Intelligent power grid power distribution method and system

Abstract

The invention relates to the technical field of intelligent power grids, and discloses an intelligent power grid power distribution method and system, and the method comprises the steps: S01, collecting the power utilization behavior characteristic data of n users; step S02, preprocessing the data in the step S01; step S03, analyzing the power consumption behavior characteristic data of the n users in the step S02, and classifying the users; step S04, constructing a scheduling optimization model based on a classification result in the step S03, and optimizing power distribution scheduling; through the step S03 and the step S04, the power consumption behavior characteristic data of n users can be analyzed, the users are classified, then a scheduling optimization model is constructed, and the power distribution scheduling is optimized in combination with the classification result in the step S03; according to the method, efficient and economical operation of the power distribution network is realized, clustering classification is performed on the users according to different power utilization behavior characteristics of different users, power distribution optimization scheduling is performed while differences among different users are reflected, and the intelligent level of the power distribution network is effectively improved.

Description

Technical Field

[0001] The present invention relates to the field of smart grid technology, and more specifically to a smart grid power distribution method and system. Background Art

[0002] With the continuous acceleration of the construction of new power systems, source-grid-load-storage load control businesses are developing rapidly. As an important bridge connecting the power generation end and the load end, the net load of the distribution network will fluctuate violently as the proportion of renewable energy increases sharply, which puts higher requirements on the flexible adjustment and rapid climbing ability of the distribution system.

[0003] Traditional distribution networks cannot effectively manage connected distributed devices, and cannot effectively optimize and dispatch the distribution resources of the distribution system. Smart grids can improve the intelligence level of distribution networks and achieve efficient and economical operation of distribution networks. At the same time, different users have different electricity consumption behavior characteristics, so their electricity demands are also different. Therefore, a smart distribution method is needed to cluster different users according to their electricity consumption behavior characteristics, reflect the differences between different users, and optimize distribution scheduling at the same time, so as to improve the intelligence level of distribution networks. Summary of the invention

[0004] In order to overcome the above-mentioned defects of the prior art, the present invention provides a smart grid power distribution method and system to solve the problems existing in the above-mentioned background technology.

[0005] The present invention provides the following technical solution: a smart grid power distribution method, comprising the following steps:

[0006] Step S01: Collecting electricity consumption behavior characteristic data of n users;

[0007] Step S02: pre-processing the data in step S01 to obtain electricity consumption behavior characteristic data that can be directly used;

[0008] Step S03: Analyze the electricity consumption behavior characteristic data of the n users in step S02 and classify the users;

[0009] Step S04: construct a scheduling optimization model based on the classification results in step S03 to optimize the power distribution scheduling.

[0010] Preferably, the electricity consumption behavior characteristic data in step S01 includes the user's electricity consumption habits, the user's electricity consumption at each time point, the peak electricity consumption, the electricity price and the daily peak-to-valley difference.

[0011] Preferably, the data preprocessing in step S02 includes but is not limited to cleaning, denoising and normalizing the data; filling missing values, removing outliers, and filling missing values ​​using an interpolation method;

[0012] The formula for normalizing the data is: Among them, S′ is the new data after normalization, S is the collected original data, minA is the minimum value of data A, and maxA is the maximum value of data A.

[0013] Preferably, the method for classifying users in step S03 adopts a k-means clustering method, specifically:

[0014] Step S11: Take the converted q word vectors as sample points, randomly select k sample points as initial cluster center points, and mark the k′ initial cluster center points in ascending order as z1, z2, z3, ... z k′ , which can be marked as z i , where i = 1, 2, 3, ... k', k' = k;

[0015] Step S12: Mark the sample points that are not used as cluster centers as calculation points, and mark the calculation points in ascending order as x1, x2, x3, ... x q-k , which can be marked as x j , where j = 1, 2, 3...qk; and calculate the distance from each calculation point to each cluster center in turn;

[0016] Step S13: establishing corresponding k″ clusters according to k′ cluster center points, k′=k″;

[0017] Step S14: calculate the point x j The distance to each cluster center is compared and the calculated point x j Assign to and calculate point x j The cluster corresponding to the nearest cluster center;

[0018] Step S15: let j=j+1, and jump back to step S14;

[0019] Step S16: repeat the above steps S14 to S15 until the loop ends when j=qk, and the qk calculation points are all assigned to the corresponding clusters;

[0020] Step S17: recalculate the new cluster center coordinates of each cluster;

[0021] Step S18: repeat the above steps S12 to S17 until the new cluster center coordinates of each cluster are recalculated in step S17 and are consistent with the new cluster center coordinates of each cluster calculated in the previous loop, and the loop ends; obtain the clusters corresponding to the k″ clusters and the corresponding sample points in each cluster;

[0022] The clusters corresponding to the k″ clusters and the corresponding sample points in each cluster are the classifications of the users.

[0023] Preferably, the formula for calculating the distance from each calculation point to each cluster center in step S12 is: Among them, D ij To calculate the point x j To cluster center point z i distance, R is the dimension of the sample point, x ju To calculate the point x j The coordinate of the uth dimension, z iu is the cluster center point z i The coordinates of the u-th dimension.

[0024] Preferably, the calculation method of the new cluster center point coordinates of each cluster in step S17 is:

[0025] Among them, y b =(y b1 ,y b2 ,…,y bR ), zc′ is the new cluster center coordinate of the cth cluster, c = 1, 2, 3...k", y b is the coordinate of the bth calculation point in the cth cluster, a is the number of calculation points in the cth cluster, b = 1, 2, 3...a, (y b1 ,y b2 ,y b3 ……y bR ) is the calculation point y b Coordinates in different dimensions.

[0026] Preferably, the specific method of constructing the scheduling optimization model in step S04 is:

[0027] Construct a scheduling optimization model with the goal of minimizing cost, and the objective function is: Where t is the scheduling period, I is the number of distributed generation sources, J is the number of energy storage devices, K is the number of controlled loads, C1(t) is the scheduling cost of distributed generation sources in the t-th scheduling period, C2(t) is the scheduling cost of energy storage devices in the t-th scheduling period, and C3(t) is the scheduling cost of controlled loads in the t-th scheduling period.

[0028] Preferably, the scheduling cost calculation formula of the distributed power source in the t-th scheduling period is: C1(t)=M1P1(t)+M2P2(t), wherein M1 is the electricity price of photovoltaic power generation, and M2 is the electricity price of wind power generation; P1(t) is the active power of photovoltaic power generation in the t-th scheduling period, and P2(t) is the active power of wind power generation in the t-th scheduling period;

[0029] The calculation formula for the dispatch cost of the energy storage device in the tth dispatch period is: C2(t) = ρP3(t)2 , where ρ is the dispatch cost parameter of the energy storage device, and P3(t) is the active power of the energy storage device in the tth dispatch period;

[0030] The scheduling cost calculation formula of the controlled load in the t-th scheduling period is: C3(t)=F1(t)+F2(t), where F1(t) is a shiftable load and F2(t) is an interruptible load.

[0031] Preferably, the calculation formula of F1(t) is: F1(t)=N1*P1′(t), wherein N1 is the unit power dispatching cost parameter of the shiftable load, and P′1(t) is the active power of the shiftable load in the tth dispatching period;

[0032] The calculation formula of F2(t) is: F2(t)=N2*P2′(t), wherein N2 is the unit power scheduling cost parameter of the interruptible load, and P′1(t) is the active power of the interruptible load in the tth scheduling period.

[0033] Smart grid distribution system, including data acquisition module, data preprocessing module, user characteristic analysis module, dispatch optimization model building module and dispatch optimization module;

[0034] The data collection module is used to collect the electricity consumption behavior characteristic data of n users and transmit it to the data preprocessing module;

[0035] The data preprocessing module is used to preprocess the data from the data acquisition module to obtain user behavior feature data that can be directly used, and transmit it to the user feature analysis module;

[0036] The user characteristic analysis module is used to receive data from the data preprocessing module, analyze the electricity consumption behavior characteristic data of n users, and classify the users;

[0037] The scheduling optimization model building module is used to build a scheduling optimization model and send a scheduling optimization instruction to the scheduling optimization module;

[0038] The scheduling optimization module is used to receive scheduling optimization instructions and optimize power distribution scheduling.

[0039] Technical effects and advantages of the present invention:

[0040] The present invention is provided with step S03 and step S04, which is conducive to analyzing the power consumption behavior characteristic data of n users, classifying the users, and then building a scheduling optimization model, and optimizing the distribution scheduling in combination with the classification result in step S03; the efficient and economical operation of the distribution network is achieved, and the users are clustered and classified according to the different power consumption behavior characteristics of different users, and the distribution optimization scheduling is performed while reflecting the differences between different users, which effectively improves the intelligence level of the distribution network. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] Figure 1 This is a flow chart of the smart grid power distribution method of the present invention.

[0042] Figure 2 This is a structural diagram of the smart grid distribution system of the present invention. DETAILED DESCRIPTION

[0043] The technical solution of the present invention will be clearly and completely described below in conjunction with the accompanying drawings in the present invention. In addition, the forms of the various structures recorded in the following embodiments are merely illustrative. The smart grid distribution method and system involved in the present invention are not limited to the various structures recorded in the following embodiments. All other implementations obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention.

[0044] Embodiment 1:

[0045] like Figure 1 As shown, this embodiment provides a smart grid power distribution method, comprising the following steps:

[0046] Step S01: Collecting electricity consumption behavior characteristic data of n users, wherein the electricity consumption behavior characteristic data includes but is not limited to the user's electricity consumption habits, the user's electricity consumption at each time point, the peak electricity consumption, the electricity price, and the daily peak-to-valley difference;

[0047] Step S02: pre-processing the data in step S01 to obtain electricity consumption behavior characteristic data that can be directly used;

[0048] Step S03: Analyze the power consumption behavior characteristic data of the n users in step S02, classify the users, and cluster the users with similar power consumption behavior characteristics, laying a foundation for subsequent power distribution optimization;

[0049] Step S04: construct a scheduling optimization model based on the classification results in step S03 to optimize the power distribution scheduling.

[0050] In this embodiment, it should be specifically explained that the data preprocessing in step S02 includes but is not limited to cleaning, denoising and normalizing the data; filling missing values ​​and removing outliers to improve data quality; using interpolation methods to fill missing values, such as using mean, median and previous and next values;

[0051] The formula for normalizing the data is: Among them, S′ is the new data after normalization, S is the collected original data, minA is the minimum value of data A, and maxA is the maximum value of data A.

[0052] In this embodiment, it should be specifically explained that the method for analyzing the electricity consumption behavior characteristic data of n users in step S03 includes:

[0053] The trained word vector model is used to convert the electricity consumption behavior feature data of n users into corresponding word vectors in sequence; wherein the training process of the word vector conversion model includes:

[0054] Collect the electricity consumption behavior feature data of multiple users in advance; use the WordPiece algorithm to build a word table containing the vocabulary of the electricity consumption behavior feature data domain; the vocabulary of the user's electricity consumption behavior feature data domain is a vocabulary collection of professional vocabulary and terminology related to the electricity consumption behavior feature data; convert the electricity consumption behavior feature data into an ID sequence through the word table, and add the position code as inputs; the inputs are deep Transformer encoded through the BERT encoder to obtain the latent sentence representation, which is the word vector; randomly select some word masks and predict the corresponding original words as the unsupervised prediction training task target; use the cross entropy to calculate the loss function, reflect the transmission optimization and modification of the word vector conversion model parameters, and complete the pre-training of the word vector conversion model; collect the user's electricity consumption behavior feature data in advance as annotations, and build a classification task; add a classification output layer based on the pre-trained word vector conversion model to determine the rate as the task target, fix the BERT encoder, and optimize and fine-tune the output layer parameters; the word vector conversion model is the BERT semantic model;

[0055] The training process of the word vector conversion model is only a conceptual training process. The BERT semantic model is an existing technology, and the specific training process of the word vector conversion model will not be described in detail in this implementation.

[0056] In this embodiment, it should be specifically explained that the method for classifying users in step S03 adopts the k-means clustering method, and the specific method is:

[0057] Step S11: Take the converted q word vectors as sample points, randomly select k sample points as initial cluster center points, and mark the k′ initial cluster center points in ascending order as z1, z2, z3, ... z k′ , which can be marked as z i , where i = 1, 2, 3, ... k', k' = k;

[0058] Step S12: Mark the sample points that are not used as cluster centers as calculation points, and mark the calculation points in ascending order as x1, x2, x3, ... x q-k , which can be marked as x j , where j = 1, 2, 3...qk; and calculate the distance from each calculation point to each cluster center in turn;

[0059] The formula for calculating the distance from each calculation point to each cluster center is: Among them, D ij To calculate the point x j To cluster center point z i distance, R is the dimension of the sample point, x ju To calculate the point x j The coordinate of the uth dimension, z iu is the cluster center point z i The coordinates of the u-th dimension;

[0060] Step S13: establishing corresponding k″ clusters according to k′ cluster center points, k′=k″;

[0061] Step S14: calculate the point x j The distance to each cluster center is compared and the calculated point x j Assign to and calculate point x j The cluster corresponding to the nearest cluster center;

[0062] Step S15: let j=j+1, and jump back to step S14;

[0063] Step S16: repeat the above steps S14 to S15 until the loop ends when j=qk, and the qk calculation points are all assigned to the corresponding clusters;

[0064] Step S17: recalculate the new cluster center coordinates of each cluster;

[0065] The calculation method of the new cluster center coordinates of each cluster is:

[0066] Among them, y b =(y b1 ,y b2 ,…,y bR), zc′ is the new cluster center coordinate of the cth cluster, c = 1, 2, 3...k", y b is the coordinate of the bth calculation point in the cth cluster, a is the number of calculation points in the cth cluster, b = 1, 2, 3...a, (y b1 ,y b2 ,y b3 ……y bR ) is the calculation point y b Coordinates in different dimensions;

[0067] Step S18: repeat the above steps S12 to S17 until the new cluster center coordinates of each cluster are recalculated in step S17 and are consistent with the new cluster center coordinates of each cluster calculated in the previous loop, and the loop ends; obtain the clusters corresponding to the k″ clusters and the corresponding sample points in each cluster;

[0068] The clusters corresponding to the k″ clusters and the corresponding sample points in each cluster are the classifications of the users.

[0069] In this embodiment, it should be specifically explained that the specific method of constructing the scheduling optimization model in step S04 is:

[0070] Construct a scheduling optimization model with the goal of minimizing cost, and the objective function is: Where t is the scheduling period, I is the number of distributed generation sources, J is the number of energy storage devices, K is the number of controlled loads, C1(t) is the scheduling cost of distributed generation sources in the t-th scheduling period, C2(t) is the scheduling cost of energy storage devices in the t-th scheduling period, and C3(t) is the scheduling cost of controlled loads in the t-th scheduling period.

[0071] In this embodiment, it should be specifically explained that the scheduling cost calculation formula of the distributed power source in the t-th scheduling period is: C1(t)=M1P1(t)+M2P2(t), where M1 is the electricity price of photovoltaic power generation, and M2 is the electricity price of wind power generation; P1(t) is the active power of photovoltaic power generation in the t-th scheduling period, and P2(t) is the active power of wind power generation in the t-th scheduling period;

[0072] The calculation formula for the dispatch cost of the energy storage device in the tth dispatch period is: C2(t) = ρP3(t) 2 , where ρ is the scheduling cost parameter of the energy storage device, and P3(t) is the active power of the energy storage device in the tth scheduling period.

[0073] In this embodiment, it should be specifically explained that the scheduling cost calculation formula of the controlled load in the t-th scheduling period is: C3(t)=F1(t)+F2(t), wherein F1(t) is a shiftable load and F2(t) is an interruptible load; the shiftable load refers to the load during the flexible scheduling operation of the power grid, which must ensure continuity and constancy, and the interruptible load refers to the load that can be reduced or cut off when the power grid operates abnormally, which has variability and intermittency;

[0074] F1(t)=N1*P1′(t), where N1 is the unit power dispatch cost parameter of the shiftable load, and P′1(t) is the active power of the shiftable load in the tth dispatch period;

[0075] F2(t)=N2*P2′(t), where N2 is the unit power dispatch cost parameter of the interruptible load, and P′1(t) is the active power of the interruptible load in the tth dispatch period;

[0076] In the distribution network, distributed power sources, energy storage devices and controlled loads are the adjustment objects, and their access positions in the distribution network are adjustable. When distributed power sources, controlled loads and energy storage devices are located at different access positions, the objective function is solved, and the optimal access position can be obtained by comparing the calculation results, and distribution optimization scheduling can be carried out.

[0077] In this embodiment, it should be specifically explained that by substituting the clustered distributed power sources, energy storage devices and controlled loads corresponding to the k″ clusters into the scheduling optimization model, the best access location corresponding to each user classification can be obtained, so as to perform corresponding distribution scheduling optimization for each user classification.

[0078] Embodiment 2:

[0079] like Figure 2 As shown, this embodiment provides a smart grid distribution system, including a data acquisition module, a data preprocessing module, a user feature analysis module, a scheduling optimization model building module and a scheduling optimization module;

[0080] The data collection module is used to collect the electricity consumption behavior characteristic data of n users and transmit it to the data preprocessing module;

[0081] The data preprocessing module is used to preprocess the data from the data acquisition module to obtain user behavior feature data that can be directly used, and transmit it to the user feature analysis module;

[0082] The user characteristic analysis module is used to receive data from the data preprocessing module, analyze the electricity consumption behavior characteristic data of n users, and classify the users;

[0083] The scheduling optimization model building module is used to build a scheduling optimization model and send a scheduling optimization instruction to the scheduling optimization module;

[0084] The scheduling optimization module is used to receive scheduling optimization instructions and optimize power distribution scheduling.

[0085] In this embodiment, it should be specifically explained that the method for classifying users by the user feature analysis module adopts the k-means clustering method, and the specific method is as follows:

[0086] Step S11: Take the converted q word vectors as sample points, randomly select k sample points as initial cluster center points, and mark the k′ initial cluster center points in ascending order as z1, z2, z3, ... z k′ , which can be marked as z i , where i = 1, 2, 3, ... k', k' = k;

[0087] Step S12: Mark the sample points that are not used as cluster centers as calculation points, and mark the calculation points in ascending order as x1, x2, x3, ... x q-k , which can be marked as x j , where j = 1, 2, 3...qk; and calculate the distance from each calculation point to each cluster center in turn;

[0088] Step S13: establishing corresponding k″ clusters according to k′ cluster center points, k′=k″;

[0089] Step S14: calculate the point x j The distance to each cluster center is compared and the calculated point x j Assign to and calculate point x j The cluster corresponding to the nearest cluster center;

[0090] Step S15: let j=j+1, and jump back to step S14;

[0091] Step S16: repeat the above steps S14 to S15 until the loop ends when j=qk, and the qk calculation points are all assigned to the corresponding clusters;

[0092] Step S17: recalculate the new cluster center coordinates of each cluster;

[0093] Step S18: repeat the above steps S12 to S17 until the new cluster center coordinates of each cluster are recalculated in step S17 and are consistent with the new cluster center coordinates of each cluster calculated in the previous loop, and the loop ends; obtain the clusters corresponding to the k″ clusters and the corresponding sample points in each cluster;

[0094] The clusters corresponding to the k″ clusters and the corresponding sample points in each cluster are the classifications of the users.

[0095] In this embodiment, it should be specifically explained that the specific manner in which the scheduling optimization model construction module constructs the scheduling optimization model is:

[0096] Construct a scheduling optimization model with the goal of minimizing cost, and the objective function is: Where t is the scheduling period, I is the number of distributed generation sources, J is the number of energy storage devices, K is the number of controlled loads, C1(t) is the scheduling cost of distributed generation sources in the t-th scheduling period, C2(t) is the scheduling cost of energy storage devices in the t-th scheduling period, and C3(t) is the scheduling cost of controlled loads in the t-th scheduling period.

[0097] In this embodiment, it should be specifically explained that the scheduling cost calculation formula of the distributed power source in the t-th scheduling period is: C1(t)=M1P1(t)+M2P2(t), where M1 is the electricity price of photovoltaic power generation, and M2 is the electricity price of wind power generation; P1(t) is the active power of photovoltaic power generation in the t-th scheduling period, and P2(t) is the active power of wind power generation in the t-th scheduling period;

[0098] The calculation formula for the dispatch cost of the energy storage device in the tth dispatch period is: C2(t) = ρP3(t) 2 , where ρ is the scheduling cost parameter of the energy storage device, and P3(t) is the active power of the energy storage device in the tth scheduling period.

[0099] In this embodiment, it should be specifically explained that the scheduling cost calculation formula of the controlled load in the t-th scheduling period is: C3(t)=F1(t)+F2(t), where F1(t) is a shiftable load and F2(t) is an interruptible load;

[0100] F1(t)=N1*P1′(t), where N1 is the unit power dispatch cost parameter of the shiftable load, and P′1(t) is the active power of the shiftable load in the tth dispatch period;

[0101] F2(t)=N2*P2′(t), where N2 is the unit power scheduling cost parameter of the interruptible load, and P′1(t) is the active power of the interruptible load in the tth scheduling period.

[0102] In the present embodiment, it should be specifically explained that the difference between the present embodiment and the prior art lies in that the present embodiment has step S03 and step S04, which analyzes the electricity consumption behavior characteristic data of n users, classifies the users, and then constructs a scheduling optimization model, and optimizes the distribution scheduling in combination with the classification result in step S03; thus, the efficient and economical operation of the distribution network is realized, and the users are clustered and classified according to the different electricity consumption behavior characteristics of different users, and the distribution optimization scheduling is performed while reflecting the differences between different users, thereby effectively improving the intelligence level of the distribution network.

[0103] Finally: The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the protection scope of the present invention.

[0104] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art who is familiar with the present technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.

Claims

1. A smart grid power distribution method, characterized in that: The following steps are involved: Step S01: Collecting electricity consumption behavior characteristic data of n users; Step S02: pre-processing the data in step S01 to obtain electricity consumption behavior characteristic data that can be directly used; Step S03: Analyze the electricity consumption behavior characteristic data of the n users in step S02 and classify the users; Step S04: construct a scheduling optimization model based on the classification results in step S03 to optimize the power distribution scheduling.

2. The smart grid power distribution method according to claim 1, characterized in that: The electricity consumption behavior characteristic data in step S01 includes the user's electricity consumption habits, the user's electricity consumption at each time point, the peak electricity consumption, the electricity price and the daily peak-to-valley difference.

3. The smart grid power distribution method according to claim 1, characterized in that: The data preprocessing in step S02 includes but is not limited to cleaning, denoising and normalizing the data; filling missing values, removing outliers, and filling missing values ​​using an interpolation method; The formula for normalizing the data is: Among them, S′ is the new data after normalization, S is the collected original data, minA is the minimum value of data A, and maxA is the maximum value of data A.

4. The smart grid power distribution method according to claim 1, characterized in that: The method for classifying users in step S03 adopts the k-means clustering method, and the specific method is as follows: Step S11: Take the converted q word vectors as sample points, randomly select k sample points as initial cluster center points, and mark the k′ initial cluster center points in ascending order as z1, z2, z3, ... z k′ , which can be marked as z i , where i = 1, 2, 3, ... k', k' = k; Step S12: Mark the sample points that are not used as cluster centers as calculation points, and mark the calculation points in ascending order as x1, x2, x3, ... x q-k , which can be marked as x j , where j = 1, 2, 3...qk; and calculate the distance from each calculation point to each cluster center in turn; Step S13: establishing corresponding k″ clusters according to k′ cluster center points, k′=k″; Step S14: calculate the point x j The distance to each cluster center is compared and the calculated point x j Assign to and calculate point x j The cluster corresponding to the nearest cluster center; Step S15: let j=j+1, and jump back to step S14; Step S16: repeat the above steps S14 to S15 until the loop ends when j=qk, and the qk calculation points are all assigned to the corresponding clusters; Step S17: recalculate the new cluster center coordinates of each cluster; Step S18: repeat the above steps S12 to S17 until the new cluster center coordinates of each cluster are recalculated in step S17 and are consistent with the new cluster center coordinates of each cluster calculated in the previous loop, and the loop ends; obtain the clusters corresponding to the k″ clusters and the corresponding sample points in each cluster; The clusters corresponding to the k″ clusters and the corresponding sample points in each cluster are the classifications of the users.

5. The smart grid power distribution method according to claim 4, characterized in that: The formula for calculating the distance from each calculation point to each cluster center in step S12 is: Among them, D ij To calculate the point x j To cluster center point z i distance, R is the dimension of the sample point, x ju To calculate the point x j The coordinate of the uth dimension, z iu is the cluster center point z i The coordinates of the u-th dimension.

6. The smart grid power distribution method according to claim 4, characterized in that: The calculation method of the new cluster center point coordinates of each cluster in step S17 is: Among them, y b =(y b1 ,y b2 ,…,y bR ), zc′ is the new cluster center coordinate of the cth cluster, c = 1, 2, 3...k", y b is the coordinate of the bth calculation point in the cth cluster, a is the number of calculation points in the cth cluster, b = 1, 2, 3...a, (y b1 ,y b2 ,y b3 ……y bR ) is the calculation point y b Coordinates in different dimensions.

7. The smart grid power distribution method according to claim 1, characterized in that: The specific method of constructing the scheduling optimization model in step S04 is: Construct a scheduling optimization model with the goal of minimizing cost, and the objective function is: Where t is the scheduling period, I is the number of distributed generation sources, J is the number of energy storage devices, K is the number of controlled loads, C1(t) is the scheduling cost of distributed generation sources in the t-th scheduling period, C2(t) is the scheduling cost of energy storage devices in the t-th scheduling period, and C3(t) is the scheduling cost of controlled loads in the t-th scheduling period.

8. The smart grid power distribution method according to claim 7, characterized in that: The scheduling cost calculation formula of the distributed power source in the t-th scheduling period is: C1(t)=M1P1(t)+M2P2(t), where M1 is the electricity price of photovoltaic power generation, M2 is the electricity price of wind power generation; P1(t) is the active power of photovoltaic power generation in the t-th scheduling period, and P2(t) is the active power of wind power generation in the t-th scheduling period; The calculation formula for the dispatch cost of the energy storage device in the tth dispatch period is: C2(t) = ρP3(t) 2 , where ρ is the dispatch cost parameter of the energy storage device, and P3(t) is the active power of the energy storage device in the tth dispatch period; The scheduling cost calculation formula of the controlled load in the t-th scheduling period is: C3(t)=F1(t)+F2(t), where F1(t) is a shiftable load and F2(t) is an interruptible load.

9. The smart grid power distribution method according to claim 8, characterized in that: The calculation formula of F1(t) is: F1(t)=N1*P1′(t), where N1 is the unit power dispatching cost parameter of the shiftable load, and P′1(t) is the active power of the shiftable load in the tth dispatching period; The calculation formula of F2(t) is: F2(t)=N2*P2′(t), wherein N2 is the unit power scheduling cost parameter of the interruptible load, and P′1(t) is the active power of the interruptible load in the tth scheduling period.

10. A smart grid power distribution system, used to implement the smart grid power distribution method according to any one of claims 1 to 7, characterized in that: It includes data acquisition module, data preprocessing module, user feature analysis module, scheduling optimization model building module and scheduling optimization module; The data collection module is used to collect the electricity consumption behavior characteristic data of n users and transmit it to the data preprocessing module; The data preprocessing module is used to preprocess the data from the data acquisition module to obtain user behavior feature data that can be directly used, and transmit it to the user feature analysis module; The user characteristic analysis module is used to receive data from the data preprocessing module, analyze the electricity consumption behavior characteristic data of n users, and classify the users; The scheduling optimization model building module is used to build a scheduling optimization model and send a scheduling optimization instruction to the scheduling optimization module; The scheduling optimization module is used to receive scheduling optimization instructions and optimize power distribution scheduling.

Images