed using the scenario matrix, and its effectiveness is evaluated through case studies testing distribution feeders on IEEE 53 bus and IEEE 123 bus. Comparative analysis with deterministic programming solutions further confirms the superiority of the proposed stochastic programming solution.

[0024] In the proposed DGIP problem, uncertainty primarily stems from the intermittency of wind turbines and photovoltaic (PV) generators, as well as variations in load demand and electricity prices. The power output of wind turbines and PV generators is not only closely related to geographical location and weather conditions but also exhibits seasonal and diurnal fluctuations. Furthermore, load demand forecasting is challenging due to fluctuations in hourly electricity prices, weather conditions, and customer preferences. Providing detailed historical scenarios for wind power, PV generation, and load demand would significantly increase computational complexity. To avoid this, a scenario matrix model is established by transforming historical scenarios into a reduced but sufficiently large set of representative scenarios. These representative scenarios possess the same degree of randomness as the historical scenarios. Figure 1 As shown, the scenario matrix consists of H representative wind power and photovoltaic power generation scenarios (i.e., ) and load demand scenarios between the corresponding upper and lower limits. 1) Heuristic Moment Matching Method: The scenario matrix modeling method in this invention obtains representative wind turbine power generation, photovoltaic power generation, and load demand scenarios by satisfying the target moments (i.e., expectation, standard deviation, skewness, and kurtosis) and correlations of historical scenarios. Unlike traditional moment matching methods, the scenario matrix modeling method in this invention avoids the computational complexity associated with difficult-to-handle high-dimensional discrete variables. Compared with Monte Carlo methods and Latin multicubic sampling techniques, the scenario matrix modeling method in this invention generates fewer representative scenarios, thereby reducing the computational burden.

[0025] The scenario matrix modeling method includes the following three steps: i) Generate an n-dimensional random matrix that follows a normal distribution. This includes independent column vectors, i.e.;

[0026] Ⅱ) Matrix transformation, as shown in equation (1), with Given an n-dimensional matrix as input, output an n-dimensional matrix. ,in It is a correlation matrix The lower triangular matrix can be determined through decomposition;

[0027] III) Cubic transformation, as shown in equation (2), transforms the column vectors that follow a normal distribution. Obtain a univariate normal random column vector with four given moments. , and It is an n-dimensional matrix, such that The transformation coefficients in equation (2) It is based on the assumed target scenario The moment equals the historical scenario The result is calculated under the condition of the target moment, as shown in equation (3), where This corresponds to the first four moments (i.e., expectation, standard deviation, skewness, and kurtosis). These refer to wind power generation, photovoltaic power generation, and load demand, respectively.

[0028] In this embodiment, the following steps are followed ( Figure 3 ), to obtain the scenario matrix; specifically including: Initialization: Based on wind power baseline data, calculate the target moments and target correlation matrix for historical scenarios. The target moments and target correlation matrices for historical scenarios refer to the target moments and target correlation matrices for historical hourly wind turbine power generation, photovoltaic power generation, and load demand. Since matrix transformation requires the expected value and standard deviation to be equal to 0 and 1 respectively, the objective moment was normalized to obtain the normalized objective moment; the formula for normalization is shown in (4);

[0029] in, and They are column vectors Normalized moments and objective moments; Randomly generated scenarios: by generating scenarios from a normal distribution random sampling Uncertainty factors Scenario, namely wind power generation Photovoltaic power generation and load demand Thus, the first matrix is ​​established. It should be noted that, This is a stochastic scenario matrix and should not be combined with the wind power generation data generated in subsequent steps. Photovoltaic power generation and load demand The representative scenarios are confused.

[0030] Matrix transformation: transform the first matrix Transform into the second matrix To meet the target relevance matrix of historical context As shown in formula (1); Cubic transformation: transform the second matrix Transformed into a third matrix, namely the normalized scenario matrix. To meet the normalization target matrix of historical context As shown in formula (2); Error criterion for verification: For the third matrix Second matrix Each error criterion is checked separately; if all error criterions meet the corresponding preset error, the inversion of the third matrix continues; if at least one error criterion fails to meet the corresponding preset error, the process returns to normalize the target moments. Specifically, the moment error ( ) and related errors ( The calculation results are shown in formula (5) and formula (6) respectively, and the upper limit of these errors is set as shown in formula (7);

[0031] in, yes The generated column vector of Rectangle, It is the correlation matrix of the generated scenario. It is a target relevance matrix of historical context; Output: For the third matrix Perform inversion to meet the target time. As shown in formula (8); the output is a scenario matrix. Generated by wind power Photovoltaic power generation and load demand Representative scenarios constitute;

[0032] Incorporating uncertainties in wind power generation, photovoltaic power generation, and electricity demand into optimization problems increases computational complexity. To ensure computational efficiency, it is necessary to describe these uncertainties by reducing the set of scenarios. Solutions based on stochastic programming can avoid conservative decision-making because the weighted sum of the objectives for all scenarios is optimized.

[0033] In this context, a stochastic DGP model is established by integrating the scenario matrix modeling method into the power flow equation. Figure 3 As shown in (9), the proposed DGP model is implemented using the YALMIP modeling language and solved in CPLEX.

[0034]

[0035] Consider candidate DG units with active and reactive power support capabilities, namely photovoltaic generators based on voltage source inverters (VSI) and wind turbines based on doubly fed induction generators (DFIG). The reactive power generation of these DG units is a function of their active power generation and power factor, as shown in (10).

[0036]

[0037] 1) Objective function The objective of the proposed stochastic DGP problem is to minimize active power loss, reactive power loss, and voltage deviation. Therefore, a multi-objective function is defined in step 4, which is a weighted sum of the active power loss index (PLI), reactive power loss index (QLI), and voltage deviation index (VDI), as shown in (11);

[0038] Considering PLI, QLI, and VDI in the objective function ensures that all power quality indices are minimized; the weighting coefficients in formula (11) and The weighting can be adjusted according to the relative importance of each power quality indicator. In this invention, it is assumed that the above weighting coefficients are equal, i.e. PLI is the ratio of the active power loss of the distribution system after DG distribution to the active power loss of the distribution system before DG distribution, as shown in formula (12). Similarly, QLI and VDI are defined as shown in formula (13) and formula (14), respectively; the lower values ​​of PLI, QLI and VDI represent the reduction of active power loss, reactive power loss and voltage deviation, respectively.

[0039] (2) Constraints Constraints (15)-(18) are based on a convex power flow model programmed with mixed integer quadratic constraints; constraint (15) balances the total active power generation (i.e., the output power of substations, wind turbines, and photovoltaic power generation) of each bus with the total active power demand (i.e., load and line losses). Similarly, constraint (16) satisfies the nodal power balance between total power generation and reactive power demand; the definitions of bus voltage and branch current are given in (17) and (18), respectively; the limits of bus voltage and branch current are defined in (19) and (20), respectively; among which, the minimum / maximum voltage deviation limit is 5%, and the maximum branch current limit is 1 pu; the maximum penetration rate of DG in the active distribution network is limited by (21); the capacity limit of a single wind turbine and photovoltaic unit on the bus is limited by equation (22). The capacity limit of a single wind turbine and photovoltaic unit on the bus has been specified in detail in equation (22).

[0040]

[0041] Figure 4 and Figure 5 The images show the 53 bus power distribution feeder and the IEEE 123 bus power distribution feeder, respectively. The peak load requirements of the 53 bus test feeder are 45.67 MW and 22.12 MVar, with maximum rated voltage and power of 18 kV and 50 MVA, respectively. The peak load requirements of the IEEE 123 bus test feeder are 77.44 MW and 37.59 MVar, with maximum rated voltage and power of 18 kV and 100 MVA, respectively.

[0042] In addition, the following assumptions were made: 1) The maximum penetration rate of total distributed generation cannot exceed 40%; 2) Candidate wind turbines and photovoltaic generators provide active and reactive power at a power factor of 0.9; 3) The minimum and maximum rated power of candidate wind turbines are 1 MW and 2 MW, respectively; 4) The minimum and maximum rated power of candidate photovoltaic modules are 0.1 MW and 2 MW, respectively; 5) The correlation between wind turbines and photovoltaic power generation in different geographical locations is not considered. This is because the geographical area covered by the distribution system is relatively small, so the wind speed and solar irradiance are the same for all buses.

[0043] To generate the scenario matrix, historical hourly data for wind power generation, photovoltaic power generation, and electricity demand over three years (2014-2016) were extracted. First, 26,280 (8,760 * 3 years) historical wind and photovoltaic power generation scenarios and load demand were normalized to their corresponding peak values. Using scenario matrix modeling techniques, normalized scenarios of varying orders of magnitude (10 to 100) can be generated to simulate and analyze various stochastic processes. Empirical studies validated the two proposed novel scenario tree generation methods. The generated scenarios showed a high degree of consistency with the cumulative distribution of historical data. The "approximately uniform method" generated the scenario cumulative distribution with the highest degree of fit to the historical distribution, and the fit increased with the number of scenarios. Furthermore, the error decreased significantly with the increase in the number of scenarios. Figure 6 As shown, this demonstrates that the scenario matrix has a fairly high accuracy in capturing the random characteristics of historical data.

[0044] By analyzing the relationship between the multi-objective index (MOI) and computation time and the number of schemes, the minimum number of schemes that achieves the best balance between computational accuracy and efficiency is selected. In a 53 bus power distribution system, Figure 7 (a) shows that the MOI tends to 1.17 when the number of schemes reaches 40. Although the MOI remains unchanged after the number of schemes exceeds 40, the computational burden increases significantly. Therefore, 40 schemes are sufficient for the DGP solution of the 53 bus system. Figure 7 As shown in (b), in the IEEE 123 bus power distribution test feeder, the MOI converges to 1.09 when the number of schemes approaches 80. Therefore, 80 schemes are selected to solve the DGP problem in the 123 bus system. Taking the normalized value of 40 schemes as an example, as... Figure 8 As shown.

[0045] CPLEX was used to solve the DGP problem in the experimental feeders of the 53 bus and the modified IEEE 123 bus, with 40 and 80 solutions respectively. According to Table 1, the DG allocation strategy using the DGP model can significantly reduce active and reactive power losses and effectively reduce voltage deviation. This conclusion is supported by related research, such as optimization methods for distribution network voltage deviation mitigation based on life-cycle cost (LCC) theory, and in-depth analysis of reactive load and losses in power systems, all demonstrating that appropriate mitigation measures can improve grid performance. The MOIs in the 53 bus and 123 bus systems are 1.17 and 1.09, respectively. Furthermore, Table 2 shows that in the 53 bus feeder, the total allocated capacity of wind turbines and photovoltaic generators is 7 MW and 9.1 MW, respectively. The total DG capacity is 16.1 MW, with a DG penetration rate of 35.8%. In the 123 bus feeder, in addition to 18.9 MW of photovoltaic units, 12 MW of wind turbine units are also allocated. The total distributed generation (DG) capacity is 30.9 MW, equivalent to a DG penetration rate of 39.9%. The reason for the over-allocation of solar photovoltaic (PV) power generation compared to wind turbines is that the historical power generation patterns of PV are more aligned with electricity demand patterns. This means that the power generation capacity factor may be a crucial consideration when allocating different types of renewable distributed generation technologies.

[0046] Table 1 Proposed DGP Solution: Power Quality Index

[0047] Table 2 Proposed DGP Solutions: Optimal Wind Turbine Capacity and Location

[0048] The proposed stochastic DGP model is based on a scenario matrix modeling method, using a smaller number of scenarios. The randomness of these scenarios is comparable to the actual historical scenarios of wind and solar power generation and load demand over the past three years. On the other hand, deterministic programming models assume perfect power generation and demand data, thus ignoring uncertainty. This invention evaluates the effectiveness of the scenario matrix modeling method solution by comparing it with deterministic solutions; specific data are shown in Table 3. It can be seen that the scenario matrix modeling method solution significantly reduces the power quality index, thereby substantially reducing active and reactive power losses and voltage deviation in the active power distribution network. In the 53 bus test feeder, total active and reactive power losses and voltage deviation were reduced by 26%, 25%, and 25.7%, respectively. Similarly, in the 123 bus test feeder, total active and reactive power losses and voltage deviation were reduced by 25%, 21%, and 26.4%, respectively. Therefore, these results confirm the superiority of the proposed DGP model based on the scenario matrix modeling method.

[0049] Table 3 Comparison of Deterministic Scheme and Proposed DGP Scheme

[0050] In the planning of active distribution networks, the uncertainties brought about by renewable distributed generation and load demand must be fully considered. Furthermore, inappropriate distributed generation allocation can lead to excessive power loss, voltage instability, power quality degradation, and protection incoordination. Against this backdrop, a stochastic distribution generation planning model based on a scenario matrix using heuristic moment-value matching technology is proposed. The scenario matrix effectively transforms massive amounts of historical data into a significantly reduced number of scenarios. A multi-objective optimization model based on indicators is considered to minimize active and reactive power losses and voltage deviations in the active distribution network. Subsequently, the performance of the proposed planning model is evaluated through case studies of IEEE 53 bus and IEEE 123 bus distribution feeders. Numerical results show that by explicitly considering system uncertainties, the proposed model provides optimal site selection and scale for wind turbines. A comparison between stochastic and deterministic programming solutions further highlights the effectiveness of the proposed method in significantly reducing active and reactive power losses and voltage deviations.

[0051] Based on the insights gained from this work, the following areas require further research: First, a quantitative comparison should be made between the proposed stochastic programming solution and robust optimization-based solutions in terms of both power quality performance and investment costs under a range of network operating conditions. Second, further attention needs to be paid to control and management methods that maximize distributed generation penetration, as the related techno-economic benefits significantly improve with higher penetration rates. Finally, investment planning for distributed generators requires further research, which will be discussed below.

[0052] Distributed generation network operators (DNOs) are constantly exploring sustainable alternatives to maximize profits, provide reliable and cost-effective services to customers, and ensure that bus voltage and power quality remain within permissible limits. To achieve these goals, distributed generation (DG) has emerged as a promising technology, with associated techno-economic benefits including reduced power losses, improved power supply reliability and security, and lower transmission costs and emissions. However, the integration of DG transforms passive distribution networks into active systems with bidirectional power flow. Furthermore, while renewable energy technologies such as wind turbines (WT) and photovoltaics (PV) are intermittent, advancements in technology, such as the application of distributed energy storage systems, can improve the stability and predictability of their energy supply. In addition to these uncertainties, load variations, demand growth, and electricity market prices introduce further sources of uncertainty. This leads to new operational and control challenges, such as voltage rise effects, increased fault levels, reduced protection capabilities, and altered transient stability. This necessitates an efficient and effective distributed generation investment planning (DGIP) methodology to compare the economic benefits of distributed generation with its investment costs, thereby ensuring a high return on investment.

[0053] Using the aforementioned scenario matrix, an innovative scenario-based stochastic model was developed for distributed generation investment planning. This model comprehensively considers the uncertainties of wind and solar power generation, as well as the impact of load demand and electricity price fluctuations. This stochastic economic model employs cost-benefit analysis to maximize net present value from the perspective of the distribution network operator. Uncertainty is characterized by a scenario matrix based on the heuristic moment matching method. Subsequently, the scenario matrix is ​​combined with deterministic power flow equations to form a stochastic programming model. The performance of the proposed model was first verified in a 53-bus distribution test feeder, and then its scalability and applicability were further verified in a 138-bus distribution network. Using a deterministic programming model as a benchmark, the superiority of the proposed planning model was also confirmed.

[0054] like Figure 9 As shown, the development of the DGIP model begins with the modeling of a scenario matrix, which encompasses multiple representative scenarios for wind power, solar power, and load demand. The scenario matrix is ​​modeled using a scenario matrix modeling method, which effectively captures the stochastic characteristics and interrelationships of wind power, solar power, and load demand. Subsequently, the scenario matrix is ​​combined with deterministic power flow equations to propose a stochastic DGIP problem, thus ensuring that the number of power flow equations equals the number of representative scenarios. Finally, the DGIP problem is solved to determine the optimal location and capacity of candidate distributed generation (DG), thereby maximizing net present value (NPV) from the perspective of distributed nomenclature (DNO). The DGIP problem is coded using the YALMIP modeling language and solved using CPLEX 12.3.

[0055] Scenario matrix modeling, based on the scenario matrix modeling method, consists of a small number of scenarios with uniform probability of occurrence. The scenario matrix modeling method uses matrix and cubic transformations to capture the correlation and random moments of historical data, respectively. Four different moments—expectation, standard deviation, skewness, and kurtosis—are usually sufficient to preserve the random characteristics of historical data. Scenario matrix modeling and its methods are detailed above. The output of this process is the scenario matrix. These include scenarios for generating wind power, photovoltaic power, and load demand.

[0056] As shown in (23), the DGIP optimization problem is to solve the scenario matrix by... With deterministic decision variables The integrated form is shown in formula (23):

[0057] Specifically, the stochastic DGIP optimization model is constructed from the perspective of the DNO (Distribution Network Operator), assuming that the DNO is the sole manager of the active distribution network. In addition to using DG units to meet load demand, the DNO can also purchase electricity from the wholesale market. Candidate DG units with active and reactive power support capabilities are considered, namely photovoltaic generators based on voltage source inverters (VSI) and wind turbines based on doubly fed induction generators (DFIG). The reactive power generation of these DG units is a function of their active power generation and power factor, as shown in (24).

[0058]

[0059] Specifically, maximizing the net present value (NPV) of DNO over the planning period is the objective function of the stochastic DGIP optimization model, as shown in Equation (25); NPV is based on the present value of revenue. Present value of costs The difference between them was calculated using a cost-benefit analysis method, as shown in formula (26);

[0060] The assessment method is to divide the annual income by the capital recovery factor ( ), as shown in formula (27); where annual revenue includes the revenue obtained annually from the sale of electricity to customers and the wholesale market, as shown in formula (28); The present value of annual cash flows is typically used to calculate the present value of annual cash flows, as shown in formula (29); the first term in formula (28) is the annual revenue earned from selling electricity to meet customer load demand, as shown in formula (30); the second term in formula (28) is the annual revenue earned from selling surplus electricity to the wholesale market, as shown in formula (31).

[0061] Since scenario H has a uniform probability of occurrence, a weighting coefficient is used. The annual revenue and costs of scenario H (i.e., in (30), (31), (39) and (40) , and ) Scaled up proportionally to the number of hours in a year; It is obtained by dividing the annualized cost by the capital recovery factor, as shown in formula (32); where the annualized cost includes the annual capital cost, annual replacement cost, annual operating cost and annual maintenance cost of the installed renewable energy generator set, as well as the cost of purchasing electricity from the wholesale market and the cost of unsupplied load, as shown in formula (33); In formula (33), the annual replacement cost is represented, which is the difference between the component replacement cost over the entire project lifecycle and the component residual value at the end of the project, as shown in formula (35); the accrued fund coefficient ( ) is used to calculate the present value of future cash flows, as shown in formula (36); in formula (33) This represents the annual maintenance cost, which is proportional to the DG installed capacity, as shown in formula (37); in formula (33) This represents the annual operating cost, which is proportional to the installed capacity of DG, as shown in formula (37). The annual cost of purchasing electricity from the wholesale market is represented in formula (33), as shown in formula (39); In (33), the annual cost of unsupplied load is represented as shown in formula (40);

[0062] Specifically, the objective function is subject to a series of constraints, as shown in formulas (41)-(48); (41)-(44) are based on a deterministic power flow model using convex mixed integer quadratic constraint programming (MIQCP); constraint (41) balances the total active power supply and demand of each bus under all representative scenarios, where the total supply includes substations, wind power and photovoltaic power generation, while the total demand is the sum of load demand and line losses; similarly, constraint (42) represents the node power balance of total reactive power supply and demand under all representative schemes; constraints (43) and (44) define the bus voltage and branch current under all cases according to active and reactive power flows, respectively.

[0063] Constraints (45) and (46) specify the lower and upper limits of the bus voltage and branch current, respectively, with the minimum / maximum voltage deviation limited to 5% and the maximum branch current limited to 1 pu;

[0064] As stated in constraint (47), the increase in the penetration rate of renewable wind turbines will reduce the stability and power quality of the distribution network, and the total installed capacity of wind turbines can be limited to a certain range of peak load demand; constraint (48) also specifies the upper and lower limits of the installed capacity of wind turbines and photovoltaic units on each bus.

[0065] Two distribution test feeders were considered: first, the performance of the proposed DGIP model was evaluated using a 53-bus distribution test feeder, and then the scalability of the model was further tested using a 138-bus distribution network. Figure 4 As shown, the 53 bus distribution test feeder consists of 61 branches, with peak demands of 45.67 MW and 22.12 Mvar. The maximum rated voltage and current of the distribution feeder are 18 kV and 50 MVA, respectively. The second distribution system consists of 138 nodes and 137 branches, with peak demands of 77.44 MW and 37.59 MVA, as shown... Figure 10 As shown. The maximum rated voltage and current of the distribution feeder are 18 kV and 100 MVA, respectively. The substations for both test feeders are located at busbar 1.

[0066] The DGIP problem was coded using the YALMIP modeling language and solved in CPLEX 12.3. The simulation settings are as follows: i) The maximum renewable energy penetration rate in the active distribution network is limited to 40% of total peak demand. The maximum DG capacity per bus is limited to 2 MW. The distribution system spans a relatively small geographical area, therefore all buses are affected by uniform wind speed and solar irradiance.

[0067] (ii) The planning period is 20 years. For simplicity, the service life of the generator and all other equipment is set at 20 years; therefore, replacement costs and residual value are not considered. The discount rate is set at 7%, which remains consistent throughout the planning period.

[0068] III) As shown in Table 1, the technical specifications of the candidate wind turbines and their corresponding investment, operation, and maintenance costs are based on... The capacity of the wind turbines is modeled as a discrete variable. In contrast, photovoltaic technology, due to its ability to provide sufficiently small modules, simulates its power generation as a continuous variable, thus significantly shortening the computation time.

[0069] IV) The cost of unsupplied load is $3,000 / MWh. The electricity prices for load levels 1-3 are $50 / MWh, $60 / MWh, and $73 / MWh, respectively, where load levels 1-3 correspond to peak demand load rates of 70%, 83%, and 100%, respectively.

[0070] Table 4 Simulation parameters of wind turbines and photovoltaic power generation devices

[0071] The performance evaluation of the DGIP model is based on the following two scenarios: Scenario A: DNO only imports / exports electricity from the wholesale electricity market.

[0072] Case B: The DNO imports / exports electricity from the wholesale electricity market and utilizes the installed DGs.

[0073] The scenario matrix is ​​generated based on three years (2014–2016) of historical hourly data on wind power generation, solar power generation, and load demand collected from the Texas Electric Reliability Council public database. Figure 11 (a) and Figure 11 As shown in (b), 26,280 (8,760 * 3 years) historical scenarios of wind power generation, photovoltaic power generation, and load demand were normalized to their peak values. Then, a scenario matrix modeling method was used to generate 10 uncertainty matrices containing 10 to 100 normalized scenarios, with a step size of 10 scenarios, where the thresholds for correlation error and moment error were both set to 5%. Figure 11 (c) shows the moment error between the scenarios generated by the scenario matrix modeling method and historical scenarios. The moment error decreases as the number of generated scenarios increases, indicating that the scenario matrix accurately captures the stochastic characteristics of historical scenarios such as wind power, photovoltaic power generation, and load demand. Figure 11 (d) shows the normalized values ​​of the 40 scenario-scenario matrices. These scenarios clearly simulate the seasonal and diurnal variations in wind power, solar power generation, and load demand. For example, in scenarios #1-8 and #33-40, solar power generation is zero, representing the nighttime period, while solar scenario #9-32 represents the daytime period.

[0074] The selection of a solution is based on minimizing the number of options, achieving the optimal trade-off between solution accuracy and computational efficiency. This can be done by comparing the objective function value (i.e., net present value) and the computation time (i.e., the computation time in CPLEX 12.3). Taking a 53 bus power distribution system as an example... Figure 12(a) shows that the net present value (NPV) stabilizes as the number of schemes increases, converging to $5.641 M$ when the number of schemes approaches 80, but the computation time is as long as 59,520 seconds (i.e., 16.5 hours). In contrast, the NPV obtained using the optimized 40 schemes is $565.3 million, demonstrating high accuracy (error of only 0.21%), while significantly improving computational efficiency, requiring only 2,160 seconds (i.e., 0.6 hours) to complete. Therefore, using 40 schemes achieves the best balance between solution accuracy and computation time. Similarly, in the test feeder of the 138 bus power distribution ( Figure 12 (b) In the analysis, the net present value (NPV) method was used to evaluate 40 schemes, yielding a NPV of $1,138 million and a computation time of only 1,836 seconds (approximately 0.51 hours). This demonstrates a significant advantage in balancing computational efficiency and result accuracy (with an error of only 0.24%). In contrast, while the 80 schemes had a higher NPV of $14.103 million, their computation time was a staggering 101,184 seconds (approximately 28.12 hours). Therefore, based on a comprehensive consideration of computational efficiency and result accuracy, the 40 schemes were selected as the DGIP solutions for the two test networks.

[0075] Table 5 shows the detailed results for 40 stochastic DGIP problems in the 53-bus distribution system. After the integration of distributed generation systems (DGs), changes in power flow distribution and topology lead to system fault components that are directly related to the output power and integration location of the DGs. Therefore, the DG penetration rate was limited to 40%, ensuring consistent overall electricity sales revenue in both test cases, with zero sales revenue in the wholesale market. The cost of unsupplied loads was zero in both test cases, demonstrating the adequacy of the proposed model. Although the capital, operation, and maintenance costs in Case B are significantly higher due to the installation of DGs, the cost of electricity purchase is significantly reduced compared to Case A. The net present value (NPV) for Case A reaches $3.158 million, however, this is significantly lower than the $5.653 million for Case B. Similarly, Table 6 shows the solutions to 40 stochastic DGIP problems in the 138-bus distribution system. Clearly, the increase in NPV from $9.181 million (Case A) to $14.18 million (Case B) represents a 59% increase in DNO profits, highlighting the economic value of the proposed DGIP model for DNOs.

[0076] Figure 13 (a) shows the optimal location and optimal size of wind and solar power generators in the 53-bus distribution network. The proposed plan allocates 10.571 MW of solar power and 7 MW of wind turbines, for a total DG capacity of 17.571 MW. Figure 13As shown in (b), 16.977 MW of photovoltaic (PV) generators and 14 MW of wind turbines are allocated in the 138 bus distribution network, resulting in a total DG capacity of 30.977 MW. The allocation of PV capacity is more prominent compared to wind turbines because PV projects overlap more significantly with load schemes than wind power projects. This means that generation capacity is an important consideration when allocating different renewable DG resources.

[0077] Table 553 Proposed DGIP Solutions for Bus Power Distribution Systems

[0078] Table 6138 Proposed DGIP Solutions for Bus Power Distribution Systems

[0079] The effectiveness of the proposed DGIP model can be evaluated through a comparative analysis with deterministic models, which ignore uncertainties by assuming perfect information on wind power, solar power generation, and load demand. To this end, a stochastic metric called "Uncertainty Ignoring Cost" (CIU) is used to measure the cost of making decisions based on deterministic solutions. For a given scenario, CIU can be used to evaluate the economics of different options by calculating the net present value (NPV) of a deterministic solution based on that scenario and comparing it with the NPV of a stochastic DGIP solution. For the 53 bus and 138 bus distribution test feeders, the NPVs based on the stochastic DGIP solution are 5.653 M$ (Table 5) and 1138 M$ (Table 6), respectively. Figure 14 (a) shows the CIU values ​​in the 53-bus distribution network, along with variations in wind power output, solar power output, load, and electricity price across different scenarios. The CIU values ​​vary significantly depending on the deterministic scenario adopted by the DGIP solution. The maximum CIU is as high as 0.892 M$ (Scenario #2), while the worst-case CIU is 0.531 M$ (Scenario #19), indicating that adopting deterministic decision-making significantly reduces the net present value. Furthermore, the expected cost ignoring uncertainty (ECIU), equivalent to the weighted sum of the CIU values ​​across all scenarios, is $0.636 billion. This ECIU represents 11.3% of the net present value of the proposed DGIP solution, equivalent to $565.3 million, and its importance cannot be ignored. For the 138-bus distribution test feeder, such as... Figure 14As shown in (b), the maximum CIU is as high as 2.884 M$ (Option #1), while the worst-case CIU is 2.158 M$ (Option #19). The calculated ECIU is $2.371 million, a significant amount as it represents 16.8% of the net present value, or $14.18 million. Therefore, the proposed DGIP model offers a superior solution compared to the deterministic programming model.

[0080] To further evaluate the performance and economic significance of the proposed DGIP model, a series of numerical experiments were conducted at different DG penetration levels (ranging from 10% to 100%). These experiments considered the impact of DG access on distribution network relay protection and how planning methods can guide the DG configuration planning of practical distribution networks. For the 53 bus distribution system, Figure 15 (a) shows a significant increase in the net present value (NPV) of the distributed network (DNO), from $195.9 M$ at a 10% DG penetration level to $1954 M$ at a 100% DG penetration level, while the ECIU also increases from $0.221 M$ to $2.187 M$. These results indicate that the economic performance of the distribution network is significantly improved with increasing DG penetration. According to the research in Reference 1, the NPV of the 138 bus distribution system was $411.4 M$ at a 10% renewable energy penetration level, increasing to $445.14 M$ at a 100% penetration level, indicating a significant improvement in the system's economics with increasing renewable energy penetration. Simultaneously, the ECIU increased from $0.687 M$ to $7.434 M$, reflecting enhanced system flexibility. Therefore, it is certain that the solution provided by the proposed model remains superior and profitable even at higher DG penetration levels.

[0081] To fully consider the various network uncertainties arising from renewable distributed generation, load demand, and electricity price dynamics, a scenario-based stochastic distributed generation investment planning model was developed using the heuristic moment matching method, aiming to maximize net present value (NPV) from the perspective of distribution network operators. The scenario matrix based on the heuristic moment matching method characterizes the uncertainties, and is then combined with deterministic power flow equations to establish a scenario-based stochastic programming model. The performance of the proposed model has been validated on distribution networks with 33, 69, 84, 136, and 417 bus lines, and its scalability has also been verified in a 138 bus distribution network. Numerical results confirm the economic significance of the proposed planning scheme from the perspective of NPV maximization. Furthermore, in-depth comparison with deterministic programming schemes reveals that the proposed model can accurately capture and address network uncertainties. The performance of the proposed model was also evaluated in light of the increasing penetration rate of distributed generation, showing that NPV will be significantly improved.

[0082] Distribution network operators can maximize profits by optimizing and integrating renewable wind turbines using the planning schemes discussed in this paper. It is worth noting that increased penetration of renewable energy generators can also negatively impact the performance of active power distribution networks, such as voltage rise effects, increased fault levels, reduced protection capabilities, and altered transient stability. Therefore, to maximize distributed generation (DG) penetration and obtain related economic benefits, distributed network operators (DNOs) need to adopt various operating strategies, such as applying advanced converter-based control technologies. Based on the insights from this work, two research directions warrant further investigation. First, it is necessary to further explore how to efficiently integrate and operate DG, as increasing DG penetration in the distribution system can significantly improve net present value. For example, high-penetration DG integration into the distribution network alters the structure of traditional distribution networks, affecting power flow distribution and short-circuit currents, requiring redesign and reconfiguration of relay protection. Furthermore, a DG planning method that considers distribution network transfer can guide the DG configuration planning of actual distribution networks, thereby increasing penetration and improving system economics.

[0083] Example 2 The difference between this embodiment and Embodiment 1 is that this embodiment provides a distributed generation scenario and cost planning system, which corresponds one-to-one with the distributed generation scenario and cost planning method in Embodiment 1; the system includes: The acquisition unit is used to acquire historical hourly wind power basic data, which includes wind turbine power generation, photovoltaic power generation, and load demand. The scenario matrix generation unit is used to generate a representative scenario matrix based on wind power basic data. The scenario matrix satisfies the expectation, standard deviation, skewness, kurtosis and correlation of historical scenarios to obtain representative scenarios of wind turbine power generation, photovoltaic power generation and load demand. The optimization model building unit is used to integrate the scenario matrix into the power flow equation to establish a stochastic DGIP optimization model; the stochastic DGIP optimization model is formed by integrating the scenario matrix with deterministic decision variables. The model solving unit is used to construct the objective function and constraints for the stochastic DGIP optimization model, solve the stochastic DGIP optimization model, and output the optimal DGIP solution.

[0084] As a further implementation, the scenario matrix generation unit includes: The calculation subunit is used to calculate the target moments and target correlation matrices for historical scenarios based on wind power basic data; the target moments and target correlation matrices for historical scenarios refer to the target moments and target correlation matrices for historical hourly wind turbine power generation, photovoltaic power generation, and load demand; The normalization sub-unit is used to normalize the target moments to obtain the normalized target moments; The matrix establishes sub-units, which are used to establish the first matrix based on the normalized target moments and the scenario of randomly generating uncertain factors by sampling from the normal distribution; The matrix transformation subunit is used to transform the first matrix to obtain the second matrix; the second matrix satisfies the target relevance matrix of the historical scenario. The cubic transformation sub-unit is used to perform a cubic transformation on the second matrix to obtain the third matrix; the third matrix is ​​the normalized target moment that satisfies the historical scenario. The inversion subunit is used to invert the third matrix to meet the target time and obtain a representative scenario matrix.

[0085] As a further implementation, the objective function and constraints for the stochastic DGIP optimization model are constructed, including: Maximizing the net present value (NPV) of DNO over the planning period is the objective function of the stochastic DGIP optimization model; the NPV is calculated using cost-benefit analysis based on the difference between the present value of revenue and the present value of costs. The constraints include: First constraint: representing the total active power supply and demand balance of each bus under all representative scenarios; Second constraint: representing the node power balance of total reactive power supply and demand under all representative scenarios; Third constraint: defining the bus voltage and branch current under all conditions based on active power flow and reactive power flow; Fourth constraint: representing the lower and upper limits of bus voltage and branch current; Fifth constraint: limiting the total installed capacity of wind turbine units within the preset range of peak load demand; Sixth constraint: representing the upper and lower limits of the installed capacity of wind turbine units and photovoltaic units on each bus.

[0086] The execution process of each unit can be carried out according to the steps of a distributed generation scenario and cost planning method in Example 1, and will not be described in detail in this example.

[0087] Meanwhile, the present invention also provides a computer device, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements the steps of the above-mentioned distributed generation scenario and cost planning method.

[0088] Meanwhile, the present invention also provides a computer-readable storage medium storing a computer program thereon, which, when executed by a processor, implements the steps of the above-described distributed generation scenario and cost planning method.

[0089] The specific embodiments described above further illustrate the purpose, technical solution, and beneficial effects of the present invention. It should be understood that the above description is only a specific embodiment of the present invention and is not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A distributed generation scenario and cost planning method, characterized in that, The method includes: Obtain historical hourly wind power basic data, including wind turbine power generation, photovoltaic power generation, and load demand; Based on the aforementioned wind power fundamental data, a representative scenario matrix is ​​generated; the scenario matrix satisfies the expectation, standard deviation, skewness, kurtosis and correlation of historical scenarios, and yields representative scenarios for wind turbine power generation, photovoltaic power generation and load demand. The scenario matrix is ​​integrated into the power flow equation to establish a stochastic DGIP optimization model; the stochastic DGIP optimization model is formed by integrating the scenario matrix with deterministic decision variables; Construct an objective function and constraints for the stochastic DGIP optimization model, solve the stochastic DGIP optimization model, and output the optimal DGIP solution.

2. The distributed generation scenario and cost planning method according to claim 1, characterized in that, Based on the aforementioned wind power baseline data, a scenario matrix is ​​generated, including: Based on the aforementioned wind power basic data, calculate the target moments and target correlation matrix for historical scenarios; The target moments are normalized to obtain normalized target moments; Based on the normalized target moments, the first matrix is ​​established by randomly generating scenarios with uncertainties from a normal distribution. The first matrix is ​​transformed to obtain the second matrix; the second matrix satisfies the target relevance matrix of the historical scenario. The second matrix is ​​cubically transformed to obtain the third matrix; the third matrix is ​​the normalized target moment that satisfies the historical scenario. The third matrix is ​​then inverted to satisfy the target time, resulting in a representative scenario matrix.

3. The distributed generation scenario and cost planning method according to claim 2, characterized in that, Based on the aforementioned wind power baseline data, a scenario matrix is ​​generated, which also includes: The third matrix and the second matrix are respectively subjected to error standard tests; If the error standard verification meets the corresponding preset error, then continue to invert the third matrix; If at least one of the error standard checks fails to meet the corresponding preset error, then return to normalize the target moment.

4. The distributed generation scenario and cost planning method according to claim 2, characterized in that, The formula for the cubic transformation is: In the formula, From column vectors that follow a normal distribution Obtain a univariate normal random column vector with four given moments; The transformation coefficients are those assumed in the target scenario. The moment equals the historical scenario The calculation is performed under the condition of the target moment; Corresponding to the first four moments: expectation, standard deviation, skewness, and kurtosis; These refer to wind turbine power generation, photovoltaic power generation, and load demand, respectively.

5. The distributed generation scenario and cost planning method according to claim 1, characterized in that, Constructing an objective function for the stochastic DGIP optimization model includes: The objective function of the stochastic DGIP optimization model is to maximize the net present value of DNO over the planning period; the net present value is calculated using a cost-benefit analysis method based on the difference between the present value of revenue and the present value of cost.

6. The distributed generation scenario and cost planning method according to claim 1, characterized in that, The constraints include: The first constraint is to indicate the balance of total active power supply and demand for each bus under all representative scenarios. The second constraint is the node power balance representing the total supply and demand of reactive power under all representative schemes. The third constraint is to define the bus voltage and branch current for all cases based on the active power flow and reactive power flow. The fourth constraint represents the lower and upper limits of the bus voltage and branch current; Fifth constraint: Limit the total installed capacity of wind turbine units to a preset range of peak load demand; The sixth constraint represents the upper and lower limits of the installed capacity of wind turbines and photovoltaic units on each bus.

7. A distributed generation scenario and cost planning system, characterized in that, The system includes: The acquisition unit is used to acquire historical hourly basic wind power data, which includes wind turbine power generation, photovoltaic power generation, and load demand. The scenario matrix generation unit is used to generate a representative scenario matrix based on the wind power basic data. The scenario matrix satisfies the expectation, standard deviation, skewness, kurtosis and correlation of historical scenarios to obtain a scenario with representative wind turbine power generation, photovoltaic power generation and load demand. The optimization model building unit is used to integrate the scenario matrix into the power flow equation to establish a stochastic DGIP optimization model; the stochastic DGIP optimization model is formed by integrating the scenario matrix with deterministic decision variables; The model solving unit is used to construct the objective function and constraints for the stochastic DGIP optimization model, solve the stochastic DGIP optimization model, and output the optimal DGIP solution.

8. The distributed generation scenario and cost planning system according to claim 7, characterized in that, The scenario matrix generation unit includes: The calculation subunit is used to calculate the target moments and target correlation matrix of historical scenarios based on the wind power basic data; A normalization subunit is used to normalize the target moments to obtain normalized target moments; The matrix establishes sub-units, which are used to establish the first matrix based on the normalized target moments and the scenario of randomly generating uncertain factors by sampling from the normal distribution; The matrix transformation subunit is used to transform the first matrix to obtain a second matrix; the second matrix satisfies the target relevance matrix of the historical scenario. A cubic transformation subunit is used to perform a cubic transformation on the second matrix to obtain a third matrix; the third matrix is ​​the normalized target moment that satisfies the historical scenario. The inversion subunit is used to invert the third matrix to satisfy the target time and obtain a representative scenario matrix.

9. A computer device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the steps of the distributed generation scenario and cost planning method according to any one of claims 1 to 6.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the steps of the distributed generation scenario and cost planning method according to any one of claims 1 to 6.

Images