A method, device, electronic device and storage medium for user privacy protection

Abstract

The present invention provides a user privacy protection method, device, electronic device and storage medium, which relates to the technical field of power system power consumption data processing. The cloud controller receives the power consumption load data of the user cluster in the historical period; processes the power consumption load data of the user cluster to determine the global objective function; generates the coordination signal of the current number of times based on the global objective function and the coordination signal of the previous number of times, and sends the coordination signal of the current number of times to each edge controller; when receiving the coordination signal of the current number of times, the edge controller processes the coordination signal of the current number of times and the power consumption load data of the local user device in the current period to obtain the local user device control strategy, so as to control the corresponding local user device. The present invention reduces the load cluster effect and improves the privacy of the power consumption load data of the user cluster on the premise that the user clusters do not trust each other and the device parameters are not shared, thereby avoiding the leakage of user sensitive information.

Description

Technical Field

[0001] The present invention relates to the technical field of power system user load data privacy protection, and in particular, to a user privacy protection method, device, electronic device, and storage medium. Background Art

[0002] With the continuous expansion of the scale of power system users and the continuous deepening of the interaction relationship among users, regional user clusters have gradually formed. The electricity consumption behavior patterns of these user clusters are affected by multi-dimensional factors such as the regional economic development level, population structure characteristics, and meteorological environment, showing a certain cluster effect. This cluster effect is mainly reflected in the time-series characteristics of the electricity load curve, including the peak-valley distribution law, fluctuation characteristics, and similarity in aspects such as load duration.

[0003] However, this regional cluster effect also brings new privacy and security challenges: malicious entities may use data mining techniques such as pattern recognition and comparative analysis to reverse-infer the individual electricity consumption behavior characteristics of target users from the load characteristics of user clusters, resulting in the leakage of user sensitive information, which highlights the importance of privacy protection for user cluster electricity load data. Summary of the Invention

[0004] In view of this, embodiments of the present invention provide a user privacy protection method, device, electronic device, and storage medium to solve the problem of leakage of user sensitive information existing in the prior art.

[0005] To achieve the above object, embodiments of the present invention provide the following technical solutions:

[0006] A first aspect of an embodiment of the present invention shows a user privacy protection method, which is applied to a user privacy protection device. The device includes a cloud controller and a plurality of edge controllers, and the cloud controller is respectively connected to the plurality of edge controllers. The method includes:

[0007] The cloud controller receives the electricity load data of the user cluster in the historical period uploaded by each of the edge controllers;

[0008] The cloud controller performs a privacy evaluation based on the electricity load data of the user cluster to determine a global objective function;

[0009] The cloud controller generates a coordination signal for the current time based on the global objective function and the coordination signal of the previous time, and sends the coordination signal for the current time to each of the edge controllers. Among them, if the previous time is the first time to send a coordination signal, the coordination signal of the previous time refers to an initial coordination signal; if the previous time is not the first time to send a coordination signal, the coordination signal of the previous time is a coordination signal obtained by processing through the global objective function and the coordination signal corresponding to its previous time;

[0010] For each edge controller, when the edge controller receives the coordination signal of the current number of times, the edge controller processes the coordination signal of the current number of times and the power consumption load data of the local user device in the current period to obtain a local user device control strategy, so as to control the corresponding local user device based on the local user device control strategy;

[0011] The edge controller obtains the new power consumption load data of the local user device after being controlled by the local user device control strategy, and returns it to the cloud controller;

[0012] When the cloud controller determines that the threshold corresponding to the coordination signal of the current number of times is less than the preset threshold, it determines that the signal convergence ends, so that each subsequent edge controller can update the local user device control strategy in real time based on the coordination signal of the current number of times.

[0013] Optionally, it further includes:

[0014] When the cloud controller determines that the threshold corresponding to the coordination signal of the current number of times is greater than or equal to the preset threshold, it receives the new power consumption load data uploaded by the edge controller, and based on the new power consumption load data uploaded by the edge controller, returns to execute the step of processing the power consumption load data of the user cluster to determine the global objective function.

[0015] Optionally, the cloud controller performs privacy evaluation based on the power consumption load data of the user cluster to determine the global objective function, including:

[0016] For each single user in the user cluster, calculate the load correlation matrix between each user based on the user load data of each user;

[0017] Construct a user cluster load network based on the load correlation matrix;

[0018] Calculate network parameters based on the user cluster load network, where the network parameters include load network aggregation degree, load network density, and load network synchronization degree;

[0019] Process the load network aggregation degree, load network density, and load network synchronization degree in the network parameters, and the power consumption load data of the user cluster to obtain the global objective function.

[0020] Optionally, constructing a user cluster load network based on the load correlation matrix includes:

[0021] Take each user in the user cluster as a node in the network;

[0022] Determine the connection relationship of the nodes corresponding to the user load correlation and the edge weights corresponding to the nodes based on the relationship between the user load correlation in the load correlation matrix and the preset correlation threshold;

[0023] Construct a user cluster load network based on the nodes, the connection relationship of the nodes corresponding to the user load correlation, and the edge weights of the nodes corresponding to the user load correlation.

[0024] Optionally, calculate network parameters based on the user cluster load network, including:

[0025] For each node in the user cluster load network, determine the sum of the edge weights of the node, the degree of the node, and the element parameters of the node in the adjacency matrix based on the user cluster load network;

[0026] Calculate the local clustering coefficient of each node based on the sum of the edge weights of the node, the degree of the node, and the element parameters of the node in the adjacency matrix;

[0027] Calculate the load network aggregation degree of the user cluster based on the local clustering coefficient of each node;

[0028] Determine the load network density of the user cluster based on the user cluster load network;

[0029] Determine the load network synchronization degree of the user cluster based on the user cluster load network.

[0030] Optionally, process the load network aggregation degree, load network density, and load network synchronization degree in the network parameters, and the electricity consumption load data of the user cluster to obtain a global objective function, including:

[0031] Construct a user cluster electricity consumption load vector matrix based on the user load data of each user in the user cluster;

[0032] Calculate the global objective function based on the load network aggregation degree, density, synchronization degree in the network parameters, and the electricity consumption load vector matrix.

[0033] Optionally, the edge controller processes the coordination signal of the current time and the electricity consumption load data of the local user equipment in the current period to obtain a local user equipment control strategy, including:

[0034] Determine the electricity consumption load data that meets the preset constraint regulation based on the electricity consumption load data of the local user equipment in the current period;

[0035] Determine the local user equipment control strategy based on the electricity consumption load data that meets the preset constraint regulation and the coordination signal of the current time.

[0036] A second aspect of the embodiments of the present invention discloses a user privacy protection device, which includes a cloud controller and a plurality of edge controllers, and the cloud controller is respectively connected to the plurality of edge controllers;

[0037] The cloud controller is configured to receive the power consumption load data of the user clusters in the historical period uploaded by each of the edge controllers; process the power consumption load data of the user clusters to determine a global objective function; generate a coordination signal for the current time based on the global objective function and the coordination signal of the previous time, and send the coordination signal for the current time to each edge controller, where if the previous time is the first time to send a coordination signal, the coordination signal of the previous time refers to an initial coordination signal, and if the previous time is not the first time to send a coordination signal, the coordination signal of the previous time is a coordination signal obtained by processing the global objective function and the coordination signal corresponding to its previous time;

[0038] For each edge controller, when receiving the coordination signal for the current time, the edge controller is configured to process the coordination signal for the current time and the power consumption load data of the local user equipment in the current period to obtain a local user equipment control strategy, so as to control the corresponding local user equipment based on the local user equipment control strategy; obtain the new power consumption load data of the local user equipment after being controlled by the local user equipment control strategy, and return it to the cloud controller;

[0039] The cloud controller is configured to determine that the signal convergence ends when determining that the threshold corresponding to the coordination signal for the current time is less than a preset threshold, so that each subsequent edge controller can update the local user equipment control strategy in real time based on the coordination signal for the current time.

[0040] A third aspect of the embodiments of the present invention discloses an electronic device, which includes a processor and a memory. The memory is used to store program codes and data for user privacy protection, and the processor is used to call the program instructions in the memory to execute any one of the user privacy protection methods shown in the first aspect of the embodiments of the present invention.

[0041] A fourth aspect of the embodiments of the present invention discloses a storage medium, which includes a stored program. When the program runs, it controls the device where the storage medium is located to execute any one of the user privacy protection methods shown in the first aspect of the embodiments of the present invention.

[0042] Based on the user privacy protection method, device, electronic device and storage medium provided by the embodiments of the present invention above, the device includes a cloud controller and a plurality of edge controllers, the cloud controller is respectively connected to the plurality of edge controllers, and the method includes: the cloud controller receives the power consumption load data of the user cluster in the historical period uploaded by each of the edge controllers; the cloud controller performs privacy evaluation based on the power consumption load data of the user cluster to determine a global objective function; the cloud controller generates a coordination signal for the current number of times based on the global objective function and the coordination signal of the previous number of times, and sends the coordination signal for the current number of times to each edge controller, wherein, if the previous number of times is the first time to send a coordination signal, the coordination signal of the previous number of times refers to an initial coordination signal, and if the previous number of times is not the first time to send a coordination signal, the coordination signal of the previous number of times is a coordination signal obtained by processing through the global objective function and the coordination signal corresponding to its previous number of times; for each edge controller, when the edge controller receives the coordination signal for the current number of times, the edge controller processes the coordination signal for the current number of times and the power consumption load data of the local user device in the current period to obtain a local user device control strategy, so as to control the corresponding local user device based on the local user device control strategy. The edge controller obtains the new power consumption load data of the local user device after being controlled by the local user device control strategy, and returns it to the cloud controller; when the cloud controller determines that the threshold value corresponding to the coordination signal for the current number of times is less than a preset threshold value, it determines that the signal convergence ends, so that each subsequent edge controller can update the local user device control strategy in real time based on the coordination signal for the current number of times. In the embodiments of the present invention, first, the power consumption load data of the user cluster uploaded by each edge controller is used to generate a spatio-temporally associated coordination signal, which is sent to each edge controller; the edge controller combines the coordination signal and the power consumption load data of the local user device in the current period to process and obtain a local user device control strategy, thereby avoiding the leakage of user sensitive information and highlighting the importance of protecting the privacy of the power consumption load data of the user cluster. Description of the Drawings

[0043] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings required for the description of the embodiments or the prior art. Obviously, the drawings in the following description are only the embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can also be obtained according to the provided drawings.

[0044] Figure 1 It is a schematic diagram of the architecture of the user privacy protection device shown in the embodiments of the present invention;

[0045] Figure 2It is a diagram of a distributed cloud-edge collaborative control architecture shown in an embodiment of the present invention;

[0046] Figure 3 It is a schematic flowchart of a user privacy protection method shown in an embodiment of the present invention;

[0047] Figure 4 It is a schematic diagram of generalized energy storage replacing traditional energy storage shown in an embodiment of the present invention;

[0048] Figure 5 It is a framework diagram of a user cluster power consumption load data privacy protection method based on distributed cloud-edge collaborative control shown in an embodiment of the present invention. Detailed implementation manners

[0049] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

[0050] The terms "first", "second", "third", "fourth", etc. (if any) in the specification and claims of this application and the above-mentioned drawings are used to distinguish similar objects, and do not have to be used to describe a specific order or sequence. It should be understood that such data can be interchanged under appropriate circumstances so that the embodiments described here can be implemented in an order different from that shown or described here. In addition, the terms "comprising" and "having" and any variations thereof are intended to cover non-exclusive inclusion. For example, a process, method, system, product or device that includes a series of steps or units does not have to be limited to those steps or units clearly listed, but may include other steps or units not clearly listed or inherent to these processes, methods, products or devices.

[0051] It should be noted that the descriptions involving "first", "second", etc. in the present invention are only for descriptive purposes, and cannot be understood as indicating or implying their relative importance or implicitly indicating the quantity of the indicated technical features. Thus, the features defined with "first" and "second" may explicitly or implicitly include at least one of such features. In addition, the technical solutions between various embodiments can be combined with each other, but it must be based on the fact that those of ordinary skill in the art can implement it. When the combination of technical solutions appears to be contradictory or unable to be implemented, it should be considered that such a combination of technical solutions does not exist and is not within the protection scope required by the present invention.

[0052] In this application, the terms "include", "comprise" or any other variant thereof are intended to cover non-exclusive inclusion, such that a process, method, article or device comprising a series of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article or device. Without further limitation, an element defined by the statement "comprising one..." does not exclude the presence of additional identical elements in the process, method, article or device comprising the said element.

[0053] As can be seen from the background art, achieving effective cluster privacy protection requires the coordinated scheduling of all users' energy storage devices within the cluster. However, these energy storage devices are distributed in physical space and have strong autonomy and independence. Secondly, operating parameters such as the state of charge, charge and discharge characteristics, device capacity, and power limit of the energy storage devices involve the commercial secrets or personal privacy of users, and once leaked, it may cause damage to the interests of users.

[0054] Therefore, the collaborative protection mechanism between user clusters must not only ensure the privacy and security of electricity load data, but also guarantee the information security at the level of users' personal devices. This characteristic of a distributed and autonomous system, combined with multi-dimensional privacy protection requirements, limits the application of traditional centralized optimization methods.

[0055] Based on this, it can be seen that the problems to be solved in this application include how to quantitatively evaluate the privacy of electricity load data of user clusters, effectively regulate distributed electricity-using devices of user clusters on the premise of non-mutual trust; and how to reduce the initial investment cost of user load data privacy protection.

[0056] The user data and other data involved in this application are all obtained with full consent and authorization, and the collection, use and processing of relevant information comply with the relevant laws, regulations and standards of relevant countries and regions.

[0057] In order to reduce the load cluster effect and improve the privacy of electricity load data of user clusters on the premise of non-mutual trust among user clusters and non-sharing of device parameters, the present invention realizes the privacy evaluation of electricity load data of user clusters based on complex network theory, uses existing generalized energy storage devices of users such as electric vehicles and HVAC to participate in privacy protection, decouples the privacy protection optimization problem through distributed cloud-edge collaborative control, and drives intelligent thermostats and charging pile actuators through edge controllers to complete load disturbance.

[0058] See Figure 1 , which is a schematic diagram of the architecture of the user privacy protection device shown in the embodiments of the present invention.

[0059] The device includes a cloud controller 10 and a plurality of edge controllers 20, and the cloud controller 10 is respectively connected to the plurality of edge controllers 20.

[0060] Based on this, the present application also shows a distributed cloud-edge collaborative control architecture diagram of the cloud controller 10 and multiple edge controllers 20, as Figure 2 shown.

[0061] Figure 2 It includes a power cyber-physical layer and a cloud-edge collaborative computing layer. The power cyber-physical layer includes the edge controller 20 and the cloud controller 10; the cloud-edge collaborative computing layer includes an edge calculator and a cloud computing center;

[0062] In Figure 2 , the edge controller 20 is represented by a solid circle, the cloud controller 10 is represented by a heptagon, the edge calculator is represented by a square, and the cloud computing center is represented by a pentagram.

[0063] Among them, the edge calculator is used to provide computing and processing capabilities for the edge controller 20, and the cloud computing center is used to provide computing and processing capabilities for the cloud controller 10.

[0064] It should be noted that the multiple edge controllers 20 are not connected to each other, and the multiple edge controllers 20 are the edge controllers 20 corresponding to each user within a user cluster.

[0065] One user corresponds to one edge controller 20. That is to say, one edge controller 20 can only process the power consumption load data of the user corresponding to it.

[0066] The cloud controller 10 is used to receive the power consumption load data of the historical period user cluster uploaded based on each of the edge controllers 20; process the power consumption load data of the user cluster to determine a global objective function; generate a coordination signal of the current number based on the global objective function and the coordination signal of the previous number, and send the coordination signal of the current number to each edge controller 20, where, if the previous number is the first time to send a coordination signal, the coordination signal of the previous number refers to an initial coordination signal, and if the previous number is not the first time to send a coordination signal, the coordination signal of the previous number is a coordination signal obtained by processing through the global objective function and the coordination signal corresponding to its previous number;

[0067] For each edge controller 20, the edge controller 20 is used to, when receiving the coordination signal of the current number, process the coordination signal of the current number and the power consumption load data of the local user device in the current period to obtain a local user device control strategy, so as to control the corresponding local user device based on the local user device control strategy; obtain the new power consumption load data of the local user device after being controlled by the local user device control strategy, and return it to the cloud controller 10;

[0068] The cloud controller 10 is configured to determine that the signal convergence ends when it determines that the threshold value corresponding to the coordination signal of the current number is less than the preset threshold value, so that each subsequent edge controller 20 can update the local user equipment control strategy in real time based on the coordination signal of the current number.

[0069] The applicant found that the similarity of the electricity consumption behaviors (air conditioner working power, electric vehicle charging power) of users within the cluster is very high, so it is easy to cause the coupled leakage of load data privacy among users within the cluster.

[0070] Therefore, in order to solve this problem of privacy leakage, the present application proposes a technology of load perturbation. After the load perturbation, the similarity of the loads of the cluster users will be significantly reduced, and the privacy will not be easily leaked.

[0071] In the embodiment of the present invention, the underlying smart meters of the users corresponding to each edge controller collect electricity consumption load data such as the total load of the corresponding users and the operation boundary data of the generalized energy storage devices. After the edge controller performs privacy desensitization processing, non-sensitive parameters such as load adjustment margins are uploaded to the cloud controller; the cloud performs processing such as security constraints and privacy evaluation based on the electricity consumption load data within the user cluster to generate a corresponding global objective function, and then generates a spatio-temporally correlated coordination signal and distributes it to each edge controller; the edge controller combines the real-time status of the local device for processing to obtain the local user equipment control strategy. The edge controller only needs to upload its own data and cannot know the data of other users, thus avoiding the leakage of user sensitive information and highlighting the importance of protecting the privacy of the electricity consumption load data of the user cluster.

[0072] Optionally, based on the user privacy protection device shown in the above embodiment of the present invention, it further includes:

[0073] When the cloud controller 10 determines that the threshold value corresponding to the coordination signal of the current number is greater than or equal to the preset threshold value, it receives the new electricity consumption load data uploaded by the edge controller 20, and returns to execute processing based on the electricity consumption load data of the user cluster based on the new electricity consumption load data uploaded by the edge controller 20 to determine the global objective function.

[0074] Optionally, based on the user privacy protection device shown in the above embodiment of the present invention, the cloud controller 10 performs privacy evaluation based on the electricity consumption load data of the user cluster to determine the global objective function, specifically for:

[0075] For each individual user within the user cluster, calculate the load correlation matrix between each user based on the user load data of each user;

[0076] Construct a user cluster load network based on the load correlation matrix;

[0077] Calculate network parameters based on the user cluster load network, where the network parameters include load network aggregation degree, load network density, and load network synchronization degree;

[0078] Process the load network aggregation degree, load network density, and load network synchronization degree in the network parameters, as well as the electricity load data of the user cluster to obtain a global objective function.

[0079] Among them, constructing a user cluster load network based on the load correlation matrix includes:

[0080] Take each user in the user cluster as a node in the network;

[0081] Based on the relationship between the user load correlation in the load correlation matrix and a preset correlation threshold, determine the connection relationship of the nodes corresponding to the user load correlation and the edge weights corresponding to the nodes;

[0082] Construct a user cluster load network based on the nodes, the connection relationship of the nodes corresponding to the user load correlation, and the edge weights of the nodes corresponding to the user load correlation.

[0083] Among them, calculating network parameters based on the user cluster load network includes:

[0084] For each node in the user cluster load network, based on the user cluster load network, determine the sum of the edge weights of the node, the degree of the node, and the element parameters of the node in the adjacency matrix;

[0085] Calculate the local clustering coefficient of each node based on the sum of the edge weights of the node, the degree of the node, and the element parameters of the node in the adjacency matrix;

[0086] Calculate the load network aggregation degree of the user cluster based on the local clustering coefficient of each node;

[0087] Determine the load network density of the user cluster based on the user cluster load network;

[0088] Determine the load network synchronization degree of the user cluster based on the user cluster load network.

[0089] Among them, processing the load network aggregation degree, load network density, and load network synchronization degree in the network parameters, as well as the electricity load data of the user cluster to obtain a global objective function includes:

[0090] Construct a user cluster electricity load vector matrix based on the user load data of each user under the user cluster;

[0091] Calculate the global objective function based on the load network aggregation degree, density, synchronization degree in the network parameters, and the electricity load vector matrix.

[0092] Optionally, based on the user privacy protection device shown in the above embodiments of the present invention, the edge controller 20 processes the coordination signal of the current number of times and the electricity load data of the local user equipment in the current period to obtain the local user equipment control strategy, specifically used for:

[0093] Determine the electricity load data that meets the preset constraint regulation based on the electricity load data of the local user equipment in the current period;

[0094] Determine the local user equipment control strategy based on the electricity load data that meets the preset constraint regulation and the coordination signal of the current number of times.

[0095] Based on the user privacy protection device shown above, the process of specifically implementing the collaborative privacy protection of electricity load is as Figure 3 shown, which is a schematic flowchart of a user privacy protection method shown in the embodiments of the present invention. The method includes:

[0096] Step S301: The cloud controller receives the electricity load data of the user clusters in the historical period uploaded by each edge controller;

[0097] Optionally, for each user, the underlying smart meter collects the total load of the user and the operation boundary data of the generalized energy storage device in the historical period. After the edge controller performs privacy desensitization processing, the electricity load data is obtained, and non-sensitive parameters such as the load adjustment margin, that is, the electricity load data, are uploaded to the cloud controller.

[0098] In the process of specifically implementing step S301, the cloud controller receives the electricity load data in the historical period uploaded by each edge controller; and combines them to obtain the electricity load data of the user clusters in the historical period.

[0099] Step S302: The cloud controller performs a privacy evaluation based on the electricity load data of the user clusters and determines the global objective function.

[0100] It should be noted that in the process of specifically implementing step S302, it includes:

[0101] Step S11: For each single user in the user cluster, calculate the load correlation matrix between each user based on the user load data of each user;

[0102] In the process of specifically implementing step S11, first, randomly select two single users within any user cluster, and represent these two users as X and Y; sort the user load data corresponding to X and Y respectively in chronological order, substitute them into formula (1), and calculate the user load correlation between every two users.

[0103] Formula (1):

[0104]

[0105] Wherein, and respectively represent and sorting values of; and respectively represent the sorting means of X and Y, n is the total number of user load data of users, and i belongs to n.

[0106] Calculate the user load correlation between every two users in the user cluster based on the above formula (1); then, construct a load correlation matrix based on the user load correlation between every two users in the user cluster.

[0107] Step S12: Construct a user cluster load network based on the load correlation matrix.

[0108] It should be noted that in the process of specifically implementing step S​​12, the following steps are included.

[0109] Step S21: Take each user in the user cluster as a node in the network.

[0110] It should be noted that the nodes in the network represent the user load sequence of a certain user.

[0111] Step S22: Based on the relationship between the user load correlation in the load correlation matrix and the preset correlation threshold, determine the connection relationship of the nodes corresponding to the user load correlation and the edge weights corresponding to the nodes.

[0112] It should be noted that the process of specifically implementing step S​​22 includes the following steps.

[0113] Step S31: For the user load correlation in each load correlation matrix, judge whether the user load correlation is greater than or equal to the preset correlation threshold. If so, execute step S32; otherwise, execute step S33.

[0114] Step S32: Determine that the connection relationship of the user nodes corresponding to the user load correlation is that there is a connection edge, and determine the edge weights of the corresponding nodes based on the user load correlation.

[0115] Step S33: Determine that the connection relationship of the user nodes corresponding to the user load correlation is a non-connected edge, and determine the edge weights of the corresponding nodes based on the user load correlation.

[0116] It should be noted that the edge weight of a user node with a connection relationship of a non-connected edge is 0.

[0117] Specifically, the calculation method of the edge weights in implementing Step S32 and Step S33 can be as shown in Formula (2).

[0118] Formula (2):

[0119]

[0120] Wherein, is the correlation coefficient of the power consumption load time series of User X and User Y, that is, the load correlation; is the preset correlation threshold.

[0121] Based on Formula (2), it can be known that when the load correlation is greater than or equal to the preset correlation threshold the load correlation is used as the edge weight of the corresponding connection edge of the corresponding node . When the load correlation is greater than the preset correlation threshold, it indicates that there is no connection edge, and at this time the edge weight is 0.

[0122] Step S23: Construct a user cluster load network based on the nodes, the connection relationships of the nodes corresponding to the user load correlation, and the edge weights of the nodes corresponding to the user load correlation. [[ID=३६]]

[0123] In the process of specifically implementing Step S23, first construct a user cluster load network with the nodes obtained in Step S21, the connection relationships of the nodes determined in Steps S32 and S33, and the edge weights, that is to say, the user cluster load network is represented by G=(N, E, W), where N is the set of nodes representing the power consumption load time series of each user, that is, the nodes obtained in Step S21; E is the set of edges in the load network, that is, the combination of the connection relationships determined in Steps S32 and S33; W is the edge weight matrix established based on cross-correlation, that is, the edge weights of the nodes determined in Steps S32 and S33.

[0124] Step S13: Calculate network parameters based on the user cluster load network, where the network parameters include the load network aggregation degree, the load network density, and the load network synchronization degree.

[0125] It should be noted that the process of specifically implementing Step S13 includes the following steps.

[0126] Step S41: For each node in the user cluster load network, based on the user cluster load network, determine the sum of the edge weights of the node, the degree of the node, and the element parameter of the node in the adjacency matrix.

[0127] In the process of specifically implementing step S41, for each node in the user cluster load network, first, obtain the edges connected to the node from the user cluster load network, that is, the edge weights corresponding to the connected edges ; accumulate them to obtain the sum of the edge weights of the node ; then, query the number of connected edges of the node from the user cluster load network and use it as the degree of the node; finally, first determine the adjacent nodes of the node from the user cluster load network, that is, the adjacent nodes; use the node and all adjacent nodes as the adjacency matrix; based on the fact that there is a connected edge between each node in the adjacency matrix and a certain adjacent node, it means that the element parameter of the node and a certain adjacent node is 1, otherwise it is 0.

[0128] It should be noted that generally, a node x has two adjacent nodes, such as node y and node k; node x, node y, and node k can form a generalized triangle, that is to say, node x, node y, and node k are used as the elements in the adjacency matrix; the connection relationship between node x, node y, and node k pairwise is the element parameter in the adjacency matrix, that is .

[0129] Step S42: Calculate the local clustering coefficient of each node based on the sum of the edge weights of the node, the degree of the node, and the element parameter of the node in the adjacency matrix.

[0130] In the process of specifically implementing step S42, substitute the sum of the edge weights of the node, the degree of the node, and the element parameter of the node in the adjacency matrix into formula (3) for calculation to obtain the weighted local clustering coefficient of the node .

[0131] Formula (3):

[0132]

[0133] Among them, is the strength of node x, that is, the sum of the weights of all connected edges; K x is the degree of node x, that is, the number of nodes directly connected to node x; are all elements in the adjacency matrix of node x. If there is an edge between node x and y, then = 1, otherwise = 0. Similarly, then determine and parameters

[0134] Determine the local clustering coefficient of each node in the network through step S41 and step S42.

[0135] Step S43: Calculate the load network aggregation degree of the user cluster based on the local clustering coefficient of each node.

[0136] In the specific process of implementing step S43, substitute the load network aggregation degree of each node into formula (4) for calculation to obtain the mean value of the local clustering coefficients of all user nodes in the user cluster That is, the load network aggregation degree of the user cluster.

[0137] It should be noted that the load network aggregation degree is defined as the mean value of the local clustering coefficients of all user nodes.

[0138] Formula (4):

[0139]

[0140] Where N is the number of nodes in the network, that is, the number of users in the user cluster;

[0141] Step S44: Determine the load network density of the user cluster based on the user cluster load network.

[0142] In the specific process of implementing step S44, first, then, count the total number E of all edges from the user cluster load network, and determine the maximum possible number of edges of the network in the user cluster load network ; Second, calculate the average value of the edge weights corresponding to all nodes ; Finally, substitute the total number E of all edges, the maximum possible number of edges of the network into formula (5) for calculation to obtain the corresponding load network density of the user cluster .

[0143] Formula (5):

[0144]

[0145] Where E is the number of edges actually existing in the graph; is the maximum possible number of edges in the graph; is the average value of the weights of all existing edges.

[0146] It should be noted that the maximum possible number of edges of the network is determined according to the number of nodes N of the graph and the type of the graph.

[0147] Average value of edge weights is the average value of the edge weights of all existing edges.

[0148] Based on this, the average value of the edge weights is calculated based on each existing connected edge E in the user cluster load network, that is, an edge with a weight greater than 0, as shown in formula (6).

[0149] Formula (6):

[0150]

[0151] Wherein, represents the edge weight corresponding to the r-th edge, r is a positive integer greater than or equal to 1 and less than or equal to E, and all the edge weights in the network are respectively .

[0152] Step S45: Determine the load network synchronization degree of the user cluster based on the user cluster load network.

[0153] In the process of specifically implementing step S45, a degree matrix and an adjacency matrix are determined from the user cluster load network; and based on the degree matrix and the adjacency matrix, the corresponding Laplacian matrix L is calculated through formula (7); then, the Laplacian matrix L is substituted into formula (8) to determine a plurality of eigenvalues ; finally, the ratio of the second smallest eigenvalue to the largest eigenvalue is determined, and this ratio is used as the load network synchronization degree , as shown in formula (9).

[0154] It should be noted that generally, a node in the network corresponds to an eigenvalue λ, so all the eigenvalues in the user cluster load network are , N is a positive integer greater than or equal to 2.

[0155] The second smallest eigenvalue is also called the algebraic connectivity, which reflects the connection tightness and anti-splitting ability of the network.

[0156] Formula (7):

[0157]

[0158] Wherein, D is the degree matrix, which is a diagonal matrix; A is the adjacency matrix, and each element in A .

[0159] Formula (8):

[0160]

[0161] Among them, is the eigenvalue, is the corresponding eigenvector, is the Laplacian matrix.

[0162] Formula (9):

[0163]

[0164] Among them, is the second smallest eigenvalue, is the N eigenvalues the maximum value among them, is the load network synchronization degree.

[0165] It should be further noted that the diagonal elements in the degree matrix are calculated as shown in the following formula (10).

[0166] Formula (10):

[0167]

[0168] Among them, is the edge weight of the edge corresponding to each node, is the diagonal element of the e-th row and e-th column.

[0169] Step S14: Based on the load network aggregation degree, load network density, and load network synchronization degree in the network parameters, and the electricity consumption load data of the user cluster, perform processing to obtain a global objective function.

[0170] Among them, the global objective function refers to the intrinsic privacy of the electricity consumption load data of the user cluster; the measurement of the intrinsic privacy can be measured by the load network aggregation degree, density, and synchronization degree of the user cluster.

[0171] It should be noted that in the process of specifically implementing step S14, the following steps are included.

[0172] Step S51: Based on the user load data of each user in the user cluster, construct a user cluster electricity consumption load vector matrix P.

[0173] In the process of specifically implementing step S51, for each user in the user cluster, obtain the basic load of user X at time t from the user's user load data , the power consumption of the generalized energy storage device of user X at time t; based on the basic load of user X at time t, substitute the power consumption of the generalized energy storage device of user X at time t into the constraint conditions of formula (11) to determine the total electricity consumption load corresponding to this user at time t ; Next, take the total power consumption load corresponding to each user at time t as the elements of the matrix to construct the power consumption load vector matrix P of the user cluster.

[0174] It should be noted that the power consumption of the generalized energy storage device of user X at time t includes the power of the HVAC and electric vehicle of user X at time t.

[0175] Formula (11):

[0176]

[0177] Among them, n is the total number of nodes in the user cluster, that is, the total number of all users; T is the current time period; is the total power consumption load of user X at time t; is the basic load of user X at time t; and are the powers of the HVAC and electric vehicle of user X at time t, respectively.

[0178] The generalized energy storage device shown in this application should meet the following conditions, specifically including:

[0179] Condition 1, having a certain energy storage and release capacity; Condition 2, being controllable and responsive during operation; Condition 3, having less impact on users' daily lives; Condition 4, having a certain power and capacity margin; Condition 5, having a high market penetration rate for the device itself.

[0180] Referring to the above conditions, the generalized energy storage device shown in this application can be HVAC (Heating, Ventilation and Air Conditioning) and electric vehicle (Electric Vehicle, EV), as Figure 4 shown. That is to say, using power-consuming devices with energy storage characteristics such as electric vehicles and HVAC as generalized energy storage to replace the traditional energy storage charging and discharging behavior reduces the initial investment cost of the privacy protection scheme and provides a feasible path for the large-scale application of such technologies on the user side.

[0181] It should be noted that in addition to the above-mentioned HVAC and EV shown, other devices that meet the above conditions can also be used, and the specific embodiments of this application are not limited thereto.

[0182] Based on this, the power consumption of the generalized energy storage device of user X at time t includes the power of the HVAC and electric vehicle of user X at time t, that is and .

[0183] The power of the HVAC and electric vehicle of user x at time t can be determined by the above formulas (12) and (13), that is and .

[0184] Formula (12):

[0185]

[0186] where and represent the indoor and outdoor temperatures at time t, respectively; is the heat transfer coefficient; is the electro-thermal conversion coefficient of the HVAC; is the operating power of the HVAC at time t. Note that when the HVAC is in heating mode, the HVAC is in cooling mode.

[0187] Furthermore, consider the maximum rated heat power constraint condition and the household user temperature comfort constraint condition of the HVAC as follows:

[0188]

[0189] where is the maximum rated heat power of the HVAC; [T min , T max is the user temperature comfort interval; the operating power of the HVAC at time t needs to be greater than or equal to 0 and less than the maximum rated heat power. And the indoor temperature at time t is greater than or equal to the minimum user temperature, or less than or equal to the maximum user temperature.

[0190] Formula (12) can be the electro-thermal mechanics principle of the HVAC.

[0191] Formula (13): As a typical controllable energy storage unit, the charging process of an electric vehicle can be described by the dynamic change of the state of charge (SOC) of the battery. In the discrete time domain, the evolution equation of SOC is shown as follows:

[0192] Formula (13):

[0193]

[0194] where is the charging power of the electric vehicle at time t; is the time step; is the rated capacity of the electric vehicle battery, is the state of charge of the electric vehicle at time t, is the state of charge of the electric vehicle at time t + 1.

[0195] Considering the safety limitations during the charging process of electric vehicles and the actual daily charging needs of users, the following constraints exist:

[0196]

[0197] Among them, is the maximum charging power of the electric vehicle; and are the upper and lower limits of the state of charge of the electric vehicle, respectively; is the state of charge of the electric vehicle when the user uses it in the morning every day; is the state of charge required by the user when using the electric vehicle in the morning every day, representing the daily charging demand of the user; is the moment when the electric vehicle starts charging after being used up in the evening every day; is the moment when the electric vehicle ends charging in the morning every day.

[0198] Based on the above formulas (12) and (13), it can be known that the operating power of the HVAC at time t can be determined by using the indoor and outdoor temperatures and the electrothermal conversion coefficient of the HVAC in the electricity load data at time t ; the power of the electric vehicle at time t can be determined by using the rated capacity of the electric vehicle battery, the state of charge SOC of the electric vehicle at time t, and the state of charge SOC of the electric vehicle at time t + 1 .

[0199] Step S52: Calculate the global objective function based on the load network aggregation degree , density , synchronization degree in the network parameters and the electricity load vector matrix P.

[0200] In the specific process of implementing step S52, substitute the electricity load vector matrix P, the load network aggregation degree , density and synchronization degree of the user cluster into formula (14) for calculation to obtain the objective function.

[0201] Formula (14):

[0202]

[0203] Among them, , and represent the load network aggregation degree, density, and synchronization degree of the user cluster, respectively; P is the electricity load vector matrix of the user cluster, and each column represents the electricity load sequence of a user; , and $\omega$ is the weight coefficient of each index, which is preset. To simplify the optimization model, generally, the weights of the three indexes can be regarded equally. $J$ is the global objective function.

[0204] Step S303: The cloud controller generates the coordination signal of the current time based on the global objective function and the coordination signal of the previous time, and sends the coordination signal of the current time to each edge controller.

[0205] Among them, if the previous time is the first time to send the coordination signal, the coordination signal of the previous time refers to the initial coordination signal; if the previous time is not the first time to send the coordination signal, the coordination signal of the previous time is the coordination signal obtained by processing the global objective function and the coordination signal corresponding to its previous time.

[0206] In the process of specifically implementing step S303, first determine the gradient of the load corresponding to the global objective function , and then substitute the gradient of the load corresponding to the global objective function , the coordination signal of the previous time and the preset step factor into formula (15) for calculation to obtain the coordination signal of the current time

[0207] Formula (15):

[0208]

[0209] In the embodiment of the present invention, the coordination signal generated each time is used to reflect the sensitivity of the overall privacy protection level of the current device to the adjustment of the electricity consumption load of each user.

[0210] Taking the cloud controller's first sending of the coordination signal as an example, that is, if the previous time is the first time, when the cloud controller receives the coordination start signal, it uses a random generation algorithm to randomly generate a coordination signal, that is, the initialization coordination signal , that is, the initial coordination signal.

[0211] Furthermore, it should be noted that after the cloud controller generates the initial coordination signal, it will first send the initial coordination signal to each edge controller, so that each edge controller can use the device control strategy generated based on the initial coordination signal and the electricity consumption load data of the local user equipment in the historical time period to control the corresponding generalized energy storage device; and collect new electricity consumption load data to update the electricity consumption load data in the historical time period and send it to the cloud controller; at this time, the cloud controller can start executing from step S301.

[0212] Taking the case where the cloud controller does not send the coordination signal for the first time, that is, if the previous time is not the first time to send the coordination signal, it means that the coordination signal of the previous time is also calculated by formula (15). It is necessary to use the updated global objective function and its previous time, that is, the coordination signal corresponding to the time before the previous time, and substitute it into formula (15) for processing to obtain, that is, the process of the current step 303.

[0213] Step S304: For each edge controller, when the edge controller receives the coordination signal of the current time, the edge controller processes the coordination signal of the current time and the power consumption load data of the local user equipment in the current period to obtain the local user equipment control strategy, so as to control the corresponding local user equipment based on the local user equipment control strategy.

[0214] It should be noted that in the process of specifically implementing step S304 where the edge controller processes the coordination signal of the current time and the power consumption load data of the local user equipment in the current period to obtain the local user equipment control strategy, the following steps are included.

[0215] Step S61: Determine the power consumption load data that meets the preset constraint regulation based on the power consumption load data of the local user equipment in the current period.

[0216] It should be noted that the preset constraints include HVAC constraints and EV constraints, which can be specifically shown in formula (16).

[0217] In the process of specifically implementing step S61, obtain the power consumption load data of the local user equipment in the current period; first, according to the data at each moment t in the current period T, make it meet the HVAC constraints and EV constraints of formula (16) to obtain the and of the edge controller a that meet the HVAC constraints and EV constraints of formula (16), that is, the power consumption load data that meets the preset constraint regulation.

[0218] The local user equipment includes the real-time status of the controllable devices such as the user's own HVAC and electric vehicle corresponding to the edge controller (including real-time device status information such as the current temperature, HVAC power, electric vehicle charging and discharging strategy, and electric vehicle battery SOC status).

[0219] Among them, the HVAC operating power of the a-th edge controller at time t and the power of the electric vehicle of the a-th edge controller at time t .

[0220] Formula (16):

[0221]

[0222]

[0223] Among them, a represents the a-th edge controller, where a is less than or equal to A and greater than or equal to 1. and respectively represent the indoor / outdoor temperature at time t in the electrical load data corresponding to the a-th edge controller; is the heat conduction coefficient of the a-th edge controller; is the HVAC electro-thermal conversion coefficient of the a-th edge controller; the HVAC operating power of the a-th edge controller at time t , and the power of the electric vehicle of the a-th edge controller at time t . needs to satisfy being greater than or equal to 0 and less than or equal to the maximum rated power of the electric vehicle of the a-th edge controller ; needs to be within , is the temperature comfort interval of the user corresponding to the a-th edge controller.

[0224] is the maximum charging power of the electric vehicle of the user corresponding to the a-th edge controller , respectively represent the upper and lower limits of the state of charge of the electric vehicle corresponding to the a-th edge controller; is the state of charge of the electric vehicle when the user corresponding to the a-th edge controller uses the electric vehicle every morning; is the state of charge required when the user corresponding to the a-th edge controller uses the electric vehicle every morning, representing the user's daily charging demand; is the time when the electric vehicle starts charging after being used up every evening; is the time when the electric vehicle finishes charging every morning.

[0225] Step S62: Determine the local user equipment control strategy based on the electrical load data that meets the preset constraints and the coordination signal of the current number of times.

[0226] It should be noted that the local user equipment control strategy here refers to the control strategy of the generalized energy storage equipment.

[0227] In the specific process of implementing step S62, the and calculated in the above step S61 for the edge controller during the time period T, and the coordination signal of the current number of times are substituted into formula (17) for calculation to determine the and when the objective function has the minimum value.Determine the local user equipment control strategy constructed by the temperature setting value adjustment instruction and the electric vehicle charging power curve based on the change.

[0228] Formula (17):

[0229]

[0230] The edge controller calculates the optimal heating, ventilation, and air conditioning (HVAC) temperature control strategy and electric vehicle charging strategy on the premise of meeting the local device constraints.

[0231] Specifically, the process of generating the specific temperature setting value adjustment instruction and the electric vehicle charging power curve includes: inputting the electricity load data of the local user equipment in the current period (such as the indoor and outdoor temperatures, operating power of the HVAC, and the comfortable interval set by the user, the current state of charge (SOC), charging power, battery capacity, and the required SOC for the next day of the electric vehicle), the coordination signal sent by the cloud controller for the current time , and the preset constraint conditions (such as power upper limit, SOC safety range, etc.) into the model constructed by Formula (16) and Formula (17); first, predict the temperature change trajectory under different power adjustments based on the electro-thermodynamic model, use the coordination signal as the dynamic weight factor, and solve the optimal power adjustment sequence through rolling horizon optimization on the premise of meeting the hard constraints of user temperature comfort, and convert the result into a fine-tuning instruction for the temperature setting value (such as ), that is, the temperature control strategy.

[0232] At the same time, predict the charging model according to the SOC evolution equation, combine the coordination signal with the constraints of the preset constraint conditions, generate a charging power curve with randomization characteristics on the basis of ensuring the SOC demand for the next day, and use the charging power curve as the charging control strategy.

[0233] Finally, combine the temperature control strategy and the charging control strategy to generate the local user equipment control strategy, so as to control the corresponding local user equipment based on the local user equipment control strategy. That is to say, drive the actuators of the user's home intelligent thermostat and charging pile through the corresponding communication interface with the temperature control strategy and the charging control strategy to complete the load disturbance.

[0234] It should be noted that load disturbance refers to adjusting the user's actual electricity consumption behavior to prevent the generation of virtual loads.

[0235] Step S305: The edge controller obtains the new electricity load data of the local user equipment after being controlled by the local user equipment control strategy and returns it to the cloud controller.

[0236] In the process of specifically implementing step S305, after the edge controller controls the local user device based on the local user device control policy, it is necessary to re-record the power consumption load data of the local user device for a period of time, that is, the new power consumption load data, and return it to the cloud controller for the cloud controller to adjust.

[0237] Step S306: The cloud controller determines whether the threshold corresponding to the current coordination signal is greater than or equal to the preset threshold. If so, execute step S307; if not, execute step S308.

[0238] In the process of specifically implementing step S306, when executing step S303 to send the current coordination signal to each edge controller, compare the threshold corresponding to the current coordination signal with the preset threshold to determine whether further convergence is required. When it is determined that the threshold corresponding to the current coordination signal is greater than or equal to the preset threshold, execute step 307; when it is determined that the threshold corresponding to the current coordination signal is less than the preset threshold, execute step S308.

[0239] Step S307: Receive the new power consumption load data uploaded by the edge controller, and based on the new power consumption load data uploaded by the edge controller, return to execute step S302.

[0240] In the process of specifically implementing step S307, it indicates that convergence has not occurred at this time. At this time, it is necessary to adjust the global objective function, and then adjust the coordination signal, that is, based on the new power consumption load data uploaded by the edge controller, return to execute step S302.

[0241] It should be noted that the cloud controller will return to execute step S302 based on the new power consumption load data uploaded by each edge controller.

[0242] Step S308: Determine that the signal convergence ends, so that each subsequent edge controller can update the local user device control policy in real time based on the current coordination signal.

[0243] In the process of specifically implementing step S308, the global objective function obtained at this time is the optimal global objective function. Each subsequent edge controller can directly use the current coordination signal as the best coordination signal to update the local user device control policy in real time.

[0244] Optionally, when it is determined that the coordination signal convergence ends, the global objective function will no longer be optimized. Each subsequent edge controller can execute step S304 based on the local power consumption device. At this time, the cloud controller can no longer interpret the power consumption load data uploaded by the edge control.

[0245] Optionally, based on the above, the present invention shows a framework diagram of a method for protecting the privacy of user cluster power consumption load data based on distributed cloud-edge collaborative control, as Figure 5 shown.

[0246] Figure 5 In it, a distribution automation master station and a distribution automation cloud platform are deployed in the cloud controller to have a data processing center and communication network facilities; thereby realizing the content described in steps S301 to S303, and steps S306 to S308.

[0247] A home energy management system and an intelligent power consumption management terminal are deployed in the edge controller, and the edge computing module and the intelligent gateway are embedded in the home energy management system and the intelligent power consumption management terminal to be used to implement step S304.

[0248] The edge controller collects the operation conditions of equipment such as heating, ventilation, air conditioning and electric vehicles through local measuring devices, processes them into the total load of the user cluster and uploads it to the cloud controller; the cloud controller processes based on the total load of the user cluster to generate a new coordination signal; the edge controller combines its own equipment operation conditions to generate a local user equipment control strategy, that is, a control signal, to control heating, ventilation, air conditioning and electric vehicles, thereby reducing the load correlation of the user cluster.

[0249] The cloud controller of the present application is regulated by the power grid, which can ensure that the user's generalized energy storage device makes adjustments within the safe operation boundary of the power grid, ensuring the scalability and security of the system; at the same time, the local calculation and optimization of the edge controller reduce the transmission range of user privacy data and allow the number of edge-side users to increase or decrease flexibly.

[0250] In the embodiment of the present invention, the underlying intelligent electricity meter corresponding to each user of each edge controller collects electricity consumption load data such as the total load of the corresponding user and the operation boundary data of the generalized energy storage device. After being subjected to privacy desensitization processing by the edge controller, non-sensitive parameters such as load adjustment margins are uploaded to the cloud controller; the cloud controller evaluates the privacy of the user cluster power consumption load data based on complex network theory, and depicts the structural characteristics of the user cluster load network and its impact on privacy from different dimensions through network indicators such as aggregation degree, density and synchronization degree to optimize the global objective function, and then generates a spatio-temporally correlated coordination signal and sends it to each edge controller; the edge controller combines the real-time state of local equipment to process and generate specific temperature setting value adjustment instructions and electric vehicle charging power curves, and finally drives the intelligent temperature controller and the charging pile actuator to complete the load disturbance through the corresponding communication interface. To reduce the load cluster effect and improve the privacy of the user cluster power consumption load data on the premise that the user cluster is not mutually trusted and the device parameters are not shared.

[0251] An embodiment of the present application provides an electronic device, which includes a processor and a memory. The memory is used to store user privacy protection program codes and data, and the processor is used to call program instructions in the memory to execute the steps implemented as shown in the user privacy protection method in the above embodiment.

[0252] An embodiment of the present invention provides a storage medium, which includes the electronic device provided in the embodiment of the present application, and this electronic device is used to execute the user privacy protection method disclosed in the embodiment of the present application.

[0253] Each embodiment in this specification is described in a progressive manner. For the same or similar parts among the embodiments, reference can be made to each other. Each embodiment focuses on the differences from other embodiments. In particular, for a system or a system embodiment, since it is basically similar to the method embodiment, the description is relatively simple, and for the relevant parts, reference can be made to the partial description of the method embodiment. The systems and system embodiments described above are merely illustrative. The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed to multiple network units. Some or all of the modules can be selected according to actual needs to achieve the purpose of the solution of this embodiment. A person of ordinary skill in the art can understand and implement it without creative work.

[0254] Those skilled in the art can further realize that the units and algorithm steps of each example described in combination with the embodiments disclosed in this article can be implemented by electronic hardware, computer software, or a combination of the two. To clearly illustrate the interchangeability of hardware and software, the composition and steps of each example have been generally described according to functions in the above description. Whether these functions are executed in a hardware or software manner depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods for each specific application to implement the described functions, but such implementation should not be considered to exceed the scope of the present invention.

[0255] The above description of the disclosed embodiments enables those skilled in the art to implement or use the present invention. Various modifications to these embodiments will be obvious to those skilled in the art, and the general principles defined herein can be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention will not be limited to the embodiments shown herein, but will be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A user privacy protection method, characterized in that, Applied to a user privacy protection device, the device includes a cloud controller and multiple edge controllers, and the cloud controller is respectively connected to the multiple edge controllers. The method includes: The cloud controller receives the power consumption load data of the user cluster in the historical period uploaded by each of the edge controllers; The cloud controller performs privacy evaluation based on the power consumption load data of the user cluster to determine the global objective function; The cloud controller generates the coordination signal of the current number based on the global objective function and the coordination signal of the previous number, and sends the coordination signal of the current number to each edge controller. Among them, if the previous number is the first time to send the coordination signal, the coordination signal of the previous number refers to the initial coordination signal. If the previous number is not the first time to send the coordination signal, the coordination signal of the previous number is the coordination signal obtained by processing the global objective function and the coordination signal corresponding to its previous number; For each edge controller, when the edge controller receives the coordination signal of the current number, the edge controller processes the coordination signal of the current number and the power consumption load data of the local user device in the current period to obtain the local user device control strategy, so as to control the corresponding local user device based on the local user device control strategy; The edge controller obtains the new power consumption load data of the local user device after being controlled by the local user device control strategy and returns it to the cloud controller; When the cloud controller determines that the threshold corresponding to the coordination signal of the current number is less than the preset threshold, it determines that the signal convergence ends, so that each subsequent edge controller can update the local user device control strategy in real time based on the coordination signal of the current number.

2. The method according to claim 1, characterized in that It further includes: When the cloud controller determines that the threshold corresponding to the coordination signal of the current number is greater than or equal to the preset threshold, it receives the new power consumption load data uploaded by the edge controller, and based on the new power consumption load data uploaded by the edge controller, returns to execute the step of processing the power consumption load data of the user cluster to determine the global objective function.

3. The method according to claim 1, characterized in that, The cloud controller performs privacy evaluation based on the power consumption load data of the user cluster to determine the global objective function, including: For each individual user in the user cluster, calculate the load correlation matrix between users based on the user load data of each user; Construct a user cluster load network based on the load correlation matrix; Calculate network parameters based on the user cluster load network, where the network parameters include load network aggregation degree, load network density, and load network synchronization degree; Process the load network aggregation degree, load network density, and load network synchronization degree in the network parameters, and the power consumption load data of the user cluster to obtain the global objective function.

4. The method according to claim 3, wherein Constructing a user cluster load network based on the load correlation matrix includes: Regarding each user in the user cluster as a node in the network; Determine the connection relationship of the nodes corresponding to the user load correlation and the edge weights corresponding to the nodes based on the relationship between the user load correlation in the load correlation matrix and the preset correlation threshold; Construct a user cluster load network based on the nodes, the connection relationship of the nodes corresponding to the user load correlation, and the edge weights of the nodes corresponding to the user load correlation.

5. The method according to claim 3, characterized in that Calculate network parameters based on the user cluster load network, including: For each node in the user cluster load network, determine the sum of the edge weights of the node, the degree of the node, and the element parameters of the node in the adjacency matrix based on the user cluster load network; Calculate the local clustering coefficient of each node based on the sum of the edge weights of the node, the degree of the node, and the element parameters of the node in the adjacency matrix; Calculate the load network aggregation degree of the user cluster based on the local clustering coefficient of each node; Determine the load network density of the user cluster based on the user cluster load network; Determine the load network synchronization degree of the user cluster based on the user cluster load network.

6. The method according to claim 3, characterized in that, Based on the load network aggregation degree, load network density, and load network synchronization degree in the network parameters, and the power consumption load data of the user cluster, perform processing to obtain a global objective function, including: Construct a user cluster power consumption load vector matrix based on the user load data of each user under the user cluster; Calculate the global objective function based on the load network aggregation degree, density, synchronization degree in the network parameters, and the power consumption load vector matrix.

7. The method according to claim 1, characterized in that, The edge controller processes the coordination signal of the current time and the power consumption load data of the local user device in the current time period to obtain a local user device control strategy, including: Determine the power consumption load data that meets the preset constraint regulation based on the power consumption load data of the local user device in the current time period; Determine the local user device control strategy based on the power consumption load data that meets the preset constraint regulation and the coordination signal of the current time.

8. A user privacy protection device, characterized in that, It includes a cloud controller and multiple edge controllers, and the cloud controller is respectively connected to the multiple edge controllers; The cloud controller is used to receive the power consumption load data of the user cluster in the historical time period uploaded by each edge controller; process the power consumption load data of the user cluster to determine the global objective function; Generate a coordination signal of the current time based on the global objective function and the coordination signal of the previous time, and send the coordination signal of the current time to each edge controller, where if the previous time is the first time to send the coordination signal, the coordination signal of the previous time refers to the initial coordination signal, and if the previous time is not the first time to send the coordination signal, the coordination signal of the previous time is the coordination signal obtained by processing the global objective function and the coordination signal corresponding to its previous time; For each edge controller, the edge controller is configured to, when receiving the coordination signal of the current number of times, process the coordination signal of the current number of times and the power consumption load data of the local user equipment in the current period to obtain a local user equipment control strategy, so as to control the corresponding local user equipment based on the local user equipment control strategy; obtain the new power consumption load data of the local user equipment after being controlled by the local user equipment control strategy, and return it to the cloud controller; The cloud controller is configured to, when determining that the threshold corresponding to the coordination signal of the current number of times is less than a preset threshold, determine that the signal convergence ends, so that subsequent edge controllers can update the local user equipment control strategy in real time based on the coordination signal of the current number of times.

9. An electronic device, characterized in that, The electronic device includes a processor and a memory. The memory is used to store program codes and data for user privacy protection, and the processor is used to call the program instructions in the memory to execute the user privacy protection method according to any one of claims 1-7.

10. A storage medium, characterized in that, The storage medium includes a stored program, wherein when the program runs, it controls the device where the storage medium is located to execute the user privacy protection method according to any one of claims 1-7.

Images