A multi-level orderly scheduling method for large-scale charging load participating in power grid peak shaving

Abstract

The application discloses a kind of multi-level orderly scheduling methods of large-scale charging load participating in power grid peak shaving, including two stages of station load distribution and station optimization scheduling;Station load distribution refers to after obtaining the adjustable capacity evaluation of each charging station, according to the load adjustment instruction issued by dispatching center and marketing department to electric vehicle aggregator, according to the adjustable capacity evaluation of each charging station Proportion, the load adjustment amount of each charging station is distributed;Station optimization scheduling refers to constructing a multi-objective station optimization scheduling model with the maximum load adjustment completion rate and the maximum self-balancing coordination capability as the target, solving by using C-MOEA / D optimization algorithm, and obtaining the optimal scheduling scheme.The application can improve the load adjustment completion rate and the self-balancing coordination capability, and is of great significance to the stable operation of power grid and the safe and efficient operation of charging station.

Description

Technical Field

[0001] The present invention belongs to the technical field of power systems and relates to a multi-level orderly dispatching method for large-scale charging loads to participate in power grid peak regulation. Background Art

[0002] With the rapid development of the electric vehicle (EV) industry, EV charging demand has become a significant issue in power systems. However, with the rapid growth in the number of EVs, the concentrated charging load, especially during peak hours, can place significant pressure on the power system, impacting grid stability. The large-scale deployment of charging facilities and the uneven distribution of charging loads complicate EV charging scheduling, making traditional load scheduling methods incapable of meeting this challenge.

[0003] Most existing charging load scheduling methods rely on time allocation strategies, such as centralized charging during off-peak hours to avoid excessive peak loads. However, this simple time division often overlooks the volatility and uncertainty of EV charging demand. The varying demands of charging users lead to unstable load scheduling. Furthermore, existing methods generally employ centralized scheduling. While this allows for global optimization, it increases computational complexity as the number of EVs increases, and requires high real-time performance, making it difficult to adapt to the dynamic changes in the power grid and charging demand.

[0004] To address this issue, numerous studies have begun to introduce smart grid technologies, combining advanced technologies such as big data, the Internet of Things, and artificial intelligence to enable real-time scheduling of charging loads. Smart grids leverage information and communications technology (ICT) to achieve efficient coordination between the power grid and charging infrastructure. Based on big data and cloud computing, charging demand forecasts can be more accurate, providing data support for charging load scheduling. However, while the application of smart grids improves the real-time and flexibility of scheduling, problems such as high computational load and poor system stability still exist. In large-scale charging networks, achieving reasonable charging load distribution and grid load balancing remains a challenge.

[0005] In recent years, multi-level scheduling methods have been gradually proposed and applied to charging load scheduling. Traditional scheduling methods, mostly based on a single level, cannot effectively handle the complex relationships between different regions and charging stations. To address this issue, strategies have been proposed that divide charging load scheduling into multiple levels. For example, regional-level scheduling optimizes regional loads by analyzing charging demand and grid carrying capacity; station-level scheduling provides refined management of loads within charging stations; and user-level scheduling flexibly adjusts charging power and time based on the charging time and needs of electric vehicle users. This multi-level scheduling approach enables more precise global charging load distribution, mitigates grid load fluctuations, and improves the utilization efficiency of charging facilities. While multi-level scheduling methods have certain advantages, achieving coordination between different levels and optimizing scheduling instructions at each level remain research hotspots. Existing scheduling methods mostly focus on optimizing local loads and lack a global perspective, resulting in problems such as concentrated charging loads and local grid overloads in large-scale charging networks. Therefore, how to combine intelligent algorithms and real-time data to make dynamic adjustments and optimize the distribution of electric vehicle charging loads under a multi-level scheduling framework remains an urgent problem to be solved in the field of charging scheduling. Summary of the Invention

[0006] To address the load regulation challenges associated with the large-scale integration of electric vehicles into distribution networks, and to effectively control and optimize the scheduling of new energy generation and energy storage devices, as well as in-station electric vehicle loads, in line with existing load regulation requirements, this paper proposes a multi-level, orderly scheduling method for large-scale charging loads participating in grid peak regulation. This method aims to achieve dynamic scheduling of charging loads through a hierarchical and regional scheduling strategy, combined with real-time data and intelligent optimization algorithms. This method not only improves the quality of grid peak regulation and electric vehicle charging efficiency, but also effectively alleviates the pressure on the grid during peak charging periods, ensuring power system stability and efficient utilization of charging facilities.

[0007] The technical solution adopted by the present invention is: a multi-level orderly scheduling method for large-scale charging loads to participate in power grid peak regulation, including two stages: inter-station load distribution and intra-station optimization scheduling;

[0008] Inter-station load allocation means allocating the load adjustment amount of each charging station in proportion to the adjustable capacity assessment value of each charging station based on the load adjustment instructions actually issued by the dispatch center and marketing department to the electric vehicle aggregator after obtaining the adjustable capacity assessment value of each charging station;

[0009] According to the evaluation value of the adjustable capacity of the charging station, the load adjustment amount is proportionally distributed among the stations. The load distribution model among the stations based on the evaluation value of the adjustable capacity is shown in formula (1):

[0010] (1);

[0011] Where, Allocate the total actual load regulation for electric vehicle aggregators during period t; is the evaluation value of the adjustable capacity of the i-th charging station in period t; is the total adjustable capacity of charging stations in the scheduling area during period t; The actual load regulation amount after inter-station load distribution for the i-th charging station in period t;

[0012] When the adjustable capacity evaluation value of each charging station cannot be obtained before the dispatch center issues a dispatch instruction, the actual load reduction is distributed among the stations in proportion according to the load forecast value of each charging station during the dispatch period. The load distribution model among the stations based on the load forecast value is shown in formula (2):

[0013] (2);

[0014] Where, is the load forecast value of the i-th charging station in period t; is the sum of the load forecast values ​​of the charging stations in the dispatch area during period t;

[0015] In-station optimal scheduling refers to building a multi-objective in-station optimal scheduling model with the goal of maximizing the load regulation completion rate and the self-balancing coordination ability, and using the C-MOEA / D optimization algorithm to solve it and obtain the optimal scheduling plan.

[0016] Further preferably, the implementation process of inter-station load distribution is as follows:

[0017] The forecasting department reports the electric vehicle load forecast for each period of the next day to the dispatching center every day, and the dispatching center informs the marketing department of the load gap for the next day;

[0018] Based on the predicted value of electric vehicle load, it is used as the user electricity price response load of the charging station. By adjusting the charging electricity price, the adjustable capacity of the user electricity price response load is predicted. Then, based on the electricity price responsiveness, the load during the charging station scheduling period affected by the electricity price is obtained. Combined with the prediction of fast and slow charging loads, the fast charging load at that time is obtained. This is used as the fast charging power adjustment load, and the charging power of the fast charging pile is adjusted to obtain the fast charging power adjustment load adjustable capacity. The sum of the user electricity price response load and the fast charging power adjustment load adjustable capacity is the total adjustable capacity of the electric vehicle aggregator.

[0019] The EV aggregator reports its total adjustable capacity during the gap period to the marketing department. The marketing department then allocates load adjustment to the EV aggregator. The EV aggregator then determines the load adjustment for each charging station during the scheduling period based on the ratio of the adjustable capacity assessment values ​​of each charging station, providing a reference value for load adjustment during each period for station optimization.

[0020] Each charging station within the electric vehicle aggregator determines orderly charging measures based on the allocated load adjustment amount and reports the pre-implementation status to the dispatch center before the day;

[0021] The dispatch center agrees to execute the plan, and the marketing department executes the plan.

[0022] Further optimization, the objects of in-station optimization scheduling include ordinary charging stations and photovoltaic storage charging stations; the multi-objective in-station optimization scheduling model is:

[0023] Goal 1: Load regulation completion rate maximum;

[0024] (3);

[0025] Where, represents the actual load power of the i-th charging station in time period t after electricity price response and charging power adjustment; is the load forecast value of the i-th charging station in period t;

[0026] Goal 2: Self-balancing and coordination skills maximum;

[0027] (4);

[0028] Where, is the power purchased by the power grid during period t, is the charging power of the charging station during period t, The output of the photovoltaic power station during period t;

[0029] Ordinary charging stations only use load adjustment completion rate The goal is to maximize the solar storage charging station with self-balancing coordination capabilities. Maximum and self-balancing coordination ability The maximum is the target.

[0030] Further preferably, the constraints of the in-station optimization scheduling model of the photovoltaic storage charging station include photovoltaic power station output constraints, energy storage charging and discharging state constraints, power balance constraints, fast charging power constraints and electricity price constraints.

[0031] Further preferably, the constraints of the optimization scheduling model within the ordinary charging station include power balance constraints, fast charging power constraints and electricity price constraints.

[0032] Further optimization, the process of using C-MOEA / D optimization algorithm to solve is as follows:

[0033] Step 1. Initialization operation: Generate reference points on the hyperplane according to the variable dimension, and then form an initial population of N , laying the foundation for subsequent evolution;

[0034] Step 2. Generate offspring and merge populations: Perform crossover and mutation operations to form a progeny population ; Combine and Form a mixed population with a population size of 2N , expand the search scope;

[0035] Step 3. Fixed constraint screening: Based on the established fixed constraint conditions, the mixed population Screening is performed to eliminate individuals that do not meet the constraints and retain solutions that meet the fixed constraints;

[0036] Step 4. Fitness ranking process: Calculate the mixed population The size constraints between the solutions are then used to analyze the mixed population based on the non-dominated method and the constraints. Conduct comprehensive ranking to clarify individual advantages and disadvantages;

[0037] Step 5. Front Layer individual selection: from the sorted mixed population Before the selection Layer individuals, get the set , as a preliminary solution;

[0038] Step 6. Population adjustment:

[0039] judge Is the population size exactly N? If not, execute the following sub-process:

[0040] Determine from The number of individuals selected in the layer is , The selected front obtained in the previous iteration Layer individual collection;

[0041] Adaptive normalization: First, the minimum value of each dimension is selected as the ideal point, and the ideal point is translated to the coordinate origin. Then, the extreme point of the population solution on each axis is determined, and finally the population solution is normalized to make the data at an appropriate scale.

[0042] Niche preservation process: connect the origin and the reference point to form a reference vector, calculate the distance between the solution of each population and the reference vector; use PBI sorting to classify the population according to the reference point, record is the number of populations associated with the jth reference point, and then In order from small to large, each time one of the populations is selected and added to , until The population number is N, ensuring population diversity and balance;

[0043] Step 7. Iteration termination judgment: If The population size is exactly N, then As the next generation parent population, check whether the maximum number of iterations has been reached. If so, end the process. If not, return and continue iterating until the termination condition is met and output the optimal solution.

[0044] Based on the large-scale integration of electric vehicles into the distribution network and the installation of new energy power generation and energy storage devices at charging stations, this invention takes into account factors such as flexible power regulation within each charging station, greater autonomy in setting charging service fees, and rapid charging and discharging startup of energy storage devices. It constructs a multi-level, orderly scheduling process within and between stations, solving the problem using the C-MOEA / D optimization algorithm to ensure the scientific nature of the results. By using an inter-station load distribution model and a multi-objective intra-station optimization scheduling model, the load regulation completion rate and self-balancing coordination capabilities can be improved, which is of great significance to the stable operation of the power grid and the safe and efficient operation of charging stations. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] Figure 1 It is a flow chart of the method of the present invention.

[0046] Figure 2 It is a flow chart of the C-MOEA / D optimization algorithm of the present invention. DETAILED DESCRIPTION

[0047] The present invention is further described below with reference to the embodiments. It is necessary to point out that the following embodiments are only used to further illustrate the present invention and are not to be construed as limiting the scope of protection of the present invention. Non-essential improvements and adjustments made by persons skilled in the art based on the above-mentioned invention contents still fall within the scope of protection of the present invention.

[0048] Reference Figure 1 , a multi-level orderly scheduling method for large-scale charging loads to participate in grid peak regulation, including two stages: inter-station load distribution and intra-station optimization scheduling;

[0049] Inter-station load allocation refers to allocating the load adjustment amount of each charging station according to the proportion of the adjustable capacity assessment value of each charging station based on the load adjustment instructions actually issued by the dispatch center and marketing department to the electric vehicle aggregator after obtaining the adjustable capacity assessment value of each charging station. The implementation process of inter-station load allocation is as follows:

[0050] (1) The forecasting department reports the city’s 24-hour electric vehicle load forecast for the next day to the dispatching center at 4:00 p.m. every day. At the same time, the dispatching center notifies the marketing department of the load gap for the next day.

[0051] (2) The total adjustable capacity of the electric vehicle aggregator is evaluated. First, according to the electric vehicle load prediction value, the user price response load of the charging station is taken as the adjustable load, and the user price response load adjustable capacity is predicted by adjusting the charging price. Then, according to the price response degree, the load of the charging station in the dispatching period is obtained after being affected by the price, and combined with the prediction of the fast and slow charging loads, the fast charging load at this time is obtained, which is taken as the fast charging power adjustment load, the charging power of the fast charging pile is adjusted, and the fast charging power adjustment load adjustable capacity is obtained. The sum of the user price response load and the fast charging power adjustment load adjustable capacity is the total adjustable capacity of the electric vehicle aggregator.

[0052] (3) The electric vehicle aggregator reports the total adjustable capacity of the electric vehicle aggregator in the gap period to the marketing department. The marketing department allocates the load adjustment amount to the electric vehicle aggregator, and the electric vehicle aggregator determines the load adjustment amount of each charging station in the dispatching period according to the proportion of the adjustable capacity evaluation value of each charging station, and provides the load adjustment reference value of each period for the station optimization.

[0053] (4) Each charging station in the electric vehicle aggregator determines the orderly charging measures according to the allocated load adjustment amount, and reports the pre-execution situation to the dispatching center before 12:00 on the same day.

[0054] (5) The dispatching center agrees to execute the plan, and the marketing department executes the plan.

[0055] The station optimization dispatching refers to constructing a multi-objective station optimization dispatching model with the maximum load adjustment completion rate and the maximum self-balancing coordination ability as the target, and solving it by using the C-MOEA / D optimization algorithm to obtain the optimal dispatching scheme.

[0056] The objects of the station optimization dispatching include two types of ordinary charging stations and light storage charging stations. The station optimization dispatching of ordinary charging stations and light storage charging stations aims to maximize the load adjustment completion rate while meeting the load adjustment requirements; in addition, new energy and energy storage devices are also considered to improve the self-balancing coordination ability. Therefore, the charging power and charging service fee plan of the ordinary charging station and the light storage charging station are formulated to meet the requirements of the dispatching center.

[0057] The inter-station load distribution allocates the actual load adjustment amount to be responsible by each charging station according to the proportion of the adjustable capacity evaluation value. For charging stations with complete analysis data, the adjustment load is allocated in proportion. For some charging stations lacking data, it is impossible to accurately evaluate using the existing model. At this time, according to the prediction value of the electric vehicle load of the charging station, the larger the load, the more users can participate in the adjustment, so the inter-station load distribution can be carried out according to the load prediction value of each charging station. According to this principle, the inter-station load distribution model according to the adjustable capacity evaluation value and the inter-station load distribution model according to the load prediction value are established respectively.

[0058] According to the evaluation value of the adjustable capacity of the charging station, the load adjustment amount is proportionally distributed among the stations. The load distribution model among the stations based on the evaluation value of the adjustable capacity is shown in formula (1):

[0059] (1);

[0060] Where, Allocate the total actual load regulation for electric vehicle aggregators during period t; is the evaluation value of the adjustable capacity of the i-th charging station in period t; is the total adjustable capacity of charging stations in the scheduling area during period t; The actual load regulation amount after inter-station load distribution for the i-th charging station in period t.

[0061] When the charging station information collection is incomplete and the adjustable capacity evaluation value of each charging station cannot be obtained before the dispatch center issues a dispatch instruction, the actual load reduction can be distributed among the stations in proportion according to the load forecast value of each charging station during the dispatch period. The load distribution model among the stations based on the load forecast value is shown in formula (2):

[0062] (2);

[0063] Where, is the load forecast value of the i-th charging station in period t; It is the sum of the load forecast values ​​of charging stations in the scheduling area during period t.

[0064] This paper uses a multi-objective in-station optimization scheduling model based on the C-MOEA / D optimization algorithm to perform in-station optimization scheduling. This is done with the goal of meeting the load regulation completion rate and the self-balancing coordination capability of the charging station. An optimization scheduling model is established for all output components within the solar-storage charging station, and the optimal scheduling solution is obtained. The multi-objective in-station optimization scheduling model based on the C-MOEA / D optimization algorithm is as follows:

[0065] Goal 1: Load regulation completion rate maximum.

[0066] According to the load distribution results between stations, the load that each charging station is responsible for reducing is obtained, and the completion of the scheduling requirements can be reflected by the ratio of the target regulated load to the actual regulated load.

[0067] (3);

[0068] Where, represents the actual load power of the i-th charging station in time period t after electricity price response and charging power adjustment; is the load forecast value of the i-th charging station in period t.

[0069] Goal 2: Self-balancing and coordination skills maximum.

[0070] The self-balancing coordination capability of the charging station is represented by the ratio of the sum of the power purchased by the power grid and the output of the photovoltaic power station to the charging power of the charging station. The larger the value, the stronger the self-balancing coordination capability and the stronger the peak load regulation capability of the participating power grid, as shown in formula (4).

[0071] (4);

[0072] Where, is the power purchased by the power grid during period t, is the charging power of the charging station during period t, is the output of the photovoltaic power station during period t.

[0073] Constraints

[0074] (1) Output constraints of photovoltaic power stations:

[0075] (5);

[0076] (6);

[0077] Where, is the maximum output of the photovoltaic power station during period t; is the power coefficient of the photovoltaic power station, which is a function of light intensity and temperature; is the access capacity of the photovoltaic power station.

[0078] (2) Energy storage charging and discharging state constraints:

[0079] During normal operation, energy storage devices are usually in one of the following three operating states: charging state, discharging state, and non-charging / discharging state. The charging and discharging state constraints of energy storage devices can be expressed as:

[0080] (7);

[0081] (8);

[0082] (9);

[0083] (10);

[0084] In the formula, the operating state of the energy storage device is limited to one of the following states: charging state, discharging state, or non-charging and discharging state; and are the lower and upper limits of the charging power of the i-th energy storage device respectively; and are the state of charge of the i-th energy storage device during charging and discharging in period t; and are the lower and upper limits of the discharge power of the i-th energy storage device, and It is a binary indicator variable of the operating status of the energy storage device, with a value of 1 or 0. When the value is 1, it means that the i-th energy storage device is in the charging state during the t period; when the value is 0, it means that it is not in the charging state; When the value is 1, it means that the i-th energy storage device is in the discharge state during the t period; when the value is 0, it means that it is not in the discharge state. ; is the energy storage capacity of the i-th energy storage device in period t; is the charging efficiency coefficient of the i-th energy storage device; is the discharge efficiency coefficient of the i-th energy storage device, is the time step; is the maximum energy storage capacity of the i-th energy storage device, and the battery energy storage capacity is set to 10% of the total daily load of the distribution network.

[0085] (3) Power balance constraints:

[0086] (11);

[0087] Where, is the charging power of the charging station during period t; is the injected power of the energy storage device of the charging station during period t.

[0088] (4) Fast charging power constraints:

[0089] Reducing the charging power of a fast charging pile will reduce the charging rate and extend the charging time. Therefore, the reduction in charging amount within the charging time caused by reducing the charging power should not affect the user's normal travel. The charging power of a fast charging pile can be expressed as:

[0090] (12);

[0091] Where P is the charging power of the fast charging pile (kW); and These are the maximum and minimum charging power of the fast charging pile (60kW and 48kW). The minimum amount of electricity required for an electric vehicle to reach its next destination; The parking time for the user is different in different areas.

[0092] (5) Electricity price constraints:

[0093] The time-of-use electricity price of a charging station consists of electricity charges and charging service fees. Adjusting the charging electricity price range is to adjust the service fee. The electricity price constraints are:

[0094] (13);

[0095] Where, represents the electricity price for electric vehicle charging, Indicates the actual electricity cost in the charging station. Indicates charging service fee, yuan / kWh.

[0096] The physical meaning of this control model is to optimize the scheduling of charging station energy storage, photovoltaics, electric vehicles and other equipment resources, and to achieve the optimal comprehensive benefits of each charging station in the regulation area while meeting the load regulation requirements under the constraints of the fast charging power regulation power station and equipment operation.

[0097] The objective functions of the optimization scheduling model within a common charging station include:

[0098] Goal 1: Load regulation completion rate maximum.

[0099] According to the load distribution results between stations, the load that each charging station is responsible for reducing is obtained, and the completion of the scheduling requirements can be reflected by the ratio of the target regulated load to the actual regulated load.

[0100] (14);

[0101] Where, represents the actual load power of the i-th charging station in time period t after electricity price response and charging power adjustment; is the load forecast value of the i-th charging station in period t.

[0102] Constraints

[0103] (1) Power balance constraints:

[0104] (15);

[0105] Where M is the set of charging stations; for any i∈M, is the charging load of the i-th charging station during period t; The power injected into the grid for the i-th charging station during period t.

[0106] (2) Fast charging power constraints:

[0107] Reducing the charging power of a fast charging pile will reduce the charging rate and extend the charging time. Therefore, the reduction in charging amount within the charging time caused by reducing the charging power should not affect the user's normal travel. The charging power of a fast charging pile can be expressed as:

[0108] (16) ;

[0109] P is the fast charging pile charging power (kW) ; and are the maximum and minimum values of the fast charging pile charging power (60 kW and 48 kW) ; is the minimum power required for the electric vehicle to reach the next destination driving process; is the parking time of the user, which is different in different areas.

[0110] (3) Price constraint:

[0111] The time-of-use electricity price of the charging station is composed of electricity charges and charging service charges, and adjusting the charging price range is to adjust the level of service charges. The price constraint is:

[0112] (17) ;

[0113] In the formula, represents the electric vehicle charging price, represents the actual electricity charge in the charging station, represents the charging service charge, yuan / degree.

[0114] The physical meaning of the control model is to realize the optimal comprehensive benefit of the operation of the charging station and the equipment under the constraint condition of adjusting the fast charging power to meet the load adjustment requirements while running in the adjustment area.

[0115] For the multi-objective optimization model constructed by the application, the C-MOEA / D optimization algorithm is used to solve it, as shown in Figure 2 , the steps are as follows:

[0116] Step 1. Initialization operation: generate reference points on the hyperplane according to the variable dimension, and then form an initial population with a population size of N , which lays the foundation for subsequent evolution;

[0117] Step 2. Generate offspring and merge the population: perform crossover and mutation operations on the initial population to form a child population ; combine and to form a hybrid population with a population size of 2N , which expands the search range;

[0118] Step 3. Fixed constraint screening: according to the established fixed constraint condition, screen the hybrid population , eliminate individuals that do not meet the constraints, and retain solutions that meet the fixed constraint condition;

[0119] Step 4. Fitness ranking process: Calculate the mixed population The size constraints between the solutions are then used to analyze the mixed population based on the non-dominated method and the constraints. Conduct comprehensive ranking to clarify individual advantages and disadvantages;

[0120] Step 5. Front Layer individual selection: from the sorted mixed population Before the selection Layer individuals, get the set , as a preliminary solution;

[0121] Step 6. Population adjustment:

[0122] judge Is the population size exactly N? If not, execute the following sub-process:

[0123] Determine from The number of individuals selected in the layer is , The selected front obtained in the previous iteration Layer individual collection;

[0124] Adaptive normalization: First, the minimum value of each dimension is selected as the ideal point, and the ideal point is translated to the coordinate origin. Then, the extreme point of the population solution on each axis is determined, and finally the population solution is normalized to make the data at an appropriate scale.

[0125] Niche preservation process: connect the origin and the reference point to form a reference vector, calculate the distance between the solution of each population and the reference vector; use PBI sorting to classify the population according to the reference point, record is the number of populations associated with the jth reference point, and then In order from small to large, each time one of the populations is selected and added to , until The population number is N, ensuring population diversity and balance;

[0126] Step 7. Iteration termination judgment: If The population size is exactly N, then As the next generation parent population, check whether the maximum number of iterations has been reached. If so, end the process. If not, return and continue iterating until the termination condition is met and output the optimal solution.

[0127] The above merely expresses the preferred embodiments of the present application, and is not intended to limit the present application in other forms. Any person skilled in the art can make changes or modifications to the above disclosed content to obtain equivalent embodiments with equivalent changes. However, any simple modification, equivalent change and modification made to the above embodiments without departing from the technical solution of the present application and according to the technical essence of the present application still belongs to the protection scope of the technical solution of the present application.

Claims

1. A multi-level orderly dispatching method for large-scale charging loads to participate in power grid peak regulation, characterized by: It includes two stages: inter-station load distribution and intra-station optimization scheduling; Inter-station load allocation means allocating the load adjustment amount of each charging station in proportion to the adjustable capacity assessment value of each charging station based on the load adjustment instructions actually issued by the dispatch center and marketing department to the electric vehicle aggregator after obtaining the adjustable capacity assessment value of each charging station; According to the evaluation value of the adjustable capacity of the charging station, the load adjustment amount is proportionally distributed among the stations. The load distribution model among the stations based on the evaluation value of the adjustable capacity is shown in formula (1): （1）； Where, Allocate the total actual load regulation for electric vehicle aggregators during period t; is the evaluation value of the adjustable capacity of the i-th charging station in period t; is the total adjustable capacity of charging stations in the scheduling area during period t; The actual load regulation amount after inter-station load distribution for the i-th charging station in period t; When the adjustable capacity evaluation value of each charging station cannot be obtained before the dispatch center issues a dispatch instruction, the actual load reduction is distributed among the stations in proportion according to the load forecast value of each charging station during the dispatch period. The load distribution model among the stations based on the load forecast value is shown in formula (2): （2）； Where, is the load forecast value of the i-th charging station in period t; is the sum of the load forecast values ​​of the charging stations in the dispatch area during period t; The implementation process of inter-station load distribution is as follows: The forecasting department reports the electric vehicle load forecast for each period of the next day to the dispatching center every day, and the dispatching center informs the marketing department of the load gap for the next day; Based on the predicted value of electric vehicle load, it is used as the user electricity price response load of the charging station. By adjusting the charging electricity price, the adjustable capacity of the user electricity price response load is predicted. Then, based on the electricity price responsiveness, the load during the charging station scheduling period affected by the electricity price is obtained. Combined with the prediction of fast and slow charging loads, the fast charging load at that time is obtained. This is used as the fast charging power adjustment load, and the charging power of the fast charging pile is adjusted to obtain the fast charging power adjustment load adjustable capacity. The sum of the user electricity price response load and the fast charging power adjustment load adjustable capacity is the total adjustable capacity of the electric vehicle aggregator. The EV aggregator reports its total adjustable capacity during the gap period to the marketing department. The marketing department then allocates load adjustment to the EV aggregator. The EV aggregator then determines the load adjustment for each charging station during the scheduling period based on the ratio of the adjustable capacity assessment values ​​of each charging station, providing a reference value for load adjustment during each period for station optimization. Each charging station within the electric vehicle aggregator determines orderly charging measures based on the allocated load adjustment amount and reports the pre-implementation status to the dispatch center before the day; The dispatch center agrees to execute the plan, and the marketing department executes it; In-station optimal scheduling refers to building a multi-objective in-station optimal scheduling model with the goal of maximizing the load regulation completion rate and the self-balancing coordination ability, and using the C-MOEA / D optimization algorithm to solve it and obtain the optimal scheduling plan.

2. The multi-level orderly scheduling method according to claim 1 is characterized in that: The objects of in-station optimization scheduling include ordinary charging stations and photovoltaic storage charging stations. The multi-objective in-station optimization scheduling model is: Goal 1: Load regulation completion rate maximum; （3）； Where, represents the actual load power of the i-th charging station in time period t after electricity price response and charging power adjustment; is the load forecast value of the i-th charging station in period t; Goal 2: Self-balancing and coordination skills maximum; （4）； Where, is the power purchased by the power grid during period t, is the charging power of the charging station during period t, is the output of the photovoltaic power station during period t, and T represents the total duration of the scheduling period; Ordinary charging stations only use load adjustment completion rate The goal is to maximize the solar storage charging station with self-balancing coordination capabilities. Maximum and self-balancing coordination ability The maximum is the target.

3. The multi-level orderly scheduling method according to claim 2 is characterized in that: The constraints of the on-site optimization scheduling model of the photovoltaic storage charging station include photovoltaic power station output constraints, energy storage charging and discharging state constraints, power balance constraints, fast charging power constraints and electricity price constraints.

4. The multi-level orderly scheduling method according to claim 2 is characterized in that: The constraints of the in-station optimization scheduling model of ordinary charging stations include power balance constraints, fast charging power constraints and electricity price constraints.

5. The multi-level orderly scheduling method according to claim 1 is characterized in that: The solution process using the C-MOEA / D optimization algorithm is as follows: Step 1. Initialization operation: Generate reference points on the hyperplane according to the variable dimension, and then form an initial population of N , laying the foundation for subsequent evolution; Step 2. Generate offspring and merge populations: Perform crossover and mutation operations to form a progeny population ; Combine and Form a mixed population with a population size of 2N , expand the search scope; Step 3. Fixed constraint screening: Based on the established fixed constraint conditions, the mixed population Screening is performed to eliminate individuals that do not meet the constraints and retain solutions that meet the fixed constraints; Step 4. Fitness ranking process: Calculate the mixed population The size constraints between the solutions are then used to analyze the mixed population based on the non-dominated method and the constraints. Conduct comprehensive ranking to clarify individual advantages and disadvantages; Step 5. Front Layer individual selection: from the sorted mixed population Before the selection Layer individuals, get the set , as a preliminary solution; Step 6. Population adjustment: judge Is the population size exactly N? If not, execute the following sub-process: Determine from The number of individuals selected in the layer is , The selected front obtained in the previous iteration Layer individual collection; Adaptive normalization: First, the minimum value of each dimension is selected as the ideal point, and the ideal point is translated to the coordinate origin. Then, the extreme point of the population solution on each axis is determined, and finally the population solution is normalized to make the data at an appropriate scale. Niche preservation process: connect the origin and the reference point to form a reference vector, calculate the distance between the solution of each population and the reference vector; use PBI sorting to classify the population according to the reference point, record is the number of populations associated with the jth reference point, and then In order from small to large, each time one of the populations is selected and added to , until The population number is N, ensuring population diversity and balance; Step 7. Iteration termination judgment: If The population size is exactly N, then As the next generation parent population, check whether the maximum number of iterations has been reached. If so, end the process. If not, return and continue iterating until the termination condition is met and output the optimal solution.

Images