Transformer area flexible load robust optimization method and device, electronic equipment and storage medium

Abstract

The invention discloses a transformer area flexible load robust optimization method and device, electronic equipment and a storage medium, relates to the field of low-voltage transformer area flexible load scheduling, and realizes gradient utilization of flexible loads and improves the distribution transformer capacity utilization rate through a multi-time scale two-stage robust optimization model and considering distributed photovoltaic uncertainty. The method comprises the steps that an air conditioner load model, an electric vehicle charge state model and an energy storage charge state model are adopted to construct a transformer area flexible load model; establishing a multi-time-scale two-stage robust optimization model based on a robust optimization algorithm and the transformer area flexible load model; solving the day-ahead stage of the robust optimization model by adopting a column constraint generation algorithm to obtain an air conditioning load, energy storage station and charging station scheduling plan of the day-ahead stage; and based on the scheduling plan of the air conditioner load, the energy storage station and the charging station in the day-ahead stage, solving the intra-day stage of the robust optimization model by adopting a solver to obtain a real-time charging and discharging power scheduling plan of the energy storage equipment and the electric vehicle.

Description

Technical Field

[0001] The present application relates to the technical field of flexible load scheduling in low-voltage substations, and in particular to a method, device, electronic equipment and storage medium for robust optimization of flexible loads in substations. Background Art

[0002] With the global energy transition and the goal of carbon neutrality, the development and utilization of renewable energy are becoming increasingly important in the energy system. Photovoltaic power generation, due to its clean and low-carbon characteristics, has gradually become a key component of low-voltage distribution networks. Simultaneously, the rapid development of energy storage technology and the widespread adoption of electric vehicles have provided new opportunities for optimizing the operation of low-voltage distribution networks. However, factors such as the uncertainty of distributed photovoltaics, the randomness of electric vehicle load charging, and the diverse characteristics of traditional loads have led to increased power fluctuations, voltage over-limits, and load imbalances in low-voltage distribution transformers.

[0003] In the related art, the main methods for dealing with uncertainty are stochastic programming and robust optimization methods. The stochastic programming method simulates uncertainty by generating random scenarios through the probability information of photovoltaic prediction errors, and can also convert random problems into linear problems to reduce the complexity of solution. The robust optimization method does not require the uncertainty probability distribution information of photovoltaic power, and only relies on the upper and lower limits of uncertainty parameters, reducing dependence on data and modeling. However, the applicant realized that the stochastic programming method has no unified standard for the probability distribution of photovoltaic prediction errors, making it difficult to accurately characterize the error characteristics, and the reliability of scenario generation is insufficient, which affects the accuracy of optimization decisions. Robust optimization is a semi-infinite optimization problem, which is difficult to solve directly and needs to be converted into a convex optimization problem through robust equivalence. For nonlinear problems, uncertainty may lead to a non-convex feasible domain, causing conventional methods to fall into local optimality and making it difficult to guarantee global optimality. Therefore, neither method can effectively deal with the heavy load and overload problems of low-voltage distribution transformers caused by the uncertainty of distributed resources. Summary of the Invention

[0004] In view of this, the present application provides a method, device, electronic device and storage medium for robust optimization of flexible load in a substation. The main purpose is to solve the problem that the existing random planning and robust optimization methods cannot effectively deal with the heavy load and overload of low-voltage substation distribution transformers caused by the uncertainty of distributed resources.

[0005] According to the first aspect of the present application, a method for robust optimization of flexible loads in a substation is provided, the method comprising:

[0006] Establishing an air conditioning load model, an electric vehicle state of charge model, and an energy storage state of charge model, and using the air conditioning load model, the electric vehicle state of charge model, and the energy storage state of charge model to construct a flexible load model for the substation;

[0007] Obtaining a robust optimization algorithm, and establishing a multi-time-scale two-stage robust optimization model based on the robust optimization algorithm and the substation flexible load model;

[0008] Obtaining a column constraint generation algorithm, and using the column constraint generation algorithm to solve the day-ahead phase of the multi-time-scale two-stage robust optimization model to obtain a scheduling plan for air conditioning loads, energy storage stations, and charging stations in the day-ahead phase;

[0009] Based on the air conditioning load, energy storage station and charging station scheduling plan in the day-ahead stage, a solver is used to optimize and solve the intraday stage of the multi-time-scale two-stage robust optimization model to obtain the real-time charging and discharging power scheduling plan of the energy storage equipment and electric vehicles.

[0010] According to a second aspect of the present application, a robust optimization device for flexible loads in a substation is provided, the device comprising:

[0011] The first modeling module is used to establish an air conditioning load model, an electric vehicle state of charge model, and an energy storage state of charge model, and use the air conditioning load model, the electric vehicle state of charge model, and the energy storage state of charge model to construct a flexible load model of the substation;

[0012] A second modeling module is used to obtain a robust optimization algorithm, and based on the robust optimization algorithm, establish a multi-time scale two-stage robust optimization model using the substation flexible load model;

[0013] A first solving module is configured to obtain a column constraint generation algorithm, and employ the column constraint generation algorithm to solve the day-ahead phase of the multi-time-scale two-stage robust optimization model to obtain a scheduling plan for air conditioning load, energy storage stations, and charging stations in the day-ahead phase;

[0014] The second solving module is used to optimize and solve the intraday stage of the multi-time-scale two-stage robust optimization model based on the air-conditioning load, energy storage station and charging station scheduling plan of the day-ahead stage, and obtain the real-time charging and discharging power scheduling plan of the energy storage equipment and electric vehicles.

[0015] According to a third aspect of the present application, an electronic device is provided, comprising a memory and a processor, wherein the memory stores a computer program, and the processor implements the steps of any one of the methods described in the first aspect when executing the computer program.

[0016] According to a fourth aspect of the present application, a storage medium is provided, on which a computer program is stored. When the computer program is executed by a processor, the steps of any one of the methods in the first aspect are implemented.

[0017] By means of the above technical solution, the technical solution provided by the embodiment of the present invention has at least the following advantages:

[0018] The present application provides a method, device, electronic device and storage medium for robust optimization of flexible loads in a substation. The present application establishes an air conditioning load model, an electric vehicle state of charge model and an energy storage state of charge model, and uses the air conditioning load model, the electric vehicle state of charge model and the energy storage state of charge model to construct a substation flexible load model, obtains a robust optimization algorithm, and based on the robust optimization algorithm, uses the substation flexible load model to establish a multi-time scale two-stage robust optimization model, obtains a column constraint generation algorithm, and uses the column constraint generation algorithm to solve the day-ahead stage of the multi-time scale two-stage robust optimization model to obtain the air conditioning load, energy storage station and charging station scheduling plan of the day-ahead stage. Based on the air conditioning load, energy storage station and charging station scheduling plan of the day-ahead stage, a solver is used to optimize and solve the intraday stage of the multi-time scale two-stage robust optimization model to obtain the real-time charging and discharging power scheduling plan of the energy storage equipment and electric vehicles. The present application proposes a multi-time scale two-stage robust optimization model to realize the tiered utilization of flexible loads and improve the utilization rate of distribution transformer capacity. The multi-time-scale two-stage robust optimization model fully considers the uncertainty of distributed photovoltaics, meets energy demand while effectively coping with distributed photovoltaic power fluctuations. Compared with the deterministic optimization method, the optimal scheduling plan for flexible loads in the substation area of this application has stronger robustness.

[0019] The above description is only an overview of the technical solution of the present application. In order to more clearly understand the technical means of the present application, it can be implemented in accordance with the contents of the specification. In order to make the above and other purposes, features and advantages of the present application more obvious and easy to understand, the specific implementation methods of the present application are listed below. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] Various other advantages and benefits will become apparent to those skilled in the art upon reading the detailed description of the preferred embodiment below. The accompanying drawings are for illustration purposes only and are not to be considered as limiting the present application. The same reference symbols are used throughout the drawings to represent the same components. In the drawings:

[0021] Figure 1 A schematic flow chart of a method for robust optimization of flexible loads in a substation area provided in an embodiment of the present application is shown;

[0022] Figure 2A A schematic flow chart of another method for robust optimization of flexible loads in a substation area provided in an embodiment of the present application is shown;

[0023] Figure 2B A flowchart of a column constraint generation algorithm solution provided by an embodiment of the present application is shown;

[0024] Figure 2C The following is a diagram showing a multi-time scale flexible load robust optimization framework for a substation provided by an embodiment of the present application;

[0025] Figure 2D shows a photovoltaic power curve diagram provided by an embodiment of the present application;

[0026] Figure 2E shows a base load power curve diagram provided by an embodiment of the present application;

[0027] Figure 2F shows the energy storage charging and discharging power diagram provided by the embodiment of the present application;

[0028] Figure 2G Shows a power diagram of a single air conditioner provided in an embodiment of the present application;

[0029] Figure 2H It shows the charging and discharging power diagram of electric vehicles in the charging station provided by the embodiment of the present application;

[0030] Figure 2I The figure shows the transformer load rate diagram under different scenarios provided by the embodiment of the present application;

[0031] Figure 2J The following shows the day-ahead scheduling and operation analysis of each scenario provided in the embodiment of the present application;

[0032] Figure 2K The day-ahead operation analysis of the station area under uncertain parameters provided by the embodiment of the present application is shown;

[0033] Figure 2L Shows the day-ahead and day-intraday energy storage charging and discharging power diagram provided by an embodiment of the present application;

[0034] Figure 2M It shows the charge and discharge power diagram of electric car No. 1 provided by the embodiment of the present application in the day before and within the day;

[0035] Figure 2N The operation cost comparison and substation operation analysis under different time scales provided by the embodiment of the present application are shown;

[0036] Figure 3 A schematic diagram of a robust optimization of flexible loads in a metropolitan area provided in an embodiment of the present application is shown;

[0037] Figure 4 A schematic diagram of the device structure of an electronic device provided in an embodiment of the present application is shown. DETAILED DESCRIPTION

[0038] In the description of the present invention, it should be understood that the terms "first" and "second" are used for descriptive purposes only and should not be understood to indicate or imply relative importance or implicitly specify the number of the technical features indicated. Therefore, a feature specified as "first" or "second" may explicitly or implicitly include one or more of the features. In the description of the present invention, "plurality" means two or more, unless otherwise specifically defined.

[0039] The following describes exemplary embodiments of the present application in more detail with reference to the accompanying drawings. Although exemplary embodiments of the present application are shown in the accompanying drawings, it should be understood that the present application can be implemented in various forms and should not be limited by the embodiments set forth herein. Rather, these embodiments are provided to enable a more thorough understanding of the present application and to fully convey the scope of the present application to those skilled in the art.

[0040] The embodiment of the present application provides a method for robust optimization of flexible load in a substation area, such as Figure 1 As shown, the method includes:

[0041] 101. Establish an air conditioning load model, an electric vehicle state of charge model, and an energy storage state of charge model, and use the air conditioning load model, electric vehicle state of charge model, and energy storage state of charge model to construct a flexible load model for the substation.

[0042] In order to solve the problem of heavy load and overload of distribution transformers caused by the uncertainty of distributed photovoltaics and disorderly charging of electric vehicle loads, this application proposes a robust optimization method for flexible loads in substations, which breaks through the traditional scheduling method of using a single time scale, and proposes a two-stage robust optimization method for flexible loads in substations taking into account multiple time scales. A two-stage robust optimization model with multiple time scales is established to achieve the cascade utilization of flexible loads, improve the utilization rate of distribution transformer capacity, fully consider the uncertainty of distributed photovoltaics, meet energy demand, and effectively cope with distributed photovoltaic power fluctuations. Compared with the deterministic optimization method, the obtained optimal scheduling plan for flexible loads in substations has stronger robustness. The executor of this application may be a substation dispatching system, which relies on the computing power of the server to provide services to users. The server may be an independent server, or it may provide cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communications, middleware services, domain name services, security services, content delivery networks (CDN), and servers for basic cloud computing such as big data and artificial intelligence platforms, so that the substation dispatching system can determine accurate and optimized real-time charging and discharging power scheduling plans for energy storage equipment and electric vehicles.

[0043] In an embodiment of the present application, the substation dispatching system constructs a substation flexible load model, taking into account the differences in response time, response capacity, and response rate of the substation flexible load, and can flexibly adjust the load of air conditioners, electric vehicles, and energy storage equipment according to the output of distributed power sources, thereby smoothing power fluctuations, reducing the impact on the power grid, and improving energy utilization efficiency. The substation flexible load model can achieve refined management of the load, and improve the utilization rate of the power grid by reasonably adjusting the operating time and power of air conditioners, electric vehicles, and energy storage equipment. Moreover, the substation flexible load model takes into account the flexibility of air conditioners, electric vehicles, and energy storage equipment, and can quickly adjust the load strategy according to different uncertainty scenarios, thereby improving the ability of the power grid to cope with uncertainty.

[0044] 102. Obtain a robust optimization algorithm. Based on the robust optimization algorithm, a multi-time-scale two-stage robust optimization model is established using the flexible load model of the substation.

[0045] In the embodiment of the present application, the substation dispatch system takes into account the uncertainty of distributed photovoltaic output and load power consumption in the low-voltage substation, and adopts a robust optimization method to establish an optimization model for low-voltage substation day-ahead dispatching and intraday real-time control, that is, a multi-time-scale two-stage robust optimization model. The day-ahead dispatch stage can fully consider the operational flexibility of equipment such as air-conditioning loads, electric vehicles, and energy storage in the worst-case scenario, and obtain a 24-hour output plan for each controllable load in the substation system; the intraday dispatch stage formulates the power dispatch plan for energy storage and charging stations in the intraday stage based on the day-ahead plan, improving the model's ability to handle complex uncertainties.

[0046] 103. Obtain a column constraint generation algorithm, and use the column constraint generation algorithm to solve the day-ahead phase of the multi-time-scale two-stage robust optimization model to obtain the day-ahead scheduling plan for air-conditioning load, energy storage station, and charging station.

[0047] In an embodiment of the present application, the substation scheduling system adopts a column constraint generation algorithm to solve the day-ahead stage of the multi-time-scale two-stage robust optimization model, generates an initial scheduling plan through the main problem, and finds the worst scenario based on the uncertainty set of the subproblem. The plan is continuously revised through the main-subproblem iteration to ensure that the scheduling plan is still feasible in extreme scenarios.

[0048] Based on the day-ahead scheduling plan for air conditioning load, energy storage stations, and charging stations, a solver is used to optimize and solve the intraday phase of the multi-time-scale two-stage robust optimization model to obtain the real-time charging and discharging power scheduling plan for energy storage equipment and electric vehicles.

[0049] In the embodiment of the present application, the air-conditioning load is no longer regulated during the day, and the substation dispatching system optimizes the power of the energy storage station and the charging station, thereby smoothing the photovoltaic and load power fluctuations during the day.

[0050] The embodiment of the present application provides a method for robust optimization of flexible loads in a substation. Compared with the prior art, the embodiment of the present invention establishes an air conditioning load model, an electric vehicle state of charge model, and an energy storage state of charge model, and uses the air conditioning load model, the electric vehicle state of charge model, and the energy storage state of charge model to construct a substation flexible load model, obtains a robust optimization algorithm, and based on the robust optimization algorithm, uses the substation flexible load model to establish a multi-time scale two-stage robust optimization model, obtains a column constraint generation algorithm, and uses the column constraint generation algorithm to solve the day-ahead stage of the multi-time scale two-stage robust optimization model to obtain the air conditioning load, energy storage station, and charging station scheduling plan of the day-ahead stage. Based on the air conditioning load, energy storage station, and charging station scheduling plan of the day-ahead stage, a solver is used to optimize and solve the intraday stage of the multi-time scale two-stage robust optimization model to obtain the real-time charging and discharging power scheduling plan of the energy storage equipment and electric vehicles. The present application proposes a multi-time scale two-stage robust optimization model to achieve tiered utilization of flexible loads and improve the utilization rate of distribution transformer capacity. The multi-time-scale two-stage robust optimization model fully considers the uncertainty of distributed photovoltaics, meets energy demand while effectively coping with distributed photovoltaic power fluctuations. Compared with the deterministic optimization method, the optimal scheduling plan for flexible loads in the substation area of this application has stronger robustness.

[0051] Furthermore, as a refinement and extension of the specific implementation of the above embodiment, in order to fully illustrate the specific implementation process of this embodiment, the embodiment of the present application provides another robust optimization method for flexible loads in the substation, such as Figure 2A As shown, the method includes:

[0052] 201. Establish an air conditioning load model.

[0053] In the embodiment of the present application, in order to reduce the impact of regulating the air conditioning load on the power comfort and make the power algebraic sum before and after regulation equal, the substation dispatching system obtains a first-order thermodynamic equivalent model and air conditioning load data, and uses the first-order thermodynamic equivalent model to model the air conditioning load data to obtain an air conditioning load model, as shown in the following formula 1:

[0054] Formula 1:

[0055]

[0056]

[0057]

[0058]

[0059]

[0060] in, Indicates the indoor temperature of the area where the air conditioning load is located at time t+1, Indicates the outdoor temperature of the area where the air conditioning load is located at time t+1, represents the cooling capacity of the air conditioning load at time t, R b Indicates the thermal resistance of the area where the air conditioning load is located, C b represents the heat capacity of the area where the air conditioning load is located, Δt represents the time scale of the day-ahead stage scheduling, which is 1h. represents the temperature setting value of the air conditioning load, ΔΓ represents the maximum indoor temperature offset, represents the power of the air conditioning load at time t, δ represents the air conditioning energy efficiency ratio, represents the power of the air conditioning load cluster at time t, N ac Indicates the air conditioning load quantity, Indicates the lower limit of air conditioning load power adjustment. Indicates the upper limit of air conditioning load power adjustment. Indicates the power setting value at time t when the air conditioning load is not involved in scheduling.

[0061] 202. Establish a state of charge model for electric vehicles.

[0062] In the embodiment of the present application, the trip termination time of the household electric vehicle conforms to the normal distribution of the piecewise logarithm, as shown in the following formula 2:

[0063] Formula 2:

[0064] The daily mileage of electric vehicles also conforms to the normal distribution of piecewise logarithms, as shown in the following formula 3:

[0065] Formula 3:

[0066] Among them, μ s , σ s 、μ D , σ D They are all parameters, and they are 17.6, 3.4, 3.2, and 0.88 respectively.

[0067] In order to reduce the impact of regulating electric vehicle load on user charging demand and make the algebraic sum of power of electric vehicles equal before and after participating in scheduling, the substation dispatching system obtains electric vehicle load data and uses the electric vehicle load data to build an electric vehicle charge state model, as shown in the following formula 4:

[0068] Formula 4:

[0069]

[0070]

[0071]

[0072]

[0073]

[0074] in, represents the capacity of the kth electric vehicle at time t, represents the charging and discharging power of the kth electric vehicle at time t, η ev represents the charging and discharging efficiency of electric vehicles, Δt represents the time scale of the day-ahead scheduling, represents the charging and discharging power of the charging station at time t, represents the time that the kth electric vehicle stays at the charging station at time t, represents the charging power of the electric vehicle, represents the discharge power of the electric vehicle, represents the charging and discharging status of the kth electric vehicle at time t, 0 indicates the discharge state. 1 indicates charging status, Indicates the lower limit of battery capacity of electric vehicles, Indicates the upper limit of battery capacity of electric vehicles, represents the target power of the k-th electric vehicle, represents the initial charge of the kth electric vehicle, Indicates the power value at time t when the charging station is not involved in scheduling.

[0075] 203. Establish an energy storage state of charge model.

[0076] In the embodiment of the present application, the substation dispatching system obtains energy storage load data and uses the energy storage load data to construct an energy storage state of charge model, such as the following formula 5:

[0077] Formula 5:

[0078]

[0079]

[0080]

[0081]

[0082] in, represents the energy storage capacity of the jth energy storage device at time t, represents the energy storage charging and discharging power of the jth energy storage device at time t, η esrepresents the energy storage charging and discharging efficiency, Δt represents the time scale of the day-ahead scheduling, represents the charging and discharging power of the energy storage station at time t, Indicates the lower limit of energy storage charging and discharging power, N es Indicates the number of energy storage devices in the energy storage station, Indicates the upper limit of energy storage charging and discharging power, represents the working status of the jth energy storage device at time t, represents the charge and discharge status of the jth energy storage device at time t, 0 indicates the discharge state. 1 indicates charging status, Indicates the lower limit of energy storage battery capacity, Indicates the upper limit of energy storage battery capacity. Indicates the initial capacity state of the energy storage device, Indicates the final capacity state of the energy storage device.

[0083] 204. Use air conditioning load model, electric vehicle state of charge model, and energy storage state of charge model to construct a flexible load model for the substation.

[0084] In this embodiment of the present application, the substation dispatch system considers the differences in response time, response capacity, and response rate of the substation's flexible loads and uses an air conditioning load model, an electric vehicle state of charge model, and an energy storage state of charge model to construct a substation flexible load model. Because the air conditioning load has the characteristics of virtual energy storage and frequent adjustments and adjustments affect user air conditioning usage needs, the air conditioning load is only adjusted during the day-ahead phase. The minimum scheduling period during the day-ahead phase is 1 hour, and the minimum scheduling period during the intraday phase is 15 minutes.

[0085] 205. Obtain transformer load data, photovoltaic output data, and basic load data, and use a robust optimization algorithm to calculate the flexible load model of the substation, transformer load data, photovoltaic output data, and basic load data to obtain a robust model for the day-ahead phase.

[0086] In the embodiment of the present application, in the day-ahead phase, taking into account the uncertainty of distributed photovoltaic output and aiming at minimizing the substation operation cost, the substation dispatching system uses the substation flexible load model, transformer load data, photovoltaic output data, and basic load data to determine the first objective function, as shown in the following formula 6:

[0087] Formula 6:

[0088]

[0089] Among them, minF1 represents the first objective function, represents the charging station control cost at time t, q csrepresents the charging station control cost coefficient, represents the charging and discharging power of the charging station at time t, represents the air conditioning load control cost at time t, q ac represents the air conditioning load control cost coefficient, represents the power of the air conditioning load cluster at time t, represents the photovoltaic operation cost at time t, q pv represents the photovoltaic operation cost coefficient, represents the actual photovoltaic power output at time t, represents the operating cost of the energy storage station at time t, q es represents the energy storage station operation cost coefficient, represents the charging and discharging power of the energy storage station at time t, represents the grid electricity cost at time t, represents the grid electricity price at time t, represents the load power of the transformer in the substation at time t, represents the transformer load penalty cost at time t, q s represents the load rate penalty coefficient, β t represents the transformer load rate at time t, β represents the transformer economic load rate, S T Indicates the rated capacity of the transformer, It represents the power factor after transformer compensation, usually taken as 0.9, and T1 represents the total scheduling period in the day-ahead stage. The objective function integrates the charging station control cost, air conditioning load control cost, photovoltaic operation cost, energy storage station operation cost, grid electricity cost and transformer load rate penalty cost, comprehensively covering all kinds of costs involved in the operation of the substation, with the goal of "minimizing the substation operation cost" to achieve comprehensive optimization scheduling of multiple devices. The objective function drives the control of flexible loads (air conditioning, energy storage, charging stations, etc.), so that the scheduling plan can still maintain economic optimization in the scenario of photovoltaic output fluctuations, and enhance the adaptability of substation operation to the uncertainty of new energy. In addition, this application introduces the transformer load rate penalty item Constrained transformer load factor β t Stay close to the economic load rate β, avoid overload or inefficient operation, ensure equipment safety and improve the economy of the power grid.

[0090] Next, the uncertainty set is determined using photovoltaic output data and base load data, as shown in the following formula 7:

[0091] Formula 7:

[0092] Where Ω is the set of uncertainties u, represents the predicted power of the photovoltaic process at time t, represents the predicted power of the base load at time t, represents the actual power of photovoltaic output at time t, represents the actual power of the base load at time t, represents the fluctuating power of photovoltaic output at time t, represents the fluctuating power of the base load at time t, and is a binary variable, When it is 1, it means the photovoltaic output power fluctuates at time t, Γ t load When it is 1, it means the power of the base load fluctuates at time t. Indicates the lower limit of the power fluctuation of photovoltaic output at time t, Indicates the upper limit of the power fluctuation of photovoltaic output at time t, It represents the lower limit of power fluctuation of base load at time t, represents the upper limit of power fluctuation of base load at time t, Γ pv Indicates the total number of periods of fluctuation of photovoltaic output, Γ load Represents the total number of periods of base load fluctuation. By setting the upper and lower limits of the PV output fluctuation power and the base load fluctuation power, as well as the total number of fluctuation periods, the fluctuation range and time characteristics of the PV output and base load are comprehensively and accurately described, covering extreme scenarios while avoiding over-conservatism, providing a reliable uncertainty boundary for robust optimization. Introducing the binary variable Γ pv , Γ load By characterizing the power fluctuation moments and combining linear constraints to construct the uncertainty set Ω, the model conforms to the mathematical structure of the linear optimization problem, making it easier to solve using optimization methods such as the column constraint generation algorithm, thereby improving computational efficiency and model solvability.

[0093] This application defines the boundaries of the worst-case scenario to ensure that the scheduling plan generated based on it (such as energy storage charging and discharging, flexible load control strategy) can still meet the requirements of power balance, equipment operation constraints, etc. in the photovoltaic and load fluctuation scenarios, thereby improving the reliability and anti-interference ability of the substation operation.

[0094] Then, the first objective function and the uncertainty set are transformed into a compact form as shown in the following formula 8:

[0095] Formula 8:

[0096]

[0097] Among them, x represents the variable:

[0098]

[0099] The variable represented by y is:

[0100]

[0101] The variable represented by u is:

[0102]

[0103] d T represents the relevant constraint coefficient matrix of the y variable, D represents the first coefficient matrix of the corresponding constraint variable, E represents the second coefficient matrix of the corresponding constraint variable, F represents the third coefficient matrix of the corresponding constraint variable, G represents the fourth coefficient matrix of the corresponding constraint variable, H represents the fifth coefficient matrix of the corresponding constraint variable, I represents the sixth coefficient matrix of the corresponding constraint variable, L represents the seventh coefficient matrix of the corresponding constraint variable, b represents the first constant column vector, g represents the second constant column vector, γ represents the first dual variable, λ represents the second dual variable, π represents the third dual variable, χ represents the fourth dual variable, represents the charging and discharging status of the kth electric vehicle at time t, represents the charge and discharge status of the jth energy storage device at time t, represents the power of the air conditioning load cluster at time t, represents the energy storage charging and discharging power of the jth energy storage device at time t, represents the energy storage capacity of the jth energy storage device at time t, represents the charging and discharging power of the kth electric vehicle at time t, represents the capacity of the kth electric vehicle at time t, represents the load power of the transformer in the substation at time t, represents the actual power of photovoltaic output at time t, represents the actual power of the base load at time t. Finally, the first objective function and uncertainty set, converted into a compact form, are used to construct a robust model for the day-ahead phase. By presenting complex constraints within a unified mathematical framework, model redundancy is reduced, the core logic of the optimization problem is clarified, and the constructed robust model for the day-ahead phase can effectively address photovoltaic and load fluctuations, generating a dispatch plan that balances economic efficiency and reliability.

[0104] 206. Construct a real-time scheduling model for the intraday stage based on the flexible load model of the substation area and the robust model of the day-ahead stage.

[0105] In the embodiment of the present application, the daily real-time control stage fixes the air conditioning scheduling plan, the start and stop plan of the electric vehicle and the energy storage equipment in the day-ahead stage, and smoothes the fluctuation of the distributed photovoltaic output in the day stage by regulating the output of the energy storage and electric vehicles.

[0106] Specifically, the substation dispatching system obtains the intraday charging and discharging power data of electric vehicles, the intraday charging and discharging power data of energy storage stations, and the electric vehicle regulation cost coefficient. It obtains the electric vehicle charging and discharging power data in the substation flexible load model and obtains the energy storage station charging and discharging power data in the day-ahead robust model.

[0107] Next, the second objective function is calculated using the intra-day electric vehicle charging and discharging power data, the intra-day energy storage station charging and discharging power data, the electric vehicle regulation cost coefficient, the electric vehicle charging and discharging power data, and the energy storage station charging and discharging power data, as shown in the following formula 9:

[0108] Formula 9:

[0109] Among them, minF2 represents the second objective function, q ev represents the electric vehicle regulation cost coefficient, represents the charging and discharging power of the electric vehicle at time t within the day, represents the charging and discharging power of the electric vehicle at time t, represents the charging and discharging power of the energy storage station at time t within the day, represents the charging and discharging power of the energy storage station at time t, and T2 represents the total scheduling period during the intraday phase. Finally, the second objective function is used to construct a real-time scheduling model for the intraday phase. This model aims to achieve economic optimization to smooth photovoltaic fluctuations, achieving power balance while precisely controlling intraday regulation costs.

[0110] 207. Establish transformer load rate constraints.

[0111] In the embodiment of the present application, the substation dispatching system establishes a transformer load rate constraint, such as the following formula 10:

[0112] Formula 10: β min ≤β t ≤β max ,

[0113] Among them, β t represents the transformer load rate at time t, β min Indicates the lower limit of transformer load rate, β max Indicates the upper limit of transformer load rate.

[0114] 208. Establish power balance constraints.

[0115] In the embodiment of the present application, the substation dispatching system establishes a power balance constraint, such as the following formula 11:

[0116] Formula 11:

[0117] in, represents the grid electricity price at time t, represents the actual power of the base load at time t, represents the charging and discharging power of the charging station at time t, represents the power of the air conditioning load cluster at time t, represents the actual photovoltaic power output at time t, Represents the charging and discharging power of the energy storage station at time t.

[0118] 209. A multi-time-scale two-stage robust optimization model is constructed by using the day-ahead robust model, the intraday real-time dispatch model, the transformer load factor constraint, and the power balance constraint.

[0119] In the embodiment of the present application, the substation dispatching system takes into account the uncertainty of distributed photovoltaic output and load power consumption in the low-voltage substation, and the present invention adopts a robust optimization method to establish an optimization model for low-voltage substation day-ahead dispatching-intraday real-time regulation. In the worst scenario, the day-ahead dispatching stage fully considers the operating flexibility of equipment such as air-conditioning loads, electric vehicles, and energy storage, and obtains a 24-hour output plan for each controllable load in the substation system. In the intraday stage, the power dispatching plan for energy storage and charging stations in the intraday stage is formulated on the basis of complying with the day-ahead plan. In this way, by fixing the air-conditioning dispatching plan, electric vehicle and energy storage start-stop plan in the day-ahead stage, focusing on the output regulation of energy storage and electric vehicles in the intraday stage, a coordinated mechanism of day-ahead strategy determination and intraday deviation adjustment is formed to avoid repeated optimization, improve dispatching efficiency, and at the same time ensure the continuity of the day-ahead plan and the pertinence of the intraday adjustment.

[0120] 210. Obtain a column constraint generation algorithm, and use the column constraint generation algorithm to solve the day-ahead phase of the multi-time-scale two-stage robust optimization model to obtain the day-ahead scheduling plan for air-conditioning load, energy storage station, and charging station.

[0121] In the embodiment of the present application, the column-and-constraint generation (C&CG) algorithm is used to solve the problem in the day-ahead phase. Specifically, the substation dispatch system obtains the day-ahead robust model corresponding to the day-ahead phase of the multi-time scale two-stage robust optimization model. The day-ahead robust model is a min-max-min two-stage optimization problem. First, the two-stage robust optimization objective function is decoupled to obtain the master problem (MP). The master problem is as shown in the following formula 12:

[0122] Formula 12: minη

[0123]

[0124] The decomposed sub-problem (Slave Problem, SP) is as follows:

[0125] Formula 13:

[0126] Among them, η represents the objective function variable of the main problem, d T represents y k The relevant constraint coefficient matrix of the variable, D represents the first coefficient matrix corresponding to the constraint variable, E represents the second coefficient matrix corresponding to the constraint variable, F represents the third coefficient matrix corresponding to the constraint variable, G represents the fourth coefficient matrix corresponding to the constraint variable, H represents the fifth coefficient matrix corresponding to the constraint variable, I represents the sixth coefficient matrix corresponding to the constraint variable, L represents the seventh coefficient matrix corresponding to the constraint variable, b represents the first constant column vector, g represents the second constant column vector, y k represents the y variable in the kth iteration, u k represents the u variable in the kth iteration, u represents the uncertainty parameter, and Ω represents the uncertainty set.

[0127] Next, the first objective function and uncertainty set converted to a compact form are obtained in the robust model of the day-ahead phase, and the strong duality theory is obtained. Based on the strong duality theory, the main problem, subproblems, the first objective function converted to a compact form, and the uncertainty set are used for calculation. These are converted to a max form and merged with the outer max problem to obtain the dual problem, as shown in the following formula 14:

[0128] Formula 14:

[0129] Where γ represents the first dual variable, λ represents the second dual variable, π represents the third dual variable, and χ represents the fourth dual variable. In Formula 14, there is a bilinear term u T λ, u corresponding to the optimal solution of the dual problem * is a point in the uncertainty set Ω. That is, when Equation 13 reaches its maximum value, the value of the uncertain variable u should be the boundary of the fluctuation range described by Equation 14. A column constraint generation algorithm is then used to iteratively calculate the primary and dual problems, resulting in the day-ahead scheduling plan for air conditioning load, energy storage stations, and charging stations.

[0130] In an optional implementation scheme, when the fluctuating power of the photovoltaic output is equal to the upper limit of the photovoltaic output power fluctuation and the fluctuating power of the base load is equal to the upper limit of the base load power fluctuation, it is the worst scenario. The substation dispatching system determines a new uncertainty set, as shown in the following formula 15:

[0131] Formula 15:

[0132] Among them, Ω is the set of new uncertainties u, represents the predicted power of photovoltaic processing at time t, P tload,pre represents the predicted power of the base load at time t, express, represents the actual power of the base load at time t, Γ t pv and Γ t load is a binary variable, Γ t pv When it is 1, it means the photovoltaic output power fluctuates at time t, Γ t load When it is 1, it means the power of the base load fluctuates at time t. Indicates the upper limit of the power fluctuation of photovoltaic output at time t, ΔP t load,max represents the upper limit of power fluctuation of base load at time t, Γ pv Indicates the total number of periods of fluctuation of photovoltaic output, Γ load represents the total number of periods of base load fluctuation. Then, it is linearized by introducing auxiliary variables and related constraints, that is, obtaining continuous auxiliary variables. The dual problem is linearized using continuous auxiliary variables to obtain the linearized formula, as shown in the following formula 16:

[0133] Formula 16:

[0134] Among them, Z′ represents a continuous auxiliary variable, Indicates the auxiliary adjustment amount of photovoltaic output, Indicates the base load auxiliary adjustment amount, represents the upper bound of the second dual variable and is a sufficiently large positive number. Finally, the column constraint generation algorithm is used to iteratively calculate the main problem and the dual problem to obtain the day-ahead air conditioning load, energy storage station and charging station scheduling plan, ensuring the robustness of the scheduling plan under photovoltaic and load fluctuation scenarios, and achieving a balance between economy and reliability. Figure 2B As shown, first initialize and set the initial worst scenario of the uncertain variable u, define the lower bound LB = -∞, the upper bound UB = +∞, the number of iterations k = 1 and the convergence deviation ε. After entering the loop, based on the current worst scenario u k* Solve the main problem and get the optimal solution (x k* ,η k* ,y k* ) and update the lower bound LB = η k* ; Then solve the main problem x k* Substitute into formula 16 to solve the subproblem and obtain the subproblem objective function value SP(y k* ) and the new worst-case scenario (k+1)* , update the upper bound UB=min(UB,c T x k* +sP(yk* )). Then, determine whether UB - LB ≤ ε. If not, set k = k + 1 and continue the iteration. If so, end the process and output the final scheduling plan. By iteratively solving the main and subproblems, gradually narrow the gap between the upper and lower bounds until the convergence condition is met, completing the optimization.

[0135] Based on the day-ahead scheduling plan for air conditioning load, energy storage stations, and charging stations, a solver is used to optimize and solve the intraday phase of the multi-time-scale two-stage robust optimization model to obtain the real-time charging and discharging power scheduling plan for energy storage equipment and electric vehicles.

[0136] In this embodiment, air conditioning load is no longer regulated during the day. Instead, the power of energy storage and charging stations is optimized and dispatched to smooth out intraday fluctuations in photovoltaic and load power. The substation dispatch system uses the GUROBI solver for optimization and solves, generating real-time charging and discharging power dispatch plans for energy storage devices and electric vehicles.

[0137] In summary, the embodiment of the present application proposes a multi-time scale flexible load robust optimization framework, such as Figure 2C As shown in the figure, first, the base load, energy storage parameters, air conditioning load, photovoltaic forecast data, and charging station load are used as the input data for the research; then, the two-stage robust optimization is carried out on the day-ahead, with a time scale of 1 hour, to optimize the photovoltaic output, base load, charging station, etc., with the goal of minimizing the operating cost of the substation, and outputting the energy storage output plan, charging station scheduling plan, and air conditioning load scheduling plan; finally, the intraday rolling optimization scheduling is carried out on a time scale of 15 minutes, based on the intraday short-term forecast of photovoltaic output, base load, and charging station data, with the goal of minimizing the flexible load control cost, and outputting the energy storage adjustment amount and the charging station adjustment amount. Through the "day-ahead + intraday" multi-time scale collaboration, the comprehensive optimization scheduling of the substation is achieved.

[0138] The effectiveness of this application is verified by a specific simulation case. The simulation case is based on the Matlab platform and is used to verify the flexible load robust optimization method of the application in a mixed office and residential area. The transformer capacity is 3150KVA, the maximum deviation of photovoltaic and load power fluctuations is 15% and 10% of the predicted value, and the uncertainty parameter Γ t pv , Γ t load Take 6 and 12 respectively, air conditioning load parameters: δ is 5.31, C b 5.56℃ / kW, R b is 0.18kWh / ℃, ΔT is 3℃, N AC is 1000, power is 1.1kW; energy storage parameters: 250kW, 200, -250kW, =1000, η ch is 0.95, capacity is 1000kWh, η dis 0.95, N ES The initial data of electric vehicles was extracted using Monte Carlo method, with a total of 100 electric vehicle data in residential and office areas. The electric vehicle parameters are: 60kW, N(0.4, 0.1), -60kW, N ev =100, η ch is 0.95, the arrival time in residential area is N(19.21, 1.31), η dis is 0.95, the time of leaving the residential area is N (7.24, 1.40), is 24, the arrival time at the office is N(8.34, 0.67), =120, the time of leaving the office is N(17.53, 0.785); the time-of-use electricity price of the power grid: the peak period is 10:00-12:00 and 15:00-21:00, the electricity purchase price is 1.3, the normal period is 7:00-10:00, 12:00-15:00, 21:00-00:00, the electricity purchase price is 0.75, the valley period is 0:00-7:00, the electricity purchase price is 0.4; the photovoltaic power curve is as follows Figure 2D The shaded area is the photovoltaic predicted power range, showing the predicted fluctuation range. The solid line represents the photovoltaic predicted power, reflecting the expected output. The line with an asterisk is the photovoltaic worst-case scenario power, which is used to simulate the power value under extreme conditions. The line with a square is the photovoltaic actual power. The basic load power curve is as follows: Figure 2E The gray area is the load forecast power range, reflecting the predicted power fluctuation range. The solid line is the load forecast power, showing the expected load level. The line with asterisks is the load worst-case scenario power, simulating the load power under extreme conditions. The line with squares represents the actual load power. The flexible load robust optimization method of the application is used for calculation, and the optimization results of the controllable load scheduling of each area are obtained as follows: Figure 2F 、 2G , 2H, where Figure 2F is the energy storage charging and discharging power, energy storage charging is positive and discharging is negative; Figure 2G The gray bar graph shows the output of the air conditioner in each time period. The dotted curve represents the outdoor temperature, which shows a trend of first rising and then falling. The asterisked curve represents the indoor temperature, and its fluctuation is controlled within the range of the solid line (upper temperature limit) and the dotted line (lower temperature limit) in the figure. Figure 2HIt represents the charging and discharging power and the stay period of electric vehicles at the charging station. The height and color changes of the surface show the charging and discharging power of electric vehicles with different numbers in each period. The depth of color corresponds to the power value (such as the color scale on the right, covering the range of -57.00kW to 57.00kW). The office area and residential area are marked to further distinguish the regional characteristics, and intuitively display the dynamic distribution, time period differences and regional characteristics of the charging and discharging power of multiple electric vehicles at the charging station throughout the day.

[0139] To verify the effectiveness of the robust optimization method for day-ahead scheduling proposed in this application, three scenarios are set up and the day-ahead economic scheduling results are analyzed:

[0140] Scenario 1: Electric vehicle loads in the substation are charging disorderly, and other loads are not involved in regulation;

[0141] Scenario 2: Economic dispatch is performed based on the controllable load dispatch cost in the substation, without considering the uncertainty of load power and PV output, and without considering the economic load factor of the transformer;

[0142] Scenario 3: Considering the uncertainty of the substation load and photovoltaic output as well as the economic load rate of the transformer, a robust optimization method is used to economically dispatch the controllable load in the substation.

[0143] Figure 2I Indicates the transformer load rate under different scenarios. The three curves of scenario 1, scenario 2, and scenario 3 show the dynamic changes of the transformer load rate under different scenarios, and mark the transformer light load threshold (dashed line) and overload threshold (solid line). The load rate in scenario 1 fluctuates significantly, and approaches or exceeds the overload threshold in some periods; the load rate in scenario 2 also shows high fluctuations in specific periods; the load rate in scenario 3 is relatively stable, and is generally maintained near the light load threshold. The operation analysis results of each scenario substation are as follows: Figure 2J As shown in the figure, in scenario 1, electric vehicles are charged and discharged in an disorderly manner, and the controllable load is not regulated, resulting in a transformer overload time of 4 hours and a transformer light load time of 15 hours; compared with scenario 1, scenario 2 considers the economic scheduling of controllable loads such as electric vehicles, energy storage, and air-conditioning equipment in the substation, and the operating cost is reduced by 28.7%. The transformer light load time is reduced by 4 hours, and the overload time is reduced by 4 hours, but the transformer heavy load time is increased by 6 hours; scenario 3 takes into account the uncertainty of the substation load and photovoltaic output as well as the economic load rate of the transformer. The operating cost increases by 19.4% compared with scenario 2, but the peak-to-valley difference is reduced by 75%. The transformer light load, heavy load, and overload time are all 0 hours, and it operates within the economic load rate range for 20 periods, effectively improving the operating life of the transformer.

[0144] from Figure 2KIt can be seen that when the uncertainty parameters are all set to 0, the two-stage robust optimization model is equivalent to the deterministic optimization model. As the uncertainty parameters increase, the substation operation cost also increases accordingly, but the resulting solution also becomes more conservative and has a stronger ability to resist intraday uncertainty risks.

[0145] Figure 2L The charging and discharging power of the two energy storage devices in the day-ahead and day-intraday stages are displayed respectively: the positive value on the vertical axis represents the discharging power, and the negative value represents the charging power, which intuitively presents the charging and discharging dynamics of the energy storage devices at different times. Figure 2M The charging and discharging power changes of electric vehicle No. 1 during the day and intraday stages are displayed, and the charging and discharging strategies and time distribution characteristics of the electric vehicle at different stages are clearly compared. Figure 2L and Figure 2M It can be seen that in order to balance the real-time fluctuations of photovoltaic output, the energy storage and electric vehicle charging and discharging power are adjusted in real time during the day. Figure 2N The operation cost comparison of substations under different time scales is analyzed. The multi-time scale optimization reduces the operating cost by 7% compared with the single time scale, which fully demonstrates that the multi-time scale can achieve multi-level utilization of controllable loads and reduce operating costs.

[0146] The embodiment of the present application provides a method for robust optimization of flexible loads in a substation. Compared with the prior art, the embodiment of the present invention establishes an air conditioning load model, an electric vehicle state of charge model, and an energy storage state of charge model, and uses the air conditioning load model, the electric vehicle state of charge model, and the energy storage state of charge model to construct a substation flexible load model, obtains a robust optimization algorithm, and based on the robust optimization algorithm, uses the substation flexible load model to establish a multi-time scale two-stage robust optimization model, obtains a column constraint generation algorithm, and uses the column constraint generation algorithm to solve the day-ahead stage of the multi-time scale two-stage robust optimization model to obtain the air conditioning load, energy storage station, and charging station scheduling plan of the day-ahead stage. Based on the air conditioning load, energy storage station, and charging station scheduling plan of the day-ahead stage, a solver is used to optimize and solve the intraday stage of the multi-time scale two-stage robust optimization model to obtain the real-time charging and discharging power scheduling plan of the energy storage equipment and electric vehicles. The present application proposes a multi-time scale two-stage robust optimization model to achieve tiered utilization of flexible loads and improve the utilization rate of distribution transformer capacity. The multi-timescale two-stage robust optimization model fully considers the uncertainty of distributed photovoltaics, meeting energy demand while effectively responding to distributed photovoltaic power fluctuations. Compared with deterministic optimization methods, the optimal dispatch plan for flexible loads in the substation area of this application has stronger robustness. In addition, the multi-timescale two-stage robust optimization model takes into account the optimization of economic load rate. While taking into account the needs of photovoltaic consumption and electric vehicle charging, it effectively reduces the light load, heavy load and overload time of the transformer, thereby improving the operating life of the transformer and ensuring the safety and stability of the substation distribution network.

[0147] Further, as Figure 1 The specific implementation of the method, the embodiment of the present application provides a flexible load robust optimization device for the area, such as Figure 3 As shown, the device includes: a first modeling module 301, a second modeling module 302, a first solution module 303 and a second solution module 304.

[0148] The first modeling module 301 is used to establish an air conditioning load model, an electric vehicle state of charge model, and an energy storage state of charge model, and use the air conditioning load model, the electric vehicle state of charge model, and the energy storage state of charge model to construct a flexible load model of the substation;

[0149] The second modeling module 302 is used to obtain a robust optimization algorithm, and based on the robust optimization algorithm, use the substation flexible load model to establish a multi-time scale two-stage robust optimization model;

[0150] A first solving module 303 is configured to obtain a column constraint generation algorithm, and employ the column constraint generation algorithm to solve the day-ahead phase of the multi-time-scale two-stage robust optimization model to obtain a scheduling plan for air conditioning load, energy storage stations, and charging stations in the day-ahead phase;

[0151] The second solving module 304 is used to optimize and solve the intraday phase of the multi-time-scale two-stage robust optimization model based on the air conditioning load, energy storage station and charging station scheduling plan of the day-ahead phase, thereby obtaining a real-time charging and discharging power scheduling plan for the energy storage equipment and electric vehicles.

[0152] In a specific application scenario, the first modeling module 301 is used to obtain a first-order thermodynamic equivalent model and air conditioning load data, and use the first-order thermodynamic equivalent model to model the air conditioning load data to obtain the air conditioning load model.

[0153]

[0154] in, represents the indoor temperature of the area where the air conditioning load is located at time t+1, Indicates the outdoor temperature of the area where the air conditioning load is located at time t+1, represents the cooling capacity of the air conditioning load at time t, R b Indicates the thermal resistance of the area where the air conditioning load is located, C b represents the heat capacity of the area where the air conditioning load is located, Δt represents the time scale of the day-ahead scheduling, Indicates the temperature setting value of the air conditioning load, ΔT indicates the maximum indoor temperature offset, P t ac,s represents the power of the air conditioning load at time t, δ represents the air conditioning energy efficiency ratio, represents the power of the air conditioning load cluster at time t, N ac Indicates the air conditioning load quantity, Indicates the lower limit of air conditioning load power adjustment. Indicates the upper limit of air conditioning load power adjustment, P t ac,set represents the power setting value at time t when the air conditioning load does not participate in the scheduling; obtain the electric vehicle load data, and use the electric vehicle load data to build the electric vehicle charge state model,

[0155]

[0156] in, represents the capacity of the kth electric vehicle at time t, represents the charging and discharging power of the kth electric vehicle at time t, η ev represents the charging and discharging efficiency of electric vehicles, Δt represents the time scale of the day-ahead scheduling, represents the charging and discharging power of the charging station at time t, represents the time that the kth electric vehicle stays at the charging station at time t, represents the charging power of the electric vehicle, represents the discharge power of the electric vehicle, represents the charging and discharging status of the kth electric vehicle at time t, 0 indicates the discharge state. 1 indicates charging status, Indicates the lower limit of battery capacity of electric vehicles, Indicates the upper limit of battery capacity of electric vehicles, represents the target power of the k-th electric vehicle, represents the initial charge of the kth electric vehicle, P t cs,set represents the power value at time t when the charging station does not participate in scheduling; obtain energy storage load data, and use the energy storage load data to build the energy storage charge state model,

[0157]

[0158] in, represents the energy storage capacity of the jth energy storage device at time t, represents the energy storage charging and discharging power of the jth energy storage device at time t, η es represents the energy storage charging and discharging efficiency, Δt represents the time scale of the day-ahead scheduling, represents the charging and discharging power of the energy storage station at time t, Indicates the lower limit of energy storage charging and discharging power, N es Indicates the number of energy storage devices in the energy storage station, Indicates the upper limit of energy storage charging and discharging power, represents the working status of the jth energy storage device at time t, represents the charge and discharge status of the jth energy storage device at time t, 0 indicates the discharge state. 1 indicates charging status, Indicates the lower limit of energy storage battery capacity, Indicates the upper limit of energy storage battery capacity. Indicates the initial capacity state of the energy storage device, Indicates the final capacity state of the energy storage device.

[0159] In a specific application scenario, the second modeling module 302 is used to obtain transformer load data, photovoltaic output data and basic load data, and use the robust optimization algorithm to calculate the flexible load model of the substation, the transformer load data, the photovoltaic output data and the basic load data to obtain a robust model for the day-ahead phase; based on the flexible load model of the substation and the robust model for the day-ahead phase, a real-time scheduling model for the intraday phase is constructed; and a transformer load rate constraint is established.

[0160] β min ≤β t ≤β max ,

[0161] Among them, β t represents the transformer load rate at time t, β min Indicates the lower limit of transformer load rate, β max Indicates the upper limit of transformer load rate; establishes power balance constraint,

[0162]

[0163] in, represents the grid electricity price at time t, represents the actual power of the base load at time t, represents the charging and discharging power of the charging station at time t, represents the power of the air conditioning load cluster at time t, represents the actual photovoltaic power output at time t, represents the charging and discharging power of the energy storage station at time t; the multi-time-scale two-stage robust optimization model is constructed by using the day-ahead robust model, the intraday real-time scheduling model, the transformer load rate constraint, and the power balance constraint.

[0164] In a specific application scenario, the second modeling module 302 is used to determine a first objective function using the flexible load model of the substation, the transformer load data, the photovoltaic output data, and the basic load data.

[0165]

[0166] Among them, minF1 represents the first objective function, represents the charging station control cost at time t, q cs represents the charging station control cost coefficient, represents the charging and discharging power of the charging station at time t, represents the air conditioning load control cost at time t, q ac represents the air conditioning load control cost coefficient, represents the power of the air conditioning load cluster at time t, represents the photovoltaic operation cost at time t, q pv represents the photovoltaic operation cost coefficient, represents the actual photovoltaic power output at time t, represents the operating cost of the energy storage station at time t, q es represents the energy storage station operation cost coefficient, represents the charging and discharging power of the energy storage station at time t, represents the grid electricity cost at time t, represents the grid electricity price at time t, represents the load power of the transformer in the substation at time t, represents the transformer load penalty cost at time t, q s represents the load rate penalty coefficient, β t represents the transformer load rate at time t, β represents the transformer economic load rate, S T Indicates the rated capacity of the transformer, represents the power factor after transformer compensation, T1 represents the total scheduling period of the day-ahead stage; the uncertainty set is determined using the photovoltaic output data and the basic load data,

[0167]

[0168] Where Ω is the set of uncertainties u, represents the predicted power of the photovoltaic process at time t, represents the predicted power of the base load at time t, represents the actual power of photovoltaic output at time t, represents the actual power of the base load at time t, represents the fluctuating power of photovoltaic output at time t, represents the fluctuating power of the base load at time t, and Γ t load is a binary variable, Γ t pvWhen it is 1, it means the photovoltaic output power fluctuates at time t, Γ t load When it is 1, it means the power of the base load fluctuates at time t. Indicates the lower limit of the power fluctuation of photovoltaic output at time t, Indicates the upper limit of the power fluctuation of photovoltaic output at time t, ΔP t load,min Indicates the lower limit of power fluctuation of base load at time t, ΔP t load,max represents the upper limit of power fluctuation of base load at time t, Γ pv Indicates the total number of periods of fluctuation of photovoltaic output, Γ load represents the total number of fluctuation periods of the base load; converting the first objective function and the uncertainty set into a compact form,

[0169]

[0170] Among them, x represents the variable:

[0171]

[0172] The variable represented by y is:

[0173]

[0174] The variable represented by u is:

[0175]

[0176] d T represents the relevant constraint coefficient matrix of the y variable, D represents the first coefficient matrix of the corresponding constraint variable, E represents the second coefficient matrix of the corresponding constraint variable, F represents the third coefficient matrix of the corresponding constraint variable, G represents the fourth coefficient matrix of the corresponding constraint variable, H represents the fifth coefficient matrix of the corresponding constraint variable, I represents the sixth coefficient matrix of the corresponding constraint variable, L represents the seventh coefficient matrix of the corresponding constraint variable, b represents the first constant column vector, g represents the second constant column vector, γ represents the first dual variable, λ represents the second dual variable, π represents the third dual variable, χ represents the fourth dual variable, represents the charging and discharging status of the kth electric vehicle at time t, represents the charge and discharge status of the jth energy storage device at time t, represents the power of the air conditioning load cluster at time t, represents the energy storage charging and discharging power of the jth energy storage device at time t, represents the energy storage capacity of the jth energy storage device at time t, represents the charging and discharging power of the kth electric vehicle at time t, represents the capacity of the kth electric vehicle at time t, represents the load power of the transformer in the substation at time t, represents the actual power of photovoltaic output at time t, represents the actual power of the base load at time t; and constructs the day-ahead robust model using the first objective function converted into a compact form and the uncertainty set.

[0177] In a specific application scenario, the second modeling module 302 is used to obtain the intraday phase electric vehicle charging and discharging power data, the intraday phase energy storage station charging and discharging power data, and the electric vehicle regulation cost coefficient, obtain the electric vehicle charging and discharging power data in the flexible load model of the substation, and obtain the energy storage station charging and discharging power data in the day-ahead robust model; use the intraday phase electric vehicle charging and discharging power data, the intraday phase energy storage station charging and discharging power data, the electric vehicle regulation cost coefficient, the electric vehicle charging and discharging power data, and the energy storage station charging and discharging power data to calculate the second objective function.

[0178]

[0179] Among them, minF2 represents the second objective function, q ev represents the electric vehicle regulation cost coefficient, P t ev,r represents the charging and discharging power of the electric vehicle at time t within the day, represents the charging and discharging power of the electric vehicle at time t, P t es,r represents the charging and discharging power of the energy storage station at time t within the day, represents the charging and discharging power of the energy storage station at time t, and T2 represents the total scheduling period in the intra-day stage; the second objective function is used to construct the real-time scheduling model in the intra-day stage.

[0180] In a specific application scenario, the first solution module 303 is used to obtain a day-ahead robust model corresponding to the day-ahead phase of the multi-time-scale two-stage robust optimization model; decompose the day-ahead robust model into a main problem and sub-problems, where the main problem is:

[0181]

[0182] The sub-problems are:

[0183]

[0184] Among them, η represents the objective function variable of the main problem, d T represents y kThe relevant constraint coefficient matrix of the variable, D represents the first coefficient matrix corresponding to the constraint variable, E represents the second coefficient matrix corresponding to the constraint variable, F represents the third coefficient matrix corresponding to the constraint variable, G represents the fourth coefficient matrix corresponding to the constraint variable, H represents the fifth coefficient matrix corresponding to the constraint variable, I represents the sixth coefficient matrix corresponding to the constraint variable, L represents the seventh coefficient matrix corresponding to the constraint variable, b represents the first constant column vector, g represents the second constant column vector, y k represents the y variable in the kth iteration, u k represents the u variable in the k-th iteration, u represents the uncertainty parameter, and Ω represents the uncertainty set; obtains the first objective function and uncertainty set converted into a compact form in the day-ahead robust model; obtains a strong duality theory, and based on the strong duality theory, uses the main problem, the subproblem, the first objective function converted into a compact form, and the uncertainty set to perform calculations to obtain the dual problem,

[0185]

[0186] Wherein, γ represents the first dual variable, λ represents the second dual variable, π represents the third dual variable, and X represents the fourth dual variable. The column constraint generation algorithm is used to iteratively calculate the main problem and the dual problem to obtain the air conditioning load, energy storage station, and charging station scheduling plan for the day-ahead period.

[0187] In a specific application scenario, the first solution module 303 is used to determine a new uncertainty set when the fluctuating power of the photovoltaic output is the upper limit of the photovoltaic output power fluctuation and the fluctuating power of the base load is the upper limit of the base load power fluctuation.

[0188]

[0189] Among them, Ω is the set of new uncertainties u, represents the predicted power of photovoltaic processing at time t, P t load,pre represents the predicted power of the base load at time t, express, represents the actual power of the base load at time t, Γ t pv and Γ t load is a binary variable, Γ t pv When it is 1, it means the photovoltaic output power fluctuates at time t, Γ t load When it is 1, it means the power of the base load fluctuates at time t. Indicates the upper limit of the power fluctuation of photovoltaic output at time t, ΔP tload,max represents the upper limit of power fluctuation of base load at time t, Γ pv Indicates the total number of periods of fluctuation of photovoltaic output, Γ load represents the total number of fluctuation periods of the base load; obtaining continuous auxiliary variables, using the continuous auxiliary variables to linearize the dual problem, and obtaining a linearized formula,

[0190]

[0191] Among them, Z′ represents a continuous auxiliary variable, Indicates the auxiliary adjustment amount of photovoltaic output, Indicates the base load auxiliary adjustment amount, represents the upper bound of the second dual variable; and iteratively calculating the main problem and the dual problem using the column constraint generation algorithm to obtain a scheduling plan for the air-conditioning load, energy storage station, and charging station in the day-ahead phase.

[0192] The embodiment of the present application provides a robust optimization device for flexible loads in a substation. Compared with the prior art, the embodiment of the present invention establishes an air conditioning load model, an electric vehicle state of charge model, and an energy storage state of charge model, and uses the air conditioning load model, the electric vehicle state of charge model, and the energy storage state of charge model to construct a substation flexible load model, obtains a robust optimization algorithm, and based on the robust optimization algorithm, uses the substation flexible load model to establish a multi-time scale two-stage robust optimization model, obtains a column constraint generation algorithm, and uses the column constraint generation algorithm to solve the day-ahead stage of the multi-time scale two-stage robust optimization model to obtain the air conditioning load, energy storage station, and charging station scheduling plan of the day-ahead stage. Based on the air conditioning load, energy storage station, and charging station scheduling plan of the day-ahead stage, a solver is used to optimize and solve the intraday stage of the multi-time scale two-stage robust optimization model to obtain the real-time charging and discharging power scheduling plan of the energy storage equipment and electric vehicles. The present application proposes a multi-time scale two-stage robust optimization model to achieve tiered utilization of flexible loads and improve the utilization rate of distribution transformer capacity. The multi-timescale two-stage robust optimization model fully considers the uncertainty of distributed photovoltaics, meeting energy demand while effectively responding to distributed photovoltaic power fluctuations. Compared with deterministic optimization methods, the optimal dispatch plan for flexible loads in the substation area of this application has stronger robustness. In addition, the multi-timescale two-stage robust optimization model takes into account the optimization of economic load rate. While taking into account the needs of photovoltaic consumption and electric vehicle charging, it effectively reduces the light load, heavy load and overload time of the transformer, thereby improving the operating life of the transformer and ensuring the safety and stability of the substation distribution network.

[0193] It should be noted that for other corresponding descriptions of the functional units involved in the flexible load robust optimization device for a substation provided in the embodiment of the present application, please refer to Figure 1 and Figure 2AThe corresponding description in will not be repeated here.

[0194] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, stored data, displayed data, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties.

[0195] The technical features of the above embodiments can be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0196] The above-described embodiments merely represent several implementation methods of the present application. While the descriptions are relatively specific and detailed, they should not be construed as limiting the scope of the present application. It should be noted that a person of ordinary skill in the art may make various modifications and improvements without departing from the spirit of the present application, and these modifications and improvements fall within the scope of protection of the present application. Therefore, the scope of protection of the present application shall be determined by the appended claims.

[0197] In an exemplary embodiment, see Figure 4 Also provided is an electronic device comprising a bus, a processor, a memory, and a communication interface. The electronic device may also include an input / output interface and a display device, wherein the various functional units can communicate with each other via the bus. The memory stores a computer program, and the processor is configured to execute the program stored in the memory and implement the robust optimization method for flexible loads in a substation area in the above-described embodiment.

[0198] A storage medium stores a computer program, which, when executed by a processor, implements the steps of the robust optimization method for flexible loads in a substation.

[0199] Through the description of the above implementation methods, those skilled in the art can clearly understand that the present application can be implemented through hardware or by means of software plus a necessary general hardware platform. Based on this understanding, the technical solution of the present application can be embodied in the form of a software product, which can be stored in a non-volatile storage medium (which can be a CD-ROM, a USB flash drive, a mobile hard disk, etc.), including a number of instructions for enabling a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in each implementation scenario of the present application.

[0200] Those skilled in the art will understand that the accompanying drawings are only schematic diagrams of a preferred implementation scenario, and the modules or processes in the accompanying drawings are not necessarily required to implement the present application.

[0201] Those skilled in the art will appreciate that the modules in the devices in the implementation scenario can be distributed in the devices of the implementation scenario according to the implementation scenario description, or can be modified accordingly and located in one or more devices different from the implementation scenario. The modules in the above implementation scenario can be combined into one module or further split into multiple submodules.

[0202] The above application serial numbers are for description only and do not represent the advantages or disadvantages of the implementation scenarios.

[0203] The above disclosure only describes several specific implementation scenarios of the present application. However, the present application is not limited thereto, and any changes that can be conceived by those skilled in the art should fall within the scope of protection of the present application.

Claims

1. A robust optimization method for flexible load in a substation, characterized by: include: Establishing an air conditioning load model, an electric vehicle state of charge model, and an energy storage state of charge model, and using the air conditioning load model, the electric vehicle state of charge model, and the energy storage state of charge model to construct a flexible load model for the substation; Obtaining a robust optimization algorithm, and establishing a multi-time-scale two-stage robust optimization model based on the robust optimization algorithm and the substation flexible load model; Obtaining a column constraint generation algorithm, and using the column constraint generation algorithm to solve the day-ahead phase of the multi-time-scale two-stage robust optimization model to obtain a scheduling plan for air conditioning loads, energy storage stations, and charging stations in the day-ahead phase; Based on the air conditioning load, energy storage station and charging station scheduling plan in the day-ahead stage, a solver is used to optimize and solve the intraday stage of the multi-time-scale two-stage robust optimization model to obtain the real-time charging and discharging power scheduling plan of the energy storage equipment and electric vehicles.

2. The method according to claim 1, characterized in that The establishment of the air conditioning load model, the electric vehicle state of charge model, and the energy storage state of charge model includes: Obtaining a first-order thermodynamic equivalent model and air conditioning load data, and using the first-order thermodynamic equivalent model to model the air conditioning load data to obtain the air conditioning load model, in, represents the indoor temperature of the area where the air conditioning load is located at time t+1, Indicates the outdoor temperature of the area where the air conditioning load is located at time t+1, represents the cooling capacity of the air conditioning load at time t, R b Indicates the thermal resistance of the area where the air conditioning load is located, C b represents the heat capacity of the area where the air conditioning load is located, Δt represents the time scale of the day-ahead scheduling, Indicates the temperature setting value of the air conditioning load, ΔT indicates the maximum indoor temperature offset, represents the power of the air conditioning load at time t, δ represents the air conditioning energy efficiency ratio, represents the power of the air conditioning load cluster at time t, N ac Indicates the air conditioning load quantity, Indicates the lower limit of air conditioning load power adjustment. Indicates the upper limit of air conditioning load power adjustment. Indicates the power setting value at time t when the air conditioning load is not involved in scheduling; Acquire electric vehicle load data, and use the electric vehicle load data to construct the electric vehicle state of charge model, in, represents the capacity of the kth electric vehicle at time t, represents the charging and discharging power of the kth electric vehicle at time t, η ev represents the charging and discharging efficiency of electric vehicles, Δt represents the time scale of the day-ahead scheduling, represents the charging and discharging power of the charging station at time t, represents the time that the kth electric vehicle stays at the charging station at time t, represents the charging power of the electric vehicle, represents the discharge power of the electric vehicle, represents the charging and discharging status of the kth electric vehicle at time t, 0 indicates the discharge state. 1 indicates charging status, Indicates the lower limit of battery capacity of electric vehicles, Indicates the upper limit of battery capacity of electric vehicles, represents the target power of the k-th electric vehicle, represents the initial charge of the kth electric vehicle, Indicates the power value at time t when the charging station does not participate in scheduling; Acquire energy storage load data, and use the energy storage load data to construct the energy storage state of charge model, in, represents the energy storage capacity of the jth energy storage device at time t, represents the energy storage charging and discharging power of the jth energy storage device at time t, η es represents the energy storage charging and discharging efficiency, Δt represents the time scale of the day-ahead scheduling, represents the charging and discharging power of the energy storage station at time t, Indicates the lower limit of energy storage charging and discharging power, N es Indicates the number of energy storage devices in the energy storage station, Indicates the upper limit of energy storage charging and discharging power, represents the working status of the jth energy storage device at time t, represents the charge and discharge status of the jth energy storage device at time t, 0 indicates the discharge state. 1 indicates charging status, Indicates the lower limit of energy storage battery capacity, Indicates the upper limit of energy storage battery capacity. Indicates the initial capacity state of the energy storage device, Indicates the final capacity state of the energy storage device.

3. The method according to claim 1, characterized in that The robust optimization algorithm is based on the flexible load model of the substation to establish a multi-time scale two-stage robust optimization model, including: Obtaining transformer load data, photovoltaic output data, and basic load data, and using the robust optimization algorithm to calculate the flexible load model of the substation, the transformer load data, the photovoltaic output data, and the basic load data to obtain a day-ahead robust model; Constructing a real-time scheduling model for the intraday phase based on the flexible load model of the substation area and the day-ahead robust model; Establish transformer load factor constraints, β min ≤β t ≤β max , Among them, β t represents the transformer load rate at time t, β min Indicates the lower limit of transformer load rate, β max Indicates the upper limit of transformer load rate; Establish power balance constraints, in, represents the grid electricity price at time t, represents the actual power of the base load at time t, represents the charging and discharging power of the charging station at time t, represents the power of the air conditioning load cluster at time t, represents the actual photovoltaic power output at time t, represents the charging and discharging power of the energy storage station at time t; The multi-time-scale two-stage robust optimization model is constructed by adopting the day-ahead stage robust model, the intra-day stage real-time scheduling model, the transformer load rate constraint, and the power balance constraint.

4. The method according to claim 3, characterized in that The robust optimization algorithm is used to calculate the flexible load model of the substation, the transformer load data, the photovoltaic output data, and the basic load data to obtain a day-ahead robust model, including: The first objective function is determined by using the flexible load model of the substation, the transformer load data, the photovoltaic output data, and the basic load data. Among them, minF1 represents the first objective function, represents the charging station control cost at time t, q cs represents the charging station control cost coefficient, represents the charging and discharging power of the charging station at time t, represents the air conditioning load control cost at time t, q ac represents the air conditioning load control cost coefficient, represents the power of the air conditioning load cluster at time t, represents the photovoltaic operation cost at time t, q pv represents the photovoltaic operation cost coefficient, represents the actual photovoltaic power output at time t, represents the operating cost of the energy storage station at time t, q es represents the energy storage station operation cost coefficient, represents the charging and discharging power of the energy storage station at time t, represents the grid electricity cost at time t, represents the grid electricity price at time t, represents the load power of the transformer in the substation at time t, represents the transformer load penalty cost at time t, q s represents the load rate penalty coefficient, β t represents the transformer load rate at time t, β represents the transformer economic load rate, S T Indicates the rated capacity of the transformer, It represents the power factor after transformer compensation, and T1 represents the total scheduling period of the day-ahead stage; The photovoltaic output data and the basic load data are used to determine an uncertainty set, Where Ω is the set of uncertainties u, represents the predicted power of the photovoltaic process at time t, represents the predicted power of the base load at time t, represents the actual power of photovoltaic output at time t, represents the actual power of the base load at time t, represents the fluctuating power of photovoltaic output at time t, represents the fluctuating power of the base load at time t, and Γ t load is a binary variable, When it is 1, it means the photovoltaic output power fluctuates at time t, Γ t load When it is 1, it means the power of the base load fluctuates at time t. Indicates the lower limit of the power fluctuation of photovoltaic output at time t, Indicates the upper limit of the power fluctuation of photovoltaic output at time t, It represents the lower limit of power fluctuation of base load at time t, represents the upper limit of power fluctuation of base load at time t, Γ pv represents the total number of periods of fluctuation in photovoltaic output, Γ load Indicates the total number of periods of base load fluctuation; Converting the first objective function and the uncertainty set into a compact form, Among them, x represents the variable: The variable represented by y is: The variable represented by u is: d T represents the relevant constraint coefficient matrix of the y variable, D represents the first coefficient matrix of the corresponding constraint variable, E represents the second coefficient matrix of the corresponding constraint variable, F represents the third coefficient matrix of the corresponding constraint variable, G represents the fourth coefficient matrix of the corresponding constraint variable, H represents the fifth coefficient matrix of the corresponding constraint variable, I represents the sixth coefficient matrix of the corresponding constraint variable, L represents the seventh coefficient matrix of the corresponding constraint variable, b represents the first constant column vector, g represents the second constant column vector, γ represents the first dual variable, λ represents the second dual variable, π represents the third dual variable, χ represents the fourth dual variable, represents the charging and discharging status of the kth electric vehicle at time t, represents the charge and discharge status of the jth energy storage device at time t, represents the power of the air conditioning load cluster at time t, represents the energy storage charging and discharging power of the jth energy storage device at time t, represents the energy storage capacity of the jth energy storage device at time t, represents the charging and discharging power of the kth electric vehicle at time t, represents the capacity of the kth electric vehicle at time t, represents the load power of the transformer in the substation at time t, represents the actual power of photovoltaic output at time t, represents the actual power of the base load at time t; The day-ahead robust model is constructed using the first objective function converted into a compact form and the uncertainty set.

5. The method according to claim 3, characterized in that The constructing of the intraday real-time scheduling model based on the flexible load model of the substation area and the day-ahead robust model includes: Obtaining intraday electric vehicle charging and discharging power data, intraday energy storage station charging and discharging power data, and electric vehicle regulation cost coefficients, obtaining electric vehicle charging and discharging power data in the flexible load model of the substation, and obtaining energy storage station charging and discharging power data in the day-ahead robust model; The second objective function is calculated using the intra-day electric vehicle charging and discharging power data, the intra-day energy storage station charging and discharging power data, the electric vehicle regulation cost coefficient, the electric vehicle charging and discharging power data, and the energy storage station charging and discharging power data. Among them, minF2 represents the second objective function, q ev represents the electric vehicle regulation cost coefficient, represents the charging and discharging power of the electric vehicle at time t within the day, represents the charging and discharging power of the electric vehicle at time t, represents the charging and discharging power of the energy storage station at time t within the day, represents the charging and discharging power of the energy storage station at time t, and T2 represents the total scheduling period within the day; The second objective function is used to construct the intra-day real-time scheduling model.

6. The method according to claim 1, wherein The column constraint generation algorithm is used to solve the day-ahead phase of the multi-time-scale two-stage robust optimization model to obtain a scheduling plan for air conditioning load, energy storage station, and charging station in the day-ahead phase, including: Obtaining a day-ahead robust model corresponding to the day-ahead phase of the multi-time-scale two-stage robust optimization model; The day-ahead robust model is decomposed into a main problem and sub-problems, where the main problem is: minη The sub-problems are: Among them, η represents the objective function variable of the main problem, d T represents y k The relevant constraint coefficient matrix of the variable, D represents the first coefficient matrix corresponding to the constraint variable, E represents the second coefficient matrix corresponding to the constraint variable, F represents the third coefficient matrix corresponding to the constraint variable, G represents the fourth coefficient matrix corresponding to the constraint variable, H represents the fifth coefficient matrix corresponding to the constraint variable, I represents the sixth coefficient matrix corresponding to the constraint variable, L represents the seventh coefficient matrix corresponding to the constraint variable, b represents the first constant column vector, g represents the second constant column vector, y k represents the y variable in the kth iteration, u k represents the u variable in the k-th iteration, u represents the uncertainty parameter, and Ω represents the uncertainty set; Obtaining a first objective function and uncertainty set converted into a compact form in the day-ahead robust model; Obtain a strong duality theory, and based on the strong duality theory, perform calculations using the main problem, the subproblem, the first objective function converted into a compact form, and the uncertainty set to obtain a dual problem. Where γ represents the first dual variable, λ represents the second dual variable, π represents the third dual variable, and χ represents the fourth dual variable; The column constraint generation algorithm is used to iteratively calculate the main problem and the dual problem to obtain the air conditioning load, energy storage station and charging station scheduling plan for the day-ahead period.

7. The method according to claim 6, characterized in that The method further comprises: When the fluctuating power of photovoltaic output is the upper limit of photovoltaic output power fluctuation and the fluctuating power of base load is the upper limit of base load power fluctuation, a new uncertainty set is determined. Among them, Ω is the set of new uncertainties u, represents the predicted power of the photovoltaic process at time t, represents the predicted power of the base load at time t, express, represents the actual power of the base load at time t, and Γ t load is a binary variable, When it is 1, it means the photovoltaic output power fluctuates at time t, Γ t load When it is 1, it means the power of the base load fluctuates at time t. Indicates the upper limit of the power fluctuation of photovoltaic output at time t, represents the upper limit of power fluctuation of base load at time t, Γ pv represents the total number of periods of fluctuation in photovoltaic output, Γ load Indicates the total number of periods of base load fluctuation; Obtain continuous auxiliary variables, use the continuous auxiliary variables to linearize the dual problem, and obtain a linearized formula. Among them, Z′ represents a continuous auxiliary variable, Indicates the auxiliary adjustment amount of photovoltaic output, Indicates the base load auxiliary adjustment amount, represents the upper bound of the second dual variable; The column constraint generation algorithm is used to iteratively calculate the main problem and the dual problem to obtain the air conditioning load, energy storage station and charging station scheduling plan for the day-ahead period.

8. A robust optimization device for flexible load in a substation, characterized in that: include: The first modeling module is used to establish an air conditioning load model, an electric vehicle state of charge model, and an energy storage state of charge model, and use the air conditioning load model, the electric vehicle state of charge model, and the energy storage state of charge model to construct a flexible load model of the substation; A second modeling module is used to obtain a robust optimization algorithm, and based on the robust optimization algorithm, establish a multi-time scale two-stage robust optimization model using the substation flexible load model; A first solving module is configured to obtain a column constraint generation algorithm, and employ the column constraint generation algorithm to solve the day-ahead phase of the multi-time-scale two-stage robust optimization model to obtain a scheduling plan for air conditioning load, energy storage stations, and charging stations in the day-ahead phase; The second solving module is used to optimize and solve the intraday stage of the multi-time-scale two-stage robust optimization model based on the air-conditioning load, energy storage station and charging station scheduling plan of the day-ahead stage, and obtain the real-time charging and discharging power scheduling plan of the energy storage equipment and electric vehicles.

9. An electronic device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 7 are implemented.

10. A storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 7 are implemented.

Images