Distribution network side optical storage network load collaborative planning method and system based on massive scene dimension reduction

Abstract

The invention relates to the technical field of photovoltaic energy storage resource planning, and provides a distribution network side optical storage network load collaborative planning method and system based on massive scene dimensionality reduction, and the method comprises the following steps: obtaining equipment parameters and historical scene data of a photovoltaic power supply, a local load and energy storage of a target region, and constructing an optical storage network load scene database of the target region; performing dimension reduction processing on the optical storage network load scene data through a principal component analysis method, and performing clustering through a Gaussian hybrid clustering algorithm to obtain a dimension-reduced optical storage network load scene set; constructing an optical storage network load collaborative planning model by taking the minimum total planning cost as a target; and solving the optical storage network load collaborative planning model under the consideration of the optical storage network load constraint, and generating a planning result. The method is used for solving the matching problem between the mass operation data of the photovoltaic power supply in the distribution network and the load storage collaborative planning of the distribution network side optical storage network, can effectively reduce the computing resources, and improves the computing speed and the planning efficiency.

Description

Technical Field

[0001] The present invention relates to the technical field of photovoltaic energy storage resource planning, and more specifically, to a distribution network-side photovoltaic energy storage network-load collaborative planning method and system based on massive scene dimensionality reduction. Background Art

[0002] In order to reduce dependence on fossil energy, electrochemical energy storage and distributed photovoltaic power sources are now widely used on the distribution network side. Photovoltaic storage grid load, also known as "source grid load storage integration", is a new energy system operation mode that covers four aspects: power supply, power grid, load and energy storage. This mode coordinates and optimizes the operation of power supply, power grid, load and energy storage to achieve efficient energy utilization, safe and stable operation of the power grid and large-scale consumption of renewable energy, thereby promoting energy transformation and sustainable development.

[0003] Since photovoltaic power generation is random, it has a great impact on the results of distribution network planning. Therefore, when planning the coordinated planning of photovoltaic, storage, network and load, it is necessary to consider the mutual influence between various devices at the same time to maximize the utilization of resources. At present, a multi-objective coordinated planning method for source, network, load and storage of distribution network has been proposed. It is proposed to dynamically reconstruct the distribution network in different time periods based on the load status on both sides of the source and load of the distribution network to obtain the demand response prediction probability, and then optimize the control strategy in real time based on the demand response prediction probability. However, with the increase in the access proportion of distributed power sources and the improvement of the degree of automation of the power system, the amount of data in the distribution network has exploded, resulting in higher time resolution and larger data dimensions for data collection, which in turn leads to the existence of massive high-dimensional scene data in the photovoltaic, storage, network and load database. When planning the coordinated planning of photovoltaic, storage, network and load, a large amount of computing resources are required, and there are problems such as slow computing speed and low planning efficiency. Summary of the invention

[0004] In order to overcome the defects of slow calculation speed and low planning efficiency caused by the presence of massive high-dimensional scene data in the coordinated planning of photovoltaic, storage, network and load in the above-mentioned prior art, the present invention provides a distribution network-side coordinated planning method and system for photovoltaic, storage, network and load based on massive scene dimensionality reduction.

[0005] In order to solve the above technical problems, the technical solution of the present invention is as follows:

[0006] A distribution network-side photovoltaic storage network-load coordination planning method based on massive scene dimensionality reduction includes the following steps:

[0007] Obtain equipment parameters and historical scenario data of photovoltaic power sources, local loads and energy storage in the target area, and build a photovoltaic, storage, grid and load scenario database for the target area;

[0008] The dimensionality reduction of the photovoltaic, storage, grid and load scenario data was performed by principal component analysis, and then clustered by Gaussian mixture clustering algorithm to obtain a dimensionality-reduced photovoltaic, storage, grid and load scenario set.

[0009] With the goal of minimizing the total planning cost, a synergistic planning model of photovoltaic storage, grid and load is constructed;

[0010] The photovoltaic-storage-grid-load collaborative planning model is solved under the photovoltaic-storage-grid-load constraints to generate a planning result.

[0011] Furthermore, the present invention also proposes a distribution network side photovoltaic storage network load collaborative planning system based on massive scene dimensionality reduction, and applies the distribution network side photovoltaic storage network load collaborative planning method proposed in the present invention. The system includes:

[0012] The data processing module is used to obtain the equipment parameters and historical scenario data of the photovoltaic power source, local load and energy storage in the target area, and build a photovoltaic storage grid-load scenario database for the target area;

[0013] The dimension reduction processing module is used to perform dimension reduction processing on the PV-storage-grid-load scenario data by using the principal component analysis method, and then clustering it by using the Gaussian mixture clustering algorithm to obtain a reduced dimension PV-storage-grid-load scenario set;

[0014] The planning module is equipped with a photovoltaic storage network load collaborative planning model constructed with the goal of minimizing the total planning cost, and is used to solve the photovoltaic storage network load collaborative planning model under the consideration of photovoltaic storage network load constraints and output the planning results.

[0015] Furthermore, the present invention also proposes a device, including a memory and a processor, wherein the memory stores computer-readable instructions, wherein when the computer-readable instructions are executed by the processor, the processor executes all or part of the steps of the distribution network side photovoltaic storage network load collaborative planning method as described in the present invention.

[0016] Furthermore, the present invention also proposes a storage medium on which computer-readable instructions are stored, wherein when the computer-readable instructions are executed by a processor, all or part of the steps of the distribution network side photovoltaic storage network load collaborative planning method as described in the present invention are implemented.

[0017] Compared with the prior art, the technical solution of the present invention has the following beneficial effects:

[0018] The present invention is used to solve the coordination problem between the massive operating data of photovoltaic power sources in the distribution network and the photovoltaic-storage-grid-load collaborative planning on the distribution network side. The principal component-Gaussian mixture clustering algorithm is used to reduce the dimension of the massive high-dimensional scene data of photovoltaic-storage-grid-load, effectively reducing the data volume and data dimension. The photovoltaic-storage-grid-load collaborative planning model based on the minimum total planning cost is further used for efficient solution, which can effectively reduce computing resources and improve computing speed and planning efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] Figure 1The present invention is a flow chart of a method for planning the coordination of photovoltaic power generation, storage network and load at the distribution network side according to an embodiment of the present invention.

[0020] Figure 2 The present invention is an architectural diagram of a distribution network-side photovoltaic-storage-grid-load collaborative planning system according to an embodiment of the present invention. DETAILED DESCRIPTION

[0021] Exemplary embodiments will be described in detail herein, examples of which are shown in the accompanying drawings. When the following description refers to the drawings, the same numbers in different drawings represent the same or similar elements unless otherwise indicated. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with the present invention. Instead, they are merely examples of devices and methods consistent with some aspects of the present invention as detailed in the appended claims.

[0022] The terms used in the present invention are only for the purpose of describing specific embodiments and are not intended to limit the present invention. The singular forms "a", "the" and "the" used in the present invention and the appended claims are also intended to include plural forms unless the context clearly indicates other meanings. It should also be understood that the term "and / or" used herein refers to and includes any or all possible combinations of one or more associated listed items.

[0023] It should be understood that although the terms first, second, third, etc. may be used in the present invention to describe various information, these information should not be limited to these terms. These terms are only used to distinguish the same type of information from each other. For example, without departing from the scope of the present invention, the first information may also be referred to as the second information, and similarly, the second information may also be referred to as the first information. Depending on the context, the word "if" as used herein may be interpreted as "at the time of" or "when" or "in response to determining".

[0024] The present invention is described in detail below with reference to the accompanying drawings and specific embodiments.

[0025] Example 1

[0026] This embodiment proposes a distribution network-side photovoltaic storage network-load coordination planning method based on massive scene dimensionality reduction, such as Figure 1 As shown, it is a flow chart of the distribution network side photovoltaic storage network load coordinated planning method of this embodiment.

[0027] The distribution network side photovoltaic storage network load coordination planning method proposed in this embodiment includes the following steps:

[0028] S1. Obtain equipment parameters and historical scenario data of photovoltaic power sources, local loads and energy storage in the target area, and build a photovoltaic, storage, grid and load scenario database for the target area;

[0029] S2. Perform dimensionality reduction processing on the PV-storage-grid-load scenario data by principal component analysis, and then perform clustering by Gaussian mixture clustering algorithm to obtain a dimensionality-reduced PV-storage-grid-load scenario set;

[0030] S3. With the goal of minimizing the total planning cost, a synergistic planning model for photovoltaic storage, grid and load is constructed;

[0031] S4. Solve the PV-storage-grid-load collaborative planning model under the PV-storage-grid-load constraints to generate a planning result.

[0032] This embodiment is used to solve the coordination problem between the massive operating data of photovoltaic power sources in the distribution network and the photovoltaic-storage-grid-load collaborative planning on the distribution network side. The principal component-Gaussian mixture clustering algorithm is used to reduce the dimension of the massive high-dimensional scene data of photovoltaic-storage-grid-load, and further use the photovoltaic-storage-grid-load collaborative planning model based on the minimum total planning cost for efficient solution, which can effectively reduce computing resources and improve computing speed and planning efficiency.

[0033] In an optional embodiment, in step S2, the dimensionality reduction processing of the photovoltaic energy storage grid load scenario data is performed by principal component analysis method, including the following steps:

[0034] For the N T-dimensional original photovoltaic output vectors in the photovoltaic storage grid-load scenario data, they are normalized as follows:

[0035]

[0036]

[0037]

[0038] Among them, S nt represents the original photovoltaic power output vector of the nth t dimension, represents the normalized photovoltaic output vector; is the nth original photovoltaic output vector S n The first moment of n is the nth original photovoltaic output vector S n The second moment of

[0039] Calculate the covariance coefficient after data normalization and construct the covariance matrix R; its expression is:

[0040]

[0041] Among them, r ij represents the normalized ith and jth photovoltaic output vectors The covariance between

[0042] Based on the eigenvalue λ of the covariance matrix R, a linear transformation is performed on its orthogonal matrix U, and the original photovoltaic output vector is mapped to the principal component space using the orthogonal matrix U to generate the principal component matrix Z; its expression is:

[0043] λ=U T RU=diag(λ1,λ2,…,λ N )

[0044] Z=U T S

[0045] Among them, λ n is the nth eigenvalue of the covariance matrix R, n=1, 2, ..., N; S is the original photovoltaic output vector set.

[0046] In this embodiment, principal component analysis (PCA) is a commonly used data dimensionality reduction technology, which converts high-dimensional data into low-dimensional data through linear transformation, while retaining the information of the original data as much as possible. In this embodiment, the optical storage grid load scenario data is reduced in dimension through PCA technology, which can reduce the data dimension while maintaining as much variance as possible, thereby achieving dimensionality reduction processing of high-dimensional data.

[0047] Further, in an optional embodiment, clustering is performed by using a Gaussian mixture clustering algorithm in step S2, including:

[0048] Calculate the weight φ of the PV-storage-grid-load scenario data after dimensionality reduction w , mean μ w and covariance D w , and construct the posterior probability γ(z nw ); its expression is:

[0049]

[0050] Among them, z nw Represents data point X n Whether it belongs to the wth Gaussian distribution; x w represents the w-th Gaussian distributed random variable; the subscript w represents the number of Gaussian distributions;

[0051] Based on the posterior probability γ(z nw ), for the weight φ w , mean μ w and covariance D w Update; its expression is:

[0052]

[0053] in, Respectively represent the updated weight, mean and covariance; X n represents the nth sample data point; ω is the intermediate variable;

[0054] Construct the log-likelihood function of the Gaussian mixture model, and check whether the parameters or log-likelihood function converge. If not, re-cluster until convergence; if yes, output the clustering result to obtain a reduced-dimensional optical-storage-grid-load scenario set; the expression of the log-likelihood function of the Gaussian mixture model is:

[0055]

[0056] Where X is the observed data point; μ is the mean vector of each Gaussian distribution; D represents the covariance matrix of each Gaussian distribution; φ represents the weight of each Gaussian distribution; Gw(·) represents the wth Gaussian distribution function.

[0057] This embodiment further combines the Gaussian mixture clustering algorithm to cluster the PV-storage-grid-load scenario data that has been processed by the principal component analysis method, wherein the Gaussian mixed model (GMM) is composed of a linear combination of W Gaussian distributions, and the main parameters are the random variable x, weight φ, mean μ, and covariance D. GMM uses the expectation-maximization (EM) algorithm to fit the mixed Gaussian distribution, and each iteration of the EM algorithm is divided into two steps: the E-step and the M-step.

[0058] Before applying the EM algorithm, it is necessary to estimate the maximum likelihood estimator of the parameters, taking the mean μ of the w-th Gaussian distribution as w For example, using G w Represents the w-th Gaussian distribution function, takes the logarithm of its general form, and then w Take the derivative and set it to zero, that is:

[0059]

[0060] Furthermore, in an optional embodiment, step S2 further includes the following steps:

[0061] Based on the dispersion B and the intra-class concentration L, a comprehensive clustering index is used to evaluate the quality of the clustering results; its expression is:

[0062]

[0063] Among them, R is the number of clustering results; L r is the inter-class concentration of the r-th clustering result; F is the number of all scenes; E r is the number of scenes contained in the rth cluster.

[0064] The larger the comprehensive clustering index O(R), the more compact the scenes in each category are, the clearer the boundaries between classes are, and the better the clustering effect is.

[0065] In this embodiment, the Calinski-Harabasz index (CH index) is selected to evaluate the clustering effect, which measures the ratio of the dispersion between clusters to the concentration within the clusters.

[0066] Further optionally, the current clustering effect is evaluated based on the comprehensive clustering index O(R). If the comprehensive clustering index O(R) is greater than a preset threshold, it means that the current clustering effect meets expectations, and step S3 is executed; otherwise, it means that the current clustering effect does not meet expectations, and step S2 is re-executed.

[0067] In an optional embodiment, the photovoltaic storage grid-load collaborative planning model includes an objective function with the goal of minimizing the total planning cost; its expression is:

[0068] m i n C=C I +C O

[0069]

[0070]

[0071] Among them, C I is the investment cost; b is the discount rate, y is the useful life of the equipment, c sub 、c PVG 、c ESS are the unit capacity investment costs of substation, photovoltaic and energy storage, respectively. is the planned capacity of the i-th substation, are the installed capacity of photovoltaic and energy storage at node i, respectively, sub ,Ω PVG ,Ω ESS are the collection of substation, photovoltaic and energy storage nodes to be installed; c line is the investment cost per unit length of line, l ij is the length of line ij, Ω line C is the collection of lines to be constructed; O is the operating cost; s is the number of days in quarter s; Λ PVG is the set of nodes connected to photovoltaic power generation. The operation and maintenance fee for each unit of electricity generated by photovoltaic power. is the active power of photovoltaic output at node i in period t of quarter s; sub Assemble for substation, is the time-of-use electricity price in period t, is the amount of electricity purchased by the substation during period s in the quarter.

[0072] In this embodiment, a photovoltaic, energy storage, grid and load collaborative planning model is constructed by taking into account the minimum investment cost and operating cost, so as to achieve coordinated optimization of the operation of power supply, grid, load and energy storage.

[0073] Furthermore, in an optional embodiment, the photovoltaic storage grid load constraints include substation planning capacity constraints, new energy and energy storage single-node installation capacity constraints, new energy and energy storage maximum installation capacity constraints, flow constraints, system safety constraints, photovoltaic operation constraints and / or energy storage operation constraints.

[0074] The substation planning capacity constraint is expressed as:

[0075]

[0076] in, They represent the minimum planned capacity and maximum planned capacity of the i-th substation respectively; K is a positive integer; is the transformer capacity of the i-th substation.

[0077] The single-node installation capacity constraint of new energy and energy storage is expressed as:

[0078]

[0079] in, is the maximum installed capacity of photovoltaic power, The maximum installed capacity of energy storage.

[0080] The maximum installed capacity constraint of new energy and energy storage is expressed as:

[0081]

[0082] in It is the maximum installed capacity of new energy.

[0083] The power flow constraint is expressed as:

[0084]

[0085]

[0086]

[0087]

[0088] Among them, u(j) is the node set with j as the end node, w(j) is the node set with j as the head node; x ij Indicates whether line ij is constructed; Pij，t , P jk，t represents the active power of line ij and line jk in period t, R ij is the resistance of line ij, I ij，t represents the circuit of line ij during period t; P j，t is the active power flowing into node j during period t; Q ij，t , Q jk，t represents the reactive power of line ij and line jk during period t, X ij is the reactance of line ij, Q j，t is the reactive power flowing into node j during period t; U j，t is the voltage amplitude of node j in period t, It represents the square of the impedance modulus, which indicates the total resistance to the flow of current; Indicates whether photovoltaic and energy storage are constructed at node j; represents the active power and reactive power of photovoltaic output of node j in period t, represents the active power output of energy storage at node j during period t; represents the active power and reactive power of the load at node j.

[0089] The system safety constraints are expressed as:

[0090]

[0091] Among them, U i，min , U i，max Respectively represent the lower and upper limits of the voltage amplitude of node i; I ij，max Represents the maximum current of line ij in time period t.

[0092] The photovoltaic operation constraint is expressed as:

[0093]

[0094] in, represents the maximum output power of the photovoltaic at node i during period t, is the photovoltaic power factor angle.

[0095] The energy storage operation constraint is expressed as:

[0096]

[0097] in, They represent the charging power and discharging power of the energy storage at node i in time period t respectively; i，t is the charging and discharging state of energy storage at node i during time period t, is the number of energy storage modules installed at node i, P ESS is the rated capacity of a single energy storage module; represents the energy storage charge at node i in period t, and Δt is the period between adjacent operating times; They represent the charging efficiency and discharging efficiency of energy storage at node i respectively; S i，min , S i，max They represent the lower and upper limits of the state of charge of the energy storage module at node i respectively.

[0098] Considering the photovoltaic-storage-grid-load coordinated planning model under photovoltaic-storage-grid-load constraints, there are non-convex constraints and it cannot be solved directly.

[0099] Further, in an optional embodiment, solving the photovoltaic-storage-grid-load collaborative planning model considering the photovoltaic-storage-grid-load constraints includes:

[0100] Variable substitution and big-M method are used to transform the non-convex constraints in the model into convex ones, where:

[0101] make Then the power flow constraint, operation safety constraint and energy storage constraint are transformed into:

[0102] ||[2P ij，t 2Q ij，t (l ij，t -u i，t )] T ||2≤l ij，t +u i，t

[0103]

[0104]

[0105]

[0106] in, represents the square of the voltage amplitude of node i in time period t; x ij is the reactance of line ij; M is the penalty factor coefficient;

[0107] The Gurobi solver is used to solve the objective function after convex processing transformation to obtain the planning result.

[0108] In this embodiment, variable substitution, big-M method and second-order cone relaxation technology are used to convert the non-convex constraints in the model into convex ones, and then the Gurobi solver is used to solve the optimal solution to obtain the photovoltaic storage network load coordination planning result that meets the minimum total planning cost.

[0109] Example 2

[0110] This embodiment applies the distribution network side photovoltaic storage network load collaborative planning method proposed in Example 1, and proposes a distribution network side photovoltaic storage network load collaborative planning system based on massive scene dimensionality reduction. Figure 2 As shown, it is the architecture diagram of the distribution network side photovoltaic storage network load collaborative planning system of this embodiment.

[0111] The distribution network side photovoltaic storage network load coordination planning system proposed in this embodiment includes:

[0112] The data processing module is used to obtain the equipment parameters and historical scenario data of the photovoltaic power source, local load and energy storage in the target area, and build a photovoltaic storage grid-load scenario database for the target area;

[0113] The dimension reduction processing module is used to perform dimension reduction processing on the PV-storage-grid-load scenario data by using the principal component analysis method, and then clustering it by using the Gaussian mixture clustering algorithm to obtain a reduced dimension PV-storage-grid-load scenario set;

[0114] The planning module is equipped with a photovoltaic storage network load collaborative planning model constructed with the goal of minimizing the total planning cost, and is used to solve the photovoltaic storage network load collaborative planning model under the consideration of photovoltaic storage network load constraints and output the planning results.

[0115] It can be understood that the system of this embodiment corresponds to the method of the above-mentioned embodiment 1, and the options in the above-mentioned embodiment 1 are also applicable to this embodiment, so they will not be described repeatedly here.

[0116] Example 3

[0117] This embodiment proposes a computer device, including a memory and a processor, wherein the memory stores computer-readable instructions, and when the computer-readable instructions are executed by the processor, the processor executes all or part of the steps of the distribution network side photovoltaic storage network load collaborative planning method proposed in Example 1.

[0118] Example 4

[0119] This embodiment proposes a storage medium on which computer-readable instructions are stored, wherein the computer-readable instructions, when executed by a processor, implement all or part of the steps of the distribution network-side photovoltaic-storage-grid-load collaborative planning method proposed in Embodiment 1.

[0120] Exemplarily, the storage medium includes, but is not limited to, a USB flash drive, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk, and other media that can store program codes.

[0121] Exemplarily, the instructions, programs, code sets or instruction sets may be implemented using conventional programming languages.

[0122] Exemplarily, the processor includes but is not limited to a smart phone, a personal computer, a server, a network device, etc., and is used to execute all or part of the steps of the distribution network side photovoltaic storage network load collaborative planning method described in Example 1.

[0123] Each embodiment of the present invention is described in a progressive manner, and the same or similar parts between the embodiments can be referred to each other, and each embodiment focuses on the differences from other embodiments. In particular, for the device embodiment, since it is basically similar to the method embodiment, the description is relatively simple, and the relevant parts can be referred to the partial description of the method embodiment. The device embodiment described above is merely exemplary, in which the modules described as separate components may or may not be physically separated, and the functions of each module can be implemented in the same or one or more software and / or hardware when implementing the scheme of the present invention. It is also possible to select some or all of the modules according to actual needs to achieve the purpose of the scheme of this embodiment.

[0124] Obviously, the above embodiments of the present invention are merely examples for clearly illustrating the present invention, and are not intended to limit the embodiments of the present invention. For those skilled in the art, other different forms of changes or modifications can be made based on the above description. It is not necessary and impossible to list all the embodiments here. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present invention should be included in the protection scope of the claims of the present invention.

Claims

1. A distribution network-side photovoltaic storage network-load collaborative planning method based on massive scene dimensionality reduction, characterized in that: The following steps are involved: Obtain equipment parameters and historical scenario data of photovoltaic power sources, local loads and energy storage in the target area, and build a photovoltaic, storage, grid and load scenario database for the target area; The dimensionality reduction of the photovoltaic, storage, grid and load scenario data was performed by principal component analysis, and then clustered by Gaussian mixture clustering algorithm to obtain a reduced-dimensional photovoltaic, storage, grid and load scenario set. With the goal of minimizing the total planning cost, a synergistic planning model of photovoltaic, energy storage, grid and load is constructed; The photovoltaic-storage-grid-load collaborative planning model is solved under the photovoltaic-storage-grid-load constraints to generate a planning result.

2. The method for coordinated planning of light-storage-grid-load at the distribution network side according to claim 1 is characterized in that: The principal component analysis method is used to reduce the dimension of the photovoltaic storage grid-load scenario data, including the following steps: For the N T-dimensional original photovoltaic output vectors in the photovoltaic storage grid-load scenario data, they are normalized as follows: Among them, S Nt represents the original photovoltaic power output vector of the nth t dimension, represents the normalized photovoltaic output vector; is the nth original photovoltaic power output vector S n The first moment of M n is the nth original photovoltaic output vector S n The second moment of Calculate the covariance coefficient after data normalization and construct the covariance matrix R; its expression is: Among them, r ij represents the normalized ith and jth photovoltaic output vectors The covariance between Based on the eigenvalue λ of the covariance matrix R, a linear transformation is performed on its orthogonal matrix U, and the original photovoltaic output vector is mapped to the principal component space using the orthogonal matrix U to generate the principal component matrix Z; its expression is: λ=U T RU=diag(λ1,λ2,…,λ N ) Z=U T S Among them, λ n is the nth eigenvalue of the covariance matrix R, n=1,2,...,N; S is the set of original photovoltaic output vectors.

3. The method for planning the coordinated photovoltaic, storage, and load distribution network side according to claim 1 is characterized in that: Clustering is performed using the Gaussian mixture clustering algorithm, including: Calculate the weight φ of the PV-storage-grid-load scenario data after dimensionality reduction w , mean μ w and covariance D w , and construct the posterior probability γ(z nw ); its expression is: Among them, z nw Represents data point X n Whether it belongs to the wth Gaussian distribution; x w represents the w-th Gaussian distributed random variable; the subscript w represents the w-th Gaussian distribution; Based on the posterior probability γ(z nw ), for the weight φ w , mean μ w and covariance D w Update; its expression is: in, Respectively represent the updated weight, mean and covariance; X n represents the nth sample data point; ω is the intermediate variable; Construct the log-likelihood function of the Gaussian mixture model, and check whether the parameters or log-likelihood function converge. If not, re-cluster until convergence; if yes, output the clustering result to obtain a reduced-dimensional optical-storage-grid-load scenario set; the expression of the log-likelihood function of the Gaussian mixture model is: Where X is the observed data point; μ is the mean vector of each Gaussian distribution; D represents the covariance matrix of each Gaussian distribution; φ represents the weight of each Gaussian distribution; G w (·) represents the w-th Gaussian distribution function.

4. The method for planning the coordinated photovoltaic, storage, and load distribution network side according to claim 3 is characterized in that: The method further includes: based on the dispersion B and the intra-class concentration L, using a comprehensive clustering index to evaluate the quality of the clustering result; the expression is: Among them, R is the number of clustering results; L r is the inter-class concentration of the r-th clustering result; F is the number of all scenes; E r is the number of scenes contained in the rth cluster; the larger the comprehensive clustering index O(R), the better the clustering effect.

5. The distribution network side photovoltaic storage grid-load coordinated planning method according to any one of claims 1 to 4, characterized in that: The photovoltaic storage grid-load collaborative planning model includes an objective function with the goal of minimizing the total planning cost; Its expression is: minC=C I +C O Among them, C I is the investment cost; b is the discount rate, y is the useful life of the equipment, c sub 、c PVG 、c ESS are the unit capacity investment costs of substation, photovoltaic and energy storage, respectively. is the planned capacity of the i-th substation, are the installed capacity of photovoltaic and energy storage at node i, respectively, sub ,Ω PVG ,Ω ESS are the collection of substation, photovoltaic and energy storage nodes to be installed; c line is the investment cost per unit length of line, l ij is the length of line ij, Ω line C is the collection of lines to be constructed; O is the operating cost; s is the number of days in quarter s; Λ PVG is the set of nodes connected to photovoltaic power generation. The operation and maintenance fee for each unit of electricity generated by photovoltaic power. is the active power of photovoltaic output at node i in period t of quarter s; sub Assemble for substation, is the time-of-use electricity price in period t, is the amount of electricity purchased by the substation during period s in the quarter.

6. The method for planning the coordination of light, storage, grid and load at the distribution network side according to claim 5 is characterized in that: The photovoltaic storage grid load constraints include substation planning capacity constraints, new energy and energy storage single node installation capacity constraints, new energy and energy storage maximum installation capacity constraints, power flow constraints, system safety constraints, photovoltaic operation constraints and / or energy storage operation constraints; wherein: The substation planning capacity constraint is expressed as: in, They represent the minimum planned capacity and maximum planned capacity of the i-th substation respectively; K is a positive integer; is the transformer capacity of the i-th substation; The single-node installation capacity constraint of new energy and energy storage is expressed as: in, is the maximum installed capacity of photovoltaic power, The maximum installed capacity of energy storage; The maximum installed capacity constraint of new energy and energy storage is expressed as: in, The maximum installed capacity for renewable energy; The power flow constraint is expressed as: Among them, u(j) is the node set with j as the end node, w(j) is the node set with j as the head node; x ij Indicates whether line ij is constructed; P ij,t , P jk,t represents the active power of line ij and line jk in period t, R ij is the resistance of line ij, I ij,t represents the circuit of line ij during period t; P j,t is the active power flowing into node j during period t; Q ij,t , Q jk,t represents the reactive power of line ij and line jk during period t, X ij is the reactance of line ij, Q j,t is the reactive power flowing into node j during period t; U j,t is the voltage amplitude of node j in period t, It represents the square of the impedance modulus, which indicates the total resistance to the flow of current; Indicates whether photovoltaic and energy storage are constructed at node j; represents the active power and reactive power of photovoltaic output of node j in period t, represents the active power output of energy storage at node j during period t; represents the active power and reactive power of the load at node j; The system safety constraints are expressed as: Among them, U i,min , U i,max Respectively represent the lower and upper limits of the voltage amplitude of node i; I ij,max represents the maximum current of line ij during time period t; The photovoltaic operation constraint is expressed as: in, represents the maximum output power of the photovoltaic at node i during period t, is the photovoltaic power factor angle; The energy storage operation constraint is expressed as: in, They represent the charging power and discharging power of the energy storage at node i in time period t respectively; i,t is the charging and discharging state of energy storage at node i in time period t, is the number of energy storage modules installed at node i, P ESS is the rated capacity of a single energy storage module; represents the energy storage charge at node i in period t, and Δt is the period between adjacent operating times; They represent the charging efficiency and discharging efficiency of energy storage at node i respectively; S i,min , S i,max They represent the lower and upper limits of the state of charge of the energy storage module at node i respectively.

7. The method for planning the coordinated photovoltaic, storage, and load distribution network side according to claim 6, characterized in that: The solving of the photovoltaic storage grid-load collaborative planning model considering the photovoltaic storage grid-load constraints includes: Variable substitution and big-M method are used to transform the non-convex constraints in the model into convex ones, where: make Then the power flow constraint, operation safety constraint and energy storage constraint are transformed into: in, represents the square of the voltage amplitude of node i in time period t; X ij is the reactance of line ij; M is the penalty factor coefficient; The Gurobi solver is used to solve the objective function after convex processing transformation to obtain the planning result.

8. A distribution network side photovoltaic storage network load collaborative planning system based on massive scene dimensionality reduction, applying the distribution network side photovoltaic storage network load collaborative planning method according to any one of claims 1 to 7, characterized in that: include: The data processing module is used to obtain the equipment parameters and historical scenario data of the photovoltaic power source, local load and energy storage in the target area, and build a photovoltaic storage grid-load scenario database for the target area; The dimension reduction processing module is used to perform dimension reduction processing on the PV-storage-grid-load scenario data by using the principal component analysis method, and then clustering it by using the Gaussian mixture clustering algorithm to obtain a reduced dimension PV-storage-grid-load scenario set; The planning module is equipped with a photovoltaic storage network load collaborative planning model constructed with the goal of minimizing the total planning cost, and is used to solve the photovoltaic storage network load collaborative planning model under the consideration of photovoltaic storage network load constraints and output the planning results.

9. A device comprising a memory and a processor, wherein the memory stores computer-readable instructions, characterized in that: When the computer-readable instructions are executed by the processor, the processor executes all or part of the steps of the distribution network side photovoltaic storage network load coordination planning method as described in any one of claims 1 to 7.

10. A storage medium having computer-readable instructions stored thereon, characterized in that: When the computer-readable instructions are executed by the processor, all or part of the steps of the distribution network side photovoltaic storage network load coordination planning method as described in any one of claims 1 to 7 are implemented.

Images