Power distribution network parameter intelligent setting calculation method considering distributed photovoltaic access

Abstract

The invention discloses a power distribution network parameter intelligent setting calculation method considering distributed photovoltaic access, and relates to the field of power distribution network parameter calculation, and the method comprises the steps: S1, collecting power distribution network operation data and distributed photovoltaic power generation data, and building a power distribution network operation scene data set considering distributed photovoltaic access; s2, according to the scene data set, establishing a power distribution network parameter intelligent setting model based on a generative adversarial network algorithm; and S3, based on the intelligent setting model, outputting a power distribution network parameter intelligent setting effect intelligent evaluation index, and by introducing a generative adversarial network algorithm, carrying out parameter intelligent optimization on the power distribution network containing a large amount of distributed photovoltaic. According to the method, the operation states of the power distribution network under different photovoltaic access conditions can be simulated and predicted more accurately, and more accurate and dynamic parameter adjustment is provided, so that the response capability and the overall efficiency of the power grid are remarkably improved. Parameter setting is optimized through an intelligent method, and adaptability of the power distribution network to renewable energy fluctuation is enhanced.

Description

Technical Field

[0001] The present invention relates to the field of distribution network parameter calculation, and in particular to a distribution network parameter intelligent setting calculation method considering distributed photovoltaic access. Background Art

[0002] With the transformation of the global energy structure and the rapid development of renewable energy technology, distributed photovoltaic systems have been widely used as a clean energy solution around the world. In this context, the widespread access to photovoltaic power generation has brought new challenges to the operation and management of traditional distribution networks. Since photovoltaic power generation has the characteristics of large output fluctuations and unpredictability, the distribution network needs to adapt to this highly dynamic and highly uncertain energy access method to ensure the stability and reliability of the power grid. Therefore, studying how to effectively set the distribution network parameters when distributed photovoltaics are widely accessed is of great significance to ensuring the safe operation of the power grid and improving energy efficiency.

[0003] However, the existing distribution network parameter setting technologies are mostly based on traditional, steady-state grid operation modes, and fail to fully consider the randomness and intermittent characteristics of distributed photovoltaic power generation. As a result, in practical applications, when photovoltaic systems are connected to the grid on a large scale, existing methods are difficult to accurately reflect the actual impact of photovoltaic power generation on the distribution network, which in turn affects the optimal scheduling and safe and stable operation of the grid. In addition, these traditional methods are inefficient when processing large data sets and are difficult to meet the rapidly changing grid operation requirements. Summary of the invention

[0004] In order to overcome the deficiencies in the prior art, the present invention proposes a distribution network parameter intelligent setting calculation method considering distributed photovoltaic access. Through this intelligent setting calculation, not only can the adaptability of the distribution network to the fluctuation and uncertainty of photovoltaic power generation be improved, but also the operation efficiency and reliability of the power grid can be significantly optimized, the operation and maintenance costs can be reduced, and the economy and environmental friendliness of the power grid can be enhanced.

[0005] In order to achieve the above object, the present invention proposes a distribution network parameter intelligent setting calculation method considering distributed photovoltaic access, comprising the following steps:

[0006] S1: Collect distribution network operation data and distributed photovoltaic power generation data to establish a distribution network operation scenario dataset considering distributed photovoltaic access;

[0007] S2: According to the scenario data set, an intelligent setting model of distribution network parameters based on the generative adversarial network algorithm is established;

[0008] S3: Based on the intelligent setting model, output the intelligent evaluation indicators of the intelligent setting effect of distribution network parameters, and the evaluation indicators include voltage deviation, power loss and system stability.

[0009] Step S1: Establishing a distribution network operation scenario data set considering distributed photovoltaic access. The specific steps include:

[0010] S101: Collect distribution network operation data and distributed photovoltaic power generation data;

[0011] S102: Distributed photovoltaic power generation data preprocessing, standardizing the collected data;

[0012] S103: extracting characteristic variables of distributed photovoltaic power generation data, and extracting characteristics such as daily average value and fluctuation range based on the time series photovoltaic power generation data;

[0013] S104: Generate distribution network operation scenario data.

[0014] The standardization process in step S102 is as follows:

[0015]

[0016] Among them, X(t) is the original data of distributed photovoltaic, μ X ,σ X They represent the mean and standard deviation of distributed photovoltaic power generation data respectively, and t represents the current moment.

[0017] The daily average value of photovoltaic power generation data in step S103 is:

[0018]

[0019] in, is the daily average value of photovoltaic power generation data; P PV is photovoltaic power generation; T is the length of time.

[0020] The fluctuation range of photovoltaic power generation data in step S103 is:

[0021] ΔP PV =max(P PV (t))-min(P PV (t))

[0022] Where ΔP PV This is the fluctuation range of photovoltaic power generation data.

[0023] In step S104, the distribution network operation scenario data is formed, specifically:

[0024] D={(L i (t),P PV (t),W(t))|t=1,2,...,T;i=1,2,...,N}

[0025] Where N is the number of distribution network nodes; P PVis photovoltaic power generation; t represents the current moment; L i (t) is the voltage of the ith node; W(t) is the set of distribution network switch states at time t.

[0026] Step S2 designs a distribution network parameter intelligent setting model based on a generative adversarial network algorithm, including a distribution network parameter intelligent setting generator and a distribution network parameter intelligent setting discrimination model.

[0027] The distribution network parameter intelligent setting generator is trained by optimizing the following objective function:

[0028]

[0029] Among them, L G is the loss function of the distribution network parameter intelligent setting generator; D(G(z)) represents the evaluation probability of the distribution network parameter intelligent setting discrimination model for generating distribution network samples containing distributed photovoltaics, The distribution of random noise vector z of the distribution network containing distributed photovoltaics is p z (z).

[0030] The intelligent setting and discrimination model of distribution network parameters maximizes the discrimination accuracy of the real data of the distribution network containing distributed photovoltaics, and its loss function is expressed as:

[0031]

[0032] Among them, L D is the loss function of the distribution network parameter intelligent setting discrimination model; D(G(z)) represents the evaluation probability of the distribution network parameter intelligent setting discrimination model for generating distribution network samples containing distributed photovoltaics; For x, the distribution is p data (x); p data (x) is the distribution of x; D(x) is the evaluation probability of the actual input distribution network sample x containing distributed photovoltaics; x is the actual input of the distribution network sample containing distributed photovoltaics.

[0033] Step S3 designs the intelligent evaluation index system for the intelligent setting effect of distribution network parameters, including voltage deviation, power loss and system stability.

[0034] Voltage deviation indicates the degree of deviation of voltage at sample nodes of the distribution network containing distributed photovoltaics, which is defined as:

[0035]

[0036] Among them, V i The voltage of the ith node, V ref is the reference voltage, N is the total number of nodes;

[0037] Power loss refers to the active power loss in the distribution network sample containing distributed photovoltaics, and its calculation formula is:

[0038]

[0039] Among them, R ij is the resistance of the line between nodes i and j, I ij is the current flowing through the line, E is the set of lines in the distribution network;

[0040] System stability measures the stability of the system by evaluating the node voltage fluctuations, which is defined as:

[0041]

[0042] in, is the average value of the node voltage.

[0043] The present invention has the following beneficial effects:

[0044] The present invention introduces a generative adversarial network algorithm to perform intelligent parameter optimization for a distribution network containing a large number of distributed photovoltaics. This method can more accurately simulate and predict the operating status of the distribution network under different photovoltaic access conditions, provide more accurate and dynamic parameter adjustment, and thus significantly improve the responsiveness and overall efficiency of the power grid. By optimizing parameter setting through intelligent methods, not only can the adaptability of the distribution network to renewable energy fluctuations be enhanced, but also the economic operation efficiency of the power grid and the quality of power supply to users can be improved. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] The present invention will be further described and explained below in conjunction with the accompanying drawings.

[0046] Figure 1 This is a comparison diagram before and after parameter setting of the distribution network containing distributed photovoltaics proposed by the present invention.

[0047] Figure 2 It is the convergence process of solving the distribution network parameter intelligent setting model based on the generative adversarial network algorithm. DETAILED DESCRIPTION

[0048] The technical solution of the present invention will be more clearly and completely explained below through description of preferred embodiments of the present invention in combination with the accompanying drawings.

[0049] A method for intelligently setting and calculating distribution network parameters considering distributed photovoltaic access includes the following steps:

[0050] S1: Collect distribution network operation data and distributed photovoltaic power generation data to establish a distribution network operation scenario dataset considering distributed photovoltaic access.

[0051] Specifically, the following steps are included:

[0052] S101: Collect distribution network operation data and distributed photovoltaic power generation data.

[0053] The main data types include: load data: the load conditions of each node in the distribution network; photovoltaic power generation data: recording the power generation of distributed photovoltaic systems at different times; environmental data: mainly involving weather data, such as solar radiation, temperature and other factors affecting photovoltaic power generation.

[0054] S102: Distributed photovoltaic power generation data preprocessing, standardizing the collected data.

[0055] The standardization process in step S102 is as follows:

[0056]

[0057] Among them, X(t) is the original data of distributed photovoltaic, μ X ,σ X They represent the mean and standard deviation of distributed photovoltaic power generation data respectively, and t represents the current moment.

[0058] S103: extracting characteristic variables of distributed photovoltaic power generation data, and extracting characteristics such as daily average value and fluctuation range based on the time series photovoltaic power generation data;

[0059] The daily average value of photovoltaic power generation data in step S103 is:

[0060]

[0061] in, is the daily average value of photovoltaic power generation data; P PV is photovoltaic power generation; T is the length of time; t represents the current moment.

[0062] The fluctuation range of photovoltaic power generation data is:

[0063] ΔP PV =max(P PV (t))-min(P PV (t))

[0064] Where ΔP PV is the fluctuation range of photovoltaic power generation data; P PV is photovoltaic power generation; t represents the current moment.

[0065] S104: Generate distribution network operation scenario data.

[0066] In step S104, the distribution network operation scenario data is formed, specifically:

[0067] D={(L i (t),P PV (t),W(t))|t=1,2,...,T;i=1,2,...,N}

[0068] Where N is the number of distribution network nodes; P PV is photovoltaic power generation; t represents the current moment; L i (t) is the voltage of the ith node; W(t) is the set of distribution network switch states at time t.

[0069] S2: Based on the scenario data set, an intelligent setting model of distribution network parameters based on the generative adversarial network algorithm is established.

[0070] Step S2 designs a distribution network parameter intelligent setting model based on a generative adversarial network algorithm, including a distribution network parameter intelligent setting generator and a distribution network parameter intelligent setting discrimination model.

[0071] The task of the distribution network parameter intelligent setting generator is to generate samples similar to the real distribution network parameters from the noise distribution. The output of the distribution network parameter intelligent setting generator can be expressed as G(z), where z is the random noise vector of the distribution network containing distributed photovoltaics, which obeys the standard normal distribution. The distribution network parameter intelligent setting generator is trained by optimizing the following objective function:

[0072]

[0073] Among them, L G is the loss function of the distribution network parameter intelligent setting generator; D(G(z)) represents the evaluation probability of the distribution network parameter intelligent setting discrimination model for generating distribution network samples containing distributed photovoltaics; The distribution of random noise vector z of the distribution network containing distributed photovoltaics is p z (z); p z (z) is the distribution of z.

[0074] The task of the distribution network parameter intelligent setting discrimination model is to distinguish the parameters generated by the distribution network parameter intelligent setting generator containing distributed photovoltaics from the real parameters. The output of the distribution network parameter intelligent setting discrimination model is a scalar D(x), which represents the probability that the input parameter x comes from the real data of the distribution network containing distributed photovoltaics. The distribution network parameter intelligent setting discrimination model improves the quality of the distribution network samples containing distributed photovoltaics generated by the generator by maximizing the accuracy of the discrimination of the real data of the distribution network containing distributed photovoltaics. Its loss function is expressed as:

[0075]

[0076] Among them, L Dis the loss function of the distribution network parameter intelligent setting discrimination model; D(G(z)) represents the evaluation probability of the distribution network parameter intelligent setting discrimination model for generating distribution network samples containing distributed photovoltaics; For x, the distribution is p data (x); p data (x) is the distribution of x; D(x) is the evaluation probability of the actual input distribution network sample x containing distributed photovoltaics; x is the actual input of the distribution network sample containing distributed photovoltaics.

[0077] The distribution network parameter intelligent setting generator and the distribution network parameter intelligent setting discrimination model minimize the adversarial loss function through the adversarial process. When the distribution network parameter intelligent setting generator and the distribution network parameter intelligent setting discrimination model reach Nash equilibrium, the distribution network sample containing distributed photovoltaics generated by the distribution network parameter intelligent setting generator is realistic enough, so that the distribution network parameter intelligent setting discrimination model cannot distinguish true from false. At this time, it is considered that the distribution network parameters containing distributed photovoltaics generated by the distribution network parameter intelligent setting generator are in line with the actual situation of the distribution network.

[0078] S3: Based on the intelligent setting model, output the intelligent evaluation indicators of the intelligent setting effect of distribution network parameters, and the evaluation indicators include voltage deviation, power loss and system stability.

[0079] Step S3 designs an intelligent evaluation index system for the intelligent setting effect of distribution network parameters, including voltage deviation, power loss and system stability.

[0080] Voltage deviation indicates the degree of deviation of voltage at sample nodes of the distribution network containing distributed photovoltaics, which is defined as:

[0081]

[0082] Among them, V i The voltage of the ith node, V ref is the reference voltage, N is the total number of nodes;

[0083] Power loss refers to the active power loss in the distribution network sample containing distributed photovoltaics, and its calculation formula is:

[0084]

[0085] Among them, R ij is the resistance of the line between nodes i and j, I ij is the current flowing through the line, E is the set of lines in the distribution network;

[0086] System stability measures the stability of the system by evaluating the node voltage fluctuations, which is defined as:

[0087]

[0088] in, is the average value of the node voltage.

[0089] In order to verify the effectiveness of the present invention, a distribution network area was selected for experimental verification. The area contains 10 main nodes and several distributed photovoltaic power stations, and the photovoltaic capacity accounts for 35% of the total load. The operating data of the area includes the load data of the nodes, photovoltaic power generation data and related environmental factors. The number of iterations of the generative adversarial network is 10,000.

[0090] Figure 1 The comparison diagram before and after the parameter setting of the distribution network containing distributed photovoltaics is shown. Among them, the blue is the calculation result of the parameter setting of the distribution network containing distributed photovoltaics based on the traditional method; the orange is the calculation result of the intelligent setting of the distribution network parameters containing distributed photovoltaics based on the present invention. It can be seen from the figure that the errors of the voltage deviation (Voltage Deviation) and power loss (Power Loss) of the calculation result of the intelligent setting of the distribution network parameters containing distributed photovoltaics based on the present invention are smaller, that is, the distribution network parameters set by the present invention can be closer to the actual operating state of the actual distribution network.

[0091] Figure 2 The convergence process of solving the distribution network parameter intelligent setting model based on the generative adversarial network algorithm is demonstrated. It can be seen that the Loss of the distribution network parameter intelligent setting generator and the Loss of the distribution network parameter intelligent setting discriminant model can converge synchronously, which shows that the proposed method can iterate the distribution network parameter intelligent setting model to the optimal solution through real-time joint game behavior.

[0092] It will be apparent to those skilled in the art that the invention is not limited to the details of the exemplary embodiments described above and that the invention can be implemented in other specific forms without departing from the spirit or essential features of the invention. Therefore, the embodiments should be considered exemplary and non-limiting in all respects, and the scope of the invention is defined by the appended claims rather than the foregoing description, and it is intended that all variations falling within the meaning and scope of the equivalent elements of the claims be included in the invention. Any reference numeral in a claim should not be considered as limiting the claim to which it relates.

Claims

1. A method for intelligent setting and calculation of distribution network parameters considering distributed photovoltaic access, characterized in that: The following steps are involved: S1: Collect distribution network operation data and distributed photovoltaic power generation data to establish a distribution network operation scenario dataset considering distributed photovoltaic access; S2: According to the scenario data set, an intelligent setting model of distribution network parameters based on the generative adversarial network algorithm is established; S3: Based on the intelligent setting model, output intelligent evaluation indicators of the intelligent setting effect of distribution network parameters, and the evaluation indicators include voltage deviation, power loss and system stability.

2. According to claim 1, a method for intelligent setting and calculation of distribution network parameters considering distributed photovoltaic access is characterized in that: The specific steps of step S1 to establish a distribution network operation scenario data set considering distributed photovoltaic access include: S101: Collect distribution network operation data and distributed photovoltaic power generation data; S102: Distributed photovoltaic power generation data preprocessing, standardizing the collected data; S103: extracting characteristic variables of distributed photovoltaic power generation data, and extracting characteristics such as daily average value and fluctuation range based on the time series photovoltaic power generation data; S104: Generate distribution network operation scenario data.

3. The method for intelligently setting and calculating distribution network parameters considering distributed photovoltaic access according to claim 2 is characterized in that: The standardization process in step S102 is as follows: Among them, X(t) is the original data of distributed photovoltaic, μ X ,σ X They represent the mean and standard deviation of distributed photovoltaic power generation data respectively, and t represents the current moment.

4. The method for intelligently setting and calculating distribution network parameters considering distributed photovoltaic access according to claim 2 is characterized in that: The daily average value of photovoltaic power generation data in step S103 is: in, is the daily average value of photovoltaic power generation data; P PV is photovoltaic power generation; T is the length of time; t represents the current moment.

5. The method for intelligently setting and calculating distribution network parameters considering distributed photovoltaic access according to claim 2 is characterized in that: The fluctuation range of photovoltaic power generation data in step S103 is: ΔP PV =max(P PV (t))-min(P PV (t)) Among them, ΔP PV is the fluctuation range of photovoltaic power generation data; P PV is photovoltaic power generation; t represents the current moment.

6. The method for intelligently setting and calculating distribution network parameters considering distributed photovoltaic access according to claim 1, characterized in that: The step S104 forms the distribution network operation scenario data, specifically: D={(L i (t),P PV (t),W(t))∣t=1,2,...,T;i=1,2,...,N} Where N is the number of distribution network nodes; P PV is photovoltaic power generation; t represents the current moment; L i (t) is the voltage of the ith node; W(t) is the set of distribution network switch states at time t.

7. The method for intelligently setting and calculating distribution network parameters considering distributed photovoltaic access according to claim 1, characterized in that: The step S2 designs a distribution network parameter intelligent setting model based on a generative adversarial network algorithm, including a distribution network parameter intelligent setting generator and a distribution network parameter intelligent setting discrimination model.

8. The method for intelligently setting and calculating distribution network parameters considering distributed photovoltaic access according to claim 7 is characterized in that: The distribution network parameter intelligent setting generator is trained by optimizing the following objective function: Among them, L G is the loss function of the distribution network parameter intelligent setting generator; G(z) is the output of the distribution network parameter intelligent setting generator; z is the random noise vector of the distribution network containing distributed photovoltaics; D(G(z)) represents the evaluation probability of the distribution network parameter intelligent setting discriminant model for generating distribution network samples containing distributed photovoltaics; The distribution of random noise vector z of the distribution network containing distributed photovoltaics is p z (z); p z (z) is the distribution of z.

9. The method for intelligently setting and calculating distribution network parameters considering distributed photovoltaic access according to claim 7, characterized in that: The distribution network parameter intelligent setting discrimination model maximizes the discrimination accuracy of the real data of the distribution network containing distributed photovoltaics, and its loss function is expressed as: Among them, L D is the loss function of the distribution network parameter intelligent setting discrimination model; D(G(z)) represents the evaluation probability of the distribution network parameter intelligent setting discrimination model for generating distribution network samples containing distributed photovoltaics; For x, the distribution is p data (x); p data (x) is the distribution of x; D(x) is the evaluation probability of the actual input distribution network sample x containing distributed photovoltaics; x is the actual input of the distribution network sample containing distributed photovoltaics.

10. The method for intelligently setting and calculating distribution network parameters considering distributed photovoltaic access according to claim 1, characterized in that: The step S3 designs an intelligent evaluation index system for the intelligent setting effect of distribution network parameters, including voltage deviation, power loss and system stability. The voltage deviation represents the degree of deviation of the voltage of the sample node of the distribution network containing distributed photovoltaics, which is defined as: Among them, V i The voltage of the ith node, V ref is the reference voltage, N is the total number of nodes; The power loss represents the active power loss in the distribution network sample containing distributed photovoltaics, and its calculation formula is: Among them, R ij is the resistance of the line between nodes i and j, I ij is the current flowing through the line, E is the set of lines in the distribution network; The system stability is measured by evaluating the node voltage fluctuation, which is specifically defined as: in, is the average value of the node voltage.

Images