An alternating direction method of multipliers based cloud-edge collaborative smart grid operation and maintenance carbon reduction method

Abstract

The application provides a cloud-edge collaborative smart grid operation and maintenance carbon reduction method based on an alternating direction multiplier method, comprising: designing a cloud-edge collaborative grid operation and maintenance system, the cloud-edge collaborative grid operation and maintenance system comprising a terminal device, a cloud computing center and an edge node; modeling related time delays of the cloud-edge collaborative grid operation and maintenance system; modeling comprehensive carbon emissions of the cloud-edge collaborative grid operation and maintenance system according to the modeling of the related time delays; and constructing an objective function for a carbon emission minimization problem of the cloud-edge collaborative grid operation and maintenance system according to the modeling of the comprehensive carbon emissions and obtaining a global optimal solution by using an alternating direction multiplier method. The application provides effective technical support for realizing low-carbon operation and carbon emission management of a power grid and effectively reduces comprehensive carbon emissions while reducing business time delays.

Description

Technical Field

[0001] This invention relates to the field of cloud-edge collaboration and smart grid operation and maintenance, specifically a cloud-edge collaborative smart grid operation and maintenance carbon reduction method based on the alternating direction multiplier method. Background Technology

[0002] In the current field of smart grid operation and maintenance, there are already some methods to reduce carbon emissions by optimizing the power structure and improving the level of intelligence. These methods typically rely on resource allocation or service scheduling to reduce the overall energy consumption and carbon emissions of the power grid. Although the above-mentioned efforts can effectively reduce network energy consumption and carbon emissions, they all ignore different service requirements. Faced with the massive amount of data generated by the power grid, their optimization and modeling methods are already difficult to meet the needs of large-scale power grid operation and maintenance.

[0003] With the rapid development of information technology and artificial intelligence, cloud-edge collaboration technology has been gradually introduced into the field of smart grid operation and maintenance. This technology, through close collaboration between the cloud and the edge, can better meet the needs of various application scenarios, thereby amplifying the application value of both. Edge computing, located close to terminal devices, serves as a data acquisition and preprocessing unit, supporting cloud applications. Cloud computing, with its powerful processing capabilities, can perform big data analysis, optimization, and model training. The trained models or business rules are then distributed to the edge, where edge computing operates based on these new models or rules. This technology can compensate for the shortcomings of traditional methods for reducing carbon emissions. In cloud-edge collaboration technology, the task offloading optimization algorithm based on the Alternative Direction Method of Multipliers (ADMM) is a key algorithm, and its effectiveness directly determines the overall carbon emissions of power grid operation and maintenance.

[0004] However, in the power system, how to apply cloud-edge collaborative technology based on the alternating direction multiplier method to smart grid operation and maintenance to process the massive amounts of data generated, and how to establish mathematical models and provide optimization methods, are issues that need further discussion. Summary of the Invention

[0005] To address the aforementioned shortcomings, this invention provides a cloud-edge collaborative power grid operation and maintenance carbon reduction method based on the alternating direction multiplier method. The method analyzes and mathematically models the carbon emissions of power grid operation and maintenance under the cloud-edge collaborative architecture, and optimizes the mathematical model using the alternating direction multiplier method. This addresses the impact of the massive amount of data generated by the power grid on large-scale power grid operation and maintenance, a problem that existing technologies cannot solve.

[0006] To achieve the above objectives, the present invention adopts the following technical solution:

[0007] A cloud-edge collaborative smart grid operation and maintenance carbon reduction method based on the alternating direction multiplier method includes the following steps:

[0008] S1: Design a cloud-edge collaborative power grid operation and maintenance system, which includes terminal equipment, a cloud computing center, and edge nodes; the terminal equipment is responsible for collecting different sensing data in the industrial internet and aggregating the collected data to the corresponding edge nodes for processing through wireless or wired data links; the cloud computing center is used to provide computing services; the edge nodes provide wireless communication infrastructure for wireless communication with the terminal equipment and data aggregation, as well as small servers for data processing and computing services.

[0009] S2: Model the relevant latency of the cloud-edge collaborative power grid operation and maintenance system;

[0010] S3. Based on the modeling of relevant delays in step S2, model the comprehensive carbon emissions of the cloud-edge collaborative power grid operation and maintenance system;

[0011] S4: Based on the modeling of comprehensive carbon emissions in step S3, construct an objective function for minimizing the carbon emissions of the cloud-edge collaborative power grid operation and maintenance system, and use the alternating direction multiplier method to obtain the global optimal solution.

[0012] Furthermore, in step S1, the cloud-edge collaborative power grid operation and maintenance system includes M edge nodes and 1 cloud computing center, where the cloud computing center is represented by C, and the edge nodes are represented by a set S = {S1, ..., S2}. m ,…,S M} represents the index number of the edge node; the number of terminal devices is N, represented by the set D = {D1, ..., D2}. n ,…,D N} indicates that n represents the index number of the terminal device;

[0013] With γ m Represents edge node S m The proportion of data task units received that are offloaded to local processing, where 0 ≤ γ m ≤; when γ m When = 1, it indicates that the edge node S m It will use its local computing resources to process all data task units; when γ m When = 0, it indicates an edge node S. m All received data task units are forwarded to cloud computing center C for processing.

[0014] Furthermore, step S2 includes:

[0015] S21: Edge node service latency modeling without local unloading

[0016] Edge node S mWhen all received data task units are directly forwarded to cloud computing center C via the wired backbone network for processing, the edge computing network is equivalent to a traditional cloud computing network, meaning all computing tasks are completed by cloud computing center C. m =0;

[0017] The processing latency of a data task unit is linearly related to the number of data task units processed. The number of data task units processed per unit time in cloud computing center C is v. Without local offloading, the edge node S... m Service latency is expressed as

[0018]

[0019] in, This represents the terminal device and its associated edge node S. m Round-trip data transmission latency; For edge node S m All data task units S m Round-trip transmission latency between the cloud computing center; s m Indicates the total number of data task units;

[0020] S22: Edge node service latency modeling during full local unloading

[0021] Edge node S m When selecting to offload all received data task units to local storage for processing, the cloud computing center will not process any data task units from edge nodes. m =1; An M / M / 1 queuing system is used to model the data task units to be processed at the edge nodes, and μ is used. m Represents edge node S m The maximum number of data task units that local computing resources can process per unit of time; when all local resources are unloaded, the edge node S m The latency model for providing services to its associated user equipment is as follows:

[0022]

[0023] Where μ m >s m ;

[0024] S23: Latency Modeling of Cloud-Edge Collaborative Offloading Services

[0025] In the case of cloud-edge collaborative offloading services, each edge node S m Only a portion of the data task unit it receives is processed using its own local computing resources, using γ m s mThis indicates that the remaining data received by the task unit will be forwarded to the cloud, using (1-γ) m )s m This indicates that when partially unloading from local storage, edge node S... m The latency model for providing services to its associated user equipment is as follows:

[0026]

[0027] Where 0≤γ m ≤1, and γ m s m <μ m .

[0028] Furthermore, step S3 includes:

[0029] S31: Carbon Emission Modeling of Edge Nodes

[0030] Total power P consumed by edge nodes m Represented as:

[0031]

[0032] Where η m The power utilization efficiency is represented by dividing the total annual power consumption of a node by the total annual power consumption of the node's information technology equipment. and β m They represent edge nodes S respectively m The static power consumption and the dynamic power consumption generated by the data processing unit;

[0033] Edge node S m In the data task unit γ that processes its local unloading m s m Queuing delay is Edge node S m carbon emissions E m (γ m The model is as follows:

[0034]

[0035] S32: Carbon Emission Modeling for Cloud Computing Centers

[0036] The total number of data task units received by a cloud computing center per unit time is expressed as:

[0037]

[0038] Where γ=<γ1,γ2,...,γ M >;

[0039] The response time of cloud computing center services is Total power consumption P C (γ) is represented as:

[0040]

[0041] Carbon emissions from cloud computing centers C (γ) is modeled as:

[0042]

[0043] S33: Carbon Emission Modeling for Backbone Network Data Transmission

[0044] Use β L η represents the dynamic power consumption of a single data transmission task unit in the backbone network. L The power efficiency of the edge node S represents the energy utilization efficiency during data transmission in the backbone network. m The received data task unit s m (1-γ) m When the data is forwarded to the cloud data center, the transmission latency is... Total power consumption P L (γ m ) is represented as:

[0045] P L (γ m )=η L β L (1-γ m )s m (3-6)

[0046] Use ∈ L The carbon intensity of power resources used for data transmission in the backbone network is represented by the following: When an edge node forwards some data task units to the cloud data center through the backbone network, the carbon emissions generated are expressed as follows:

[0047]

[0048] Furthermore, in step S4, when the cloud-edge collaborative network consists of one cloud computing center and M edge nodes, the objective function for minimizing carbon emissions in the cloud-edge collaborative power grid operation and maintenance system is as follows:

[0049]

[0050] Wherein, γ includes the proportion of the business volume offloaded by M edge nodes to the total number of businesses of all industrial internet terminals; E C (γ), E m (γ m ) and E L (γ mThese represent the computational carbon emissions of the cloud computing center, the computational carbon emissions of the m-th edge node, and the transmission carbon emissions generated by the industrial internet terminal corresponding to the m-th edge node uploading business data, respectively.

[0051] Furthermore, in step S4, when using the alternating direction multiplier method to obtain the global optimal solution for the objective function established in step S41:

[0052] First, the iterative formula of the alternating direction multiplier method is simplified based on the augmented Lagrange expression. Specifically,

[0053] Using indicator function I C Adding the feasible region to the objective function (.) transforms the objective function into the following problem:

[0054] min[E C (γ)+1 T (E(γ)+E L (γ))]+I C (γ) (4-2)

[0055] Among them I C (γ) is the indicator function of the feasible region, that is:

[0056]

[0057] Introducing the constraint z = γ, the objective function is transformed into:

[0058]

[0059] Where f(y) = E C (y)+1 T (E(y)+E L (y)),h(z)=I C (z);

[0060] After the above equivalent transformation, the objective function is converted into the standard alternating direction multiplier method form;

[0061] Next, we establish the augmented Lagrange multiplier formulas for the next step of the solution:

[0062]

[0063] Where ρ>0 are the augmented Lagrange coefficients, and λ∈R M As dual variables;

[0064] Based on the above Lagrange multiplier formula, the alternating direction multiplier method consists of the following iterative steps:

[0065] y k+1 =argminL ρ(y,z k ,λ k (4-6)

[0066]

[0067] λ k+1 =λ k +ρ(y k+1 -z k+1 (4-8)

[0068] The iterative formula is simplified by combining the linear and quadratic terms in the augmented Lagrange and scaling the dual variable. The scaled dual variable is defined as follows: The alternating direction multiplier method is equivalent to the following expression:

[0069]

[0070] u k+1 :=u k +ρ(y k+1 -z k+1 (4-11)

[0071] Secondly, the stopping criterion for the Alternating Direction Multiplier Method (ADMM) is determined. Specifically, the original feasibility residual and the dual feasibility residual for the above problem are defined:

[0072] r k =||y k -z k ||2 (4-12)

[0073] s k =||z k -z k-1 ||2 (4-13)

[0074] Where r k For the original feasibility residual, s k It is a dual feasibility residual, and the stopping criterion is to determine whether the two residuals are sufficiently small.

[0075] Furthermore, the determination of whether the two residuals are sufficiently small as a stopping criterion specifically includes:

[0076]

[0077] Where, ∈ pri >0, ∈ dual >0 indicates the feasibility error of the suboptimal solution of the objective function. The suboptimal solution obtained by the alternating direction multiplier method is used as the final solution of the objective function.

[0078] Furthermore, ∈ abs and ∈rel 10 respectively -4 and 10 -2 .

[0079] Compared with the prior art, the present invention has the following beneficial effects:

[0080] 1. The cloud-edge collaborative optimization task offloading method proposed in this invention can take into account the advantages of cloud computing centers and edge nodes. The application of cloud-edge collaborative technology in power grid carbon emission management can improve data processing efficiency, reduce latency, enhance data security, and support accurate monitoring and metering. It provides effective technical support for realizing low-carbon operation and carbon emission management of the power grid and effectively reduces the overall carbon emissions while reducing business latency.

[0081] 2. The mathematical analysis and simulation results of this invention confirm that the cloud-edge collaborative architecture and collaborative task optimization offloading method can effectively reduce the total carbon emissions of computing networks and transmission systems while providing more flexible business performance guarantees. Therefore, it is expected to become an important enabling technology for green and low-carbon power grid operation and maintenance systems. Attached Figure Description

[0082] Figure 1 This is a flowchart illustrating the cloud-edge collaborative smart grid operation and maintenance carbon reduction method based on the alternating direction multiplier method of the present invention.

[0083] Figure 2 This is the flowchart of the ADMM algorithm.

[0084] Figure 3 compares the actual carbon emissions and energy consumption with the minimum total energy consumption as the objective function when the number of edge node tasks is different. (a) compares the actual carbon emissions with the minimum total energy consumption as the objective function when the number of edge node tasks is different, and (b) compares the actual energy consumption with the minimum total energy consumption as the objective function when the number of edge node tasks is different. Detailed Implementation

[0085] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0086] Please see Figure 1 This invention provides a cloud-edge collaborative power grid operation and maintenance carbon reduction method based on the alternating direction multiplier method, comprising the following steps:

[0087] S1: Design a cloud-edge collaborative power grid operation and maintenance system, which includes terminal devices, a cloud computing center, and edge nodes. The terminal devices are responsible for collecting various sensing data from the industrial internet and aggregating the collected data to the corresponding edge nodes for processing via wireless or wired data links. The cloud computing center provides high-quality, high-performance computing services with abundant available computing and power resources. Edge nodes provide wireless communication infrastructure that can wirelessly communicate with terminal devices and aggregate data, such as wireless communication base stations, access points, or aggregation gateways, and also include small servers that can provide data processing and computing services.

[0088] Specifically, consider a cloud-edge collaborative power grid operation and maintenance system comprising M edge nodes and one cloud computing center, where the cloud computing center is denoted by C, and the edge nodes are represented by a set S = {S1, ..., S2}. m ,…,S M} represents the index number of the edge node; the number of terminal devices is N, represented by the set D = {D1, ..., D2}. n ,…,D N} represents the index number of the terminal device. Using γ m Represents edge node S m The proportion of data task units received that are offloaded to local processing, where 0 ≤ γ m ≤; when γ m When = 1, it indicates that the edge node S m It will use its local computing resources to process all data task units; when γ m When = 0, it indicates an edge node S. m All received data task units are forwarded to cloud computing center C for processing.

[0089] S2: Model the relevant latency of the cloud-edge collaborative power grid operation and maintenance system.

[0090] The latency of power grid services is one of the key indicators for measuring the response speed, stability and reliability of the power grid. It has a crucial impact on the overall service quality of the power grid and, more importantly, is closely related to the carbon emissions of power grid operation and maintenance.

[0091] Considering the modeling of latency related to cloud-edge collaborative power grid operation and maintenance systems, this embodiment of the invention uses T m Indicates the relationship with edge node S m The service response latency of connected terminal devices includes the round-trip time of data task units and the queuing latency for processing. The terminal device and its associated edge node S m The round-trip data transmission delay between them can be considered a constant. express.

[0092] Step S2 includes the following sub-steps:

[0093] S21: Edge node service latency modeling without local unloading.

[0094] Edge node S m Alternatively, all received data task units can be directly forwarded to cloud computing center C for processing via a wired backbone network. In this case, the edge computing network can be equivalent to a traditional cloud computing network, meaning all computing tasks are completed by cloud computing center C. m =0. Since cloud computing centers typically have ample computing resources, the queuing latency of data task units is negligible compared to the processing latency. Assuming a linear relationship between the processing latency and the number of data task units, the number of data task units that cloud computing center C can process per unit time is v. In summary, without local offloading, edge node S... m The service latency can be expressed as:

[0095]

[0096] in, For edge node S m All data task units s m Round-trip transmission latency between the cloud computing center; s m This indicates the total number of data task units.

[0097] S22: Edge node service latency modeling during full local unloading.

[0098] Edge node S m The option is to offload all received data task units to the local machine for processing. In this case, the cloud computing center will not process any data task units from the edge nodes. m =1. However, due to physical limitations and cost control, the local computing resources of edge nodes are relatively limited, so the queuing latency of data processing task units cannot be ignored. In this invention, the embodiments of this invention adopt a universal queuing model, namely the M / M / 1 queuing system, to model the data task units to be processed at the edge nodes, and use μ m Represents edge node S m The maximum number of data task units that local computing resources can process per unit of time. In this case, embodiments of the present invention can offload all local resources, allowing edge node S to... m The latency model for providing services to its associated user equipment is as follows:

[0099]

[0100] Where μ m >sm

[0101] S23: Latency modeling for cloud-edge collaborative offloading services.

[0102] In the case of cloud-edge collaborative offloading services, each edge node S m Only a portion of the data task unit that processes the received data using its own local computing resources is available, γ. m s m This indicates that the remaining data received by the task unit will be forwarded to the cloud, using (1-γ) m )s m This indicates that, in embodiments of the present invention, when a portion of the local data is unloaded, the edge node S... m The latency model for providing services to its associated user equipment is as follows:

[0103]

[0104] Where 0≤γ m ≤1, and γ m s m <μ m .

[0105] S3. Based on the modeling of relevant delays in step S2, model the comprehensive carbon emissions of the cloud-edge collaborative power grid operation and maintenance system.

[0106] The comprehensive carbon emissions of the power grid operation and maintenance system mainly include: carbon emissions generated by terminal equipment collecting data and transmitting it to the corresponding base station via wireless channels; carbon emissions generated by edge nodes connected to the base station processing locally offloaded data task units; carbon emissions generated by cloud computing centers processing received data task units; and carbon emissions generated by data transmission between edge nodes and cloud computing centers via the backbone network.

[0107] Step S3 includes the following sub-steps:

[0108] S31: Carbon emission modeling of edge nodes.

[0109] Edge node S m Once deployed, its carbon intensity is determined by the combined sources of electricity supply and can be considered a constant, denoted by ∈ m Therefore, its carbon emissions are primarily determined by total operating power and processing latency. The total power consumed by an edge node depends on Power Usage Effectiveness (PUE), static and dynamic power consumption. Power Usage Effectiveness is a widely used energy efficiency indicator for data centers, defined as the node's total annual power consumption divided by the node's total annual power consumption of its IT equipment, expressed as η. mStatic power consumption, also known as leakage power consumption, is mainly caused by leakage current in the system and is unrelated to the computing resource usage of each edge node; while dynamic power consumption is mainly determined by the activity of computing resources. In this embodiment of the invention, it is respectively represented by... and β m Represents edge node S m The total power P is calculated by considering the static power consumption and the dynamic power consumption generated by the data processing unit. m It can be represented as:

[0110]

[0111] Furthermore, as mentioned earlier, edge node S m In the data task unit γ that processes its local unloading m s m Queuing delay is Therefore, its carbon emissions E m (γ m It can be modeled as:

[0112]

[0113] S32: Carbon emission modeling for cloud computing centers.

[0114] Similar to edge nodes, the total power of a cloud computing center is determined by energy efficiency, static power consumption, and dynamic power consumption. However, unlike edge nodes, its static power consumption is significantly higher. Dynamic power consumption β generated by processing a unit of data task c and energy efficiency η c The carbon intensity is relatively high. Meanwhile, because cloud computing centers consume enormous amounts of electricity, they are typically deployed in regions where clean energy accounts for a high percentage of the electricity supply. Compared to edge nodes that are deployed to network terminals, their carbon intensity is significantly lower. m Typically small. Since the cloud computing center can receive and process data task units forwarded by all edge nodes, the total number of data task units it receives per unit time can be expressed as:

[0115]

[0116] Where γ=<γ1,γ2,...,γ M Accordingly, as mentioned earlier, due to the abundant computing resources in cloud computing centers, the queuing latency of data task units is negligible compared to the processing latency, and its corresponding service latency is... And its total power consumption P C (γ) can be represented as:

[0117]

[0118] In summary, the carbon emissions E of cloud computing centers C (γ) can be modeled as:

[0119]

[0120] S33: Carbon emission modeling for backbone network data transmission.

[0121] In the cloud-edge collaborative power grid operation and maintenance system considered in this invention, the backbone network mainly provides high-speed wired data transmission services between edge servers and cloud computing centers. The total power consumption in the backbone network can also be divided into static power consumption and dynamic power consumption. Since the static power consumption of the backbone network is usually composed of the networking equipment of edge nodes or cloud computing centers, in this embodiment of the invention, the static power consumption of the backbone network is considered as a part of the static power consumption of edge nodes or cloud computing centers, i.e., included in... and In this study, the focus is on the dynamic power consumption during data transmission. Since the dynamic power consumption and transmission latency of data transmission in the backbone network depend on the number of data task units being transmitted, we take the example of a single edge node forwarding a portion of the data task units it receives to the cloud computing center to analyze the carbon emission model of backbone network data transmission.

[0122] Use β L η represents the dynamic power consumption of a single data transmission task unit in the backbone network. L The power efficiency of the edge node S represents the energy utilization efficiency during data transmission in the backbone network. m The received data task unit s m (1-γ) m When the data is forwarded to the cloud data center, the transmission latency is... Total power consumption P L (γ m This can be represented as:

[0123] P L (γ m )=η L β L (1-γ m )s m (3-6)

[0124] Therefore, using ∈ L If we represent the carbon intensity of power resources used for data transmission in the backbone network, then the carbon emissions generated when edge nodes forward some data task units to the cloud data center through the backbone network can be expressed as:

[0125]

[0126] S4: Based on the modeling of comprehensive carbon emissions in step S3, construct and optimize the objective function for minimizing the carbon emissions of the cloud-edge collaborative power grid operation and maintenance system.

[0127] Specifically, it is assumed that all power grid equipment terminals must first upload their generated services to the nearest base station via a wireless network, and then decide whether to offload them to the edge node connected to the base station or upload them to the cloud computing center via the backbone data transmission network.

[0128] Step S4 includes the following sub-steps:

[0129] S41: Construct an objective function for minimizing the carbon emissions of the cloud-edge collaborative power grid operation and maintenance system.

[0130] When the cloud-edge collaborative network consists of a cloud computing center and M edge nodes, the embodiment of the present invention can model the problem of minimizing carbon emissions of the cloud-edge collaborative power grid operation and maintenance system as follows:

[0131]

[0132] Here, γ includes the proportion of the business volume offloaded by M edge nodes to the total number of businesses across all industrial internet terminals. E C (γ), E m (γ m ) and E L (γ m The carbon emissions from the cloud computing center, the carbon emissions from the m-th edge node, and the carbon emissions from the transmission of business data uploaded by the industrial internet terminal corresponding to the m-th edge node are respectively the computational carbon emissions of the cloud computing center, the computational carbon emissions of the m-th edge node, and the transmission carbon emissions generated by the industrial internet terminal corresponding to the m-th edge node uploading business data. This embodiment of the invention notes that different edge nodes may receive different types of services; therefore, the computational carbon emissions from the cloud computing center need to comprehensively consider the service type of each edge node m and the corresponding offloading volume γ for that service type. m .

[0133] S42: The global optimal solution is obtained by using the alternating direction multiplier method to obtain the objective function established in step S41.

[0134] First, the iterative formula of the alternating direction multiplier method is simplified based on the augmented Lagrange expression.

[0135] The problem described above, within the allowable parameter range, is a convex optimization problem with inequality constraints, which can be solved using traditional optimization methods, such as the interior-point method. In particular, if the constraints are transformed into part of the objective function by introducing an indicator function, the global optimum can be obtained using the Alternative Direction Method of Multipliers (ADMM), the process of which is as follows: Figure 2As shown. Observation reveals that in the objective function of the above problem, the carbon emissions E of the edge node network... m (γ m Carbon emissions from backbone network data transmission (E) L (γ m The sum of the carbon emissions E of the edge node network is independent of each edge node, meaning there is no coupling between them; similarly, each constraint is also independent and uncoupled with respect to each edge node. Therefore, the carbon emissions E of the edge node network... m (γ m Carbon emissions from backbone network data transmission (E) L (γ m The minimization part of the problem can be solved independently and in a distributed manner across the edge nodes, while the cloud computing center can coordinate the solution process of the sub-problems among the edge nodes and E. C The solution minimizes (γ). The Alternating Direction Multiplier Method (ADMM) described above can achieve a task forwarding strategy that minimizes system-level carbon emissions while protecting the privacy of information such as computing power, power, and communication resources among edge nodes.

[0136] Since the constraints of the above model are all convex sets, and the feasible region is the intersection of the four constraints, it can be concluded that the feasible region is a convex set. Therefore, the indicator function I can be used. C Adding the feasible region to the objective function (.) transforms the problem into the following:

[0137] min[E C (γ)+1 T (E(γ)+E L (γ))]+I C (γ) (4-2)

[0138] Among them I C (γ) is the indicator function of the feasible region, that is:

[0139]

[0140] To transform the problem into the standard ADMM form, this embodiment of the invention introduces the constraint z = γ, so the problem becomes:

[0141]

[0142] Where f(y) = E C (y)+1 T (E(y)+E L (y)),h(z)=I C (z). After the above equivalent transformation, the problem is converted into the standard ADMM form. Then, the augmented Lagrange multiplier formula is established for the next step of solving:

[0143]

[0144] Where ρ>0 are the augmented Lagrange coefficients, and λ∈R M Let be the dual variables. Based on the above Lagrange multipliers, the ADMM algorithm can be composed of the following iterative steps:

[0145] y k+1 =argminL ρ (y,z k ,λ k (4-6)

[0146]

[0147] λ k+1 =λ k +ρ(y k+1 -z k+1 (4-8)

[0148] To improve computational efficiency and provide a concise expression, this invention simplifies the iterative formula by combining linear and quadratic terms in the augmented Lagrange multiplier and scaling the dual variable. The scaled dual variable is defined as follows: Therefore, ADMM can be equivalent to the following expression:

[0149]

[0150] u k+1 :=u k +ρ(y k+1 -z k+1 (4-11)

[0151] Secondly, the stopping criteria and algorithm steps of the Alternating Direction Multiplier Method (ADMM) are determined.

[0152] Since the above problem is a constrained optimization problem, the convergence criterion of the algorithm should rely on the KKT conditions of the constrained optimization problem. Therefore, this invention first defines the original feasibility residual and the dual feasibility residual of the above problem:

[0153] r k =||y k -z k ||2 (4-12)

[0154] s k =||z k -z k-1 ||2 (4-13)

[0155] Where r k For the original feasibility residual, s kThese are the dual feasibility residuals. Based on the primal and dual feasibility conditions of the KKT conditions for a constrained optimization problem, the values ​​of both residuals should be 0 when ADMM converges. However, in practical applications, the stopping criterion is often whether the two residuals are sufficiently small.

[0156]

[0157] Where, ∈ pri >0, ∈ dual >0 indicates the feasibility error of the suboptimal solution to the above problem, and the embodiments of the present invention set ∈ abs and ∈ rel 10 respectively -4 and 10 -2 Based on the above formula, the suboptimal solution of the algorithm can be obtained as the final solution to the problem.

[0158] like Figure 2 As shown, this invention uses the alternating direction multiplier method to obtain the global optimal solution for the objective function established in step S41. The detailed algorithm steps are as follows:

[0159] First, edge servers m initialize their respective decision variables, and the parameter coordinator initializes auxiliary variables, dual variables, and the maximum number of iterations MAX_ITER. Next, it checks if the iteration count i is less than the maximum number of iterations. If so, all edge servers update simultaneously, updating the decision variables according to formula (4-9) and feeding the results back to the parameter coordinator, waiting for the coordinator to send the auxiliary and dual variables. Upon receiving these, the parameter coordinator updates the auxiliary variables according to formula (4-10) and the dual variables according to formula (4-11). Then, it checks if the stopping criterion is met. If not, iteration continues; if it is met, the corresponding auxiliary and dual variables are sent to the corresponding edge server m, and the process ends.

[0160] Please refer to Figure 3. This embodiment of the invention simulates a cloud-edge computing network consisting of 60 power grid terminals, 6 edge nodes, and 1 cloud computing center. The performance and carbon emissions of three offloading schemes are mainly compared, including (1) full offloading, where all tasks uploaded from industrial internet terminals are processed by edge nodes, (2) no-task offloading, where all tasks are uploaded to the cloud computing center for processing, and (3) cloud-edge collaborative optimization offloading, which is the cloud-edge collaborative optimal task offloading method solved by convex optimization method.

[0161] To compare the performance of the three schemes in reducing carbon emissions from the power grid, the objective function is to minimize the carbon emissions defined in formula (4-1), and to measure the actual carbon emissions and energy consumption as the total number of tasks received by each edge node increases. Figure 3(a) shows that the proposed cloud-edge collaborative optimization offloading scheme achieves the lowest carbon emissions from the power grid compared to the cases of no local offloading and all local offloading, and increases slowly with the increase in task load. Figure 3(b) shows a significant difference between the changes in energy consumption and carbon emissions. Although the system and total energy consumption are lower when all tasks are offloaded to local edge nodes, the large differences in carbon intensity coefficients among edge nodes result in substantial actual carbon emissions.

[0162] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. An alternating direction method of multipliers based cloud-edge collaborative smart grid operation and maintenance carbon reduction method, characterized in that, Comprise the following steps: S1: design cloud edge collaborative power grid operation and maintenance system, the cloud edge collaborative power grid operation and maintenance system includes terminal equipment, cloud computing center and edge node;The terminal equipment is responsible for the collection of different sensing data in industrial internet, and the collected data is converged to the corresponding edge node for processing through wireless or wired data link;The cloud computing center is used for providing computing service;The edge node provides wireless communication infrastructure for wireless communication with terminal equipment and data aggregation and small server for data processing and computing service; S2: model the related time delay of the cloud edge collaborative power grid operation and maintenance system; S3, according to the modeling of related time delay in step S2, model the comprehensive carbon emission of the cloud edge collaborative power grid operation and maintenance system; S4: according to the modeling of comprehensive carbon emission in step S3, the carbon emission minimization problem of the cloud edge collaborative power grid operation and maintenance system is constructed to obtain the global optimal solution by using alternating direction multiplier method; The step S3 includes: S31: carbon emission modeling of edge node Total power consumed by the edge node is represented as: （3-1）； wherein represents the electrical energy utilization efficiency, defined as the total annual electricity consumption of the node divided by the total annual electricity consumption of the node's information technology equipment; and respectively represent the static power consumption and the dynamic power consumption generated by the edge node for processing a unit data task; Edge node In processing its locally offloaded data task units The queuing delay at the edge node is modeled as: The carbon emissions of the edge node are modeled as: （3-2）； S32: carbon emission modeling of cloud computing center the total amount of data task units accepted by cloud computing center in unit time is expressed as: （3-3）； wherein ; The corresponding time delay of the cloud computing center service is The total power consumption is expressed as: （3-4）； Carbon emissions of cloud computing centers Modeling is: （3-5）； S33: Carbon emission modeling of backbone network data transmission represents the dynamic power consumption caused by the transmission of a unit data task unit in the backbone network, represents the energy utilization efficiency of the backbone network during data transmission, and the edge node transmits the data task unit received by it to the cloud data center, and the transmission delay is , and the total power consumption is represented as: （3-6）； With denoting the carbon intensity of the power resource of data transmission in the backbone network, the carbon emissions generated when the edge node forwards part of the data task units to the cloud data center through the backbone network are expressed as: （3-7）； In the step S4, when the cloud-edge collaborative network is composed of one cloud computing center and The objective function of the cloud-edge collaborative power grid operation and maintenance system carbon emission minimization problem is constructed as follows: （4-1）； Among them, 𝜸 includes The proportion of business volume offloaded by each edge node to the total number of business transactions of all industrial internet terminals; , and These are the computational carbon emissions of cloud computing centers, the first The calculated carbon emissions of the first edge node and the first The carbon emissions generated by the industrial internet terminals corresponding to each edge node uploading business data.

2. The cloud-edge collaborative smart grid operation and maintenance carbon reduction method based on the alternating direction multiplier method according to claim 1, wherein: In step S1, the cloud-edge collaborative power grid operation and maintenance system includes M edge nodes and 1 cloud computing center, wherein the cloud computing center is denoted by C, the edge nodes are denoted by a set S = { ,… , …, }, and m represents the index number of the edge node; the number of terminal devices is N, denoted by a set D = { , , …, }, and n represents the index number of the terminal device; with denotes the edge node the proportion of the received data task units that are offloaded to the local for processing, wherein when = 1, denotes the edge node will process all data task units with its local computing resources; when = 0, denotes the edge node will forward all received data task units to the cloud computing center C for processing.

3. The cloud-edge collaborative smart grid operation and carbon reduction method based on the alternating direction multiplier method according to claim 2, characterized in that: The step S2 includes: S21: edge node service delay modeling edge node without local offloading When the received all data task units are directly forwarded to the cloud computing center C for processing through the wired backbone network, the edge computing network is equivalent to a traditional cloud computing network, that is, all computing tasks are completed by the cloud computing center C, = 0; The processing time delay of the data task unit is linearly related to the processing quantity, and the quantity of the data task units processed by the cloud computing center C in a unit time is , the service time delay of the edge node without local offloading is represented as , and the service time delay of the edge node with local offloading is represented as : （2-1）； wherein, denotes the round-trip data transmission latency between the terminal device and the edge node with which it is associated; denotes the round-trip data transmission latency between the terminal device and the edge node with which it is associated; denotes the round-trip transmission latency between the edge node and the cloud computing center; denotes the number of all data task units of the edge node; denotes the round-trip transmission latency between the edge node and the cloud computing center; denotes the number of all data task units of the edge node; S22: Edge node service latency modeling edge node when all local offloading When the received data task unit is selected to be offloaded to the local and processed, the cloud computing center will not process any data task unit from the edge node, = 1; adopt The queuing system models the data task units to be processed at the edge node, and uses To represent the maximum number of data task units that the edge node Can process per unit of time of local computing resources; when all local offloading, the edge node The corresponding latency of the service provided to its associated user equipment is modeled as: （2-2）； wherein ; S23: Cloud-edge collaborative offloading service latency modeling In the case of cloud-edge collaborative offloading service, each edge node processes a part of its received data task units using only its own local computing resources, denoted by , and forwards the rest of its received data task units to the cloud, denoted by . When partially offloading locally, the edge node models the corresponding latency of providing service to its associated user equipment as: （2-3）； wherein .

4. The cloud-edge collaborative smart grid operation and maintenance carbon reduction method based on the alternating direction multiplier method according to claim 1, wherein: In the step S4, when the global optimal solution of the objective function established in step S4 is obtained by using alternating direction multiplier method: Firstly, based on the augmented Lagrange expression, the iteration formula of alternating direction multiplier method is simplified, specifically, with the indicator function Adding the feasible region to the objective function, the objective function is transformed to the following problem: （4-2）； wherein is an indicator function of the feasible region, i.e.: （4-3）； Introducing constraints The objective function is transformed to: （4-4）； wherein ; After the above equivalent conversion, the objective function is converted into the standard form of alternating direction multiplier method; Then, the augmented Lagrange multiplier formula is established, and the next step is solved: （4-5）； wherein are Lagrange multipliers, are dual variables; Based on the above Lagrange multiplier formula, the alternating direction multiplier method is composed of the following iteration steps: （4-6）； （4-7）； （4-8）； The simplification of the iterative formula is made by combining the linear and quadratic terms in the augmented Lagrangian and rescaling the dual variables The alternating direction method of multipliers is equivalent to the following expression （4-9）； （4-10）； （4-11）； Secondly, the stop criterion of alternating direction multiplier method ADMM is determined, specifically, Define the original feasibility residual and dual feasibility residual of the above problem: （4-12）； （4-13）； wherein is the original feasibility residual, is the dual feasibility residual, and the two residuals are judged to be sufficiently small as a stopping criterion.

5. The cloud-edge collaborative smart grid operation and carbon reduction method based on the alternating direction multiplier method according to claim 4, characterized in that: The judgment of whether the two residuals are sufficiently small is used as the stop criterion, specifically including: （4-14）； wherein > 0, > 0 indicates the feasibility error of the suboptimal solution of the objective function, and the suboptimal solution of the alternating direction multiplier method is obtained as the final solution of the objective function.

6. The cloud-edge collaborative smart grid operation and carbon reduction method based on the alternating direction multiplier method according to claim 5, characterized in that: and are and .

Images