A distributed energy collaborative optimization method based on cloud edge architecture

Abstract

The application relates to a distributed energy collaborative optimization method based on a cloud edge end along architecture and belongs to the technical field of intelligent micro-grid. A technical scheme is as follows: S1. Defining a cloud edge end along four-layer architecture; S2. Information transmission link: information is reported upwards, and is transferred and delivered between the same levels; S3. Demand optimization link: demand is decomposed downwards, and is collaboratively optimized between the same levels; and S4. Energy flow link: energy is aggregated upwards, and is interacted and communicated between the same levels. Through the designed cloud edge end along four-layer architecture, the distributed energy can be converted from centralized regulation and control to active participation in demand response. On this basis, a distributed energy collaborative optimization method based on the cloud edge end along architecture is further designed, the global collaborative optimization of the distributed energy is realized, the response error of the distributed energy is first reduced, the response time of the distributed energy is effectively reduced, the problem of excessively high computing power of traditional centralized regulation and control is solved, the operation cost of the distributed energy is reduced, and the problem of high operation cost of the distributed energy is solved.

Description

Technical Field

[0001] This invention relates to a distributed energy collaborative optimization method based on a cloud-edge-device architecture, belonging to the field of smart microgrid technology. Background Technology

[0002] First, traditional centralized control models have potential weaknesses such as numerous steps, significant time delays, and slow response times. With tens or even hundreds of millions of distributed energy resources as new controllable resources, uploading all information to the main station for unified optimization and decision-making would lead to enormous decision-making delays, making it difficult to cope with the real-time, ever-changing, and heterogeneous operating environment of distributed energy. Second, distributed energy suffers from a severe lack of global coordination capabilities and insufficient response accuracy, resulting in weak dynamic tracking capabilities and poor dispatchability. This limits the grid's effective utilization of distributed energy and also affects the overall operating efficiency and safety level of the power system.

[0003] With the rapid increase in the installed capacity of distributed renewable energy, the grid regulation requirements of existing technologies are shifting from "centralized control" to "collective self-consistency." For example, Chinese patent CN116544941B discloses a distributed energy application strategy control system and method based on a cloud-edge collaborative architecture. Although it achieves economic operation of distributed energy to a certain extent, the cloud computing power is still too high, resulting in a long response time for distributed energy and a large response error. Chinese patent CN120416258B proposes a distributed renewable energy data scheduling and processing method based on cloud-edge-device collaboration, designing a three-layer cloud-edge-device architecture. Although it reduces the response time of distributed energy to a certain extent, the operating cost of distributed energy is high, and it still relies on the traditional centralized scheduling power command transmission from the upper layer to the lower layer. Therefore, the reduction in cloud computing power is limited. Summary of the Invention

[0004] This invention proposes a distributed energy collaborative optimization method based on a cloud-edge-device architecture. It establishes a regulation architecture where distributed energy actively participates in demand response, and further designs a global collaborative optimization method for distributed energy based on this architecture. First, it addresses the problem of insufficient response accuracy of distributed energy, reducing response errors. Second, it effectively solves the problem of excessive cloud computing power in traditional centralized control modes, reducing the response time of distributed energy. Finally, it effectively addresses the problem of high operating costs of distributed energy, reducing operating costs and ensuring the safe, stable, and economical operation of the distribution network after massive distributed energy is connected to the network. This invention provides a distributed energy collaborative optimization method based on a cloud-edge-device architecture. It not only designs a four-layer distributed energy regulation architecture based on cloud-edge-device architecture, but also further designs a global collaborative optimization method for distributed energy based on this. Finally, simulations verify the effectiveness of the method, solving the aforementioned technical problems existing in existing technologies.

[0005] The technical solution of this invention is:

[0006] A distributed energy collaborative optimization method based on a cloud-edge-device architecture establishes a regulation architecture in which distributed energy actively participates in demand response and performs collaborative optimization of distributed energy to improve the response accuracy, reduce response error, shorten response time, and lower operating costs of distributed energy, achieving global collaborative optimization of distributed energy operation. The method includes the following steps:

[0007] S1. Define a four-layer architecture: cloud, edge, and endpoint.

[0008] S2. Information transmission process: Information is reported upwards and circulated among peers;

[0009] S3. Requirements Optimization Phase: Decompose requirements downwards and optimize collaboratively among peers;

[0010] S4. Energy flow link: Energy aggregates upwards and interacts and communicates with each other at the same level.

[0011] In step S1, the definitions and functions of the four-layer architecture along the cloud-edge-device are as follows:

[0012] The cloud layer is used to receive the demands of the control center and the electricity market, and contains one or more distributed energy aggregation clusters beneath it;

[0013] The edge layer refers to the distributed energy aggregation group, which is used to sense cloud demand, optimize power adjustment commands between distributed energy groups, and complete mutual assistance between distributed energy groups to meet cloud demand. It contains one or more distributed energy aggregation zones.

[0014] The edge layer refers to the distributed energy aggregation area, which is used to sense the edge layer demand, optimize the power adjustment command of the distributed energy interval, and complete the mutual assistance of the distributed energy interval for the edge layer demand. It contains one or more distributed energy individuals.

[0015] "Along the layer" refers to the use of distributed energy individuals to sense end-layer demand, optimize power regulation commands among distributed energy individuals, and complete mutual assistance among distributed energy individuals to meet end-layer demand. These distributed energy individuals include distributed photovoltaics, distributed wind power, battery energy storage systems, and electric vehicles, among others.

[0016] The specific steps of the information transmission stage in step S2 are as follows:

[0017] S21. Distributed energy individuals along the layer report predicted new energy output, predicted uncontrollable load, battery energy storage system output limit, and electric vehicle output limit to the distributed energy zone at the end layer, and send copies to other distributed energy individuals in the same distributed energy zone;

[0018] S22. The end-layer distributed energy zone aggregates the predicted new energy output, predicted uncontrollable load, battery energy storage system output limit, and electric vehicle output limit information of the distributed energy individuals within the zone to form a first-level aggregation result. The first-level aggregation result is then reported to the edge-layer distributed energy group and copied to other distributed energy zones within the same distributed energy group.

[0019] S23. The edge layer distributed energy group re-aggregates the aggregation results of the distributed energy regions within the group to form a secondary aggregation result, and sends the secondary aggregation result to other distributed energy groups.

[0020] The specific steps of the demand optimization step in step S3 are as follows:

[0021] S31. Cloud layer acquisition of control center and electricity market demand;

[0022] S32. Each distributed energy group in the edge layer senses the cloud layer demand and, in conjunction with the secondary aggregation results of each distributed energy group, completes the optimization of the power regulation command of each distributed energy group, thereby realizing the proactive collaborative optimization of the power regulation command of the edge layer.

[0023] S33. Each distributed energy zone in the end layer senses the power regulation command of the edge layer, and combines the first-level aggregation results of each distributed energy zone to complete the optimization of the power regulation command of each distributed energy zone, thereby realizing the proactive collaborative optimization of the power regulation command of the end layer.

[0024] S34. Each distributed energy individual along the layer senses the power regulation command along the layer, and combines the predicted new energy output, predicted uncontrollable load, battery energy storage system output limit, and electric vehicle output limit information of each distributed energy individual to complete the optimization of the power regulation command of each distributed energy individual, thereby realizing the proactive collaborative optimization of the power regulation command along the layer.

[0025] The specific steps of the energy flow process in step S4 are as follows:

[0026] S41. Distributed energy individuals along the layer respond to their optimized power regulation commands along the layer and sense the response results of other distributed energy individuals in the same distributed energy area in real time.

[0027] S42. The end-layer distributed energy zone summarizes the response results of the distributed energy individuals within the zone to form a first-level response result, and then communicates the first-level response results to other distributed energy zones within the same distributed energy group.

[0028] S43. The edge layer distributed energy group summarizes the updated first-level response results of the distributed energy areas within the group to form second-level response results, and then exchanges the second-level response results to send the second-level response results to other distributed energy groups;

[0029] S44. The cloud layer summarizes the secondary response results of all distributed energy groups, completes the demand regulation of the control center and the electricity market, and realizes the overall coordinated optimization of distributed energy.

[0030] The objective function and constraints for the proactive collaborative optimization of power regulation commands for each distributed energy group in step S32 are as follows:

[0031] Objective function: The objective is to minimize the operating cost of all distributed energy clusters. The calculation formula is as follows:

[0032] (1)

[0033] In the formula: and Let be the electricity sales revenue and electricity purchase cost of the m-th distributed energy group at time t, respectively. The operating cost of the battery storage system and electric vehicle for the m-th distributed energy cluster at time t is calculated as follows:

[0034] (2)

[0035] (3)

[0036] (4)

[0037] In the formula: and Let be the wind power and photovoltaic power of the m-th distributed energy cluster at time t, respectively. and Let be the charging and discharging power of the battery storage system and the electric vehicle of the m-th distributed energy cluster at time t, respectively, with charging represented by positive values ​​and discharging by negative values. and Let be the battery storage system capacity and electric vehicle capacity of the m-th distributed energy cluster, respectively. Let be the load power of the m-th distributed energy group at time t. and These are the electricity sales price and the electricity purchase price, respectively. and These are the unit costs of battery energy storage systems and electric vehicle operation, respectively. For scheduling intervals;

[0038] Constraints:

[0039] (5)

[0040] and These represent the maximum charging and discharging power of the battery storage system in the m-th distributed energy cluster and the maximum charging and discharging power of the electric vehicle, respectively. Let t be the cloud's required power. Let be the interaction power between the m-th distributed energy group and other distributed energy groups at time t. The external input is positive and the external output is negative.

[0041] The objective function and constraints for the proactive collaborative optimization of power regulation commands for each distributed energy zone in step S33 are as follows:

[0042] Objective function: The objective is to minimize the operating cost of all distributed energy zones. The calculation formula is as follows:

[0043] (6)

[0044] In the formula: and Let be the electricity sales revenue and electricity purchase cost of the h-th distributed energy zone at time t, respectively. The calculation method for the operating cost of the battery energy storage system and electric vehicle in the h-th distributed energy zone at time t is as follows:

[0045] (7)

[0046] (8)

[0047] (9)

[0048] In the formula: and These represent the wind power and photovoltaic power of the h-th distributed energy zone at time t, respectively. and Let be the charging and discharging power of the battery storage system and the electric vehicle in the h-th distributed energy zone at time t, respectively. Charging is represented by positive values, and discharging by negative values. and These represent the battery storage system capacity and electric vehicle capacity of the h-th distributed energy zone, respectively. Let be the load power of the h-th distributed energy zone at time t. and These are the electricity sales price and the electricity purchase price, respectively. and These are the unit costs of battery energy storage systems and electric vehicle operation, respectively. For scheduling intervals;

[0049] Constraints:

[0050] (10)

[0051] In the formula: and These represent the maximum charging and discharging power of the battery storage system and the maximum charging and discharging power of the electric vehicle in the h-th distributed energy zone, respectively. Let be the edge layer power requirement at time t. Let be the interaction power between the h-th distributed energy zone and other distributed energy zones at time t. The external input is positive and the external output is negative.

[0052] The objective function and constraints for the proactive collaborative optimization of power regulation commands for each distributed energy individual along the layer in step S34 are as follows:

[0053] Objective function: The goal is to minimize the operating cost of all individual distributed energy resources. The formula for calculating the operating cost of a distributed energy zone is as follows:

[0054] (11)

[0055] In the formula: and Let be the electricity sales revenue and electricity purchase cost of a certain distributed energy zone at time t, respectively. Let t be the operating cost of the battery storage system and electric vehicle in a certain distributed energy zone at time t. The calculation method is as follows:

[0056] (12)

[0057] (13)

[0058] (14)

[0059] In the formula: and Let be the wind power and photovoltaic power of a certain distributed energy zone at time t, respectively. and Let be the charging and discharging power of the battery storage system and the electric vehicle in a distributed energy zone at time t, respectively. Charging is represented by positive values, and discharging by negative values. and These represent the battery storage system capacity and electric vehicle capacity of a specific distributed energy zone. Let t be the load power of a certain distributed energy zone at time t. and These are the electricity sales price and the electricity purchase price, respectively. and These are the unit costs of battery energy storage systems and electric vehicle operation, respectively. For scheduling intervals;

[0060] Constraints:

[0061] (15)

[0062] In the formula: and These represent the maximum charging and discharging power of a battery storage system and the maximum charging and discharging power of an electric vehicle in a specific distributed energy zone. Let t be the power required by the end layer at time t.

[0063] The cloud-edge-end four-layer architecture in step S1:

[0064] Each distributed energy cluster in the edge layer is equipped with the same smart terminal and has the same White Whale optimization algorithm with all parameters embedded in it to complete the active collaborative optimization of the power adjustment command of each distributed energy cluster in the edge layer as described in step S32.

[0065] Each distributed energy zone in the end layer is equipped with the same smart terminal and has the same White Whale optimization algorithm with all parameters embedded in it to complete the active collaborative optimization of the power regulation command of each distributed energy zone in the end layer as described in step S33.

[0066] Each distributed energy individual along the layer is configured with the same smart terminal and implanted with the same White Whale optimization algorithm with all parameters, which is used to complete the active collaborative optimization of the power adjustment command of each distributed energy individual along the layer as described in step S34.

[0067] The beneficial effects of this invention are as follows: Through the designed cloud-edge-device-edge four-layer architecture, this invention enables distributed energy to shift from traditional centralized control to active participation in demand response. Furthermore, a distributed energy collaborative optimization method based on the cloud-edge-device-edge architecture is designed, achieving full-domain collaborative optimization of distributed energy. This firstly reduces the response error of distributed energy and improves its response accuracy; secondly, it effectively reduces the response time of distributed energy, solving the latency problem of traditional centralized control mode; and thirdly, it reduces the operating cost of distributed energy, addressing the high operating cost of distributed energy.

[0068] This invention effectively solves the problem of excessive cloud computing power caused by the optimized operation of massive distributed energy resources by proactively optimizing the lower layer to perceive the needs of the upper layer, thus achieving global collaborative optimization of distributed energy resources. Furthermore, the invention has a reliable design principle, a clear and simple structure, and very broad application prospects. Therefore, compared with existing technologies, this invention has outstanding substantive features and significant progress, and its beneficial effects are obvious. Attached Figure Description

[0069] Figure 1 This is a flowchart of the present invention;

[0070] Figure 2This is a schematic diagram of the cloud-edge-device four-layer architecture of the distributed energy system of the present invention.

[0071] Figure 3 This is a flowchart of the power regulation command optimization process for each layer of distributed energy based on the white whale optimization algorithm of the present invention.

[0072] Figure 4 This represents the optimized operating cost results for all distributed energy clusters in this invention.

[0073] Figure 5 This is the response result of the cloud layer in this invention;

[0074] Figure 6 The response results of the distributed energy cluster of this invention;

[0075] Figure 7 This is the response result of the distributed energy zone of the present invention;

[0076] Figure 8 The response results of individual distributed energy sources within the distributed energy zone 1 of this invention;

[0077] Figure 9 The response results of individual distributed energy sources within the distributed energy zone 2 of this invention;

[0078] Figure 10 The response results of individual distributed energy sources within the distributed energy zone 3 of this invention;

[0079] Figure 11 The response results of individual distributed energy sources within the distributed energy zone 4 of this invention;

[0080] Figure 12 The response results of individual distributed energy sources within the distributed energy zone 5 of this invention. Detailed Implementation

[0081] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention.

[0082] This invention proposes a distributed energy collaborative optimization method based on a cloud-edge-device architecture, with appended... Figure 1 The flowchart of this invention includes the following detailed steps.

[0083] S1. Define a four-layer architecture: cloud, edge, endpoint, and endpoint:

[0084] The distributed energy cloud-edge-device four-layer architecture defined in this invention is shown in the appendix. Figure 2 As shown, where:

[0085] The cloud layer is used to receive the demands of the control center and the electricity market, and contains one or more distributed energy aggregation clusters beneath it;

[0086] The edge layer refers to the distributed energy aggregation group, which is used to sense cloud demand, optimize power adjustment commands between distributed energy groups, and complete mutual assistance between distributed energy groups to meet cloud demand. It contains one or more distributed energy aggregation zones.

[0087] The edge layer refers to the distributed energy aggregation area, which is used to sense the edge layer demand, optimize the power adjustment command of the distributed energy interval, and complete the mutual assistance of the distributed energy interval for the edge layer demand. It contains one or more distributed energy individuals.

[0088] The term "along layer" refers to distributed energy entities such as distributed photovoltaics, distributed wind power, battery energy storage systems, and electric vehicles. It is used to sense end-layer demand, optimize power regulation commands between distributed energy entities, and complete mutual support between distributed energy entities to meet end-layer demand.

[0089] S2. Information Transmission Stage: Information is reported upwards and circulated among peers; the specific steps in this stage are as follows:

[0090] S21. Distributed energy individuals along the layer report predicted new energy output, predicted uncontrollable load, battery energy storage system output limit, and electric vehicle output limit to the distributed energy zone at the end layer, and send copies to other distributed energy individuals in the same distributed energy zone;

[0091] S22. The end-layer distributed energy zone aggregates the predicted new energy output, predicted uncontrollable load, battery energy storage system output limit, and electric vehicle output limit information of the distributed energy individuals within the zone to form a first-level aggregation result. The first-level aggregation result is then reported to the edge-layer distributed energy group and copied to other distributed energy zones within the same distributed energy group.

[0092] S23. The edge layer distributed energy group re-aggregates the aggregation results of the distributed energy regions within the group to form a secondary aggregation result, and sends the secondary aggregation result to other distributed energy groups.

[0093] S3. Requirements Optimization Phase: Requirements are decomposed downwards, and peer-to-peer collaborative optimization is performed. The specific steps in this phase are as follows:

[0094] S31. Cloud layer acquisition of control center and electricity market demand;

[0095] S32. Each distributed energy group in the edge layer senses the cloud layer demand and, in conjunction with the secondary aggregation results of each distributed energy group, completes the optimization of the power regulation command of each distributed energy group, thereby realizing the proactive collaborative optimization of the power regulation command of the edge layer.

[0096] The objective function and constraints for the proactive collaborative optimization of power regulation commands for each distributed energy cluster in step S32 are as follows:

[0097] Objective function: The objective is to minimize the operating cost of all distributed energy clusters. The calculation formula is as follows:

[0098] (16)

[0099] In the formula: and Let be the electricity sales revenue and electricity purchase cost of the m-th distributed energy group at time t, respectively. The operating cost of the battery storage system and electric vehicle for the m-th distributed energy cluster at time t is calculated as follows:

[0100] (17)

[0101] (18)

[0102] (19)

[0103] In the formula: and Let be the wind power and photovoltaic power of the m-th distributed energy cluster at time t, respectively. and Let be the charging and discharging power of the battery storage system and the electric vehicle of the m-th distributed energy cluster at time t, respectively, with charging represented by positive values ​​and discharging by negative values. and Let be the battery storage system capacity and electric vehicle capacity of the m-th distributed energy cluster, respectively. Let be the load power of the m-th distributed energy group at time t. and These are the electricity sales price and the electricity purchase price, respectively. and These are the unit costs of battery energy storage systems and electric vehicle operation, respectively. For scheduling intervals;

[0104] Constraints:

[0105] (20)

[0106] and These represent the maximum charging and discharging power of the battery storage system in the m-th distributed energy cluster and the maximum charging and discharging power of the electric vehicle, respectively. Let t be the cloud's required power. Let be the interaction power between the m-th distributed energy group and other distributed energy groups at time t. The external input is positive and the external output is negative.

[0107] In step S32, each distributed energy cluster in the edge layer is configured with the same smart terminal and implanted with the Whale Optimization Algorithm, which has the same parameters. This algorithm is used to actively and collaboratively optimize the power regulation commands of each distributed energy cluster in the edge layer as described in step S32. The optimization process of the Whale Optimization Algorithm is shown in the attached figure. Figure 3 As shown, the detailed optimization steps are as follows:

[0108] 1) Define the parameters of the beluga optimization algorithm

[0109] Set the population size Z and the maximum number of iterations n for the beluga optimization algorithm. max ;

[0110] 2) Initialization

[0111] (twenty one)

[0112] In the formula, Let i be the position of the i-th beluga whale in the j-th dimension. and For finding the optimal upper and lower bounds in the j-th dimension, These are random numbers that follow a uniform distribution in the range [0,1].

[0113] 3) Determine the optimization process

[0114] The beluga optimization algorithm is based on a balance factor. To determine the optimization process of an individual beluga whale, when At that time, the beluga whale was in the overall exploration phase, when At that time, the beluga whale was in a partial development stage; balancing factor The calculation formula is as follows:

[0115] (twenty two)

[0116] In the formula: The number is a random number between (0, 1), and n is the current iteration number;

[0117] 4) Global Exploration Process

[0118] The method for updating the location of individual beluga whales during global exploration is as follows:

[0119] (twenty three)

[0120] In the formula: For the first A beluga whale in The updated position in 3D space; A randomly selected integer; For the first A beluga whale in Position in 3D space; For the r-th randomly selected beluga whale in The position in dimensional space (r is a random integer between 1 and N); and It is a random number, and its value range is 100%. .

[0121] 5) Local Development Process

[0122] The method for updating the location of individual beluga whales during partial development is as follows:

[0123] (twenty four)

[0124] In the formula: For the first The updated location of the beluga whale; and It is a random number, and its value range is 100%. ; For all beluga whale individuals in the first The global optimal solution at the next iteration; and The first A beluga whale and a randomly selected first The first beluga whale The position at the next iteration; and These are the random jump strength and the Levy flight function, which measure the Levy flight force, respectively, and their calculation formulas are as follows:

[0125] (25)

[0126] In the formula: and Random numbers that conform to a normal distribution; Let be a constant, and take it as 1.5.

[0127] 6) The process of the beluga whale falling

[0128] The beluga whale's position during the fall was calculated as follows:

[0129] (26)

[0130] In the formula: , and It is a random number, and its value range is 100%. ; The formula for calculating the step length of a beluga whale when it falls is as follows:

[0131] (27)

[0132] In the formula: , These represent the upper and lower boundaries of the optimization variable, respectively. The step factor, whose value is related to the probability of fall behavior and population size, is calculated using the following formula:

[0133] (28)

[0134] In the formula: The probability of a beluga whale falling to its death is expressed by the following formula:

[0135] (29)

[0136] S33. Each distributed energy zone in the end layer senses the power regulation command of the edge layer, and combines the first-level aggregation results of each distributed energy zone to complete the optimization of the power regulation command of each distributed energy zone, thereby realizing the proactive collaborative optimization of the power regulation command of the end layer.

[0137] The objective function and constraints for the proactive collaborative optimization of power regulation commands for each distributed energy zone in step S33 are as follows:

[0138] Objective function: The objective is to minimize the operating cost of all distributed energy zones. The calculation formula is as follows:

[0139] (30)

[0140] In the formula: and Let be the electricity sales revenue and electricity purchase cost of the h-th distributed energy zone at time t, respectively. The calculation method for the operating cost of the battery energy storage system and electric vehicle in the h-th distributed energy zone at time t is as follows:

[0141] (31)

[0142] (32)

[0143] (33)

[0144] In the formula: and These represent the wind power and photovoltaic power of the h-th distributed energy zone at time t, respectively. and Let be the charging and discharging power of the battery storage system and the electric vehicle in the h-th distributed energy zone at time t, respectively. Charging is represented by positive values, and discharging by negative values. and These represent the battery storage system capacity and electric vehicle capacity of the h-th distributed energy zone, respectively. Let be the load power of the h-th distributed energy zone at time t. and These are the electricity sales price and the electricity purchase price, respectively. and These are the unit costs of battery energy storage systems and electric vehicle operation, respectively. For scheduling intervals;

[0145] Constraints:

[0146] (34)

[0147] In the formula: and These represent the maximum charging and discharging power of the battery storage system and the maximum charging and discharging power of the electric vehicle in the h-th distributed energy zone, respectively. Let be the edge layer power requirement at time t. Let be the interaction power between the h-th distributed energy zone and other distributed energy zones at time t. The external input is positive and the external output is negative.

[0148] S34. Each distributed energy individual along the layer senses the power regulation command along the layer, and combines the predicted new energy output, predicted uncontrollable load, battery energy storage system output limit, and electric vehicle output limit information of each distributed energy individual to complete the optimization of the power regulation command of each distributed energy individual, thereby realizing the proactive collaborative optimization of the power regulation command along the layer.

[0149] The objective function and constraints for the proactive collaborative optimization of power regulation commands for each distributed energy source along the layer in step S34 are as follows:

[0150] Objective function: The goal is to minimize the operating cost of all individual distributed energy resources. The formula for calculating the operating cost of a distributed energy zone is as follows:

[0151] (35)

[0152] In the formula: and Let be the electricity sales revenue and electricity purchase cost of a certain distributed energy zone at time t, respectively. Let t be the operating cost of the battery storage system and electric vehicle in a certain distributed energy zone at time t. The calculation method is as follows:

[0153] (36)

[0154] (37)

[0155] (38)

[0156] In the formula: and Let be the wind power and photovoltaic power of a certain distributed energy zone at time t, respectively. and Let be the charging and discharging power of the battery storage system and the electric vehicle in a distributed energy zone at time t, respectively. Charging is represented by positive values, and discharging by negative values. and These represent the battery storage system capacity and electric vehicle capacity of a specific distributed energy zone. Let t be the load power of a certain distributed energy zone at time t. and These are the electricity sales price and the electricity purchase price, respectively. and These are the unit costs of battery energy storage systems and electric vehicle operation, respectively. For scheduling intervals;

[0157] Constraints:

[0158] (39)

[0159] In the formula: and These represent the maximum charging and discharging power of a battery storage system and the maximum charging and discharging power of an electric vehicle in a specific distributed energy zone. Let t be the power required by the end layer at time t.

[0160] S4. Energy Flow Stage: Energy converges upwards and interacts with each other at the same level; the specific steps of this stage are as follows:

[0161] S41. Distributed energy individuals along the layer respond to their optimized power regulation commands along the layer and sense the response results of other distributed energy individuals in the same distributed energy area in real time.

[0162] S42. The end-layer distributed energy zone summarizes the response results of the distributed energy individuals within the zone to form a first-level response result, and then communicates the first-level response results to other distributed energy zones within the same distributed energy group.

[0163] S43. The edge layer distributed energy group summarizes the updated first-level response results of the distributed energy areas within the group to form second-level response results, and then exchanges the second-level response results to send the second-level response results to other distributed energy groups;

[0164] S44. The cloud layer summarizes the secondary response results of all distributed energy groups, completes the demand regulation of the control center and the electricity market, and realizes the overall coordinated optimization of distributed energy.

[0165] To further understand this invention and verify the effectiveness of the proposed distributed energy collaborative optimization method based on a cloud-edge-device architecture, simulations were conducted using distributed energy resources within the jurisdiction of a certain city. The relevant parameters of the distributed energy resources within the jurisdiction are shown in Table 1 below.

[0166] Table 1 Relevant parameters of distributed energy within the jurisdiction

[0167]

[0168] In the simulation, the unit price of electricity sales The price is 0.372 yuan / kWh; the electricity purchase price per unit. Based on peak-valley electricity pricing, the price is 0.249 yuan / kWh from 0:00 to 7:59, 0.682 yuan / kWh from 8:00 to 9:59, 13:00 to 17:59, and 22:00 to 23:59, and 1.321 yuan / kWh from 10:00 to 12:59 and 18:00 to 21:59; the unit price of operating costs for battery energy storage systems and electric vehicles... and Each deep charge-discharge cycle costs 400 yuan. For the White Whale optimization algorithm embedded in each smart terminal, its parameters are uniformly set to a maximum of 300 iterations and a number of optimized individuals of 50.

[0169] The optimization results for the operating costs of all distributed energy clusters are attached. Figure 4 As shown, the lowest cost of 116,840 yuan is reached after 139 iterations. At this time, the operating cost of distributed energy group 1 is 57,580 yuan, the operating cost of distributed energy group 2 is 12,990 yuan, and the operating cost of distributed energy group 3 is 46,270 yuan.

[0170] The response results of all distributed energy sources to the cloud are attached. Figure 5 As shown in the attached figure, the response results of the distributed energy cluster are as follows. Figure 6 As shown in the attached figure, the response results of the distributed energy zone are as follows. Figure 7 As shown in the attached figure, the individual response results of distributed energy resources within distributed energy zone 1 are as follows. Figure 8 As shown in the attached figure, the individual response results of distributed energy resources within distributed energy zone 2 are as follows. Figure 9 As shown in the attached figure, the individual response results of distributed energy resources within distributed energy zone 3 are as follows. Figure 10 As shown in the attached figure, the individual response results of distributed energy resources within distributed energy zone 4 are as follows. Figure 11As shown in the attached figure, the individual response results of distributed energy resources within distributed energy zone 5 are as follows. Figure 12 As shown. (From the appendix) Figure 5 It can be seen that the response error of the distributed energy source under the method proposed in this invention is 4.5441kW, which verifies that the present invention can effectively reduce the response error of the distributed energy source. In addition, the response time of the distributed energy source under the method proposed in this invention is 12.73s, indicating that the present invention can also effectively reduce the response time of the distributed energy source.

[0171] To further highlight the advantages of this invention, a distributed energy collaborative optimization method based on a cloud-edge-device architecture is compared with three other distributed energy collaborative optimization methods. Method 1 is a distributed energy collaborative optimization method based on blockchain technology, Method 2 is a distributed energy collaborative optimization method based on deep reinforcement learning, and Method 3 is a distributed energy multi-agent collaborative control method based on an event-triggered mechanism. The response error and response time of the distributed energy under different methods are shown in Table 2 below.

[0172] Table 2 Comparison of Distributed Energy Response Results under Different Methods

[0173]

[0174] The above comparison shows that the method designed in this invention has the lowest distributed energy response error, the fastest response time, and the lowest operating cost. This verifies the superiority of this invention in terms of distributed energy response accuracy, solving the distributed energy response delay problem, and reducing the operating cost of distributed energy.

Claims

1. A distributed energy collaborative optimization method based on cloud-edge-device architecture, characterized in that: Establishing a regulatory framework for distributed energy resources to actively participate in demand response and to perform collaborative optimization of distributed energy resources, thereby improving the response accuracy, reducing response error, shortening response time, and lowering operating costs, and achieving full-domain collaborative optimization of distributed energy resources, includes the following steps: S1. Define a four-layer architecture: cloud, edge, and endpoint. In step S1, the definitions and functions of the four-layer architecture along the cloud-edge-device are as follows: The cloud layer is used to receive the demands of the control center and the electricity market, and contains one or more distributed energy aggregation clusters beneath it; The edge layer refers to the distributed energy aggregation group, which is used to sense cloud demand, optimize power adjustment commands between distributed energy groups, and complete mutual assistance between distributed energy groups to meet cloud demand. It contains one or more distributed energy aggregation zones. The edge layer refers to the distributed energy aggregation area, which is used to sense the edge layer demand, optimize the power adjustment command of the distributed energy interval, and complete the mutual assistance of the distributed energy interval for the edge layer demand. It contains one or more distributed energy individuals. "Along the layer" refers to the use of distributed energy individuals to sense end-layer demand, optimize power regulation commands between distributed energy individuals, and complete mutual assistance between distributed energy individuals to meet end-layer demand. S2. Information transmission process: Information is reported upwards and circulated among peers; The specific steps of the information transmission stage in step S2 are as follows: S21. Distributed energy individuals along the layer report predicted new energy output, predicted uncontrollable load, battery energy storage system output limit, and electric vehicle output limit to the distributed energy zone at the end layer, and send copies to other distributed energy individuals in the same distributed energy zone; S22. The end-layer distributed energy zone aggregates the predicted new energy output, predicted uncontrollable load, battery energy storage system output limit, and electric vehicle output limit information of the distributed energy individuals within the zone to form a first-level aggregation result. The first-level aggregation result is then reported to the edge-layer distributed energy group and copied to other distributed energy zones within the same distributed energy group. S23. The edge layer distributed energy group re-aggregates the aggregation results of the distributed energy regions within the group to form a secondary aggregation result, and sends the secondary aggregation result to other distributed energy groups; S3. Requirements Optimization Phase: Decompose requirements downwards and optimize collaboratively among peers; S4. Energy flow link: Energy aggregates upwards and interacts and communicates with each other at the same level.

2. The distributed energy collaborative optimization method based on cloud-edge-device architecture according to claim 1, characterized in that: The specific steps of the demand optimization step in step S3 are as follows: S31. Cloud layer acquisition of control center and electricity market demand; S32. Each distributed energy group in the edge layer senses the cloud layer demand and, in conjunction with the secondary aggregation results of each distributed energy group, completes the optimization of the power regulation command of each distributed energy group, thereby realizing the proactive collaborative optimization of the power regulation command of the edge layer. S33. Each distributed energy zone in the end layer senses the power regulation command of the edge layer, and combines the first-level aggregation results of each distributed energy zone to complete the optimization of the power regulation command of each distributed energy zone, thereby realizing the proactive collaborative optimization of the power regulation command of the end layer. S34. Each distributed energy individual along the layer senses the power regulation command along the layer, and combines the predicted new energy output, predicted uncontrollable load, battery energy storage system output limit, and electric vehicle output limit information of each distributed energy individual to complete the optimization of the power regulation command of each distributed energy individual, thereby realizing the proactive collaborative optimization of the power regulation command along the layer.

3. A distributed energy collaborative optimization method based on a cloud-edge-end architecture as described in claim 1 or 2, characterized in that: The specific steps of the energy flow process in step S4 are as follows: S41. Distributed energy individuals along the layer respond to their optimized power regulation commands along the layer and sense the response results of other distributed energy individuals in the same distributed energy area in real time. S42. The end-layer distributed energy zone summarizes the response results of the distributed energy individuals within the zone to form a first-level response result, and then communicates the first-level response results to other distributed energy zones within the same distributed energy group. S43. The edge layer distributed energy group summarizes the updated first-level response results of the distributed energy areas within the group to form second-level response results, and then exchanges the second-level response results to send the second-level response results to other distributed energy groups; S44. The cloud layer summarizes the secondary response results of all distributed energy groups, completes the demand regulation of the control center and the electricity market, and realizes the overall coordinated optimization of distributed energy.

4. The distributed energy collaborative optimization method based on cloud-edge-end architecture according to claim 2, characterized in that: The objective function and constraints for the proactive collaborative optimization of power regulation commands for each distributed energy group in step S32 are as follows: Objective function: The objective is to minimize the operating cost of all distributed energy clusters. The calculation formula is as follows: (1) In the formula: and Let be the electricity sales revenue and electricity purchase cost of the m-th distributed energy group at time t, respectively. The operating cost of the battery storage system and electric vehicle for the m-th distributed energy cluster at time t is calculated as follows: (2) (3) (4) In the formula: and Let be the wind power and photovoltaic power of the m-th distributed energy cluster at time t, respectively. and Let be the charging and discharging power of the battery storage system and the electric vehicle of the m-th distributed energy cluster at time t, respectively, with charging represented by positive values ​​and discharging by negative values. and Let be the battery storage system capacity and electric vehicle capacity of the m-th distributed energy cluster, respectively. Let be the load power of the m-th distributed energy group at time t. and These are the electricity sales price and the electricity purchase price, respectively. and These are the unit costs of battery energy storage systems and electric vehicle operation, respectively. For scheduling intervals; Constraints: (5) and These represent the maximum charging and discharging power of the battery storage system in the m-th distributed energy cluster and the maximum charging and discharging power of the electric vehicle, respectively. Let t be the cloud's required power. Let be the interaction power between the m-th distributed energy group and other distributed energy groups at time t. The external input is positive and the external output is negative. The objective function and constraints for the proactive collaborative optimization of power regulation commands for each distributed energy zone in step S33 are as follows: Objective function: The objective is to minimize the operating cost of all distributed energy zones. The calculation formula is as follows: (6) In the formula: and Let be the electricity sales revenue and electricity purchase cost of the h-th distributed energy zone at time t, respectively. The calculation method for the operating cost of the battery energy storage system and electric vehicle in the h-th distributed energy zone at time t is as follows: (7) (8) (9) In the formula: and These represent the wind power and photovoltaic power of the h-th distributed energy zone at time t, respectively. and Let be the charging and discharging power of the battery storage system and the electric vehicle in the h-th distributed energy zone at time t, respectively. Charging is represented by positive values, and discharging by negative values. and These represent the battery storage system capacity and electric vehicle capacity of the h-th distributed energy zone, respectively. Let be the load power of the h-th distributed energy zone at time t. and These are the electricity sales price and the electricity purchase price, respectively. and These are the unit costs of battery energy storage systems and electric vehicle operation, respectively. For scheduling intervals; Constraints: (10) In the formula: and These represent the maximum charging and discharging power of the battery storage system and the maximum charging and discharging power of the electric vehicle in the h-th distributed energy zone, respectively. Let be the edge layer power requirement at time t. Let be the interaction power between the h-th distributed energy zone and other distributed energy zones at time t. The external input is positive and the external output is negative. The objective function and constraints for the proactive collaborative optimization of power regulation commands for each distributed energy individual along the layer in step S34 are as follows: Objective function: The goal is to minimize the operating cost of all individual distributed energy resources. The formula for calculating the operating cost of a distributed energy zone is as follows: (11) In the formula: and Let be the electricity sales revenue and electricity purchase cost of a certain distributed energy zone at time t, respectively. Let t be the operating cost of the battery storage system and electric vehicle in a certain distributed energy zone at time t. The calculation method is as follows: (12) (13) (14) In the formula: and Let be the wind power and photovoltaic power of a certain distributed energy zone at time t, respectively. and Let be the charging and discharging power of the battery storage system and the electric vehicle in a distributed energy zone at time t, respectively. Charging is represented by positive values, and discharging by negative values. and These represent the battery storage system capacity and electric vehicle capacity of a distributed energy cluster, respectively. Let t be the load power of a certain distributed energy zone at time t. and These are the electricity sales price and the electricity purchase price, respectively. and These are the unit costs of battery energy storage systems and electric vehicle operation, respectively. For scheduling intervals; Constraints: (15) In the formula: and These represent the maximum charging and discharging power of a battery storage system and the maximum charging and discharging power of an electric vehicle in a specific distributed energy zone. Let t be the power required by the end layer at time t.

5. The distributed energy collaborative optimization method based on cloud-edge-device architecture according to claim 2, characterized in that: The cloud-edge-end four-layer architecture in step S1: Each distributed energy cluster in the edge layer is equipped with the same smart terminal and has the same White Whale optimization algorithm with all parameters embedded in it to complete the active collaborative optimization of the power adjustment command of each distributed energy cluster in the edge layer as described in step S32. Each distributed energy zone in the end layer is equipped with the same smart terminal and has the same White Whale optimization algorithm with all parameters embedded in it to complete the active collaborative optimization of the power regulation command of each distributed energy zone in the end layer as described in step S33. Each distributed energy individual along the layer is configured with the same smart terminal and implanted with the same White Whale optimization algorithm with all parameters, which is used to complete the active collaborative optimization of the power adjustment command of each distributed energy individual along the layer as described in step S34.

Images