Multi-stage distributed energy aggregation scheduling method based on cloud-edge collaboration

Abstract

The invention discloses a multilevel distributed energy aggregation scheduling method based on cloud edge collaboration. The method comprises the following steps: carrying out information collection through an edge end; the cloud calculates the minimum operation cost of the power distribution network system; the load requirement is met in the operation range of the aggregator, and the power distribution network operates safely; calculating an optimal operation scheme of all load aggregators of the power distribution network in a certain region in each time period; and combining the operation schemes of each time period so as to form an integral optimal scheme. According to the technical scheme, the operation cost of the power distribution network is taken as an objective function, the safe operation of the power distribution network and the output constraint of each distributed energy source are taken as constraint conditions, and the objective optimization of each time period is carried out, so that the distributed energy source aggregation scheduling is completed, and multiple distributed new energy source loads in one region or among regions are aggregated; and unified optimization scheduling is realized.

Description

Technical Field

[0001] The present invention relates to the field of new energy and energy storage technology, and in particular to a multi-level distributed energy aggregation scheduling method based on cloud-edge collaboration. Background Art

[0002] In the process of continuous construction of new power systems, more and more distributed new energy resources are integrated and connected to the grid on the power generation side and the user side, such as small wind power generation devices, rooftop solar photovoltaic (PV) systems, gas combustion micro turbines, electric vehicles (EV), battery energy storage systems (BESS), etc., leading to the transformation of the operation mode of the power grid and the market at the same time. In the new power system with new energy as the main body, traditional technical means and production models can no longer adapt to the operation requirements of the power grid with a high proportion of new energy, which brings more challenges to the stability, continuity and adjustability of the power grid operation.

[0003] At present, distributed renewable energy is not allowed to directly participate in traditional power grid dispatching and trading due to its small capacity and behind-the-meter characteristics, so load aggregators of distribution systems have emerged. Distributed renewable energy in a specific area is gathered together by aggregators to form a unified operation mode. Load aggregators conduct energy transactions or provide ancillary services in the spot market.

[0004] Data show that the output of distributed renewable energy is random, intermittent, and generally small in scale. It is geographically dispersed and difficult to coordinate uniformly. At the same time, the aggregation characteristics of distributed renewable energy in the region (such as output upper and lower limits, ramp rate, operating costs, etc.) are highly uncertain. Therefore, the load aggregation dispatch center increases the difficulty of optimizing the dispatch of distributed renewable energy.

[0005] Chinese patent document CN107671414A discloses a "dispatching method for distributed energy storage systems to improve the consumption of new energy". It includes a four-layer structure of unit layer, unit control layer, regional coordination layer, and main network centralized control layer. The main network centralized control layer is connected to the regional coordination layer through a regional controller. The regional coordination layer realizes cross-regional electricity trading through inter-regional interconnection lines. Through the hierarchical optimization of the regional coordination layer and the main network centralized control layer, multi-level regulation of distributed power sources and distributed energy storage systems is realized, and an energy dispatching mechanism for the overall coordinated consumption of the power grid and mutual cooperation is constructed. The above technical solution lacks optimization for the goals of each time period. Summary of the invention

[0006] The present invention mainly solves the technical problem that the original technical solution lacks optimization of targets for each time period, and provides a multi-level distributed energy aggregation scheduling method based on cloud-edge collaboration. By taking the distribution network operation cost as the objective function and the safe operation of the distribution network and the output constraints of each distributed energy source as constraints, the target of each time period is optimized, thereby completing distributed energy aggregation scheduling, realizing the aggregation of multiple distributed new energy loads within a region or between regions, and realizing unified optimization scheduling.

[0007] The above technical problem of the present invention is mainly solved by the following technical solution: The present invention comprises the following steps: S1 collects information through the edge end; S2 cloud computing distribution network system minimum operating cost; S3 ensures that load demand is met within the aggregator’s operating area and that the distribution network operates safely; S4 calculates the optimal operation plan for all load aggregators in a certain regional distribution network at each time period; S5 combines the operation plans for each period to form the best overall plan.

[0008] By taking the distribution network operation cost as the objective function and the safe operation of the distribution network and the output constraints of each distributed energy source as the constraints, the target optimization for each time period is carried out to complete the distributed energy aggregation scheduling, realize the aggregation of multiple distributed new energy loads within a region or between regions, and realize unified optimization scheduling.

[0009] Preferably, the edge end includes distributed energy within the operating range of each load aggregator in the distribution network, and the edge end collects information while performing local control. The distributed new energy is mainly used as the edge end, and its main responsibility is to collect data, perform local control, etc. The cloud computing center mainly calculates the collected data, calculates the optimal scheduling strategy, and sends the strategy to each distributed energy aggregator.

[0010] Preferably, the information collected includes the output level, output ramping and load that can be adjusted of photovoltaic, wind and gas generators, as well as the working status of electric vehicle charging stations and energy storage stations, as well as the basic load and weather conditions within the range. The data collected every day can be used as historical data to predict model training data.

[0011] Preferably, the step S2 specifically includes calculating the operating cost combination of multiple load aggregators in the distribution network under safe operation conditions, the operating cost of each load aggregator is composed of distributed energy components and grid costs within its operating scope, the cost of clean energy is taken as 0, and the costs include gas-fired generators, grid power purchases and participation in demand response costs.

[0012] Preferably, the objective function formula of the cost is expressed as: Where T is the set of time periods in a day, one hour is a time period, and there are 24 time periods in a day. C(*) is the cost function of different types of distributed energy. P gas (t), P bg (t), P dr (t) represents the total load of gas generators, power purchase from the grid, and participation in demand response in time period t, and θ1, θ2, and θ3 are weight coefficients. The cost of clean energy is regarded as 0, so the cost mainly includes the cost of gas generators, power purchase from the grid, and participation in demand response.

[0013] Preferably, the step S3 specifically includes that each aggregator must meet the load demand within its operating range and ensure the safe operation of the distribution network flow, and the constraint conditions are: If there is no distributed energy or load characteristics within the load aggregator's operating range, this item is zero, P bload , P c , P v , P w They are the distribution network usage load, energy storage system charging and discharging load, distributed photovoltaic, and wind power load. α is -1, 0, and 1, representing the energy storage system discharging, no, and charging states, respectively.

[0014] Preferably, the step S4 specifically includes, the cloud computing the best operation plan for all load aggregators in a regional distribution network in each time period according to the electrical characteristic constraints of various types of distributed energy sources.

[0015] Preferably, the electrical characteristic constraints of each type of distributed energy are as follows: p gas，min ≤p gas (t)≤p gas，max p der (t+1)-p der (t)|≤Δp der 0≤p v (t)≤p v，max 0≤p w (t)≤p w，max 0≤p dr (t)≤p dr,max Among them, P gas，min and P gas，max are the lower and upper limits of the gas generator set output, ΔPder is the ramp rate of distributed energy units, der = {gas turbine, wind turbine, photovoltaic} =, P v，max , P w，max They are the maximum output limits of distributed photovoltaic and wind energy, P dr，max The maximum load value that can participate in demand response regulation.

[0016] Preferably, the step S5 combines the operation schemes of each time period to form an overall optimal scheme. The specific calculation method is to use the simplex method to calculate the optimal solution for each time period, i.e., each hour. This method is a conventional method, which can call the algorithm in the cvxopt algorithm package of Python, and is mainly divided into the following steps: 1) According to the target formula, determine the initial feasible basis and the initial basis feasible solution, and establish the initial simplex table. This is the starting point of the solution process. It is necessary to determine an initial basis feasible solution and construct the initial simplex table based on it.

[0017] 2) Optimality test. Check whether the current basic feasible solution is the optimal solution. If the test numbers of all non-basic variables are non-positive, the current solution is optimal; if there is a positive test number, it means that the current solution is not optimal.

[0018] 3) Transform from one basis feasible solution to another basis feasible solution with a larger objective function value. If the current solution is not optimal, an improved basis feasible solution needs to be found. This usually involves selecting an input basis variable and an output basis variable, and then performing a basis transformation.

[0019] 4) Update the simplex table. After completing the basis transformation, the simplex table needs to be updated, including recalculating the test numbers of the basis variables and non-basis variables.

[0020] 5) Repeat steps 2 and 3 until the optimal solution is found or the problem is determined to be unsolvable. This process may involve multiple iterations until all test numbers are non-positive or the problem is determined to be unbounded.

[0021] 6) Termination condition. When all test numbers are non-positive, it can be declared that the optimal solution to the problem has been found; if there is a positive test number, the problem may have no solution or an unbounded solution.

[0022] The beneficial effects of the present invention are as follows: by taking the distribution network operation cost as the objective function, and taking the safe operation of the distribution network and the output constraints of each distributed energy source as the constraints, the target optimization of each time period is performed, thereby completing the distributed energy aggregation scheduling, realizing the aggregation of multiple distributed new energy loads within a region or between regions, and realizing unified optimization scheduling. BRIEF DESCRIPTION OF THE DRAWINGS

[0023] Figure 1 It is a system overall architecture diagram of the present invention.

[0024] Figure 2 It is a flow chart of the present invention.

[0025] Figure 3 This is a typical daily photovoltaic output and large user load diagram of the present invention. DETAILED DESCRIPTION

[0026] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the technical solution of the present invention is further described in detail below through embodiments and in combination with the accompanying drawings. It should be understood that the specific implementation method described here is only an optimal embodiment of the present invention, which is only used to explain the present invention and does not limit the scope of protection of the present invention. All other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention.

[0027] Before discussing the exemplary embodiments in more detail, it should be mentioned that some exemplary embodiments are described as processes or methods depicted as flow charts. Although the flow charts describe the operations (or steps) as sequential processes, many of the operations (or steps) therein can be implemented in parallel, concurrently, or simultaneously. In addition, the order of the operations can be rearranged. The process can be terminated when its operation is completed, but can also have additional steps not included in the drawings; the process can correspond to a method, function, procedure, subroutine, subprogram, etc.

[0028] The technical solution of the present invention is further specifically described below through embodiments and in conjunction with the accompanying drawings.

[0029] Embodiment: This embodiment provides a multi-level distributed energy aggregation scheduling method based on cloud-edge collaboration, such as Figure 2 As shown, the following steps are included: S1 collects information through the edge end. The edge end includes distributed energy within the operating range of each load aggregator in the distribution network, and the edge end collects information while performing local control. Figure 1 As shown in the figure, distributed new energy is mainly used as the edge end, and its main responsibility is to collect data, perform local control, etc. The cloud computing center mainly calculates the collected data, calculates the optimal scheduling strategy, and sends the strategy to each distributed energy aggregator.

[0030] Information collected includes the output level, output ramping and load that can be adjusted for photovoltaic, wind and gas generators, as well as the working status of electric vehicle charging stations and energy storage stations, as well as the basic load and weather conditions within the range. The data collected every day can be used as historical data to predict model training data.

[0031] S2 cloud-based computing distributes the minimum operating cost of the network system, specifically, calculating the operating cost combination of multiple load aggregators in the distribution network under safe operation conditions. The operating cost of each load aggregator is composed of distributed energy components and grid costs within its operating scope. The cost of clean energy is taken as 0, and the costs include gas-fired generators, grid power purchases, and participation in demand response costs.

[0032] The objective function formula of cost is expressed as: Where T is the set of time periods in a day, one hour is a time period, and there are 24 time periods in a day. C(*) is the cost function of different types of distributed energy. P gas (t), P bg (t), P dr (t) represents the total load of gas generators, power purchase from the grid, and participation in demand response in time period t, and θ1, θ2, and θ3 are weight coefficients. The cost of clean energy is regarded as 0, so the cost mainly includes the cost of gas generators, power purchase from the grid, and participation in demand response.

[0033] S3 ensures that the load demand is met within the aggregator's operating range and that the distribution network operates safely. Specifically, each aggregator's operating range must meet the load demand and ensure the safe operation of the distribution network flow. The constraints are: If there is no distributed energy or load characteristics within the load aggregator's operating range, this item is zero, P bload , P c , P v , P w They are the distribution network usage load, energy storage system charging and discharging load, distributed photovoltaic, and wind power load. α is -1, 0, and 1, representing the energy storage system discharging, no, and charging states, respectively.

[0034] S4 calculates the optimal operation plan for all load aggregators in a certain area's distribution network at each time period. Specifically, the cloud calculates the optimal operation plan for all load aggregators in a certain area's distribution network at each time period based on the electrical characteristics constraints of various types of distributed energy.

[0035] The electrical characteristics constraints of each type of distributed energy are as follows: p gas，min ≤p gas (t)≤p gas，max p der (t+1)-p der (t)|≤Δp der 0≤p v (t)≤pv，max 0≤p w (t)≤p w，max 0≤p dr (t)≤p dr,max Among them, P gas，min and P gas，max are the lower and upper limits of the gas generator set output, ΔP der is the ramp rate of distributed energy units, der = {gas turbine, wind turbine, photovoltaic} =, P v，max , P w，max They are the maximum output limits of distributed photovoltaic and wind energy, P dr，max The maximum load value that can participate in demand response regulation.

[0036] S5 combines the operation plans of each period to form the best overall plan. The specific calculation method is to use the simplex method to calculate the optimal solution for each period, that is, each hour. This method is a conventional method that can call the algorithm in the cvxopt algorithm package of Python. It is mainly divided into the following steps: 1) According to the target formula, determine the initial feasible basis and the initial basis feasible solution, and establish the initial simplex table. This is the starting point of the solution process. It is necessary to determine an initial basis feasible solution and construct the initial simplex table based on it.

[0037] 2) Optimality test. Check whether the current basic feasible solution is the optimal solution. If the test numbers of all non-basic variables are non-positive, the current solution is optimal; if there is a positive test number, it means that the current solution is not optimal.

[0038] 3) Transform from one basis feasible solution to another basis feasible solution with a larger objective function value. If the current solution is not optimal, an improved basis feasible solution needs to be found. This usually involves selecting an input variable and an output variable, and then performing a basis transformation.

[0039] 4) Update the simplex table. After completing the basis transformation, the simplex table needs to be updated, including recalculating the test numbers of the basis variables and non-basis variables.

[0040] 5) Repeat steps 2 and 3 until the optimal solution is found or the problem is determined to be unsolvable. This process may involve multiple iterations until all test numbers are non-positive or the problem is determined to be unbounded.

[0041] 6) Termination condition. When all test numbers are non-positive, it can be declared that the optimal solution to the problem has been found; if there is a positive test number, the problem may have no solution or an unbounded solution.

[0042] By taking the distribution network operation cost as the objective function and the safe operation of the distribution network and the output constraints of each distributed energy source as the constraints, the target optimization for each time period is carried out to complete the distributed energy aggregation scheduling, realize the aggregation of multiple distributed new energy loads within a region or between regions, and realize unified optimization scheduling.

[0043] When working: By using distributed new energy as the edge end, the main responsibility is to collect data, local control, etc. The cloud computing center mainly calculates the collected data, calculates the optimal scheduling strategy, and sends the strategy to each distributed energy aggregator. The overall system block diagram is as follows Figure 1 .

[0044] The method mainly takes the distribution network operation cost as the objective function, the distribution network safe operation and the output constraints of each distributed energy source as the constraints, and optimizes the target of each time period to complete the distributed energy aggregation scheduling. The overall flow chart of the method is as follows: Figure 2 .

[0045] First, information will be collected through the edge, including the output level of distributed energy sources within the operating range of each load aggregator in the distribution network, including photovoltaic, wind power, gas generators, etc., output ramp conditions, load capacity that can be adjusted, working status of electric vehicle charging stations, working status of energy storage stations, as well as basic load conditions within the range and weather conditions. The data collected every day can be used as historical data to predict model training data.

[0046] The cloud calculates the minimum operating cost of the distribution network system. First, the operating cost combination of multiple load aggregators in the distribution network under safe operation is calculated. The operating cost of each load aggregator is composed of distributed energy components and grid costs within its operating range. The cost of clean energy is regarded as 0. Therefore, the cost mainly includes gas generators, grid power purchases, and demand response costs. The objective function formula is expressed as: Where T is a set of time periods in a day, one hour is a time period, and there are 24 time periods in a day. C(*) is the cost function of different types of distributed energy, which is usually fixed. P gas (t), P bg (t), P dr (t) represent the total load of gas-fired power generation units, power grid purchases, and demand response in time period t, θ1, θ2, and θ3 are weight coefficients, and the coefficients usually default to 1.

[0047] At the same time, each aggregator must meet the load demand within its operating range and ensure the safe operation of the distribution network. The constraints are: If there is no distributed energy or load characteristics within the operation range of some load aggregators, this item can be zero. bload , P c , P v , P w They are the distribution network usage load, energy storage system charging and discharging load, distributed photovoltaic, and wind power load. α can be -1, 0, and 1, representing the energy storage system discharging, no, and charging states, respectively.

[0048] The electrical characteristics constraints of each type of distributed energy are as follows: p gas,min ≤p gas (t)≤p gas,max (4) p der (t+1)-p der (t)|≤Δp der (5) 0≤p v (t)≤p u,max (6) 0≤p w (t)≤p w,max (7) O≤p dr (t)≤p dr,max (8) Among them, P gas，min and P gas，max are the lower and upper limits of the gas generator set output, ΔP der is the ramp rate of distributed energy units, der = {gas turbine, wind turbine, photovoltaic} =, P v，max , P w，max They are the maximum output limits of distributed photovoltaic and wind energy, P dr，max The maximum load value that can participate in demand response regulation.

[0049] The above linear optimization model is a conventional linear optimization calculation problem. The cloud calculates the optimal operation plan for all load aggregators in a certain area's distribution network in each period based on the above model, and then combines the operation plans for each period to form the overall optimal plan.

[0050] The specific calculation method is to use the simplex method to calculate the optimal solution for each time period, that is, each hour. This method is a conventional method that can call the algorithm in the cvxopt algorithm package of Python. It is mainly divided into the following steps: 1) According to the target formula, determine the initial feasible basis and the initial basis feasible solution, and establish the initial simplex table. This is the starting point of the solution process. It is necessary to determine an initial basis feasible solution and construct the initial simplex table based on it.

[0051] 2) Optimality test. Check whether the current basic feasible solution is the optimal solution. If the test numbers of all non-basic variables are non-positive, the current solution is optimal; if there is a positive test number, it means that the current solution is not optimal.

[0052] 3) Transform from one basis feasible solution to another basis feasible solution with a larger objective function value. If the current solution is not optimal, an improved basis feasible solution needs to be found. This usually involves selecting an input variable and an output variable, and then performing a basis transformation.

[0053] 4) Update the simplex table. After completing the basis transformation, the simplex table needs to be updated, including recalculating the test numbers of the basis variables and non-basis variables.

[0054] 5) Repeat steps 2 and 3 until the optimal solution is found or the problem is determined to be unsolvable. This process may involve multiple iterations until all test numbers are non-positive or the problem is determined to be unbounded.

[0055] 6) Termination condition. When all test numbers are non-positive, it can be declared that the optimal solution to the problem has been found; if there is a positive test number, the problem may have no solution or an unbounded solution.

[0056] Example First, information is collected through the edge, including the output level and output ramp of distributed energy sources such as photovoltaic, wind, and gas generators within the operating range of each load aggregator in the distribution network, which can participate in load regulation, the working status of electric vehicle charging stations, the working status of energy storage stations, and the basic load conditions within the range and weather conditions. The data collected every day can be used as historical data to predict model training data.

[0057] Furthermore, the cloud calculates the minimum operating cost of the distribution network system and the operating cost combination of multiple load aggregators in the distribution network under safe operation conditions. The operating cost of each load aggregator is composed of distributed energy components and grid costs within its operating scope. The cost of clean energy is regarded as 0, so the cost mainly includes gas-fired generators, power purchases from the grid, and participation in demand responses.

[0058] Further, according to the cost objective function, we take the classic daily data of a certain region as an example. There is a load aggregator for dispatch in this region, which includes 3 gas turbines with outputs of P and gas1 , P gas2 , P gas3 ; One photovoltaic, output is P v , and a large user load P bload .

[0059] Among them, the output range of gas units and photovoltaics is as follows: Serial number type Output range 1 Gas unit 1 [0,330] 2 Gas unit 2 [0,480] 3 Gas unit 3 [0,500] 4 Photovoltaic [0,160] The climbing rate of each generator set is as follows: Serial number type Climbing rate (kW / min) 1 Gas unit 1 50 2 Gas unit 2 75 3 Gas unit 3 50 4 Photovoltaic 20 According to the objective function, the cost calculation formula is: Among them, θ1, θ2, θ3 are all 2, θ4 is 1.8, C1, C2, C3, C4 are 0.5, 0.7, 0.8, 0.5 respectively.

[0060] Furthermore, the typical daily photovoltaic output and large user load are as follows: Figure 3 As shown, when the data is actually used, the predicted value is used, in which the difference between the user load and the photovoltaic power is mainly calculated through the gas turbine and the power grid purchase P bg To make up for it, the minimum cost of multi-energy aggregation scheduling is calculated and allocated according to each time period.

[0061] Furthermore, based on the results of calculating the linear optimal solution for each time period using the simplex method, the output of each unit is obtained as shown in the following table, where the power unit is kW. Timing pgas1 pgas2 pgas3 PBG 0 300 450 0 170 1 330 480 0 109 2 330 480 0 112 3 330 480 0 113 4 330 480 0 115 5 330 480 0 113 6 330 480 0 120 7 330 480 0 123 8 330 480 0 124 9 330 480 0 93 10 330 480 0 182 11 330 480 300 75 12 330 480 500 17 13 330 480 500 45 14 330 480 500 61 15 330 480 500 60 16 330 480 500 50 17 330 480 500 42 18 330 480 500 74 19 330 480 500 29 20 330 480 450 41 21 330 480 400 23 22 330 480 200 91 23 330 480 0 206

[0062] A person of ordinary skill in the art can understand that all or part of the steps in the various methods of the above embodiments can be completed by instructing related hardware through a program, and the program can be stored in a computer-readable storage medium, and the storage medium may include: a read-only memory (ROM), a random access memory (RAM), a disk or an optical disk, etc.

[0063] The specific embodiments described herein are merely examples of the spirit of the present invention. The above embodiments only express several implementation methods of the present invention, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the scope of the patent of the present invention. It should be pointed out that technicians in the technical field to which the present invention belongs can make various modifications or supplements to the described specific embodiments or replace them in a similar manner, but will not deviate from the spirit of the present invention or exceed the scope defined in the attached claims. For ordinary technicians in this field, multiple deformations and improvements can also be made without departing from the concept of the present invention. Therefore, the scope of protection of the patent of the present invention shall be based on the attached claims.

Claims

1. A multi-level distributed energy aggregation scheduling method based on cloud-edge collaboration, characterized in that: The following steps are involved: S1 collects information through the edge; S2 cloud computing distribution network system minimum operating cost; S3 ensures that load demand is met within the aggregator’s operating area and that the distribution network operates safely; S4 calculates the optimal operation plan for all load aggregators in a certain regional distribution network at each time period; S5 combines the operation plans for each period to form the best overall plan.

2. According to the multi-level distributed energy aggregation scheduling method based on cloud-edge collaboration according to claim 1, it is characterized in that: The edge end includes distributed energy within the operating range of each load aggregator in the distribution network, and the edge end performs local control while collecting information.

3. A multi-level distributed energy aggregation scheduling method based on cloud-edge collaboration according to claim 1 or 2, characterized in that: The information collected includes the output level, output ramping and load that can be adjusted of photovoltaic, wind and gas-fired generator sets, as well as the working status of electric vehicle charging stations and energy storage stations, as well as the basic load conditions and weather conditions within the range.

4. According to the multi-level distributed energy aggregation scheduling method based on cloud-edge collaboration according to claim 1, it is characterized in that: The step S2 specifically includes calculating the operating cost combination of multiple load aggregators in the distribution network under safe operation conditions, the operating cost of each load aggregator is composed of distributed energy components and grid costs within its operating scope, the cost of clean energy is taken as 0, and the costs include gas-fired generator sets, grid power purchases and participation in demand response costs.

5. According to the multi-level distributed energy aggregation scheduling method based on cloud-edge collaboration according to claim 1, it is characterized in that: The objective function formula of the cost is expressed as: Where T is the set of time periods in a day, one hour is a time period, and there are 24 time periods in a day. C(*) is the cost function of different types of distributed energy. P gas (t), P bg (t), P dr (t) represent the total load of gas-fired power generation units, power grid purchases, and demand response in time period t, and θ1, θ2, and θ3 are weight coefficients respectively.

6. The multi-level distributed energy aggregation scheduling method based on cloud-edge collaboration according to claim 1 is characterized in that: The step S3 specifically includes that each aggregator must meet the load demand within its operating scope and ensure the safe operation of the distribution network flow, and its constraint conditions are: If there is no distributed energy or load characteristics within the load aggregator's operating range, this item is zero, P bload , P c , P v , P w They are the distribution network usage load, energy storage system charging and discharging load, distributed photovoltaic, and wind power load. α is -1, 0, and 1, representing the energy storage system discharging, no, and charging states, respectively.

7. The multi-level distributed energy aggregation scheduling method based on cloud-edge collaboration according to claim 1 is characterized in that: The step S4 specifically includes, the cloud computing the best operation plan for all load aggregators in a regional distribution network in each time period according to the electrical characteristic constraints of various types of distributed energy sources.

8. The multi-level distributed energy aggregation scheduling method based on cloud-edge collaboration according to claim 7 is characterized in that: The electrical characteristic constraints of each type of distributed energy are as follows: p gas，min ≤p gas (t)≤p gas，max p der (t+1)-p der (t)|≤Δp der 0≤p v (t)≤p v，max 0≤p w (t)≤p w，max 0≤p dr (t)≤p dr,max Among them, P gas，min and P gas，max are the lower and upper limits of the gas generator set output, ΔP der is the ramp rate of distributed energy units, der = {gas turbine, wind turbine, photovoltaic} =, P v，max , P w，max They are the maximum output limits of distributed photovoltaic and wind energy, P dr，max The maximum load value that can participate in demand response regulation.

9. The multi-level distributed energy aggregation scheduling method based on cloud-edge collaboration according to claim 1 is characterized in that: The step S5 combines the operation plans of each time period to form an overall optimal plan. The specific calculation method is to use the simplex method to calculate the optimal solution for each time period, that is, each hour.

Images