New energy electric field intelligent scheduling method and system based on edge calculation

Abstract

The invention discloses a new energy electric field intelligent scheduling method and system based on edge computing, and the method comprises the steps: constructing a digital twin scheduling network based on a physical entity network, and a cloud edge-end collaborative data processing and scheduling optimization model based on edge computing, which are used for supporting the medium and long term planning and intelligent scheduling optimization of a power grid; a multi-target random optimization problem of data processing delay minimization and virtual-real synchronization delay error minimization is constructed, the random optimization problem is converted into a deterministic optimization problem through a Lyapunov optimization technology, and finally, a non-dominated sorting genetic algorithm is adopted to obtain an optimization strategy of data scheduling and resource allocation. By fusing edge calculation and digital twinborn technologies, automation of data scheduling in a new energy electric field is realized, the response speed and stability of a power grid system are improved, and technical support is provided for efficient operation of power grid planning and new energy power generation.

Description

Technical Field

[0001] The present invention relates to the technical field of mobile edge computing, and in particular, to an intelligent scheduling method and system for new energy power fields based on edge computing. Background Art

[0002] With the rapid development of renewable energy, the scale of new energy power fields (such as wind farms, photovoltaic power fields, etc.) is constantly expanding, and the amount of data generated by equipment in the power field is increasing exponentially. How to efficiently manage and schedule this data has become an important challenge for ensuring the stable operation and effective management of new energy power fields. Especially in new energy power fields, the data generated by key equipment (such as power generation equipment, meteorological monitoring equipment, and user equipment), including meteorological data (such as wind speed, wind direction, temperature, humidity, and air pressure), power generation data (such as real-time power generation and equipment status), and user data (such as operation data and interaction information generated during work), is of great significance for the planning and scheduling of the power grid and the prediction of new energy power generation capacity.

[0003] Traditional data processing methods mainly rely on a centralized cloud computing architecture, that is, all data is transmitted to the cloud center for centralized processing. However, this method has significant limitations when facing large-scale power fields: on the one hand, the transmission of massive data brings problems of high latency and high energy consumption; on the other hand, it is difficult for the cloud center to perform real-time processing on dynamically changing data, resulting in a lag in scheduling decisions, thereby affecting the operation efficiency and stability of the power field.

[0004] To solve these problems, edge computing technology provides a solution for efficiently processing data in new energy power fields. Edge computing deploys computing resources on edge servers close to the data source, offloads some data processing tasks from the cloud center to the edge servers for processing, effectively reducing data transmission latency and cloud computing pressure. However, how to efficiently schedule and process multi-source heterogeneous data in new energy power fields remains a difficult problem that needs to be solved urgently. Especially in complex scenarios with a wide variety of equipment and latency limitations in task processing, traditional scheduling methods are difficult to balance real-time performance and global optimality. Summary of the Invention

[0005] In view of the above existing problems, the present invention is proposed.

[0006] Therefore, in view of the fact that the prior art cannot achieve efficient intelligent scheduling of new energy power fields, the present invention proposes an intelligent scheduling method and system for new energy power fields based on edge computing, which optimizes data scheduling and resource allocation by combining edge computing and digital twin technology to achieve multiple objectives of minimizing data processing latency and minimizing the virtual-real synchronization error.

[0007] To achieve the above object, the present invention provides the following technical solutions:

[0008] In a first aspect, the present invention provides an intelligent scheduling method for a new energy power field based on edge computing, including: obtaining the location information of various devices in the new energy power field and building a physical entity network; constructing a digital twin scheduling network based on the physical entity network to virtually map the location information and resource status of various devices in the new energy power field; constructing a cloud-edge-end collaborative data processing and scheduling optimization model based on the digital twin scheduling network; constructing a multi-objective stochastic optimization problem with minimized data processing delay and minimized virtual-real synchronization delay error according to the cloud-edge-end collaborative data processing and scheduling optimization model; using an optimization technique to handle the long-term load balancing constraint problem of the edge server in the multi-objective stochastic optimization problem, and transforming the original problem into a deterministic optimization problem without long-term constraints within a single time slot; applying a non-dominated sorting genetic algorithm to solve the deterministic optimization problem to obtain an optimized strategy for data scheduling and resource allocation.

[0009] As a preferred solution of the intelligent scheduling method for a new energy power field based on edge computing according to the present invention, where: the obtaining the location information of various devices in the new energy power field and building a physical entity network includes:

[0010] Obtaining the location information of power generation devices, user devices, meteorological monitoring devices and edge servers in the new energy power field, and then obtaining a device set Edge server set where the three-dimensional coordinates of the device are represented as The three-dimensional coordinates of the edge server are represented as

[0011] Building a physical entity network according to the location information of relevant devices and edge servers in the new energy power field.

[0012] As a preferred solution of the intelligent scheduling method for a new energy power field based on edge computing according to the present invention, where: the virtual mapping of the location information and resource status of various devices in the new energy power field includes:

[0013] Based on the physical entity network, the entire activity cycle of the system is divided into T time slots t with equal length and Δ, expressed as

[0014] It is recorded that within the t-th time slot, the task data generated by the i-th device is represented as where D i (t) represents the size of the task data volume, C i (t) represents the number of CPU cycles required to process 1 bit of data, T i (t) represents the delay limit of the processed task, and then the digital twin corresponding to the device is represented as

[0015] The digital twin corresponding to the edge server is denoted as where f j (t) represents the estimated value of the DT for the actual computing frequency of the edge server, representing the error between the actual computing frequency and the DT estimated value.

[0016] As a preferred solution of the intelligent scheduling method for a new energy electric field based on edge computing according to the present invention, wherein: the construction of the cloud-edge-end collaborative data processing and scheduling optimization model based on edge computing includes:

[0017] In the new energy electric field, the data transmission mechanism between each device and the edge server adopts OFDMA. Assuming there are S available subcarriers for wireless transmission; B bw is the bandwidth of each subcarrier, and the subcarrier allocation matrix is where w i,s,j (t)=1 means that in the t-th time slot, device i establishes communication with edge server j through subcarrier s and offloads the task data to edge server j; in each time slot t, each subcarrier can only be selected by one device, and each device only offloads to one edge server;

[0018] f i min (t)=(D i (t)C i (t)) / T i (t) represents the minimum computing frequency required to process the task data. If f i min (t)≥f j (t), the data is directly transmitted to the cloud center for processing without being processed after being offloaded to the edge server; since device i exclusively occupies a subcarrier, the communication rate between device i and edge server j in the t-th time slot is expressed as:

[0019]

[0020] where the channel gain matrix of the subcarrier is expressed as represents the channel gain when device i transmits to edge server j, α0 represents the unit channel power gain, represents the three-dimensional Euclidean distance between device i and edge server j, p e represents the transmission power of the device, and δ represents Gaussian white noise, following a Gaussian distribution with zero expectation;

[0021] Based on the above communication offloading decision, it is recorded that device i offloads data within the t-th time slot and ensures that the data task is offloaded within the t-th time slot; a collaborative computing strategy is adopted to process the tasks offloaded by the device, that is, the edge server j simultaneously processes the task data offloaded by multiple devices and dynamically allocates computing resources f i,j (t), and the constraint of the allocated computing resources is:

[0022] Set the load threshold ε and require that the long-term average load of each edge server does not exceed the load threshold, which is expressed by the formula:

[0023]

[0024] where, represents the total amount of data processed by the edge server j within the t-th time slot;

[0025] The computing delay required for the edge server j to calculate the data offloaded by device i within the t-th time slot is:

[0026]

[0027] The processing delay requirement for device i to offload data is expressed as:

[0028]

[0029] The total computing delay of the edge server j within the t-th time slot is expressed as:

[0030]

[0031] The total task processing delay of the edge server j within the t-th time slot is expressed as:

[0032]

[0033] The data processing delay of the scheduling system is expressed as:

[0034]

[0035] The error between the actual computing delay and the DT estimated delay of the edge server j within the t-th time slot is expressed as:

[0036]

[0037] The virtual-real synchronization delay error of the scheduling system is expressed as:

[0038] ΔT(t) = max{ΔT1(t),..., ΔT B (t)}.

[0039] As a preferred solution of the intelligent scheduling method for new energy power fields based on edge computing according to the present invention, wherein: the construction of the multi-objective stochastic optimization problem includes:

[0040] P1:

[0041] s.t.C1:

[0042] C2:

[0043] C3:

[0044] C4:

[0045] C5:

[0046] C6:

[0047] C7:

[0048] Wherein, F(t) respectively represent the computing task scheduling decision and resource allocation strategy. The constraint C1 represents the offloading decision variable of the device. The constraints C2 and C3 respectively represent that each subcarrier is only selected by one device within the t-th time slot, and each device is only offloaded to one edge server. The constraints C4 and C5 respectively represent the single and total constraints for the edge server to allocate computing resources. The constraint C6 represents the processing delay constraint for device i to offload data. The constraint C7 represents the long-term load balancing constraint of the edge server.

[0049] As a preferred solution of the intelligent scheduling method for new energy power fields based on edge computing according to the present invention, wherein: the transformation of the original problem into a deterministic optimization problem without long-term constraints within a single time slot includes:

[0050] Construct a dynamic load virtual queue Q j (t);

[0051] Define the Lyapunov function to describe the sum of squares of the backlogs of all dynamic load virtual queues within the t-th time slot;

[0052] Construct the Lyapunov drift according to the expectation of the difference between two adjacent time slots of the Lyapunov function;

[0053] Obtain the Lyapunov drift-plus-penalty function through the fusion of the drift-plus-penalty strategy and Furthermore, transform the multi-objective stochastic optimization problem into a deterministic optimization problem P2 without long-term constraints, expressed as:

[0054] P2:

[0055]

[0056] s.t.C1-C6

[0057] As a preferred solution of the intelligent scheduling method for a new energy electric field based on edge computing according to the present invention, wherein: solving the deterministic optimization problem by using the non-dominated sorting genetic algorithm includes:

[0058] Initializing parameters, including the maximum number of iterations H, the population size K, the crossover probability q cs , the mutation probability q vn , the simulated binary crossover parameter β, the polynomial mutation parameter μ, and the task data cache size Q(t);

[0059] Generating an initial population p 0,k =[w 1,1,1 (t),...,w N,S,B (t),f 1,1 (t),...,f N,B (t)],k = 1,...,K;

[0060] For each individual in the population, calculating its fitness according to the objective functions of data processing delay and virtual-real synchronization delay error;

[0061] Calculating the violation value CV j of each individual in p0, which represents the part of the time consumed to actually complete the task exceeding the specified time;

[0062] Based on the fitness calculation results, performing fast non-dominated sorting on the individuals in the population, dividing them into multiple non-dominated levels, and calculating the crowding distance of the individuals in each non-dominated level to measure the distribution of individuals in the solution space;

[0063] According to the violation value, non-dominated level, and crowding distance, preferentially selecting individuals, and performing crossover and mutation operations according to the probabilities q cs and q vn to generate new individuals, and repeating this step until the generated quantity reaches K;

[0064] According to the calculated fitness of the offspring and the violation value, merging the parent generation and the offspring to form a new population with a population size of 2K; according to the calculated non-dominated level and crowding distance of the new population, and based on the violation value, non-dominated level, and crowding distance, selecting the next parent population;

[0065] Repeating the above steps until the maximum number of iterations H is reached, and outputting the final Pareto optimal solution set F.

[0066] In a second aspect, the present invention provides an intelligent scheduling system for a new energy power field based on edge computing, including:

[0067] An information fitting module, configured to obtain the location information of various devices in the new energy power field, build a physical entity network, construct a digital twin scheduling network based on the physical entity network, and perform virtual mapping on the location information and resource status of various devices in the new energy power field;

[0068] A model construction module, configured to construct a cloud-edge-end collaborative data processing and scheduling optimization model based on edge computing according to the digital twin scheduling network;

[0069] An optimization problem construction and processing module, configured to construct a multi-objective stochastic optimization problem of minimizing data processing delay and minimizing the virtual-real synchronization delay error according to the cloud-edge-end collaborative data processing and scheduling optimization model, and adopt an optimization technique to process the long-term load balancing constraint problem of the edge server in the multi-objective stochastic optimization problem, and transform the original problem into a deterministic optimization problem without long-term constraints within a single time slot;

[0070] A problem solving module, configured to solve the deterministic optimization problem by applying a non-dominated sorting genetic algorithm to obtain an optimization strategy for data scheduling and resource allocation.

[0071] In a third aspect, the present invention provides an electronic device, including:

[0072] A memory and a processor;

[0073] The memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions, and when the computer-executable instructions are executed by the processor, the steps of the intelligent scheduling method for a new energy power field based on edge computing are implemented.

[0074] In a fourth aspect, the present invention provides a computer-readable storage medium, which stores computer-executable instructions, and when the computer-executable instructions are executed by a processor, the steps of the intelligent scheduling method for a new energy power field based on edge computing are implemented.

[0075] Compared with the prior art, the beneficial effects of the present invention are as follows: The present invention provides an intelligent scheduling method and system for a new energy power field based on edge computing. By integrating edge computing and digital twin technology, a digital twin scheduling network based on a physical entity network and a cloud-edge-end collaborative data processing and scheduling optimization model are constructed to realize the intelligence and automation of data scheduling in the new energy power field; a multi-objective stochastic optimization problem of minimizing data processing delay and minimizing virtual-real synchronization delay error is constructed, and the long-term load balancing constraint is dynamically processed through Lyapunov optimization technology, so as to gradually approach the global optimal solution and balance system performance and objectives; the NSGA-II algorithm is used to obtain a globally optimal scheduling decision and resource allocation scheme, effectively improving the real-time performance and reliability of the system; the computing resources of edge servers are reasonably allocated, and through efficient data processing and optimized scheduling, the resource utilization rate and scheduling efficiency are improved, providing important technical support for the efficient operation of the new energy power field and the medium- and long-term planning and scheduling of the power grid. BRIEF DESCRIPTION OF THE DRAWINGS

[0076] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following will briefly introduce the drawings required for the description of the embodiments. Obviously, the following drawings are only some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can be obtained based on these drawings.

[0077] Figure 1 It is a schematic diagram of the overall process logic of the intelligent scheduling method for a new energy power field based on edge computing according to an embodiment of the present invention;

[0078] Figure 2 It is a schematic diagram of the intelligent scheduling optimization model of cloud-edge-end collaboration in a new energy power field for the intelligent scheduling method for a new energy power field based on edge computing according to an embodiment of the present invention;

[0079] Figure 3 It is a flowchart of the NSGA-II algorithm for the intelligent scheduling method for a new energy power field based on edge computing according to an embodiment of the present invention;

[0080] Figure 4 It is a convergence performance curve diagram of the NSGA-II algorithm for solving data scheduling and resource allocation for the intelligent scheduling method for a new energy power field based on edge computing according to an embodiment of the present invention;

[0081] Figure 5 It is a Pareto front distribution diagram for the intelligent scheduling method for a new energy power field based on edge computing according to an embodiment of the present invention;

[0082] Figure 6Comparison chart of each objective of different algorithms of the intelligent scheduling method for new energy power fields based on edge computing according to an embodiment of the present invention under different numbers of devices. Detailed implementation manners

[0083] To make the above objects, features, and advantages of the present invention more obvious and understandable, the following will describe in detail the specific implementation manners of the present invention with reference to the accompanying drawings of the specification. Obviously, the described embodiments are part of the embodiments of the present invention, rather than all embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the scope of protection of the present invention.

[0084] Embodiment 1

[0085] Referring to Figures 1-3 An embodiment of the present invention provides an intelligent scheduling method for a new energy power field based on edge computing. As Figure 2 shown, it consists of N devices, B edge servers, and 1 cloud center, aiming to optimize and schedule the data generated by devices in the new energy power field to achieve multiple objectives of minimizing data processing delay and minimizing the virtual-real synchronization delay error. As Figure 1 shown, it specifically includes the following steps:

[0086] S100: Obtain the location information of various devices in the new energy power field and build a physical entity network;

[0087] S200: Based on the physical entity network, construct a digital twin scheduling network to virtually map the location information and resource status of various devices in the new energy power field;

[0088] S300: Construct a cloud-edge-end collaborative data processing and scheduling optimization model based on the digital twin scheduling network;

[0089] S400: Construct a multi-objective stochastic optimization problem of minimizing data processing delay and minimizing the virtual-real synchronization delay error according to the cloud-edge-end collaborative data processing and scheduling optimization model;

[0090] S500: Adopt an optimization technique to handle the long-term load balancing constraint problem of edge servers in the multi-objective stochastic optimization problem, and transform the original problem into a deterministic optimization problem without long-term constraints within a single time slot;

[0091] S600: Apply the non-dominated sorting genetic algorithm to solve the deterministic optimization problem and obtain an optimization strategy for data scheduling and resource allocation.

[0092] It should be noted that the present invention provides an intelligent scheduling method and system for a new energy power field based on edge computing. By integrating edge computing and digital twin technologies, a digital twin scheduling network based on a physical entity network and a cloud-edge-end collaborative data processing and scheduling optimization model are constructed to realize the intelligence and automation of data scheduling in the new energy power field; a multi-objective stochastic optimization problem of minimizing data processing delay and virtual-real synchronization delay error is constructed, and the long-term load balancing constraint is dynamically processed through Lyapunov optimization technology, so as to gradually approach the global optimal solution and balance system performance and objectives; the NSGA-II algorithm is used to obtain the globally optimal scheduling decision and resource allocation scheme, effectively improving the real-time performance and reliability of the system; the computing resources of edge servers are reasonably allocated, and through efficient data processing and optimized scheduling, the resource utilization rate and scheduling efficiency are improved, providing important technical support for the efficient operation of the new energy power field and the medium- and long-term planning and scheduling of the power grid.

[0093] In the embodiment of the present application, in step S100, based on the data collected by a third party, the location information of various devices in the new energy power field is obtained, including power generation devices, user devices, meteorological monitoring devices, and edge servers, and then a physical entity network is built, which specifically includes:

[0094] S101: Based on the data collected by a third party, the location information of power generation devices, user devices, meteorological monitoring devices, and edge servers in the new energy power field is obtained, and then a device set is obtained Edge server set Where the three-dimensional coordinates of the device are expressed as The three-dimensional coordinates of the edge server are expressed as

[0095] S102: Based on the location information of relevant devices and edge servers in the new energy power field, a physical entity network is built.

[0096] In the embodiment of the present application, in step S200, a digital twin scheduling network is constructed in the cloud center based on the information of the physical entity network, and the location information and resource status of various devices in the new energy power field are virtually mapped to realize the synchronization of physical devices and the digital twin scheduling network, which specifically includes:

[0097] S201: Based on the physical entity network, the entire activity cycle of the system is divided into T time slots t with equal length and Δ, which is expressed as

[0098] S202: Denote that in the t-th time slot, the task data generated by the i-th device is expressed as Where D i (t) represents the size of the task data volume, C i(t) represents the number of CPU cycles required to process 1 bit of data, T i (t) represents the latency limit of the processed task, and thus the digital twin corresponding to the device is denoted as To enable the digital twin network to operate smoothly, all digital twins are constructed in the cloud center;

[0099] S203: The digital twin corresponding to the edge server is denoted as where f j (t) represents the estimated value of the actual computing frequency of the edge server by the DT, represents the error between the actual computing frequency and the DT estimated value.

[0100] In the embodiments of the present application, the above step S300 constructs an edge computing-based cloud-edge-end collaborative data processing and scheduling optimization model according to the digital twin scheduling network, specifically including:

[0101] S301: Since the computing resources of each device are limited and cannot meet the processing latency requirements of computing tasks, edge computing is introduced to construct a data task offloading model, a communication model, and a latency model for the intelligent scheduling of new energy power grids;

[0102] S302: In the new energy power grid, the data transmission mechanism between each device and the edge server adopts OFDMA. It is assumed that there are S available subcarriers for wireless transmission; B bw is the bandwidth of each subcarrier, and the subcarrier allocation matrix is where w i,s,j (t) = 1 indicates that in the t-th time slot, device i establishes communication with edge server j through subcarrier s and offloads task data to edge server j; in each time slot t, each subcarrier can only be selected by one device, that is each device only offloads to one edge server, that is

[0103] Each device-generated task data has a computing latency limit, f i min (t) = (D i (t)C i (t)) / T i (t) represents the minimum computing frequency required to process the task data. If f i min (t) ≥ f j(t), the data is directly transmitted to the cloud center for processing without being processed after being unloaded to the edge server; since device i exclusively occupies a subcarrier, the interference between devices can be ignored; according to the Shannon formula, the communication rate between device i and edge server j within the t-th time slot is expressed as:

[0104]

[0105] where the channel gain matrix of the subcarrier is expressed as represents the channel gain when device i transmits to edge server j, α0 represents the unit channel power gain, represents the three-dimensional Euclidean distance between device i and edge server j, p e represents the transmission power of the device, δ represents Gaussian white noise, which follows a Gaussian distribution with zero expectation;

[0106] S303: Based on the above communication offloading decision, it is recorded that device i offloads data within the t-th time slot and ensures that the data task is offloaded within the t-th time slot. The corresponding transmission delay can be expressed as:

[0107]

[0108] The collaborative computing strategy is adopted to process the tasks offloaded by the device, that is, edge server j simultaneously processes the task data offloaded by multiple devices and dynamically allocates computing resource f i,j (t), and the constraint of the allocated computing resource is:

[0109]

[0110] To ensure that the load of each edge server is within a reasonable range in the long term and avoid overloading or idling of some edge servers, a load threshold ε is set and it is required that the long-term average load of each edge server does not exceed the load threshold, which is expressed by the formula:

[0111]

[0112] where, represents the total amount of data processed by edge server j within the t-th time slot;

[0113] Since the CPU computing resources of the edge server are sufficient, the tasks offloaded in each time slot can be processed within that time slot. The computing delay required for edge server j to calculate the data offloaded by device i within the t-th time slot is:

[0114]

[0115] The processing delay requirement for the data offloaded by device i is expressed as:

[0116]

[0117] The total computing delay of edge server j in the t-th time slot is expressed as:

[0118]

[0119] The total task processing delay of edge server j in the t-th time slot is expressed as:

[0120]

[0121] The data processing delay of the scheduling system is expressed as:

[0122]

[0123] S304: Since there is a delay error in the data interaction between the DT and the actual device, the DT cannot fully and accurately reflect the actual state of the device. However, the error between the actual computing delay and the DT estimated delay can be obtained in advance; the error between the true computing delay and the DT estimated delay of edge server j in the t-th time slot is expressed as:

[0124]

[0125] The virtual-real synchronization delay error of the scheduling system is expressed as:

[0126] ΔT(t) = max{ΔT1(t), …, ΔT B (t)}

[0127] In the embodiment of the present application, step S400 constructs a multi-objective stochastic optimization problem for minimizing the data processing delay and the virtual-real synchronization delay error according to the cloud-edge-end collaborative data processing and scheduling optimization model, specifically including:

[0128] S401: During the entire scheduling period, under the condition that each edge server provides computing resources, try to ensure that the data processing delay and the virtual-real synchronization delay error of each device are minimized, and then construct a multi-objective stochastic optimization problem, expressed as:

[0129] P1:

[0130] s.t.C1:

[0131] C2:

[0132] C3:

[0133] C4:

[0134] C5:

[0135] C6:

[0136] C7:

[0137] Among them, F(t) represents the computing task scheduling decision and resource allocation strategy respectively:

[0138]

[0139] F(t) = [f 1,1 (t),..., f i,j (t),..., f N,B (t)]

[0140] Constraint C1 represents the offloading decision variable of the device. Constraints C2 and C3 respectively represent that each subcarrier is selected by only one device within the t-th time slot, and each device is offloaded to only one edge server. Constraints C4 and C5 respectively represent the single and total constraints for the edge server to allocate computing resources. Constraint C6 represents the processing delay constraint for device i to offload data. Constraint C7 represents the long-term load balancing constraint of the edge server.

[0141] In the embodiment of the present application, the above step S500 uses an optimization technique to handle the long-term load balancing constraint problem of the edge server in the multi-objective stochastic optimization problem, and transforms the original problem into a deterministic optimization problem without long-term constraints within a single time slot, specifically including:

[0142] S501: According to constraint C7, construct a dynamic load virtual queue Q j (t) for each edge server, and the derivation process is as follows:

[0143] Q j (t) = max{Q j (t - 1) + D j (t - 1) - κ, 0}

[0144] Among them, κ represents the load capacity allocated in a single time slot, and D j (t - 1) - κ represents the deviation of the load in the (t - 1)-th time slot;

[0145] S502: Define the Lyapunov function to describe the sum of squares of the backlogs of all dynamic load virtual queues within the t-th time slot:

[0146]

[0147] Among them, Θ(t) = {Q1(t),..., Q B (t)};

[0148] Construct the Lyapunov drift according to the expectation of the difference between two adjacent time slots of the Lyapunov function as follows:

[0149]

[0150] S503: Minimizing the Lyapunov drift aims to reduce the backlog in each dynamic load virtual queue and ensure the stable operation of the queue; obtain the Lyapunov drift-plus-penalty function by integrating the drift-plus-penalty strategy and

[0151]

[0152] where the non-negative coefficients V1 and V2 are used to measure the weight ratio between the drift and the objective function, and the upper bound of the Lyapunov drift penalty function can be expressed as:

[0153]

[0154] where is a constant;

[0155] S504: Transform the multi-objective stochastic optimization problem into a deterministic optimization problem P2 without long-term constraints, expressed as:

[0156] P2:

[0157]

[0158] s.t. C1 - C6

[0159] In the embodiments of the present application, the above step S600 applies the non-dominated sorting genetic algorithm to solve the deterministic optimization problem and obtains the optimization strategy for data scheduling and resource allocation, as Figure 3 shown, specifically including:

[0160] S601: Initialize the parameters, including the maximum number of iterations H, the population size K, the crossover probability q cs , the mutation probability q vn , the simulated binary crossover parameter β, the polynomial mutation parameter μ, and the task data cache Q(t);

[0161] S602: Generate the initial population p 0,k = [w 1,1,1 (t), …, w N,S,B (t), f 1,1 (t), …, f N,B (t)], k = 1, …, K;

[0162] S603: For each individual in the population, calculate its fitness according to the objective function of data processing delay and virtual-real synchronization delay error;

[0163] S604: Calculate the default value CV of each individual in p0 j , which represents the part where the time consumed to actually complete the task exceeds the specified time, and the calculation formula is:

[0164]

[0165] S605: Based on the fitness calculation results, perform fast non-dominated sorting on the individuals in the population and divide them into multiple non-dominated levels;

[0166] S606: Calculate the crowding distance of the individuals in each non-dominated level to measure the distribution of individuals in the solution space;

[0167] S607: Select individuals preferentially according to the default value, non-dominated level, and crowding distance, and perform crossover and mutation operations according to probabilities q cs and q vn to generate new individuals. Repeat this step until the generated quantity reaches K; among them, for integer variables, use the single-point crossover method. Randomly select a crossover point in two parent individuals and exchange the variable values at the crossover point to generate new offspring individuals; the mutation operation uses the random mutation method, randomly select a variable and assign it a random integer within the feasible range; for real variables, the crossover operation uses the simulated binary crossover method, and calculate the generated offspring variable values based on the variable values of the parent individuals and the randomly generated control parameter β; the mutation operation uses the polynomial mutation method, and generate a small-range random adjustment value according to the current variable value and the upper and lower limits of the feasible range, so as to achieve variable fine-tuning and finally generate the offspring population;

[0168] S608: According to the calculated fitness and default value of the offspring, merge the parent and offspring to form a new population with a population size of 2K; according to the calculated non-dominated level and crowding distance of the new population, and select the next parent population based on the default value, non-dominated level, and crowding distance;

[0169] S609: Repeat the above steps until the maximum number of iterations H is reached, and output the final Pareto optimal solution set F.

[0170] Therefore, in this embodiment, the digital twin technology is adopted to build an intelligent dispatching system for new energy power plants to carry out refined management of new energy power plants, which can construct a virtual mapping of equipment and data streams, and monitor and predict the operating status, task distribution, and resource usage of equipment in the power plant in real time, so as to provide support for dispatching optimization. Combining edge computing and digital twin technology, through intelligent data scheduling and resource allocation, it is possible to effectively minimize the data processing delay and the virtual-real synchronization delay error, improve the data processing efficiency and resource utilization rate, and provide technical guarantees for the efficient operation of new energy power plants and grid planning and dispatching.

[0171] Embodiment 2

[0172] Referring to Figures 4-6 , based on the previous embodiment, this embodiment provides an application example of an intelligent dispatching method and system for new energy power plants based on edge computing to verify and illustrate the technical effects adopted in this method.

[0173] Figure 4 shows the convergence performance of the NSGA-II algorithm when solving data scheduling and resource allocation. As the number of iterations increases, the hypervolume index gradually increases and tends to be stable, indicating that the distribution and approximation of the population solution set are gradually optimized. In the early rapid convergence stage of the algorithm, the hypervolume value shows an obvious upward trend, indicating that the population quickly approaches the Pareto front and the quality of the solution is gradually improved. After that, as the number of iterations further increases, the growth rate of the hypervolume value gradually decreases, indicating that the population tends to be stable and the algorithm gradually converges. It can be seen from the figure that after 400 iterations, the hypervolume value tends to be maximized and the NSGA-II algorithm reaches convergence.

[0174] Figure 5 shows the Pareto front distribution diagram of data processing delay and virtual-real synchronization delay error. Each point in the figure represents a non-dominated solution, that is, the Pareto solution obtained by the NSGA-II algorithm in the multi-objective optimization of data scheduling and resource allocation. It can be seen from the figure that as the data processing delay increases, the virtual-real synchronization delay error decreases, and each point together constitutes the Pareto front, intuitively reflecting the trade-off relationship in multi-objective optimization. This shows that the present invention can effectively explore and approach the Pareto optimal solution set through the NSGA-II algorithm, provide a variety of trade-off solutions for the optimization of data processing delay and virtual-real synchronization delay error, and verify the applicability of the method in complex scenarios.

[0175] Figure 6 shows the comparison of each objective of different algorithms under different numbers of devices. From Figure 6(a) It can be seen that the average data processing delay gradually increases with the increase in the number of devices. However, the NSGA-II algorithm is always superior to the MOEA / D and DGEA algorithms, showing the lowest average data processing delay. The MOEA / D algorithm ranks second, and the DGEA algorithm has the largest delay. In the case of a large number of devices, the advantage of NSGA-II is more obvious, reflecting a more efficient resource allocation ability. Figure 6 (b) shows the changing trend of the average virtual-real synchronization delay error. With the increase in the number of devices, the virtual-real synchronization delay errors of all algorithms gradually increase. Among them, the NSGA-II algorithm still performs the best, followed by the MOEA / D algorithm, and the DGEA algorithm has the largest virtual-real synchronization delay error. This indicates that the NSGA-II algorithm can better balance the data processing delay and the virtual-real synchronization delay error in multi-objective optimization, demonstrating higher optimization efficiency and system stability in complex scenarios.

[0176] As can be seen from the above application examples, the present invention provides an intelligent scheduling method and system for a new energy electric field based on edge computing. By integrating edge computing and digital twin technology, a digital twin scheduling network based on the physical entity network and a cloud-edge-end collaborative data processing and scheduling optimization model are constructed to realize the intelligence and automation of data scheduling in the new energy electric field; a multi-objective stochastic optimization problem of minimizing the data processing delay and the virtual-real synchronization delay error is constructed, and the long-term load balancing constraint is dynamically processed through Lyapunov optimization technology, so as to gradually approach the global optimal solution and balance the system performance and objectives; the NSGA-II algorithm is used to obtain the globally optimal scheduling decision and resource allocation scheme, effectively improving the real-time performance and reliability of the system; the computing resources of the edge server are reasonably allocated, and through efficient data processing and optimized scheduling, the resource utilization rate and scheduling efficiency are improved, providing important technical support for the efficient operation of the new energy electric field and the medium- and long-term planning and scheduling of the power grid.

[0177] Example 3

[0178] In this embodiment, an intelligent scheduling system for a new energy electric field based on edge computing is provided, including:

[0179] An information fitting module, configured to obtain the location information of various devices in the new energy electric field, build a physical entity network, construct a digital twin scheduling network based on the physical entity network, and perform virtual mapping on the location information and resource status of various devices in the new energy electric field;

[0180] A model construction module, configured to construct a cloud-edge-end collaborative data processing and scheduling optimization model based on edge computing according to the digital twin scheduling network;

[0181] An optimization problem construction and processing module, which is used to construct a multi-objective stochastic optimization problem with minimized data processing delay and minimized virtual-real synchronization delay error according to the cloud-edge-end collaborative data processing and scheduling optimization model, and adopts optimization techniques to handle the long-term load balancing constraint problem of edge servers in the multi-objective stochastic optimization problem, and transforms the original problem into a deterministic optimization problem without long-term constraints within a single time slot;

[0182] A problem solving module, which is used to solve the deterministic optimization problem by applying the non-dominated sorting genetic algorithm to obtain an optimized strategy for data scheduling and resource allocation.

[0183] It should be noted that the technical solution of the intelligent scheduling system of the new energy power plant based on edge computing belongs to the same concept as the technical solution of the above-mentioned intelligent scheduling method of the new energy power plant based on edge computing. For the details not described in detail in the technical solution of the intelligent scheduling system of the new energy power plant based on edge computing in this embodiment, reference can be made to the description of the technical solution of the intelligent scheduling method of the new energy power plant based on edge computing.

[0184] The above-mentioned unit modules can be embedded in the processor of the computer device in hardware form or independent of the processor, or stored in the memory of the computer device in software form, so that the processor can call and execute the operations corresponding to the above-mentioned modules.

[0185] This embodiment also provides an electronic device, which includes a processor, a memory, a communication interface, a display screen and an input device connected through a system bus. Among them, the processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The communication interface of the computer device is used to communicate with an external terminal in a wired or wireless manner, and the wireless manner can be realized through WIFI, a carrier network, NFC (near field communication) or other technologies. The computer program, when executed by the processor, implements an intelligent scheduling method of a new energy power plant based on edge computing. The display screen of the computer device can be a liquid crystal display screen or an electronic ink display screen, and the input device of the computer device can be a touch layer covered on the display screen, or a button, a trackball or a touchpad set on the shell of the computer device, or an external keyboard, a touchpad or a mouse, etc.

[0186] This embodiment also provides a computer-readable storage medium, on which a computer program is stored, and when the program is executed by the processor, it implements the method proposed in the above-mentioned embodiment.

[0187] The storage medium proposed in this embodiment and the method proposed in the above embodiment belong to the same inventive concept. For technical details not described in detail in this embodiment, reference can be made to the above embodiment, and this embodiment has the same beneficial effects as the above embodiment.

[0188] From the above description of the embodiments, those skilled in the art can clearly understand that the present invention can be implemented by means of software and the necessary general-purpose hardware. Of course, it can also be implemented by hardware, but in many cases the former is a better implementation. Based on such an understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as a floppy disk, read-only memory (ROM), random access memory (RAM), flash memory (FLASH), hard disk, or optical disc of a computer, including several instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the method of the embodiment of the present invention.

[0189] It should be noted that the above embodiments are only used to illustrate the technical solution of the present invention and not to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those of ordinary skill in the art should understand that the technical solution of the present invention can be modified or equivalently replaced without departing from the spirit and scope of the technical solution of the present invention, and it should be covered by the scope of the claims of the present invention.

[0190] Those skilled in the art should understand that the embodiments of the present application can be provided as a method, a system, or a computer program product. Therefore, the present application can take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware aspects. Moreover, the present application can take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code. The solutions in the embodiments of the present application can be implemented in various computer languages.

[0191] The present application is described with reference to the flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each flow and / or block in the flowchart and / or block diagram, and the combination of flows and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to the processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing devices to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing devices generate for implementing in the process Figure 1One or more processes and / or blocks Figure 1 means for the functions specified in one or more blocks.

[0192] These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instruction means that implement the functions in the process Figure 1 One or more processes and / or blocks Figure 1 the functions specified in one or more blocks.

[0193] These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer-implemented process, whereby the instructions executed on the computer or other programmable apparatus provide steps for implementing the functions in the process Figure 1 One or more processes and / or blocks Figure 1 the functions specified in one or more blocks.

[0194] Although the preferred embodiments of the present application have been described, additional changes and modifications can be made by those skilled in the art once they learn the basic creative concept. Therefore, the appended claims are intended to be construed to include the preferred embodiments as well as all changes and modifications that fall within the scope of the present application.

[0195] Obviously, those skilled in the art can make various changes and modifications to the present application without departing from the spirit and scope of the present application. Thus, if these modifications and variations of the present application fall within the scope of the claims of the present application and their equivalent technologies, the present application is also intended to include these modifications and variations.

Claims

1. A new energy electric field intelligent scheduling method based on edge computing, characterized in that: include: Obtain the location information of various devices in the new energy electric field and build a physical entity network; A digital twin dispatching network is constructed based on the physical entity network to virtually map the location information and resource status of various types of equipment in the new energy electric field; Constructing a cloud-edge-end collaborative data processing and scheduling optimization model based on edge computing according to the digital twin scheduling network; According to the cloud-edge collaborative data processing and scheduling optimization model, a multi-objective stochastic optimization problem of minimizing data processing delay and minimizing virtual-real synchronization delay error is constructed; Using optimization technology to deal with the long-term load balancing constraint problem of edge servers in the multi-objective stochastic optimization problem, the original problem is transformed into a deterministic optimization problem without long-term constraints in a single time slot; A non-dominated sorting genetic algorithm is used to solve the deterministic optimization problem and obtain an optimization strategy for data scheduling and resource allocation.

2. The method for intelligent dispatching of new energy electric fields based on edge computing according to claim 1, characterized in that: The acquisition of location information of various devices in the new energy electric field and the establishment of a physical entity network include: Obtain the location information of power generation equipment, user equipment, meteorological monitoring equipment and edge servers in the new energy power field, and then obtain the equipment collection Edge Server Collection The three-dimensional coordinates of the device are expressed as The three-dimensional coordinates of the edge server are expressed as Build a physical network based on the location information of relevant equipment and edge servers in the new energy power field.

3. The method for intelligent dispatching of new energy electric fields based on edge computing according to claim 2, characterized in that: The virtual mapping of the location information and resource status of various types of equipment in the new energy electric field includes: Based on the physical entity network, the entire activity cycle of the system is divided into T time slots t of equal length and Δ, expressed as In the tth time slot, the task data generated by the i-th device is expressed as Where D i (t) represents the size of the task data, C i (t) represents the number of CPU cycles required to process 1 bit of data, T i (t) represents the delay limit of the task being processed, and the digital twin corresponding to the device is expressed as The digital twin corresponding to the edge server is represented as where f j (t) represents DT’s estimate of the actual computing frequency of the edge server, Represents the error between the true calculated frequency and the DT estimate.

4. The method for intelligent dispatching of new energy electric fields based on edge computing according to claim 3 is characterized in that: The cloud-edge-end collaborative data processing and scheduling optimization model based on edge computing includes: In the new energy electric field, the data transmission mechanism between each device and the edge server adopts OFDMA. Assuming that there are S available subcarriers Wireless transmission; B bw is the bandwidth of each subcarrier, and the subcarrier allocation matrix is where w i,s,j (t) = 1 means that in the tth time slot, device i establishes communication with edge server j through subcarrier s and unloads task data to edge server j; in each time slot t, each subcarrier can only be selected by one device, and each device only unloads to one edge server; f i min (t)=(D i (t)C i (t)) / T i (t) represents the minimum computing frequency required to process the task data. If f i min (t)≥f j (t), the data is unloaded to the edge server without processing and is directly transmitted to the cloud center for processing; since device i occupies one subcarrier, the communication rate between device i and edge server j in the tth time slot is expressed as: Among them, the channel gain matrix of the subcarrier is expressed as represents the channel gain when device i transmits to edge server j, α0 represents the unit channel power gain, represents the three-dimensional Euclidean distance between device i and edge server j, p e represents the transmission power of the device, δ represents Gaussian white noise, which obeys zero-expectation Gaussian distribution; Based on the above communication offloading decision, device i performs data offloading in the tth time slot and ensures that the data task is offloaded in the tth time slot; the collaborative computing strategy is used to process the task offloaded by the device, that is, the edge server j processes the task data offloaded by multiple devices at the same time and dynamically allocates computing resources f to them. i,j (t), the constraints on the allocated computing resources are: The load threshold ε is set and the long-term average load of each edge server is required not to exceed the load threshold. The formula is expressed as: in, represents the total amount of data processed by edge server j in the tth time slot; The computational delay required by edge server j to calculate the data offloaded by device i in the tth time slot is: The processing delay requirement of the data offloaded by device i is expressed as: The total computation delay of edge server j in the tth time slot is expressed as: The total delay of task processing by edge server j in the tth time slot is expressed as: The data processing delay of the scheduling system is expressed as: The error between the actual calculated delay of edge server j and the DT estimated delay in the tth time slot is expressed as: The virtual and real synchronization delay error of the scheduling system is expressed as: ΔT(t)=max{ΔT1(t),...,ΔT B (t)}。 5. The method for intelligent dispatching of new energy electric fields based on edge computing according to claim 4, characterized in that: The construction of the multi-objective stochastic optimization problem includes: in, F(t) represents the computing task scheduling decision and resource allocation strategy respectively, constraint C1 represents the device offloading decision variable, constraints C2 and C3 respectively represent that each subcarrier is only selected by one device in the tth time slot, and each device is only offloaded to one edge server, constraints C4 and C5 respectively represent the single and total constraints of edge server allocation of computing resources, constraint C6 represents the processing delay constraint of device i offloading data, and constraint C7 represents the long-term load balancing constraint of the edge server.

6. The method for intelligent dispatching of new energy electric fields based on edge computing according to claim 5, characterized in that: The transformation of the original problem into a deterministic optimization problem without long-term constraints in a single time slot includes: Construct a dynamic load virtual queue Q for each edge server j (t); Define the Lyapunov function to describe the sum of squares of the backlog of all dynamic load virtual queues in the tth time slot; Constructing a Lyapunov drift according to the expectation of the difference between two adjacent time slots of the Lyapunov function; The Lyapunov drift penalty function is obtained by integrating the drift penalty strategy and Then, the multi-objective stochastic optimization problem is transformed into a deterministic optimization problem P2 without long-term constraints, which can be expressed as:

7. The method for intelligent dispatching of new energy electric fields based on edge computing according to claim 6, characterized in that: The applying of the non-dominated sorting genetic algorithm to solve the deterministic optimization problem includes: Initialization parameters, including the maximum number of iterations H, population size K, and crossover probability q cs , mutation probability q vn , simulated binary crossover parameter β, polynomial variation parameter μ, task data cache size Q(t); Generate the initial population p 0,k =[w 1,1,1 (t),...,w N,S,B (t),f 1,1 (t),...,f N,B (t)], k = 1, ..., K; For each individual in the population, its fitness is calculated based on the objective function of data processing delay and virtual-real synchronization delay error; Calculate the default value CV of each individual in p0 j , which means the time actually consumed to complete the task exceeds the specified time; Based on the fitness calculation results, the individuals in the population are quickly non-dominated sorted and divided into multiple non-dominated levels. The crowding distance of each non-dominated level is calculated to measure the distribution of individuals in the solution space. According to the default value, non-dominated level and crowding distance, the individual is selected according to the probability q cs and q vn Perform crossover and mutation operations to generate new individuals, and repeat this step until the number of generated individuals reaches K; According to the calculated offspring fitness and default value, merge the parent and offspring to form a new population with a population size of 2K; according to the calculated non-dominated level and crowding distance of the new population, select the next parent population based on the default value, non-dominated level and crowding distance; Repeat the above steps until the maximum number of iterations H is reached, and output the final Pareto optimal solution set F.

8. A system using the method for intelligent dispatching of new energy electric fields based on edge computing as claimed in any one of claims 1 to 7, characterized in that: include: An information fitting module is used to obtain the location information of various types of equipment in the new energy electric field, build a physical entity network, build a digital twin scheduling network based on the physical entity network, and virtually map the location information and resource status of various types of equipment in the new energy electric field; A model construction module, used to construct a cloud-edge-end collaborative data processing and scheduling optimization model based on edge computing according to the digital twin scheduling network; An optimization problem construction and processing module is used to construct a multi-objective stochastic optimization problem of minimizing data processing delay and minimizing virtual-real synchronization delay error according to the cloud-edge collaborative data processing and scheduling optimization model, and use optimization technology to process the long-term load balancing constraint problem of the edge server in the multi-objective stochastic optimization problem, and convert the original problem into a deterministic optimization problem without long-term constraints in a single time slot; The problem solving module is used to solve the deterministic optimization problem by applying a non-dominated sorting genetic algorithm to obtain an optimization strategy for data scheduling and resource allocation.

9. An electronic device, comprising a memory and a processor, characterized in that: The memory is used to store computer executable instructions, and the processor is used to execute the computer executable instructions. When the computer executable instructions are executed by the processor, the steps of the method according to claims 1 to 7 are implemented.

10. A computer-readable storage medium having computer-executable instructions stored thereon, characterized in that: When the computer executable instructions are executed by a processor, the steps of the method according to claims 1 to 7 are implemented.

Images