A cloud edge end collaborative high concurrency access method and system

Abstract

The application belongs to the technical field of power systems, and discloses a cloud-edge-end cooperative high-concurrency access method and system. The method comprises the following steps: collecting power service data of a service terminal, analyzing the power service data, determining the edge gateway access type of the power service data, and determining the access channel priority of the service terminal accessing the edge gateway according to the edge gateway access type; the access channel priority is sorted in descending order to obtain a first sorting queue, and the service terminal meeting the access channel pre-allocation judgment threshold is accessed to the corresponding edge gateway according to the first sorting queue; the profit of the edge gateway accessing the cloud server is determined according to the multi-time scale terminal high-concurrency access scene of the power system, and the edge gateways not accessing the cloud server are sorted in descending order according to the profit to obtain a second sorting queue, and the edge gateways are accessed to the cloud server corresponding to the highest profit according to the second sorting queue. The application meets the differentiated access demand of a large number of service terminals.

Description

Technical Field

[0001] This invention relates to the field of power system technology, and in particular to a cloud-edge-device collaborative high-concurrency access method and system. Background Technology

[0002] Cloud-edge-device collaboration combines the advantages of cloud computing and edge computing, significantly reducing the bandwidth requirements for edge-cloud data transmission and improving data processing efficiency and business responsiveness. However, with the large-scale integration of high-proportion renewable energy sources, new business models such as source-grid-load-storage coordinated scheduling, distributed power source management, and grid situational awareness have high concurrent access demands and large data collection volumes. To meet the high-concurrency access requirements of different businesses, it is necessary to rationally allocate communication resources and perform channel pre-allocation and load balancing. However, traditional cloud-edge-device collaboration lacks effective edge-side channel pre-allocation and edge-cloud data access mechanisms. Under the constraints of limited communication, computing, and storage resources, it cannot meet the high-concurrency access and data processing requirements, resulting in problems such as unbalanced cloud-edge load and inability to guarantee business needs.

[0003] Patent No. 202010582955.8 proposes a channel resource allocation system and method. Based on terminal access requests, resources are allocated in edge devices. If these requests cannot be fulfilled, resources are allocated in the cloud service center at different times according to the service level of the request. From the perspective of resource service matching and integration, service categories and levels are distinguished, and different cloud-edge collaborative resources are allocated according to different service types and levels, ensuring optimized allocation of edge channel resources and thus improving the reliability of large-scale terminal access. Patent No. 202210635189.6 proposes a cloud platform load balancing method. It uses an active acquisition method to obtain network response latency, obtains access load information based on the network response latency value, and connects to a load balancing server for processing to achieve cloud platform load balancing. However, the above technologies only optimize from the perspective of terminal access management or load balancing, and still have the following drawbacks.

[0004] 1. Existing technologies require a large amount of frequent signaling interaction between the terminal and the edge gateway, ignoring the edge layer queue status and the differentiated access success rate constraints of the end layer, and cannot adapt to scenarios with a large number of terminals accessing at high concurrency.

[0005] 2. Existing technologies ignore the service priority of terminals, making it difficult to meet the differentiated needs of different services. At the same time, they do not consider the evolution of cloud-edge node queue backlog and spatiotemporal coupling, as well as the impact of the queue backlog difference between cloud-edge nodes on high-concurrency access decisions, resulting in severe data backlog on some network nodes, which cannot meet the timely response requirements of services under high-concurrency access of massive terminals. Summary of the Invention

[0006] This invention provides a cloud-edge-device collaborative high-concurrency access method and system to solve the aforementioned technical problems in the prior art.

[0007] To provide a basic understanding of some aspects of the disclosed embodiments, a brief summary is given below. This summary is not intended as a general commentary, nor is it intended to identify key / important components or to describe the scope of protection of these embodiments. Its sole purpose is to present some concepts in a simple form as a prelude to the detailed description that follows.

[0008] According to a first aspect of the present invention, a cloud-edge-device collaborative high-concurrency access method is provided, which is applied to a high-concurrency access scenario of multiple time scale terminals in a power system.

[0009] In one embodiment, the cloud-edge-device collaborative high-concurrency access method includes:

[0010] Collect power service data from service terminals, analyze the power service data, determine the edge gateway access type of the power service data, and determine the access channel priority of the service terminal to access the edge gateway based on the edge gateway access type;

[0011] The access channels are sorted in descending order of priority to obtain a first sorting queue, and service terminals that meet the access channel pre-allocation judgment threshold are connected to the corresponding edge gateway according to the first sorting queue.

[0012] Based on the high-concurrency access scenario of multiple time-scale terminals in the power system, the profit of edge gateways accessing cloud servers is determined. Based on the profit, edge gateways that are not connected to cloud servers are sorted in descending order to obtain a second sorting queue. Based on the second sorting queue, edge gateways are connected to the cloud server corresponding to the highest profit.

[0013] In one embodiment, the high-concurrency access scenario of multi-timescale terminals in the power system includes a terminal layer, an edge layer, and a cloud layer.

[0014] The terminal layer includes N service terminals, and the set of service terminals is D = {d1,...,d2}. n ,...,d N The edge layer includes J edge gateways, and the set of edge gateways is E = {e1,...,e...}. j ,...,e J The set of service terminals within the communication range of the edge gateway is: The cloud layer includes a high-concurrency access management platform and K cloud servers, the set of which is S = {s1,...,s...} k ,...,s K};

[0015] Furthermore, the service terminal transmits the collected data to the edge gateway via a power line carrier communication network; the edge gateway transmits the data to D within its communication range. j Each terminal is pre-allocated an access channel and has D j Each data queue is used to store uncomputed business terminal data; the high-concurrency access management platform dynamically adjusts the edge gateway data access decision based on the queue backlog of the edge gateway and cloud server;

[0016] The terminal layer, the edge layer, and the cloud layer adopt a multi-time scale model to divide the data transmission time into I time periods, and each time period contains T0 time slots with a time slot length of τ.

[0017] In one embodiment, the service terminal and the edge gateway have a first data queue backlog evolution model, and the formula of the first data queue backlog evolution model is:

[0018]

[0019]

[0020]

[0021] The edge gateway and the cloud server have a second data queue backlog evolution model, and the formula for the second data queue backlog evolution model is:

[0022]

[0023]

[0024] In the formula, The edge gateway caches the data queue of service terminals. Let t be the amount of data uploaded by the service terminal to the edge gateway in time slot t. This refers to the amount of data uploaded from the edge gateway to the cloud server regarding the business terminal. Pre-assign indicator variables to the channel in the i-th time period. This indicates that a transmission channel has been pre-allocated for the service terminal; otherwise... The upload speed from the service terminal to the edge gateway. Minimum data collection constraint for the business terminal; This refers to the upload speed from the edge gateway to the cloud server. x represents the amount of data uploaded to the cloud server by the service terminals in the edge gateway. j,k (t) is the indicator variable for selecting the cloud server in the t-th time slot, x j,k(t) = 1 indicates that the edge gateway uploads data to the cloud server for computation; otherwise, x j,k (t)=0, Z j，k (t) represents the data queue of the cloud server caching the edge gateway. Y represents the amount of data uploaded from the edge gateway to the cloud server in time slot t. j,k (t) represents the amount of data processed by the cloud server from the edge gateway in the t-th time slot. The cloud server in time slot t is used to process the data from the business terminals in the edge gateway. The computing resources required to process each bit of service terminal data.

[0025] In one embodiment, the edge gateway includes a first load balancing model for load balancing the data queues of service terminals, and the formula for the first load balancing model is:

[0026]

[0027] The cloud server has a second load balancing model for load balancing the data queues of the edge gateway. The formula for the second load balancing model is:

[0028]

[0029] In the formula, For the average queue backlog of edge gateways, The edge gateway caches the data queue of service terminals. Z represents the average queue backlog for cloud servers. j，k (t) represents the data queue of the cloud server caching the edge gateway. For load balancing of data queues of service terminals in the edge gateway, This refers to the load balancing of data queues in the edge gateway of the cloud server.

[0030] In one embodiment, analyzing the power service data to determine the edge gateway access type of the power service data includes:

[0031] Analyze the power business data at each time point to determine whether the business terminal corresponding to the power business data is a newly connected business terminal;

[0032] If the judgment result is yes, the service data feature vector of the power service data and the data traffic feature vector of the service terminal are compared to calculate the service matching degree, and the edge gateway access type is determined based on the service matching degree; if the judgment result is no, the edge gateway access type is determined by using the historical service classification results.

[0033] The formula for calculating the business matching degree is as follows:

[0034]

[0035] In the formula, X is the data traffic feature vector of the business terminal. m Let m be the data feature vector of the m-th type of business. Matching to business needs.

[0036] In one embodiment, determining the access channel priority for a service terminal to access the edge gateway based on the edge gateway access type includes:

[0037] Based on the access type and historical power service data, the average queue deviation of the service terminal and the access success rate deviation between the service terminal and the edge gateway are calculated; wherein, the average queue deviation includes the deviation of queue backlog, queue input, queue output from the average value of similar service terminals;

[0038] The access channel priority for a service terminal to access the edge gateway is calculated based on the average queue deviation and the access success rate deviation.

[0039] In one embodiment, the formula for calculating the average deviation of the queue is:

[0040]

[0041]

[0042]

[0043] In the formula, For queue backlog deviation, Input deviation for the queue, For queue output deviation, The edge gateway caches the data queue of service terminals. Let t be the amount of data uploaded by the service terminal to the edge gateway in time slot t. The amount of data about the business terminal uploaded from the edge gateway to the cloud server, |S m (i)| represents the set of business types S m The number of elements in (i).

[0044] In one embodiment, the formula for calculating the access success rate deviation is:

[0045]

[0046] In the formula, To account for the deviation in access success rate, c n For the minimum expected constraints of service terminals and edge gateways, Pre-assign indicator variables to the channel in the i-th time period, where l is the summation index.

[0047] In one embodiment, the formula for calculating the access channel priority is:

[0048]

[0049] In the formula, Priority for access channels, and A larger value indicates a greater likelihood of pre-allocating transmission channels to service terminals in the next time period. For queue backlog deviation, Input deviation for the queue, For queue output deviation, This is to account for the deviation in access success rate.

[0050] In one embodiment, the formula for calculating the profit of an edge gateway accessing a cloud server in a high-concurrency access scenario involving multiple time-scale terminals in a power system is as follows:

[0051]

[0052] In the formula, χ j，k (t) represents the profit of the edge gateway accessing the cloud server, p j,k (t) represents the bidding cost that the edge gateway needs to pay to access the cloud server. The business priority weight for the business terminal. For load balancing of data queues of service terminals in the edge gateway, This refers to the load balancing of data queues in the edge gateway of the cloud server.

[0053] According to a second aspect of the present invention, a cloud-edge-device collaborative high-concurrency access system is provided, which is applied to high-concurrency access scenarios of terminals with multiple time scales in power systems.

[0054] In one embodiment, the cloud-edge-device collaborative high-concurrency access system includes:

[0055] The priority determination module is used to collect power service data from the service terminal, analyze the power service data, determine the edge gateway access type of the power service data, and determine the access channel priority of the service terminal to access the edge gateway based on the edge gateway access type.

[0056] The edge access module is used to sort the access channels in descending order to obtain a first sorting queue, and according to the first sorting queue, to connect the service terminals that meet the access channel pre-allocation judgment threshold to the corresponding edge gateway.

[0057] The edge-cloud access module is used to determine the profit of edge gateways accessing cloud servers based on the high-concurrency access scenario of multi-time-scale terminals in the power system, and to sort the edge gateways that have not been connected to cloud servers in descending order according to the profit to obtain a second sorting queue. According to the second sorting queue, the edge gateways are connected to the cloud server corresponding to the highest profit.

[0058] In one embodiment, the high-concurrency access scenario of multi-timescale terminals in the power system includes a terminal layer, an edge layer, and a cloud layer.

[0059] The terminal layer includes N service terminals, and the set of service terminals is D = {d1,...,d2}. n ,...,d N The edge layer includes J edge gateways, and the set of edge gateways is E = {e1,...,e...}. j ,...,e J The set of service terminals within the communication range of the edge gateway is: The cloud layer includes a high-concurrency access management platform and K cloud servers, the set of which is S = {s1,...,s...} k ,...,s K};

[0060] Furthermore, the service terminal transmits the collected data to the edge gateway via a power line carrier communication network; the edge gateway transmits the data to D within its communication range. j Each terminal is pre-allocated an access channel and has D j Each data queue is used to store uncomputed business terminal data; the high-concurrency access management platform dynamically adjusts the edge gateway data access decision based on the queue backlog of the edge gateway and cloud server;

[0061] The terminal layer, the edge layer, and the cloud layer adopt a multi-time scale model to divide the data transmission time into I time periods, and each time period contains T0 time slots with a time slot length of τ.

[0062] In one embodiment, the service terminal and the edge gateway have a first data queue backlog evolution model, and the formula of the first data queue backlog evolution model is:

[0063]

[0064]

[0065]

[0066] The edge gateway and the cloud server have a second data queue backlog evolution model, and the formula for the second data queue backlog evolution model is:

[0067]

[0068]

[0069] In the formula, The edge gateway caches the data queue of service terminals. Let t be the amount of data uploaded by the service terminal to the edge gateway in time slot t. This refers to the amount of data uploaded from the edge gateway to the cloud server regarding the business terminal. Pre-assign indicator variables to the channel in the i-th time period. This indicates that a transmission channel has been pre-allocated for the service terminal; otherwise... The upload speed from the service terminal to the edge gateway. Minimum data collection constraint for the business terminal; This refers to the upload speed from the edge gateway to the cloud server. x represents the amount of data uploaded to the cloud server by the service terminals in the edge gateway. j,k (t) is the indicator variable for selecting the cloud server in the t-th time slot, x j,k (t) = 1 indicates that the edge gateway uploads data to the cloud server for computation; otherwise, x j,k (t)=0, Z j，k (t) represents the data queue of the cloud server caching the edge gateway. Y represents the amount of data uploaded from the edge gateway to the cloud server in time slot t. j,k (t) represents the amount of data processed by the cloud server from the edge gateway in the t-th time slot. The cloud server in time slot t is used to process the data from the business terminals in the edge gateway. The computing resources required to process each bit of service terminal data.

[0070] In one embodiment, the edge gateway includes a first load balancing model for load balancing the data queues of service terminals, and the formula for the first load balancing model is:

[0071]

[0072] The cloud server has a second load balancing model for load balancing the data queues of the edge gateway. The formula for the second load balancing model is:

[0073]

[0074] In the formula, For the average queue backlog of edge gateways, The edge gateway caches the data queue of service terminals. Z represents the average queue backlog for cloud servers. j，k (t) represents the data queue of the cloud server caching the edge gateway. For load balancing of data queues of service terminals in the edge gateway, This refers to the load balancing of data queues in the edge gateway of the cloud server.

[0075] In one embodiment, the priority determination module includes a terminal judgment module, a matching degree calculation module, and a type determination module, wherein,

[0076] The terminal judgment module is used to analyze the power business data at each time point and determine whether the business terminal corresponding to the power business data is a newly accessed business terminal.

[0077] The matching degree calculation module is used to compare the business data feature vector of the power business data with the data flow feature vector of the business terminal when the judgment result is yes, and calculate the business matching degree.

[0078] The type determination module is used to determine the edge gateway access type based on the service matching degree when the judgment result is yes; and to determine the edge gateway access type based on the historical service classification results when the judgment result is no.

[0079] The formula for calculating the business matching degree is as follows:

[0080]

[0081] In the formula, X is the data traffic feature vector of the business terminal. m Let m be the data feature vector of the m-th type of business. Matching to business needs.

[0082] In one embodiment, the priority determination module includes a queue average calculation module, an access success rate deviation calculation module, and a priority calculation module, wherein,

[0083] The queue average value calculation module is used to calculate the queue average value deviation of the service terminal based on the access type and historical power service data. The queue average value deviation includes the queue backlog, queue input, queue output and the average value deviation of the same type of service terminal.

[0084] The access success rate deviation calculation module is used to calculate the access success rate deviation between the service terminal and the edge gateway based on the access type and historical power service data.

[0085] The priority calculation module is used to calculate the access channel priority of the service terminal to access the edge gateway based on the average queue deviation and the access success rate deviation.

[0086] In one embodiment, the formula for calculating the average deviation of the queue is:

[0087]

[0088]

[0089]

[0090] In the formula, For queue backlog deviation, Input deviation for the queue, For queue output deviation, The edge gateway caches the data queue of service terminals. Let t be the amount of data uploaded by the service terminal to the edge gateway in time slot t. S represents the amount of data uploaded from the edge gateway to the cloud server regarding the business terminal. m (i) represents the set of business types S m The number of elements in (i).

[0091] In one embodiment, the formula for calculating the access success rate deviation is:

[0092]

[0093] In the formula, To account for the deviation in access success rate, c n For the minimum expected constraints of service terminals and edge gateways, Pre-assign indicator variables to the channel in the i-th time period, where l is the summation index.

[0094] In one embodiment, the formula for calculating the access channel priority is:

[0095]

[0096] In the formula, Priority for access channels, and A larger value indicates a greater likelihood of pre-allocating transmission channels to service terminals in the next time period. For queue backlog deviation, Input deviation for the queue, For queue output deviation, This is to account for the deviation in access success rate.

[0097] In one embodiment, the formula for calculating the profit of an edge gateway accessing a cloud server in a high-concurrency access scenario involving multiple time-scale terminals in a power system is as follows:

[0098]

[0099] In the formula, χ j，k (t) represents the profit of the edge gateway accessing the cloud server, p j,k (t) represents the bidding cost that the edge gateway needs to pay to access the cloud server. The business priority weight for the business terminal. For load balancing of data queues of service terminals in the edge gateway, This refers to the load balancing of data queues in the edge gateway of the cloud server.

[0100] The technical solutions provided by the embodiments of the present invention may include the following beneficial effects:

[0101] This invention classifies services by calculating the matching degree between the terminal data traffic feature vector and the service data feature vector. Based on the classification results, an access priority score is constructed according to the side queue status to determine the pre-allocation of access channels. The channel pre-allocation comprehensively considers the side queue status deviation and the terminal access success rate deviation, thereby improving the accuracy of pre-allocation, reducing terminal access latency, and meeting the differentiated needs of high-concurrency terminal access.

[0102] This invention is based on the back pressure design concept. It optimizes the cloud server access strategy of the edge gateway through a load imbalance awareness price increase mechanism, reduces the number of accesses and queue backlog, meets high concurrency constraints, and achieves dynamic adaptation of cloud server computing resources and edge gateway queue backlog through dynamic bidding, thereby realizing cloud-edge load balancing and meeting business data processing needs.

[0103] It should be understood that the above general description and the following detailed description are exemplary and explanatory only, and are not intended to limit the invention. Attached Figure Description

[0104] The accompanying drawings, which are incorporated in and form part of this specification, illustrate embodiments consistent with the invention and, together with the description, serve to explain the principles of the invention.

[0105] Figure 1 This is a flowchart illustrating a cloud-edge-device collaborative high-concurrency access method according to an exemplary embodiment;

[0106] Figure 2 This is a structural block diagram of a cloud-edge-device collaborative high-concurrency access system according to an exemplary embodiment;

[0107] Figure 3This is a schematic diagram illustrating a cloud-edge-device collaborative high-concurrency access process in a practical application, according to an exemplary embodiment.

[0108] Figure 4 This is an architecture diagram of a cloud-edge-device collaborative high-concurrency access system in a practical application, as illustrated in an exemplary embodiment.

[0109] Figure 5 This is a schematic diagram of the structure of a cloud-edge-device collaborative high-concurrency access device in a practical application, according to an exemplary embodiment. Detailed Implementation

[0110] The following description and accompanying drawings fully illustrate specific embodiments described herein to enable those skilled in the art to practice them. Some embodiments may include or substitute parts and features of other embodiments. The scope of the embodiments herein encompasses the entire scope of the claims and all available equivalents thereof. Throughout this document, the terms “first,” “second,” etc., are used only to distinguish one element from another without requiring or implying any actual relationship or order between the elements. Indeed, a first element can also be referred to as a second element, and vice versa. Furthermore, the terms “comprising,” “including,” or any other variations thereof are intended to cover non-exclusive inclusion, such that a structure, apparatus, or device that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a structure, apparatus, or device. Without further limitation, an element defined by the phrase “comprising one…” does not exclude the presence of other identical elements in the structure, apparatus, or device that includes said element. The various embodiments described herein are presented in a progressive manner, with each embodiment focusing on its differences from other embodiments; similar or identical parts between embodiments can be referred to interchangeably.

[0111] The terms "longitudinal," "lateral," "upper," "lower," "front," "rear," "left," "right," "vertical," "horizontal," "top," "bottom," "inner," and "outer" used in this document to indicate orientations or positional relationships are based on the orientations or positional relationships shown in the accompanying drawings. They are used solely for the convenience of describing the document and for simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, they should not be construed as limitations on the invention. In the description herein, unless otherwise specified and limited, the terms "installed," "connected," and "linked" should be interpreted broadly. For example, they can refer to mechanical or electrical connections, or internal connections between two elements; they can be direct connections or indirect connections through an intermediate medium. Those skilled in the art can understand the specific meaning of the above terms according to the specific circumstances.

[0112] In this document, unless otherwise stated, the term "multiple" means two or more.

[0113] In this article, the character " / " indicates that the objects before and after it are in an "or" relationship. For example, A / B means: A or B.

[0114] In this article, the term "and / or" describes an association between objects, indicating that three relationships can exist. For example, A and / or B means: A or B, or A and B.

[0115] It should be understood that although the steps in the flowchart are shown sequentially according to the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless explicitly stated herein, there is no strict order constraint on the execution of these steps, and they can be executed in other orders. Moreover, at least some steps in the diagram may include multiple sub-steps or multiple stages. These sub-steps or stages are not necessarily completed at the same time, but can be executed at different times. The execution order of these sub-steps or stages is not necessarily sequential, but can be performed alternately or in turn with other steps or at least some of the sub-steps or stages of other steps.

[0116] The modules in the apparatus or system of this application can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in or independent of the processor in a computer device in hardware form, or stored in the memory of a computer device in software form, so that the processor can call and execute the operations corresponding to each module.

[0117] Where there is no conflict, the embodiments and features in the embodiments of the present invention can be combined with each other.

[0118] Figure 1 An embodiment of a cloud-edge-device collaborative high-concurrency access method of the present invention is shown.

[0119] In this optional embodiment, the cloud-edge-device collaborative high-concurrency access method is applied to high-concurrency access scenarios of terminals across multiple time scales in power systems, including:

[0120] Step S101: Collect power service data from the service terminal, analyze the power service data, determine the edge gateway access type of the power service data, and determine the access channel priority of the service terminal to access the edge gateway based on the edge gateway access type.

[0121] Step S103: Sort the access channel priorities in descending order to obtain a first sorting queue, and according to the first sorting queue, connect the service terminals that meet the access channel pre-allocation judgment threshold to the corresponding edge gateway.

[0122] Step S105: Based on the high-concurrency access scenario of multiple time-scale terminals in the power system, determine the profit of the edge gateway accessing the cloud server, and sort the edge gateways that have not been connected to the cloud server in descending order according to the profit to obtain a second sorting queue. According to the second sorting queue, connect the edge gateway to the cloud server corresponding to the highest profit.

[0123] In this optional embodiment, when analyzing the power service data to determine the edge gateway access type of the power service data, the power service data at each time point can be analyzed to determine whether the service terminal corresponding to the power service data is a newly accessed service terminal; if the determination result is yes, the service data feature vector of the power service data and the data traffic feature vector of the service terminal are compared to calculate the service matching degree, and the edge gateway access type is determined based on the service matching degree; if the determination result is no, the historical service classification result is used to determine the edge gateway access type.

[0124] In this optional embodiment, when determining the access channel priority of a service terminal accessing the edge gateway based on the edge gateway access type, the average queue deviation of the service terminal and the access success rate deviation between the service terminal and the edge gateway can be calculated based on the access type and historical power service data; wherein, the average queue deviation includes the deviation of queue backlog, queue input, queue output from the average value of similar service terminals; the access channel priority of the service terminal accessing the edge gateway is calculated based on the average queue deviation and the access success rate deviation.

[0125] Figure 2 An embodiment of a cloud-edge-device collaborative high-concurrency access system of the present invention is shown.

[0126] In this optional embodiment, the cloud-edge-device collaborative high-concurrency access system is applied to high-concurrency access scenarios for terminals across multiple time scales in power systems, including:

[0127] The priority determination module 201 is used to collect power service data from the service terminal, analyze the power service data, determine the edge gateway access type of the power service data, and determine the access channel priority of the service terminal to access the edge gateway based on the edge gateway access type.

[0128] The edge access module 203 is used to sort the access channels in descending order to obtain a first sorting queue, and according to the first sorting queue, to connect the service terminals that meet the access channel pre-allocation judgment threshold to the corresponding edge gateway.

[0129] The edge-cloud access module 205 is used to determine the profit of edge gateways accessing cloud servers based on the high-concurrency access scenario of multi-time-scale terminals in the power system, and to sort the edge gateways that have not been connected to cloud servers in descending order according to the profit to obtain a second sorting queue. According to the second sorting queue, the edge gateways are connected to the cloud server corresponding to the highest profit.

[0130] In this optional embodiment, the priority determination module 201 may include a terminal judgment module (not shown in the figure), a matching degree calculation module (not shown in the figure), and a type determination module (not shown in the figure). The terminal judgment module is used to analyze the power service data at each time point and determine whether the service terminal corresponding to the power service data is a newly accessed service terminal. The matching degree calculation module is used to compare the service data feature vector of the power service data and the data traffic feature vector of the service terminal to calculate the service matching degree if the judgment result is yes. The type determination module is used to determine the edge gateway access type based on the service matching degree if the judgment result is yes; otherwise, it uses historical service classification results to determine the edge gateway access type.

[0131] In this optional embodiment, the priority determination module 201 may further include a queue average calculation module (not shown in the figure), an access success rate deviation calculation module (not shown in the figure), and a priority calculation module (not shown in the figure). The queue average calculation module is used to calculate the queue average deviation of the service terminal based on historical power service data according to the access type. The queue average deviation includes the deviations of queue backlog, queue input, queue output, and the average deviation of similar service terminals. The access success rate deviation calculation module is used to calculate the access success rate deviation between the service terminal and the edge gateway based on historical power service data according to the access type. The priority calculation module is used to calculate the access channel priority of the service terminal accessing the edge gateway based on the queue average deviation and the access success rate deviation.

[0132] In practical applications, the cloud-edge-device collaborative high-concurrency access method can be divided into four steps: 1. Construction of high-concurrency access scenarios for multiple time scale terminals in new power systems; 2. Construction of optimization problems for high-concurrency access for multiple time scale terminals; 3. Pre-allocation of edge access channels at large time scales based on access priority evaluation; 4. Cloud-edge load balancing at small time scales based on load imbalance awareness bidding.

[0133] The technical solution of the present invention will be described in detail below based on the four steps described above. Specifically:

[0134] 1. Construction of high-concurrency access scenarios for multi-timescale terminals in new power systems

[0135] New power system scenarios involving high-concurrency access across multiple time scales include the cloud layer, edge layer, and terminal layer. The terminal layer is defined as containing N service terminals, denoted by D = {d1,...,d...}. n ,...,d N The edge layer is defined as containing J edge gateways, set as E = {e1,...,e2}. j ,...,e J Defined in the edge gateway e j The set of terminals within the communication range is The cloud layer consists of a high-concurrency access management platform and K cloud servers, with the set of cloud servers being S = {s1,...,s...}. k ,...,s K The service terminal transmits the collected data to the edge gateway via a power line carrier communication network. The edge gateway is the D network within its communication range. j Each terminal is pre-allocated an access channel, and a D is constructed. j Each data queue stores unprocessed terminal data. The management platform dynamically adjusts edge gateway data access decisions based on queue backlogs in the edge gateway and cloud server, achieving load balancing optimization. The cloud server processes data accessed from the edge gateway, supporting power services such as distributed power management and panoramic sensing.

[0136] The new power system adopts a multi-time-scale model for high-concurrency access scenarios of multi-time-scale terminals, dividing the system into I large time scales (time periods), each time period containing T0 hourly time scales (time slots), with a time slot length of τ.

[0137] Define edge gateway e j Cached business terminals Data queue is Its backlog evolution model is as follows:

[0138]

[0139] in, Represents the service terminal in time slot t. Upload to edge gateway e j The amount of data, Indicates edge gateway e j Uploaded to cloud server for business terminals The amount of data. Pre-assign indicator variables to the large-scale channel in the i-th time period. Represented as a business terminal Pre-allocate transmission channels, otherwise

[0140] Depending on the minimum value between the edge data throughput and the amount of data collected by the terminal, it can be expressed as:

[0141]

[0142] in, Indicates the business terminal to edge gateway e j upload speed, Indicates the business terminal The minimum data collection volume constraint is related to business requirements. Similarly, Depends on the minimum of edge-cloud data throughput and terminal data queue backlog, which can be expressed as:

[0143]

[0144] in, Indicates edge gateway e j to cloud server s k upload speed, For edge gateway e j Mid-service terminal The queue is uploaded to the cloud server. k The amount of data is defined as x, which is the indicator variable for selecting a small-scale cloud server in the t-th time slot. j,k (t), x j,k (t) = 1 indicates that the edge gateway e j Upload the data to the cloud server. k Calculate, otherwise x j,k (t) = 0.

[0145] cloud server s k Upper Edge Gateway e j The evolution model of data queue backlog is as follows:

[0146]

[0147] in, For the edge gateway e in time slot t j Uploaded to cloud server k Data volume, Y j,k (t) represents the cloud server s in the t-th time slot. k Processing from edge gateway e j The amount of data depends on the cloud server. k Available computing resources and edge gateway e j The queue backlog can be represented as:

[0148]

[0149] in, Represents the cloud server s in time slot t. k Used to process edge gateway e j Mid-service terminal Data computing resources Indicates the terminal that processes each bit of service. The computing resources required for the data.

[0150] 2. Optimization of high-concurrency access for terminals at multiple time scales.

[0151] Based on the data queue backlog evolution model constructed in step 1, a data queue load balancing model for edge gateways and cloud servers is constructed.

[0152] Edge gateway e j Mid-service terminal The data queue load balancing model is as follows:

[0153]

[0154] in, Indicates edge gateway e j The average queue backlog, i.e., the number of service terminals. queue backlog The closer the queue backlog is to the average, the more balanced the load. Similarly, define a cloud server s. k Middle edge gateway e j The data queue load balancing model is as follows:

[0155]

[0156] Cloud-edge load balancing can be calculated as follows:

[0157]

[0158] in, Indicates the business terminal Business priority weight, β Z This indicates the weight of cloud load balancing.

[0159] Based on the cloud-edge load balancing model described above, the optimization problem for high-concurrency access from multiple time-scale terminals can be constructed as follows:

[0160]

[0161]

[0162] Where, q j For edge gateway e j Maximum number of allocatable channels; c n For business terminals The minimum expected constraint for pre-allocated channels, i.e., the terminal access success rate constraint; q k For cloud servers k The maximum number of queues that can be processed simultaneously; For queue backlog indicator variable, express otherwise For cloud servers k The upper limit of queue backlog.

[0163] The constraints are as follows: C1 is the constraint for access channel pre-allocation indicator variable, service terminal access indicator variable, and queue backlog indicator variable; C2 is the constraint for service terminal access channel pre-allocation; C3 is the constraint for service terminal access success rate; C4 is the constraint for edge gateway access to cloud server; C5 and C6 are the constraints for high-concurrency access of service terminals, i.e., each cloud server can process a maximum of q concurrent requests. k Each business terminal queue, and the queue backlog on the cloud server does not exceed the upper limit.

[0164] The aforementioned multi-timescale terminal high-concurrency access optimization problem can be further decomposed into a large-timescale edge access channel pre-allocation optimization problem and a small-timescale cloud-edge load balancing optimization problem.

[0165] 3. Pre-allocation of edge access channels on a large time scale based on access priority assessment.

[0166] Furthermore, step 3 includes the following steps:

[0167] Step 3.1: Terminal service classification based on traffic feature matching degree;

[0168] The terminal access type is determined at each time period. If the terminal is identified as a service terminal... For newly accessed terminals, the feature vectors of service data are compared with... The business matching degree is calculated from the data traffic feature vector to achieve business classification; otherwise, the historical business classification results are used. The business matching degree is calculated as follows:

[0169]

[0170] in, For business terminals Data traffic feature vector, X m Let be the data feature vector of the m-th type of business. The larger the value, the more likely it is to be a business terminal. The higher the matching degree between the data traffic characteristics and the m-th type of service, the better. If the service terminal... It is identified as the m-th type of service, and the service terminal will be... The set S of the m-th type of business is added. m In (i), that is

[0171] Step 3.2: Calculate the deviation between the edge layer queue status and the end layer access success rate;

[0172] Based on the terminal service classification results obtained in step 3.1, for the m-th type of service, the edge gateway e j Based on historical data from the previous period, the calculation service terminal The deviation between the queue backlog, queue input, and queue output and the average value of similar service terminals. Definition These are queue backlog deviation, queue input deviation, and queue output deviation, respectively, and are calculated using the following formulas:

[0173]

[0174]

[0175]

[0176] Among them, |S m (i)| is the set S m The number of elements in (i).

[0177] definition Minimum expectation constraint c n With business terminal The deviation in access success rate is calculated using the following formula:

[0178]

[0179] In the formula, To account for the deviation in access success rate, c n For the minimum expected constraints of service terminals and edge gateways, Pre-assign indicator variables to the channel in the i-th time period, where l is the summation index.

[0180] Step 3.3: Optimize the edge access channel pre-allocation strategy based on access priority assessment;

[0181] Based on the results obtained in step 3.2 as well as Access priority score The calculation is as follows:

[0182]

[0183] in, The larger the value, the more likely it is to be a business terminal in the next time period. Pre-allocate transmission channels and define V minA threshold is pre-assigned for the access channel, and the set of terminals that meet the threshold is: D j The terminals in (i) are arranged in descending order of their scores, with the edge gateway e... j For the first q in the sorting j The service terminal pre-allocates the access channel, that is

[0184] 4. Small-scale cloud-edge load balancing based on load imbalance awareness bidding.

[0185] Furthermore, step 4 includes the following steps:

[0186] Step 4.1: Calculate access profit based on load balancing awareness.

[0187] Based on data queue load balancing modeling, edge-cloud load balancing is perceived, and edge gateway load balancing is calculated. j Access cloud server s k Profit:

[0188]

[0189] Where, p j,k (t) represents the edge gateway e j Access cloud server s k The bidding costs that need to be paid.

[0190] Step 4.2: Iterative optimization of cloud-edge load balancing based on load imbalance awareness bidding.

[0191] Based on the access profit obtained in step 4.1, the edge gateway e that is not connected to the cloud server j Sort the access points in descending order of profit, and prioritize the cloud servers with the highest access profits. For example: cloud server s k Initiating an access request, when accessing the same cloud server s k The edge gateway e that initiated the access request j When the quantity or queue backlog meets high concurrency constraints, the management platform allows these edge gateways to access the cloud server. k Otherwise, a load imbalance-aware bidding mechanism is needed to resolve the contention issue and meet high concurrency constraints. The bidding process is described below:

[0192] The management platform improves cloud server performance through load imbalance awareness and bidding. k The cost of cloud servers. k For edge gateway e j The cost update formula is:

[0193]

[0194] Where Δp represents the step size for cost increment. The cost update approach is: the higher the priority of the terminal service and the greater the difference in backlog between the cloud and edge queues, the smaller the cost increase. After the cost increase is obtained based on the cost update formula, the edge gateways participating in the bidding recalculate the access cloud server s. k The edge gateway with low profit margin abandons the cloud server. k Instead, they initiate access requests to other cloud servers that offer higher profits. As costs continue to rise, access to cloud servers... k The number of edge gateways or queue backlog will gradually decrease until the high concurrency constraints are met.

[0195] Repeat steps 4.1 and 4.2 until all edge gateways are connected to the cloud server, at which point the current time slot optimization is complete. The edge gateways then upload data to the cloud server according to the cloud-edge access policy.

[0196] In practical applications, cloud-edge-device collaborative high-concurrency access systems can... Figure 4 As shown, it can be divided into three layers: cloud side, edge side, and terminal side. The terminal side includes a new type of power system business terminal; the edge side includes an edge-to-edge communication module, a service matching degree calculation module, a service classification module, a queue status deviation calculation module, an access success rate deviation calculation module, an access priority evaluation module, and an access channel pre-allocation module; the cloud side includes a cloud-edge communication module, a load imbalance perception bidding module, a cloud-edge access decision module, a data processing module, and a data storage module. Details are as follows:

[0197] 1. Edge layer: New type of power system business terminal: used to collect various power business data and upload them to the edge layer.

[0198] 2. Edge Layer: Edge Layer Communication Module: Receives various power service data uploaded from the terminal layer and cloud-edge access policies from the cloud layer, issues access channel pre-allocation decisions to service terminals, and uploads queue data to the cloud layer. Service Matching Degree Calculation Module: Compares service data feature vectors with terminal data traffic feature vectors, calculates the service matching degree carried by the terminal, and sends the calculation result to the service classification module. Service Classification Module: Classifies the services carried by the terminal based on the service matching degree results, and transmits the classification results to the queue status deviation calculation module and the access success rate deviation calculation module. Queue Status Deviation Calculation Module: Calculates the deviation of the terminal's queue backlog, queue input, and queue output from the average value of similar service terminals, and sends the calculation result to the access priority evaluation module. Access Success Rate Deviation Calculation Module: Calculates the deviation between the minimum expected constraint of the pre-allocated channel and the terminal access success rate, and sends the calculation result to the access priority evaluation module. Access Priority Evaluation Module: Calculates the access priority score based on the queue status deviation and access success rate deviation, and sends it to the access channel pre-allocation module. Access channel pre-allocation module: Based on the access priority scoring results, the terminals that meet the access channel pre-allocation judgment threshold are arranged in descending order of score, and the data queue is sorted according to the order to make access channel pre-allocation decisions and send them to the edge communication module.

[0199] 3. Cloud Layer: Cloud Layer Communication Module: Distributes cloud-edge access strategies to the edge layer and receives queue data uploaded by the edge layer. Cloud-Edge Access Decision Module: Calculates access profit and optimizes the cloud-edge access strategy based on the access profit. If the access strategy does not meet high concurrency constraints, it sends a bidding request to the load imbalance awareness bidding module; otherwise, it sends the obtained access strategy to the cloud-edge communication module. Load Imbalance Awareness Bidding Module: Based on the bidding request from the cloud-edge access decision module, it senses the load imbalance of the cloud edge, updates the bidding cost, and sends the updated cost to the cloud-edge access decision module. Data Processing Module: Processes the queue data uploaded by the edge layer and sends the processing results to the data storage module. Data Storage Module: Stores the data processing results.

[0200] In practical applications, cloud-edge-device collaborative high-concurrency access devices can... Figure 5 As shown, it includes a side-layer communication module, a service matching degree calculation module, a service classification module, a queue status deviation calculation module, an access success rate deviation calculation module, an access priority evaluation module, an access channel pre-allocation module, and a power supply module. The power supply module is responsible for supplying power to each module in the device.

[0201] This invention is not limited to the structures described above and shown in the accompanying drawings, and various modifications and changes can be made without departing from its scope. The scope of this invention is limited only by the appended claims.

Claims

1. A cloud-edge-device collaborative high-concurrency access method, characterized in that, The cloud-edge-device collaborative high-concurrency access method is applied to high-concurrency access scenarios for terminals across multiple time scales in power systems, including: Collect power service data from service terminals, analyze the power service data, determine the edge gateway access type of the power service data, and determine the access channel priority of the service terminal to access the edge gateway based on the edge gateway access type; The access channels are sorted in descending order of priority to obtain a first sorting queue, and service terminals that meet the access channel pre-allocation judgment threshold are connected to the corresponding edge gateway according to the first sorting queue. Based on the high-concurrency access scenario of multiple time scale terminals in the power system, the profit of edge gateways accessing cloud servers is determined. Based on the profit, edge gateways that are not connected to cloud servers are sorted in descending order to obtain a second sorting queue. Based on the second sorting queue, edge gateways are connected to the cloud server corresponding to the highest profit. The high-concurrency access scenario of multi-timescale terminals in the power system includes a terminal layer, an edge layer, and a cloud layer, wherein the terminal layer includes... A set of service terminals, the collection of service terminals is The edge layer includes There are 1 edge gateway, and the set of edge gateways is 1. The set of service terminals within the communication range of the edge gateway is The cloud layer includes a high-concurrency access management and control platform and A collection of cloud servers, the cloud server collection is Furthermore, the service terminal transmits the collected data to the edge gateway via a power line carrier communication network; the edge gateway transmits the data within its communication range... Each terminal is pre-allocated an access channel and has A data queue is used to store uncomputed business terminal data; the high-concurrency access management platform dynamically adjusts the edge gateway data access decision based on the queue backlog of the edge gateway and cloud server; a multi-timescale model is adopted between the terminal layer, the edge layer, and the cloud layer to divide the data transmission time into segments. There are 10 time periods, and each time period contains 100 time periods. There are 1 time slot, and the time slot length is 1. ; The service terminal and the edge gateway have a first data queue backlog evolution model, and the formula of the first data queue backlog evolution model is: ； ； ； The edge gateway and the cloud server have a second data queue backlog evolution model, and the formula for the second data queue backlog evolution model is: ； ； In the formula, The edge gateway caches the data queue of service terminals. For the first The amount of data uploaded from the time-slot service terminal to the edge gateway This refers to the amount of data uploaded from the edge gateway to the cloud server regarding the business terminal. For the first Channel pre-assignment indicator variables for each time period, This indicates that a transmission channel has been pre-allocated for the service terminal; otherwise... , The upload speed from the service terminal to the edge gateway. Minimum data collection constraint for the business terminal; This refers to the upload speed from the edge gateway to the cloud server. This refers to the amount of data uploaded to the cloud server by the service terminals in the edge gateway. For the first The cloud server selection indicator variable for each time slot. This indicates that the edge gateway uploads data to the cloud server for computing; otherwise... , The cloud server caches the data queue of the edge gateway. For the first The amount of data uploaded from the time-slot edge gateway to the cloud server For the first The time-slot cloud server processes the amount of data from the edge gateway. For the first Time-slot cloud servers are computing resources used to process data from business terminals in edge gateways. The computing resources required to process each bit of service terminal data.

2. The cloud-edge-device collaborative high-concurrency access method according to claim 1, characterized in that, The edge gateway includes a first load balancing model for load balancing the data queues of service terminals. The formula for the first load balancing model is as follows: ； The cloud server has a second load balancing model for load balancing the data queues of the edge gateway. The formula for the second load balancing model is: ； In the formula, For the average queue backlog of edge gateways, The edge gateway caches the data queue of service terminals. For the average queue backlog of cloud servers, The cloud server caches the data queue of the edge gateway. For load balancing of data queues of service terminals in the edge gateway, This refers to the load balancing of data queues in the edge gateway of the cloud server.

3. The cloud-edge-device collaborative high-concurrency access method according to claim 1, characterized in that, Analyzing the power business data determines the edge gateway access type for the power business data, including: Analyze the power business data at each time point to determine whether the business terminal corresponding to the power business data is a newly connected business terminal; If the judgment result is yes, the service data feature vector of the power service data and the data traffic feature vector of the service terminal are compared to calculate the service matching degree, and the edge gateway access type is determined based on the service matching degree; if the judgment result is no, the edge gateway access type is determined by using the historical service classification results. The formula for calculating the business matching degree is as follows: In the formula, This is the data traffic feature vector of the business terminal. For the first Data feature vectors of similar businesses Matching to business needs.

4. The cloud-edge-device collaborative high-concurrency access method according to claim 3, characterized in that, Based on the edge gateway access type, the priority of the access channel for a service terminal to access the edge gateway includes: Based on the access type and historical power service data, the average queue deviation of the service terminal and the access success rate deviation between the service terminal and the edge gateway are calculated; wherein, the average queue deviation includes the deviation of queue backlog, queue input, queue output from the average value of similar service terminals; The access channel priority for a service terminal to access the edge gateway is calculated based on the average queue deviation and the access success rate deviation.

5. The cloud-edge-device collaborative high-concurrency access method according to claim 4, characterized in that, The formula for calculating the average deviation of the queue is: In the formula, For queue backlog deviation, Input deviation for the queue, For queue output deviation, The edge gateway caches the data queue of service terminals. For the first The amount of data uploaded from the time-slot service terminal to the edge gateway This refers to the amount of data uploaded from the edge gateway to the cloud server regarding the business terminal. For a set of business types The number of elements in the array.

6. The cloud-edge-device collaborative high-concurrency access method according to claim 5, characterized in that, The formula for calculating the access success rate deviation is: In the formula, To account for the deviation in access success rate, For the minimum expected constraints of service terminals and edge gateways, For the first Channel pre-assignment indicator variables for each time period, For summation index.

7. The cloud-edge-device collaborative high-concurrency access method according to claim 6, characterized in that, The formula for calculating the access channel priority is as follows: In the formula, Priority for access channels, and A larger value indicates a greater likelihood of pre-allocating transmission channels to service terminals in the next time period. For queue backlog deviation, Input deviation for the queue, For queue output deviation, This is to account for the deviation in access success rate.

8. The cloud-edge-device collaborative high-concurrency access method according to claim 2, characterized in that, Based on the scenario of high-concurrency access from multiple time-scale terminals in the power system, the formula for calculating the profit of edge gateways accessing cloud servers is as follows: In the formula, For the profit of edge gateways connecting to cloud servers, The bidding cost required to connect an edge gateway to a cloud server. The business priority weight for the business terminal. For load balancing of data queues of service terminals in the edge gateway, This refers to the load balancing of data queues in the edge gateway of the cloud server.

9. A cloud-edge-device collaborative high-concurrency access system, characterized in that, The cloud-edge-device collaborative high-concurrency access system is applied to high-concurrency access scenarios for terminals across multiple time scales in power systems, including: The priority determination module is used to collect power service data from the service terminal, analyze the power service data, determine the edge gateway access type of the power service data, and determine the access channel priority of the service terminal to access the edge gateway based on the edge gateway access type. The edge access module is used to sort the access channels in descending order to obtain a first sorting queue, and according to the first sorting queue, to connect the service terminals that meet the access channel pre-allocation judgment threshold to the corresponding edge gateway. The edge-cloud access module is used to determine the profit of edge gateways accessing cloud servers based on the high-concurrency access scenario of multi-time-scale terminals in the power system, and to sort the edge gateways that have not been connected to cloud servers in descending order according to the profit to obtain a second sorting queue. According to the second sorting queue, the edge gateways are connected to the cloud server corresponding to the highest profit. The high-concurrency access scenario of multi-timescale terminals in the power system includes a terminal layer, an edge layer, and a cloud layer, wherein the terminal layer includes... A set of service terminals, the collection of service terminals is The edge layer includes There are 1 edge gateway, and the set of edge gateways is 1. The set of service terminals within the communication range of the edge gateway is The cloud layer includes a high-concurrency access management and control platform and A collection of cloud servers, the cloud server collection is Furthermore, the service terminal transmits the collected data to the edge gateway via a power line carrier communication network; the edge gateway transmits the data within its communication range... Each terminal is pre-allocated an access channel and has A data queue is used to store uncomputed business terminal data; the high-concurrency access management platform dynamically adjusts the edge gateway data access decision based on the queue backlog of the edge gateway and cloud server; a multi-timescale model is adopted between the terminal layer, the edge layer, and the cloud layer to divide the data transmission time into segments. There are 10 time periods, and each time period contains 100 time periods. There are 1 time slot, and the time slot length is 1. ; The service terminal and the edge gateway have a first data queue backlog evolution model, and the formula of the first data queue backlog evolution model is: ； ； ； The edge gateway and the cloud server have a second data queue backlog evolution model, and the formula for the second data queue backlog evolution model is: ； ； In the formula, The edge gateway caches the data queue of service terminals. For the first The amount of data uploaded from the time-slot service terminal to the edge gateway This refers to the amount of data uploaded from the edge gateway to the cloud server regarding the business terminal. For the first Channel pre-assignment indicator variables for each time period, This indicates that a transmission channel has been pre-allocated for the service terminal; otherwise... , The upload speed from the service terminal to the edge gateway. Minimum data collection constraint for the business terminal; This refers to the upload speed from the edge gateway to the cloud server. This refers to the amount of data uploaded to the cloud server by the service terminals in the edge gateway. For the first The cloud server selection indicator variable for each time slot. This indicates that the edge gateway uploads data to the cloud server for computing; otherwise... , The cloud server caches the data queue of the edge gateway. For the first The amount of data uploaded from the time-slot edge gateway to the cloud server For the first The time-slot cloud server processes the amount of data from the edge gateway. For the first Time-slot cloud servers are computing resources used to process data from business terminals in edge gateways. The computing resources required to process each bit of service terminal data.

10. The cloud-edge-device collaborative high-concurrency access system according to claim 9, characterized in that, The edge gateway includes a first load balancing model for load balancing the data queues of service terminals. The formula for the first load balancing model is as follows: ； The cloud server has a second load balancing model for load balancing the data queues of the edge gateway. The formula for the second load balancing model is: ； In the formula, For the average queue backlog of edge gateways, The edge gateway caches the data queue of service terminals. For the average queue backlog of cloud servers, The cloud server caches the data queue of the edge gateway. For load balancing of data queues of service terminals in the edge gateway, This refers to the load balancing of data queues in the edge gateway of the cloud server.

11. The cloud-edge-device collaborative high-concurrency access system according to claim 9, characterized in that, The priority determination module includes a terminal judgment module, a matching degree calculation module, and a type determination module, wherein... The terminal judgment module is used to analyze the power business data at each time point and determine whether the business terminal corresponding to the power business data is a newly accessed business terminal. The matching degree calculation module is used to compare the business data feature vector of the power business data with the data flow feature vector of the business terminal when the judgment result is yes, and calculate the business matching degree. The type determination module is used to determine the edge gateway access type based on the service matching degree when the judgment result is yes; and to determine the edge gateway access type based on the historical service classification results when the judgment result is no. The formula for calculating the business matching degree is as follows: In the formula, This is the data traffic feature vector of the business terminal. For the first Data feature vectors of similar businesses Matching to business needs.

12. The cloud-edge-device collaborative high-concurrency access system according to claim 11, characterized in that, The priority determination module includes a queue average calculation module, an access success rate deviation calculation module, and a priority calculation module, wherein... The queue average value calculation module is used to calculate the queue average value deviation of the service terminal based on the access type and historical power service data. The queue average value deviation includes the queue backlog, queue input, queue output and the average value deviation of the same type of service terminal. The access success rate deviation calculation module is used to calculate the access success rate deviation between the service terminal and the edge gateway based on the access type and historical power service data. The priority calculation module is used to calculate the access channel priority of the service terminal to access the edge gateway based on the average queue deviation and the access success rate deviation.

Images