Micro-grid dispatching method based on electric vehicle dispatching potential and travel risk

Abstract

The invention discloses a micro-grid scheduling method based on electric vehicle scheduling potential and travel risk, comprising the following steps: S1, establishing a two-stage scheduling model of an industrial micro-grid, the two-stage scheduling model comprising a day-ahead scheduling model and a real-time scheduling model; s2, initializing the two-stage scheduling model established in the step S1; s3, the day-ahead stage scheduling model is a mixed integer linear model, a Gurobi solver is used for solving, and an optimal scheduling plan of the industrial micro-grid is obtained; the real-time stage scheduling model adopts a double-layer optimization model to solve the real-time adjustment cost and the charging and discharging plan of the electric vehicle, the optimal scheduling plan of the industrial micro-grid is substituted into the real-time stage scheduling model, the objective function of the real-time stage scheduling model is solved, and the optimal charging and discharging plan of the electric vehicle is obtained. According to the method, an industrial micro-grid optimal scheduling plan is obtained through an industrial micro-grid two-stage model based on electric vehicle scheduling potential and travel risks.

Description

Technical Field

[0001] The present invention relates to the technical field of microgrid dispatching strategies, and in particular to a microgrid dispatching method based on electric vehicle dispatching potential and travel risk. Background Art

[0002] Renewable energy will play an important role in the energy transformation of the power system due to its environmental friendliness and renewability. However, the volatility, intermittency and uncertainty of renewable energy generation may have a negative impact on the stable operation of the power grid. As a new type of power system, microgrid can reduce the impact of uncertainty of renewable energy on the power grid. At the same time, with its high proportion of renewable energy generation, high degree of power automation, and customizable personalized power services, it has become one of the key research points in the transformation of the power system. However, the uncertainty of renewable energy and load will bring great challenges to the operation stability and reliability of microgrids, thereby affecting the power quality of the power grid and users. One of the effective measures to solve this problem is to use electric vehicles to connect to the grid, and reduce the burden of microgrids during peak power consumption through orderly charging and discharging of electric vehicles.

[0003] In summary, with the increasing development of electric vehicles and their penetration in the power grid, an effective means to solve the uncertainty of renewable energy in microgrids is to apply the vehicle-to-grid technology of electric vehicles to provide auxiliary services for coordinated charging and discharging for microgrids and achieve stable operation of microgrids. However, when electric vehicles provide auxiliary services, they will inevitably lead to battery loss and travel problems for electric vehicle users. How to meet the travel needs of users and increase their willingness to participate while using electric vehicles, achieve a benign interaction between electric vehicles and microgrids, and increase the enthusiasm of electric vehicle users to participate in providing auxiliary services will play an important role in the future power system industry and energy field. Therefore, analyzing the application of electric vehicle scheduling potential and travel risks in industrial microgrid scheduling will provide certain guiding significance for the stable and reliable operation of industrial microgrids. Summary of the invention

[0004] In view of the deficiencies in the prior art, the present invention provides a microgrid scheduling method based on electric vehicle scheduling potential and travel risk. The method obtains an optimal scheduling plan for the industrial microgrid through a two-stage model of an industrial microgrid based on electric vehicle scheduling potential and travel risk, including an optimal day-ahead scheduling plan for the industrial microgrid, a real-time correction plan, and a real-time optimal charging and discharging plan for electric vehicles.

[0005] In order to solve the above technical problems, the technical solution of the present invention is:

[0006] A microgrid dispatching method based on electric vehicle dispatching potential and travel risk comprises the following steps:

[0007] S1. Establish a two-stage dispatching model for industrial microgrids, wherein the two-stage dispatching model includes a day-ahead stage dispatching model and a real-time stage dispatching model.

[0008] The day-ahead scheduling model takes the minimum operation cost of the industrial microgrid as the optimization goal, and its objective function is:

[0009] min F ahaed =F1+F CVaR

[0010] Among them, F1 is the microgrid operation cost, F CVaR The conditional risk cost caused by net load uncertainty;

[0011] The real-time stage scheduling model takes the minimization of the real-time correction cost of the industrial microgrid and the maximization of the benefits of the cooperative alliance as the optimization objectives, and its objective function is:

[0012]

[0013] in, is the incentive compensation cost that the microgrid needs to give to the electric vehicle alliance at time t, that is, the total income of the alliance, is the additional penalty cost that the microgrid needs to pay to the grid at time t, The risk cost of selling charging flexibility and power for electric vehicle i at time t;

[0014] S2, initialize the two-stage scheduling model established in S1;

[0015] S3, the day-ahead scheduling model is a mixed integer linear model, which is solved using the Gurobi solver to obtain the optimal scheduling plan for the industrial microgrid;

[0016] The real-time stage scheduling model adopts a two-layer optimization model to solve the real-time adjustment cost and the charging and discharging plan of electric vehicles. The optimal scheduling plan of the industrial microgrid is substituted into the real-time stage scheduling model to solve the objective function of the real-time stage scheduling model and obtain the optimal charging and discharging plan of electric vehicles.

[0017] Preferably, the objective function of the day-ahead scheduling model is as follows:

[0018] Taking the electricity cost of industrial microgrid as F ahaed As the optimization goal, its objective function is:

[0019]

[0020] Among them, F1 is the microgrid operation cost, F CVaR is the conditional risk cost caused by net load uncertainty, They are the power purchase cost of the power grid, the photovoltaic operation cost, and the wind turbine operation cost at time t respectively; is the net power of the microgrid at time t, is the grid electricity price at time t, CVaR t is the average net load error of the microgrid under the condition of exceeding the allowable interval at time t, H(Er t ) is the penalty function for exceeding the power tolerance range, Er t is the net load error at time t, VaR t is the power tolerance interval boundary at time t, ΔEr is the jump, D is the basic penalty multiple, [x] + is the ceiling function.

[0021] Preferably, the objective function of the real-time stage scheduling model is as follows:

[0022] The real-time stage is based on 15-minute nodes. At each time node, the optimization goal is to minimize the real-time adjustment cost of the microgrid and maximize the benefits of the electric vehicle alliance. The objective function is:

[0023]

[0024] in, is the incentive compensation cost that the microgrid needs to give to the electric vehicle alliance at time t, that is, the total income of the alliance, is the additional penalty cost that the microgrid needs to pay to the grid at time t, is the actual net load error, is the actual net load of the microgrid, VaR is the allowable error, The risk cost of selling charging flexibility and electricity to electric vehicles once, that is, the revenue of participating in scheduling alone; Z is the number of members in the grand alliance.

[0025] Preferably, the day-ahead stage scheduling model is set with constraints, and the constraints set in the day-ahead stage scheduling model include microgrid power balance constraints, microgrid peak-to-valley difference constraints, microgrid interactive power constraints and EV charging constraints.

[0026] Preferably, the real-time stage scheduling model is set with constraints, and the constraints set by the real-time stage scheduling model include real-time power balance constraints, power grid constraints and EV constraints.

[0027] Preferably, in step S2, the parameters of the day-ahead scheduling model are initialized, including the industrial microgrid basic load Photovoltaic output power Fan output power Grid time-of-use electricity price, charging price, time period number T, number of electric vehicles N, rated capacity of electric vehicle batteries E, electric vehicle travel characteristics, electric vehicle charging power Charge and discharge efficiency

[0028] Preferably, in step S2, the parameters of the real-time stage scheduling model are initialized, including the time-of-use electricity price of the power grid, the charging price, the number of time periods T, the number of electric vehicles N, the rated capacity of the electric vehicle battery E, the electric vehicle travel characteristic quantity, and the electric vehicle charging and discharging power Charge and discharge efficiency Electric vehicle battery cycle life CL under nominal conditions nom , Depth of discharge DoD of electric vehicle batteries nom , Electric vehicle battery purchase and recycling prices

[0029] Preferably, in step S3, the method for solving the day-ahead scheduling model is:

[0030] Industrial microgrid base load Photovoltaic output power Fan output power Calculate the day-ahead forecast of industrial microgrid net load The expression is as follows:

[0031]

[0032] Industrial microgrid net load error Er t , the expression is as follows:

[0033]

[0034] is the historical actual net load power;

[0035] Based on the historical net load error data of industrial microgrid and the day-ahead forecast data of industrial microgrid, the conditional risk cost F caused by net load uncertainty is calculated using the CVaR method. CVaR and the net load power tolerance range, or VaR;

[0036] The number of electric vehicles N, the rated capacity of electric vehicle batteries E, the travel characteristics of electric vehicles, and the charging power of electric vehicles Charge and discharge efficiency According to the distribution of electric vehicle travel, the adjustable potential of electric vehicles is calculated using Minkowski additivity;

[0037] The dispatchable potential of electric vehicles and the conditional risk cost F CVaR, predict the net load of industrial microgrid, build a mixed integer linear model with the goal of minimizing the day-ahead electricity cost of industrial microgrid, and use Gurobi solver to solve the mixed integer linear model to obtain the optimal power purchase and optimal charging power Form the optimal scheduling plan for the day-ahead phase.

[0038] As a preference, the conditional risk cost F caused by the net load uncertainty CVaR The calculation method of the net load power tolerance range VaR is as follows:

[0039] Established the loss function in are decision variables and random variables, namely actual net load power and predicted net load power, respectively. Based on the historical data of predicted net load power, the probability density θ(Er t ), whose decision variables Determined with the threshold δ, the loss function Cumulative distribution function and loss function of as follows:

[0040]

[0041] Therefore, under a certain confidence level α∈(0,1), for a certain decision variable The expression of VaR function is as follows:

[0042]

[0043] Among them, the confidence level α is determined by the decision maker based on the internal resources of the system and the resource scheduling capacity, and R is a real number set;

[0044] Through VaR, the CVaR expression can be obtained as follows:

[0045]

[0046] Construct an auxiliary function to simplify the CVaR expression:

[0047]

[0048] By replacing the integral with sampling points, the above formula can be further simplified to:

[0049]

[0050] Among them, N is the total number of samples, the threshold δ is an auxiliary variable, and its optimal value is VaR;

[0051] Therefore, we can get the conditional risk value cost F CVaR ,as follows:

[0052]

[0053] Preferably, in step S3, the method for calculating the adjustable potential of the electric vehicle is:

[0054] The electric vehicle travel characteristics are obtained and initialized by the charging piles of the charging station in the microgrid. The electric vehicle travel characteristics include the initial SOC, arrival time, departure time reported by the electric vehicle user, and expected SOC information. The load model corresponding to the electric vehicle individual is established through the electric vehicle travel characteristics. The expression is as follows:

[0055]

[0056] in, are the charging and discharging rates of the electric vehicle at time t; They are the upper limits of the charging and discharging power of electric vehicles respectively; are Boolean variables of the charging and discharging status of the electric vehicle, is a Boolean variable indicating the arrival state of the electric vehicle. Electric vehicles can be charged and discharged at the same time; is the state of charge of the electric vehicle at time t, SOC x，min , SOC x，max They are the upper and lower limits of the state of charge of electric vehicles respectively; are the charging and discharging efficiency of electric vehicles respectively; Δt is the charging and discharging time period of electric vehicles; t ar ,t le are the arrival and departure times of electric vehicles, respectively, and x is the type of electric vehicle;

[0057] In view of the differences in arrival and departure times between individual electric vehicles, Boolean variables are considered when calculating the charging and discharging power. The access time of all electric vehicles to the industrial microgrid is unified into the same definition domain through Boolean variables, and the power of the electric vehicle aggregate at the charging station is calculated using Minkowski additivity. The expression is as follows:

[0058]

[0059] Among them, SOC n，ar , SOC n，le are the battery charge states of the electric vehicle at the time of arrival and departure respectively; X EV A collection of electric vehicle types;

[0060] Variables, parameters and dispatchable potential Ω of the charging station electric vehicle aggregate EV , the expression is as follows:

[0061]

[0062] in, They are the charging and discharging power of the electric vehicle aggregate at the charging station at time t and the amount of electricity of the electric vehicle aggregate at the charging station; They are the upper limits of charge and discharge of electric vehicle aggregates at the charging station at time t; are the upper and lower limits of the power of the electric vehicle aggregate at the charging station at time t; ΔSOC t is the change in the amount of electricity of the electric vehicle aggregate at the charging station at time t due to the driving behavior of the electric vehicle users.

[0063] Preferably, in step S3, the method for solving the real-time stage scheduling model is:

[0064] Obtain and initialize the expected battery charge of electric vehicle x through the electric vehicle charging pile The battery charge SOC at the current time t x，t , residence time T stop , the required charging time is T ch , the travel risk is modeled through the generalized bell-shaped membership function, and the travel risk of electric vehicles is calculated. The expression is as follows:

[0065]

[0066] in, SOC x，t are the travel risk and state of charge of electric vehicles at time t, is the expected SOC when the EV leaves, and σ is the current required charging time T ch With the residence time T stop The ratio of

[0067] The electric vehicle battery attenuation model is used to calculate the degree of battery power attenuation caused by the sale of electric vehicle power. The expression is as follows:

[0068]

[0069] Among them, CL nom , DoD nom is the cycle life and depth of discharge of electric vehicle batteries under nominal conditions, e t，x is the current discharge capacity, E x The maximum charge of the electric vehicle battery;

[0070] Through the electric vehicle travel risk function and the battery power attenuation degree, using the incentive-based electric vehicle compensation method, the risk cost of selling charging flexibility and power of the electric vehicle at this moment is calculated. The expression is as follows:

[0071]

[0072] in, is the power of electric vehicles participating in the dispatch at time t, The maximum charging price for industrial microgrids, For the EV battery sold separately, The recycling price of scrapped electric vehicle batteries;

[0073] Initialize the parameters related to industrial microgrid, electric vehicle, and cooperative alliance at time t, including the power gap of industrial microgrid, grid price, electric vehicle travel risk, and battery sales and recycling costs;

[0074] Set the initial number of electric vehicles in the cooperative alliance M, and the expression is as follows:

[0075]

[0076] Among them, ΔP del For the power gap of industrial microgrid, is the minimum non-zero discharge power of the cooperative alliance electric vehicle, The optimal power purchase power;

[0077] Calculate the total response power of electric vehicles based on the number of participating electric vehicles Passed into the lower model in the two-layer optimization model;

[0078] The lower model calculates the real-time adjustment cost without the participation of electric vehicles. Then, the EVs that meet the requirements are extracted according to the total power response and the number of electric vehicles. The real-time adjustment cost of the industrial microgrid is minimized as the optimization goal. The Gurobi solver is used to optimize the real-time adjustment cost of the industrial microgrid obtained by the cooperative alliance and the total benefit of the cooperative alliance, and then the results are passed to the upper model.

[0079] The upper model in the two-layer optimization model calculates the contribution of individual electric vehicles in the alliance according to the nonlinear energy mapping function, thereby calculating the benefits obtained by individual electric vehicles in the alliance. The expression is as follows:

[0080]

[0081] in, are the output power of member i and the maximum output power of the members in the alliance respectively; is the risk factor of member i; is the nonlinear energy contribution value of member i, F(x) is the nonlinear mapping function, which uses the exponential function of the natural logarithm e; γ i is the overall contribution of member i;

[0082] Determine whether the number of electric vehicles reaches the upper and lower limits. If so, stop the iteration and output the real-time adjustment cost and individual profit of electric vehicles. Otherwise, adjust the number of electric vehicles and recalculate the total response power of electric vehicles.

[0083] The present invention has the following characteristics and beneficial effects:

[0084] The above technical solution is adopted, and the two-stage model of industrial microgrid based on electric vehicle scheduling potential and travel risk is used to study the planning and optimization problems between two-stage multi-agents, so as to realize the orderly charging and discharging of electric vehicles, protect the travel needs and basic interests of electric vehicle users, reduce the operating costs of industrial microgrids, and achieve the goal of stable and reliable operation of industrial microgrids. The two-stage scheduling method of industrial microgrid can better meet the needs of the actual industrial microgrid scheduling problem, reduce the operating costs of industrial microgrids, alleviate the "peak-on-peak" phenomenon of industrial microgrids, and reduce the burden on industrial microgrids; while scheduling electric vehicles, it can meet the travel needs of electric vehicle users and reduce the charging costs of electric vehicle users. BRIEF DESCRIPTION OF THE DRAWINGS

[0085] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative labor.

[0086] Figure 1 This is a two-stage dispatching flow chart of an industrial microgrid in an embodiment of the present invention.

[0087] Figure 2 Comparison of industrial microgrid load curves before and after day-ahead scheduling in an embodiment of the present invention.

[0088] Figure 3 This is the dispatchable potential of electric vehicles in the day-ahead stage in the embodiment of the present invention.

[0089] Figure 4 It is the electric vehicle dispatching result in the embodiment of the present invention.

[0090] Figure 5 Graph showing the relationship between electric vehicle characteristic quantities and travel risks in an embodiment of the present invention.

[0091] Figure 6 Comparison of industrial microgrid load curves in the real-time correction stage in an embodiment of the present invention.

[0092] Figure 7This is a comparison chart of the results before and after the real-time stage power purchase adjustment in an embodiment of the present invention.

[0093] Figure 8 This is the overall charging and discharging status of the electric vehicle during the real-time correction phase in the embodiment of the present invention. DETAILED DESCRIPTION

[0094] It should be noted that, in the absence of conflict, the embodiments of the present invention and the features in the embodiments may be combined with each other.

[0095] In order to make the objectives, technical solutions and advantages of the present invention more clearly understood, the present invention is further described in detail below with reference to the accompanying drawings.

[0096] On the contrary, the present invention covers any substitution, modification, equivalent method and scheme made on the essence and scope of the present invention as defined by the claims. Further, in order to make the public have a better understanding of the present invention, some specific details are described in detail in the detailed description of the present invention below. Those skilled in the art can fully understand the present invention without the description of these details.

[0097] The present invention relates to an optimization method for industrial microgrid scheduling problems, which uses a two-stage scheduling model to generate a day-ahead scheduling and real-time correction plan for an industrial microgrid with the lowest operating cost. The optimization objectives of the two-stage scheduling model are to minimize the operating cost and real-time correction cost of the industrial microgrid. Figure 1 As shown. The purpose of the present invention is to propose a microgrid dispatching method based on the dispatching potential and travel risk of electric vehicles, so as to realize the control of orderly charging and discharging of electric vehicles, realize the protection of the travel needs and basic interests of electric vehicle users, realize the goal of reducing the operating cost of industrial microgrids and realizing the stable and reliable operation of industrial microgrids. To illustrate the effect of the present invention, the present invention is described in detail below with the industrial microgrid system as the implementation object of the present invention:

[0098] S1. Establish a two-stage scheduling model for an industrial microgrid, wherein the two-stage scheduling model includes a day-ahead scheduling model and a real-time scheduling model.

[0099] Ⅰ) The day-ahead dispatch model takes the minimum operation cost of the industrial microgrid as the optimization goal, and its objective function is:

[0100] min F ahaed =F1+F CVaR (2)

[0101] Among them, F1 is the microgrid operation cost, F CVaR is the conditional risk cost caused by net load uncertainty.

[0102] Specifically, the microgrid operating cost F1 generally includes the grid power purchase cost, photovoltaic operating cost, and wind turbine operating cost, and the expression is as follows:

[0103]

[0104] in, They are respectively the grid electricity purchase cost, photovoltaic operation cost, and wind turbine operation cost at time t.

[0105] Furthermore, the cost of purchasing electricity from the power grid is calculated as follows:

[0106]

[0107] in, is the net power of the microgrid at time t, is the electricity price of the power grid at time t, is the total charging power of the electric vehicle at time t, is the microgrid load in addition to the total charging power of electric vehicles.

[0108] Furthermore, the conditional risk cost caused by net load uncertainty is calculated as follows:

[0109] The net load uncertainty calculation mainly caused by photovoltaic, wind turbine and load, among which the conditional risk value CVaR of uncertainty t As described in detail in S3, the conditional risk cost is calculated by formula (5):

[0110]

[0111] in, is the net power of the microgrid at time t, is the grid electricity price at time t, CVaR t is the average net load error of the microgrid under the condition of exceeding the allowable interval at time t, H(Er t ) is the penalty function for exceeding the power tolerance range, Er t is the net load error at time t, VaR t is the power tolerance interval boundary at time t, ΔEr is the jump, D is the basic penalty multiple, [x] + is the ceiling function.

[0112] It should be noted that FCVaR includes a penalty part, so the objective function of the day-ahead scheduling model includes a penalty function.

[0113] A further arrangement of this embodiment is that the day-ahead stage scheduling model is set with constraints, and the constraints set in the day-ahead stage scheduling model include microgrid power balance constraints, microgrid peak-to-valley difference constraints, microgrid interactive power constraints and EV charging constraints.

[0114] The specific constraints are as follows:

[0115] Microgrid power balance constraints:

[0116]

[0117] in, is the power of other MG loads except EVs at time t, is the charging load of regular EV users at time t, The net load is also the power purchased by MG.

[0118] Microgrid peak-to-valley difference constraints:

[0119]

[0120] Among them, ΔP MG Optimize and solve the peak-to-valley difference of the front MG load curve, The peak-to-valley difference of the MG load curve is obtained by optimization.

[0121] Microgrid interactive power constraints:

[0122]

[0123] in, They respectively represent the upper and lower limits of electricity purchase from the power grid.

[0124] EV Charging Constraints:

[0125] In the day-ahead scheduling, only the flexibility of EV charging is considered, and EV discharge compensation is not considered. At present, the common process of scientific charging of lithium-ion batteries can be divided into three stages: pre-regulation, constant current charging and constant voltage charging. Therefore, in order to protect the charging life and charging safety of EV batteries and improve charging efficiency. The present invention constrains the EV battery charging process and charging times as follows:

[0126]

[0127] in, For the charging times, reducing the charging times and increasing the charging continuity can protect the battery charging life; The maximum charging power of the battery will change with the change of SOC, so that it matches the three-stage charging. x When ≤80%, it is in constant current charging state and can use high current charging, so the charging power is When SOC x <20% or SOC x When it is >80%, it is in the pre-voltage regulation and constant voltage charging state. To protect the battery life, the charging current is small, so the charging power is During the actual charging process, Varies with current.

[0128] II) The real-time stage dispatch model takes the minimization of the real-time correction cost of the industrial microgrid and the maximization of the benefits of the cooperative alliance as the optimization objectives. Its objective function is:

[0129]

[0130] in, is the incentive compensation cost that the microgrid needs to give to the electric vehicle alliance at time t, that is, the total income of the alliance, is the additional penalty cost that the microgrid needs to pay to the grid at time t, Selling charging flexibility and risk cost of electricity for electric vehicle i at time t.

[0131] Specifically, the penalty cost is calculated by the portion exceeding the day-ahead power purchase plan using a specific penalty method of the power grid. The specific penalty cost calculation method in the implementation scheme of the present invention is as shown in formula (11):

[0132]

[0133] Furthermore, the constraints set by the real-time stage scheduling model include real-time power balance constraints, grid constraints and EV constraints.

[0134] Real-time power balancing constraints:

[0135]

[0136] in, They are the actual total load of the system at time t, the actual net load (also the actual purchased power), the actual photovoltaic output power, and the actual wind power output power. They are the actual discharge power and actual charging power of CSEVA at time t respectively.

[0137] Grid constraints:

[0138]

[0139] Among them, when It means that MG exceeds (or falls short of) the amount of electricity purchased in the electricity purchase plan and will be subject to a penalty for exceeding the upper and lower limits.

[0140] EV Constraints:

[0141] In the real-time scheduling process, in order to ensure that the EV individuals do not leave the big alliance, it is necessary to ensure the travel needs and profitability of the EV individuals. Therefore, the specific constraints are as follows:

[0142]

[0143] in, is the actual expected SOC at the time of EV leaving, Profit for EV distribution in the major leagues.

[0144] S2. Initialize the two-stage dispatch model established in S1. The parameters of the two-stage dispatch model include the industrial microgrid basic load predicted a day ago. Photovoltaic output power Fan output power Grid time-of-use electricity price, charging price, time period number T, number of electric vehicles N, rated capacity of electric vehicle batteries E, electric vehicle travel characteristics, electric vehicle charging and discharging power Charge and discharge efficiency Electric vehicle battery cycle life CL under nominal conditions nom , Depth of discharge DoD of electric vehicle batteries nom , Electric vehicle battery purchase and recycling prices

[0145] Specifically, the parameters of the day-ahead scheduling model are initialized, including the industrial microgrid base load Photovoltaic output power Fan output power Grid time-of-use electricity price, charging price, time period number T, number of electric vehicles N, rated capacity of electric vehicle batteries E, electric vehicle travel characteristics, electric vehicle charging power Charge and discharge efficiency

[0146] Initialize the parameters of the real-time scheduling model, including the grid time-of-use electricity price, charging price, time period number T, number of electric vehicles N, electric vehicle battery rated capacity E, electric vehicle travel characteristics, electric vehicle charging and discharging power Charge and discharge efficiency Electric vehicle battery cycle life CL under nominal conditions nom , Depth of discharge DoD of electric vehicle batteries nom , Electric vehicle battery purchase and recycling prices

[0147] S3. In the two-stage dispatching model of industrial microgrid, the day-ahead dispatching model adopts the day-ahead energy dispatching model based on conditional value at risk (CVaR) considering the dispatching potential of electric vehicles. The model is a mixed integer linear model and is solved by Gurobi solver. The real-time dispatching model adopts the real-time correction model based on dynamic game considering the travel risk of electric vehicles. The real-time dispatching model adopts a two-layer optimization model to solve the real-time adjustment cost and the charging and discharging plan of electric vehicles, solve the objective function corresponding to the two-stage dispatching model, and obtain the optimal dispatching plan of industrial microgrid and the optimal charging and discharging plan of electric vehicles. It specifically includes the following sub-steps:

[0148] S3-1. Solve the day-ahead scheduling model.

[0149] S3-1-1. Initialize the industrial microgrid basic load predicted a day ago Photovoltaic output power Fan output power Grid time-of-use electricity price, charging price, time period number T, number of electric vehicles N, rated capacity of electric vehicle batteries E, electric vehicle travel characteristics, electric vehicle charging power Charge and discharge efficiency

[0150] S3-1-2, Industrial microgrid basic load Photovoltaic output power Fan output power Calculate the day-ahead forecast of industrial microgrid net load Calculate the historical net load error Er of the industrial microgrid based on the historical data of the industrial microgrid t , the calculation formula is as follows:

[0151]

[0152] in, is the actual net load power.

[0153] S3-1-3. Based on the historical net load error data of industrial microgrid and the related data of industrial microgrid predicted a day ago, the conditional risk cost F caused by net load uncertainty is calculated using the CVaR method. CVaR And the net load power tolerance range, namely VaR. The specific method is as follows:

[0154] S3-1-3-1. Establishing the loss function in are decision variables and random variables, namely actual net load power and predicted net load power, respectively. Based on the historical data of predicted net load power, the probability density θ(Er t ), whose decision variables Determined with the threshold δ, the loss function Cumulative distribution function and loss function of as follows:

[0155]

[0156] S3-1-3-2, therefore, under a certain confidence level α∈(0,1), for a certain decision variable The expression of VaR function is as follows:

[0157]

[0158] Among them, the confidence level α is determined by the decision maker based on the internal resources of the system and the resource schedulability, and R is a set of real numbers.

[0159] S3-1-3-3. Through VaR, the CVaR expression can be obtained as follows:

[0160]

[0161] S3-1-3-4. However, in the process of solving practical problems, it is too complicated to accurately solve the above expression, so we construct an auxiliary function to simplify the expression. After simplification, it becomes as follows:

[0162]

[0163] S3-1-3-5. By replacing the integral with the sampling point, the above formula (20) can be further simplified to:

[0164]

[0165] Among them, N is the total number of samples, the threshold δ is an auxiliary variable, and its optimal value is VaR.

[0166] S3-1-3-6, it can be obtained that the conditional risk value cost F CVaR ,as follows:

[0167]

[0168] S3-1-4, based on the number of electric vehicles N, the rated capacity of electric vehicle batteries E, the travel characteristics of electric vehicles, and the charging power of electric vehicles Charge and discharge efficiency According to the distribution of electric vehicle travel, the adjustable potential of electric vehicles is calculated using Minkowski additivity. The specific steps are as follows:

[0169] S3-1-4-1. The charging piles of the charging stations in the microgrid obtain and initialize the travel characteristics of electric vehicles, including: initial SOC, arrival time, departure time reported by electric vehicle users, and expected SOC. The load model corresponding to the individual electric vehicle can be established through the above information. The expression is as follows:

[0170]

[0171] in, are the charging and discharging rates of the electric vehicle at time t; They are the upper limits of the charging and discharging power of electric vehicles respectively; are Boolean variables representing the charging and discharging status of electric vehicles, i.e., the charging and discharging of a single electric vehicle cannot be performed at the same time. is a Boolean variable indicating the arrival state of the electric vehicle. Electric vehicles can be charged and discharged at the same time; is the state of charge of the electric vehicle at time t, SOC x，min , SOC x，max They are the upper and lower limits of the state of charge of electric vehicles respectively; are the charging and discharging efficiency of electric vehicles respectively; Δt is the charging and discharging time period of electric vehicles; t ar ,t le are the arrival time and departure time of the electric vehicle respectively, and x is the type of the electric vehicle.

[0172] S3-1-4-2: The arrival and departure times of individual electric vehicles are different, so the present invention takes into account the Boolean variable when calculating the charging and discharging power. The access time of all electric vehicles to the industrial microgrid is unified into the same definition domain through Boolean variables. Therefore, the individual electric vehicles have Minkowski additivity, and the power of the electric vehicle aggregate at the charging station is calculated using Minkowski additivity. The expression is as follows:

[0173]

[0174] Among them, SOC n，ar , SOC n，le are the battery charge states of the electric vehicle at the time of arrival and departure respectively; X EV A collection of electric car types.

[0175] S3-1-4-3. In summary, for the variables, parameters and dispatchable potential Ω of the charging station electric vehicle aggregate EV , the expression is as follows:

[0176]

[0177] in, are the charging and discharging power of CS electric vehicle A and the power of CS electric vehicle A at time t respectively; are the upper limits of charge and discharge of CS electric vehicle A at time t respectively; are the upper and lower limits of the power of the electric vehicle aggregate at the charging station at time t; ΔSOC t is the change in the amount of electricity of the electric vehicle aggregate at the charging station at time t due to the driving behavior of the electric vehicle users.

[0178] S3-1-5, based on the dispatchable potential of electric vehicles and conditional risk cost F CVaR , predict the net load of industrial microgrid, build a mixed integer linear model with the goal of minimizing the day-ahead electricity cost of industrial microgrid, and use Gurobi solver to solve the mixed integer linear model to obtain the optimal power purchase and optimal charging power The optimal dispatching plan is formed in the day-ahead phase, and the optimal power purchase plan is transmitted to the real-time phase.

[0179] S3-2. Solve the real-time stage scheduling model.

[0180] S3-2-1. In the real-time stage, at each unit time node, the charging price, the number of time periods T, the electric vehicle travel characteristics, including the expected battery charge of the electric vehicle x The battery charge SOC at the current time t x，t , residence time T stop , the required charging time is T ch , the generalized bell-shaped membership function is used to calculate the travel risk of the electric vehicle individual at the current time t, and the battery power decay function and the electric vehicle incentive function are used to calculate the risk cost of selling charging flexibility and power. The specific sub-steps are as follows:

[0181] S3-2-1-1. Obtain and initialize the expected battery charge of electric vehicle x through the electric vehicle charging pile The battery charge SOC at the current time t x，t , residence time T stop , the required charging time is T ch , the travel risk is modeled through the generalized bell-shaped membership function, and the travel risk of electric vehicles is calculated. The expression is as follows:

[0182]

[0183] in, SOC x，t are the travel risk and state of charge of electric vehicles at time t, is the expected SOC when the EV leaves, and σ is the current required charging time Tch With the residence time T stop ratio.

[0184] S3-2-1-2. Use the electric vehicle battery attenuation model to calculate the degree of battery power attenuation caused by the sale of electric vehicle power. The expression is as follows:

[0185]

[0186] Among them, CL nom , DoD nom is the cycle life and depth of discharge of electric vehicle batteries under nominal conditions, e t，x is the current discharge capacity, E x The maximum charge of an electric vehicle battery.

[0187] Specifically, the nominal conditions refer to: at room temperature, the battery is charged with a current of 1C, and after being fully charged, the battery is discharged with a current of 1C after waiting for 1 hour.

[0188] S3-2-1-3. Based on the electric vehicle travel risk function and the battery power attenuation degree, the risk cost of selling charging flexibility and power of the electric vehicle at that moment is calculated using the incentive-based electric vehicle compensation method. The expression is as follows:

[0189]

[0190] in, is the power of electric vehicles participating in the dispatch at time t, The maximum charging price for industrial microgrids, For the EV battery sold separately, The recycling price of scrapped electric vehicle batteries.

[0191] S3-2-2. At each time node, the cooperative game model is used to take the deviation between the actual net load and the net load of the optimal power purchase plan on the day before and the travel risk cost of the electric vehicle at that time node, and the electric vehicle charging and discharging power as the decision variable, and the minimization of the real-time adjustment cost of the microgrid and the maximization of the benefits of the electric vehicle alliance as the optimization goals. A real-time correction model based on dynamic game that takes into account the travel risk of electric vehicles is constructed, and the real-time adjustment cost and individual profit of electric vehicles are calculated through a two-layer optimization method, and the real-time electric vehicle charging and discharging power and microgrid real-time correction results are output. The specific sub-steps are as follows:

[0192] S3-2-2-1. Initialize the parameters related to industrial microgrid, electric vehicle and cooperative alliance at time t, including the power gap of industrial microgrid, grid price, electric vehicle travel risk and battery sales and recycling costs;

[0193] S3-2-2-2. Set the initial number of electric vehicles in the cooperative alliance, M, as follows:

[0194]

[0195] Among them, ΔP del For the power gap of industrial microgrid, It is the minimum non-zero discharge power of the cooperative alliance electric vehicles.

[0196] S3-2-2-3. Calculate the total response power of electric vehicles based on the number of participating electric vehicles Passed into the lower model in the two-layer optimization model;

[0197] S3-2-2-4, the lower model calculates the real-time adjustment cost without the participation of electric vehicles. Then, according to the total power response of electric vehicles and the number of electric vehicles, the EVs that meet the requirements are extracted, and the real-time adjustment cost of the industrial microgrid is minimized as the optimization goal. The Gurobi solver is used to optimize the real-time adjustment cost of the industrial microgrid obtained by the cooperative alliance and the total benefit of the cooperative alliance, and then passed to the upper model in the two-layer optimization model.

[0198] Specifically, the lower model calculates the real-time adjustment cost without the participation of electric vehicles, and then extracts EVs that meet the requirements based on the total power response of electric vehicles and the number of electric vehicles, with the minimization of the real-time adjustment cost of the industrial microgrid as the optimization goal.

[0199] S3-2-2-5. The upper model calculates the contribution of individual electric vehicles in the alliance according to the nonlinear energy mapping function, and thus calculates the benefits of individual electric vehicles in the alliance. The expression is as follows:

[0200]

[0201]

[0202] in, are the output power of member i and the maximum output power of the members in the alliance respectively; is the risk factor of member i; is the nonlinear energy contribution value of member i, F(x) is the nonlinear mapping function, which uses the exponential function of the natural logarithm e; γ i is the overall contribution of member i.

[0203] Specifically, the upper model takes the maximum benefit of the cooperative alliance as the optimization goal. According to the risk cost and the size of the benefit of the electric vehicle selling charging flexibility and electricity, it determines whether all electric vehicles participating in the cooperative alliance are profitable. If all are profitable and the benefits of the cooperative alliance are not increasing, the real-time adjustment cost and individual profit of the electric vehicle are output, otherwise the subsequent steps are executed;

[0204] S3-2-2-6. Based on the risk cost and the size of the profit of selling charging flexibility and electricity of electric vehicles, determine whether all electric vehicles participating in the cooperative alliance are profitable. If they are profitable, output the real-time adjustment cost and individual electric vehicle profit situation. Otherwise, execute S3-2-2-7.

[0205] S3-2-2-7. Determine whether the number of electric vehicles has reached the upper and lower limits. If so, stop the iteration and output the real-time adjustment cost and individual profit of electric vehicles. Otherwise, adjust the number of electric vehicles and execute S3-2-2-3.

[0206] During the optimization process, the load curve comparison results before and after the industrial microgrid dispatch are as follows: Figure 2 As shown in Figure 2, the day-ahead electric vehicle dispatch potential and electric vehicle dispatch results are shown in Figure 2. Figure 3 , Figure 4 As shown; the above picture can reflect that the present invention has achieved the purpose of alleviating the "peak on peak" phenomenon of the distribution network and reducing the load of the distribution network. At the same time, the load curve of the day before becomes smoother, the volatility of the load curve of the park is reduced, and the dispatchability of electric vehicles is fully utilized. At the same time, in the real-time correction stage, the effect of correcting the power in the real-time stage and the verification effect of the travel risk of electric vehicles with the characteristic quantity are investigated. Based on the scheduling results of the day-ahead stage, the comparison of the load curve of the industrial microgrid in the real-time correction stage, the comparison of the results before and after the adjustment of the power purchase in the real-time stage, and the comparison of the adjustment cost are obtained. Figure 5 , Figure 6 As shown in Table 1, the travel risk of electric vehicles and the characteristics of electric vehicles are confirmed, the dispatch results of electric vehicles and the charging cost of electric vehicles are compared. Figure 7 , Figure 8 As shown in Table 2. As shown in the above charts, it can be seen that the two-stage optimization model of the industrial microgrid described in the present invention not only reduces the electricity cost of the industrial power grid, but also reduces the charging cost of EV users participating in the mediation while meeting the travel needs of users, and stabilizes the load fluctuation of the microgrid.

[0207] Table 1 Comparison of system operation cost and adjustment cost

[0208]

[0209] Table 2 Comparison of charging costs for EVs participating in regulation

[0210]

[0211] The embodiments of the present invention are described in detail above with reference to the accompanying drawings, but the present invention is not limited to the described embodiments. For those skilled in the art, various changes, modifications, substitutions and variations of these embodiments including components are made without departing from the principles and spirit of the present invention, and still fall within the scope of protection of the present invention.

Claims

1. A microgrid dispatching method based on electric vehicle dispatching potential and travel risk, characterized in that: The steps include: S1. Establish a two-stage dispatching model for industrial microgrids, wherein the two-stage dispatching model includes a day-ahead stage dispatching model and a real-time stage dispatching model. The day-ahead scheduling model takes the minimum operation cost of the industrial microgrid as the optimization goal, and its objective function is: my F ahaed =F1+F CVaR Among them, F1 is the microgrid operation cost, F CVaR The conditional risk cost caused by net load uncertainty; The real-time stage scheduling model takes the minimization of the real-time correction cost of the industrial microgrid and the maximization of the benefits of the cooperative alliance as the optimization goals, and its objective function is: in, is the incentive compensation cost that the microgrid needs to give to the electric vehicle alliance at time t, that is, the total income of the alliance, is the additional penalty cost that the microgrid needs to pay to the grid at time t, The risk cost of selling charging flexibility and power for electric vehicle i at time t; S2, initialize the two-stage scheduling model established in S1; S3, the day-ahead scheduling model is a mixed integer linear model, which is solved using the Gurobi solver to obtain the optimal scheduling plan for the industrial microgrid; The real-time stage scheduling model adopts a two-layer optimization model to solve the real-time adjustment cost and the charging and discharging plan of electric vehicles. The optimal scheduling plan of the industrial microgrid is substituted into the real-time stage scheduling model to solve the objective function of the real-time stage scheduling model and obtain the optimal charging and discharging plan of electric vehicles.

2. The microgrid dispatching method based on electric vehicle dispatching potential and travel risk according to claim 1 is characterized in that: The objective function of the day-ahead scheduling model is as follows: Taking the electricity cost of industrial microgrid as F ahead As the optimization goal, its objective function is: Among them, F1 is the microgrid operation cost, F CVaR is the conditional risk cost caused by net load uncertainty, They are the power purchase cost of the power grid, the photovoltaic operation cost, and the wind turbine operation cost at time t respectively; is the net power of the microgrid at time t, is the grid electricity price at time t, CVaR t is the average net load error of the microgrid under the condition of exceeding the allowable interval at time t, H(Er t ) is the penalty function for exceeding the power tolerance range, Er t is the net load error at time t, VaR t is the power tolerance interval boundary at time t, ΔEr is the jump, D is the basic penalty multiple, [x] + is the ceiling function.

3. The microgrid dispatching method based on electric vehicle dispatching potential and travel risk according to claim 1 is characterized in that: The objective function of the real-time stage scheduling model is as follows: The real-time stage is based on 15-minute nodes. At each time node, the optimization goal is to minimize the real-time adjustment cost of the microgrid and maximize the benefits of the electric vehicle alliance. The objective function is: in, is the incentive compensation cost that the microgrid needs to give to the electric vehicle alliance at time t, that is, the total income of the alliance, is the additional penalty cost that the microgrid needs to pay to the grid at time t, is the actual net load error, is the actual net load of the microgrid, VaR is the allowable error, The risk cost of selling charging flexibility and electricity to electric vehicles once, that is, the revenue of participating in scheduling alone; Z is the number of members in the grand alliance.

4. The microgrid dispatching method based on electric vehicle dispatching potential and travel risk according to claim 2 or 3, characterized in that: The day-ahead stage scheduling model is set with constraints, and the constraints set in the day-ahead stage scheduling model include microgrid power balance constraints, microgrid peak-to-valley difference constraints, microgrid interactive power constraints and EV charging constraints.

5. The microgrid dispatching method based on electric vehicle dispatching potential and travel risk according to claim 4 is characterized in that: The real-time stage scheduling model is set with constraints, and the constraints set by the real-time stage scheduling model include real-time power balance constraints, power grid constraints and EV constraints.

6. The microgrid dispatching method based on electric vehicle dispatching potential and travel risk according to claim 5 is characterized in that: In step S2, the parameters of the day-ahead scheduling model are initialized, including the industrial microgrid basic load Photovoltaic output power Fan output power Grid time-of-use electricity price, charging price, time period number T, number of electric vehicles N, rated capacity of electric vehicle batteries E, electric vehicle travel characteristics, electric vehicle charging power Charge and discharge efficiency 7. The microgrid dispatching method based on electric vehicle dispatching potential and travel risk according to claim 6 is characterized in that: In step S2, the parameters of the real-time scheduling model are initialized, including the time-of-use electricity price of the power grid, the charging price, the number of time periods T, the number of electric vehicles N, the rated capacity of the electric vehicle battery E, the electric vehicle travel characteristic quantity, and the electric vehicle charging and discharging power Charge and discharge efficiency Electric vehicle battery cycle life CL under nominal conditions nom , Depth of discharge DoD of electric vehicle batteries nom , Electric vehicle battery purchase and recycling prices 8. The microgrid dispatching method based on electric vehicle dispatching potential and travel risk according to claim 4 is characterized in that: In step S3, the method for solving the day-ahead scheduling model is as follows: Industrial microgrid base load Photovoltaic output power Fan output power Calculate the day-ahead forecast of industrial microgrid net load The expression is as follows: Industrial microgrid net load error Er t , the expression is as follows: is the historical actual net load power; Based on the historical net load error data of industrial microgrid and the day-ahead forecast data of industrial microgrid, the conditional risk cost F caused by net load uncertainty is calculated using the CVaR method. CVaR and the net load power tolerance range, or VaR; The number of electric vehicles N, the rated capacity of electric vehicle batteries E, the travel characteristics of electric vehicles, and the charging power of electric vehicles Charge and discharge efficiency According to the distribution of electric vehicle travel, the adjustable potential of electric vehicles is calculated using Minkowski additivity; The dispatchable potential of electric vehicles and the conditional risk cost F CVaR , predict the net load of industrial microgrid, build a mixed integer linear model with the goal of minimizing the day-ahead electricity cost of industrial microgrid, and use Gurobi solver to solve the mixed integer linear model to obtain the optimal power purchase and optimal charging power Form the optimal scheduling plan for the day-ahead phase.

9. The microgrid dispatching method based on electric vehicle dispatching potential and travel risk according to claim 8, characterized in that: The conditional risk cost F caused by the net load uncertainty CVaR The calculation method of the net load power allowable range VaR is as follows: Established the loss function in are decision variables and random variables, namely actual net load power and predicted net load power, respectively. Based on the historical data of predicted net load power, the probability density θ(Er t ), whose decision variables Determined with the threshold δ, the loss function Cumulative distribution function and loss function of as follows: Therefore, under a certain confidence level α∈(0,1), for a certain decision variable The expression of VaR function is as follows: Among them, the confidence level α is determined by the decision maker based on the internal resources of the system and the resource scheduling capacity, and R is a real number set; Through VaR, the expression of CVαR can be obtained as follows: Among them, N is the total number of samples, the threshold δ is an auxiliary variable, and its optimal value is VaR; Therefore, the conditional risk value cost F CVaR ,as follows:

10. The microgrid dispatching method based on electric vehicle dispatching potential and travel risk according to claim 9, characterized in that: In step S3, the method for calculating the adjustable potential of the electric vehicle is: The electric vehicle travel characteristics are obtained and initialized by the charging piles of the charging station in the microgrid, and the load model corresponding to the individual electric vehicle is established through the electric vehicle travel characteristics. The expression is as follows: in, are the charging and discharging rates of the electric vehicle at time t; They are the upper limits of the charging and discharging power of electric vehicles respectively; are Boolean variables of the charging and discharging status of the electric vehicle, is a Boolean variable indicating the arrival state of the electric vehicle. Electric vehicles can be charged and discharged at the same time; is the state of charge of the electric vehicle at time t, SOC x，min , SOC x，max They are the upper and lower limits of the state of charge of electric vehicles respectively; are the charging and discharging efficiency of electric vehicles respectively; Δt is the charging and discharging time period of electric vehicles; t ar ,t le are the arrival and departure times of electric vehicles, respectively, and t is the type of electric vehicle; In view of the differences in arrival and departure times between individual electric vehicles, Boolean variables are considered when calculating the charging and discharging power. The access time of all electric vehicles to the industrial microgrid is unified into the same definition domain through Boolean variables, and the power and electricity of the electric vehicle aggregate at the charging station are calculated using Minkowski additivity, thereby calculating the dispatchable potential Ω EV , the expression is as follows: in, They are the charging and discharging power of the electric vehicle aggregate at the charging station at time t and the amount of electricity of the electric vehicle aggregate at the charging station; They are the upper limits of charge and discharge of electric vehicle aggregates at the charging station at time t; are the upper and lower limits of the power of the electric vehicle aggregate at the charging station at time t; ΔSOC t is the change in the amount of electricity of the electric vehicle aggregate at the charging station at time t due to the driving behavior of the electric vehicle users.

11. The microgrid dispatching method based on electric vehicle dispatching potential and travel risk according to claim 10, characterized in that: In step S3, the solution method of the real-time stage scheduling model is: Obtain and initialize the expected battery charge of electric vehicle x through the electric vehicle charging pile The battery charge SOC at the current time t x，t , residence time T stop , the required charging time is T ch , the travel risk is modeled through the generalized bell-shaped membership function to calculate the travel risk of electric vehicles; Use the electric vehicle battery attenuation model to calculate the degree of battery power attenuation caused by the electric vehicle's sales of electricity; Through the electric vehicle travel risk function and the battery power attenuation degree, using the incentive-based electric vehicle compensation method, the risk cost of selling charging flexibility and power of the electric vehicle at that moment is calculated; Initialize the parameters related to industrial microgrid, electric vehicle, and cooperative alliance at time t, including the power gap of industrial microgrid, grid price, electric vehicle travel risk, and battery sales and recycling costs; Set the initial number of electric vehicles in the cooperative alliance M, and the expression is as follows: Among them, ΔP del The power gap of industrial microgrids is is the minimum non-zero discharge power of the cooperative alliance electric vehicle, The optimal power purchase power; Calculate the total response power of electric vehicles based on the number of participating electric vehicles Passed into the lower model in the two-layer optimization model; The Gurobi solver is used to optimize the real-time adjustment cost of the industrial microgrid and the total benefit of the cooperative alliance, and then pass them into the upper model of the two-layer optimization model. The upper model in the two-layer optimization model calculates the contribution of individual electric vehicles in the alliance according to the nonlinear energy mapping function, thereby calculating the benefits obtained by the individual electric vehicles in the alliance. The expression is as follows: in, are the output power of member i and the maximum output power of the members in the alliance respectively; is the risk factor of member i; is the nonlinear energy contribution value of member i, F(x) is the nonlinear mapping function, which uses the exponential function of the natural logarithm e; γ i is the overall contribution of member i; Determine whether the number of electric vehicles reaches the upper and lower limits. If so, stop the iteration and output the real-time adjustment cost and individual profit of electric vehicles. Otherwise, adjust the number of electric vehicles and recalculate the total response power of electric vehicles.

Images