An energy optimization scheduling method for integrated energy stations based on high-dimensional multi-objectives

Abstract

This invention relates to a high-dimensional, multi-objective energy optimization and scheduling method for integrated energy stations. The method first collects integrated energy station system data; then, it inputs the data into an integrated energy station energy optimization model, solves the model using an improved NSGA-III algorithm, and outputs optimized control parameters. This method enhances the algorithm's global search capabilities and improves the distribution of settlement results within the feasible domain. Furthermore, the improved algorithm significantly accelerates computational speed while ensuring computational accuracy.

Description

Technical Field

[0001] The present invention belongs to the field of energy scheduling and is used for energy scheduling of integrated energy stations. While meeting the charging needs of electric vehicles, it pursues maximizing station revenue, minimizing station power purchases, and minimizing energy storage system losses. In particular, it relates to an energy optimization scheduling method for integrated energy stations based on high-dimensional multi-objectives. Background Art

[0002] With the rapid rise of the electric vehicle industry, traditional gas stations are facing the transformation to electric vehicle charging stations to meet the needs of electric vehicles. Existing optimization and scheduling models for electric vehicle charging stations are mostly single-objective or dual-objective optimization models. These models are generally solved using linear programming, intelligent optimization algorithms or deep learning algorithms. There are few studies on high-dimensional multi-objective scheduling models. Moreover, the optimization and scheduling of electric energy is a multi-objective, multi-stage, multi-constrained, nonlinear complex optimization problem. The traditional single-objective, two-objective or planning-type modeling methods cannot meet the scheduling requirements in terms of accuracy when calculating the results. To this end, the present invention establishes an integrated energy station optimization and scheduling model with a photovoltaic system and an energy storage system. While meeting the power demand of electric vehicles, excess energy can also be fed back to the power grid to maximize profits. In addition, while maximizing profits, it is necessary to ensure that the loss of the energy storage system is minimized and the daily power purchase is minimized.

[0003] NSGA-III is a recently popular algorithm for solving three or more objectives. It incorporates a reference point mechanism into individual selection, resulting in significant breakthroughs in solving multi-problem problems. Initial data, such as the size of the population Np and the maximum number of iterations, is input. Based on the size of population Np, the location of the reference point is generated. At the beginning of the iteration, population Np is split into two populations NP1 and NP2 of equal size. Crossover and mutation operations are performed between each pair of individuals in populations NP1 and NP2 to generate population Ns. Populations Np and Ns are merged and substituted into the objective function to determine the objective function value. A non-dominated sorting operation is then performed on populations Np∪Ns. Excellent individuals are selected using the reference point mechanism and added to population Ng until the number of individuals in population Ng equals the number of individuals in population Np. This iterative process is repeated until the exit criterion is met, and the result is output. This method can be used to calculate a solution set close to the Pareto frontier for engineering problems. Finally, different solutions within the solution set can be selected based on specific needs.

[0004] When developing energy scheduling plans, it is crucial to consider the energy interactions between various systems and establish a multi-objective scheduling model. This model must then be transformed into a mathematical model consisting solely of multiple objective functions and constraints. Conventional single- or two-objective mathematical models fail to adequately consider the inter-system interactions and thus struggle to meet the needs of in-station scheduling. Traditional planning methods require a balanced approach to solving problems, lacking a unified solution. As the number of decision variables increases, the amount of data computation and storage required increases exponentially, leading to the "curse of dimensionality" when solving the problem. Summary of the Invention

[0005] The above technical problems of the present invention are mainly solved by the following technical solutions:

[0006] An energy optimization scheduling method for an integrated energy station based on high-dimensional multi-objectives is characterized by:

[0007] Collecting integrated energy station system data;

[0008] The integrated energy station system data is input into the integrated energy station energy optimization model, and the improved NSGA-Ⅲ algorithm is used to solve it and output the optimized control parameters.

[0009] In the above method, the integrated energy station includes a photovoltaic system model, an energy storage system model, and an energy storage system loss model, wherein:

[0010] The output power of the solar photovoltaic panel in the photovoltaic system model is expressed by the following formula:

[0011]

[0012] P pv (t) represents the photovoltaic power at time t; G(t) represents the solar radiation at time t; G sr is the rated solar radiation; P pvr Indicates the rated power of the photovoltaic panel; η pv is the power generation efficiency; β T is the temperature coefficient; T(t) represents the temperature of the environment at time t; T CT Indicates the temperature of the photovoltaic panel under normal conditions; T crT Indicates the reference temperature of the photovoltaic panel.

[0013] The energy state of the energy storage system model is expressed by the following formula:

[0014]

[0015] E S (t+1) represents the remaining energy of the energy storage system at t+1; E GS (t) represents the amount of charge from the power grid to the energy storage system at time t; EPVS (t) represents the amount of charge from the photovoltaic system to the energy storage system at time t; E SG (t) represents the amount of electricity fed back to the grid by the energy storage system at time t; E SK (t) represents the amount of electricity that the energy storage system charges the car at time t; P GS (t) represents the charging power of the power grid to the energy storage system at time t; P PVS (t) represents the charging power of the photovoltaic system to the energy storage system at time t; P SG (t) represents the discharge power of the energy storage system to the grid at time t; P SK (t) represents the charging power of the energy storage system to the car at time t. SX (t) represents the discharge power of the energy storage system at time t; η c Indicates charging efficiency; η d Represents the discharge efficiency; Δt = 1h. The energy storage system loss model is expressed using the following formula:

[0016] Q c =(α×SOCi+β)exp Η ×S Z

[0017]

[0018] Qc is the battery capacity loss rate; SOC is the initial state of charge of the battery; S represents the cumulative battery throughput; α, β, E a 、R g 、R g , Z, and Η are related parameters.

[0019] In the above method, the energy optimization model of the integrated energy station includes three objective functions: maximizing the daily profit within the station, minimizing the daily power purchase within the station, and minimizing the daily loss of the energy storage system within the station, as well as three constraints: energy balance constraint, power storage constraint, and energy flow constraint.

[0020] In the above method, the maximum daily profit within the station is calculated using the following formula:

[0021]

[0022] R is the daily profit of the integrated energy station; E SK (t) is the amount of electricity provided by the energy storage system to the car during period t; E GK (t) is the amount of electricity provided by the public power grid to the car during period t; E PVG (t) is the amount of electricity that the photovoltaic system transmits to the public grid during period t; E SG (t) is the amount of electricity delivered from the energy storage system to the public grid during period t; E GS (t) is the amount of electricity from the grid to the charging pile during period t; PeXK (t) is the charging price of the car in period t; Pe GX (t) is the grid electricity price during period t; Pe XG (t) is the price at which the station sells electricity to the grid during period t.

[0023] In the above method, the minimum daily power purchase within the station is calculated using the following formula:

[0024]

[0025] E G It is the daily electricity purchase amount of the integrated energy station, which mainly comes from the charging amount of the energy storage system from the public power grid and the power supply of the public power grid to the charging piles.

[0026] In the above method, the following formula is used for calculation:

[0027]

[0028] Q represents the total loss rate of the energy storage system in one day. C (t) is the loss rate of the energy storage system during period t.

[0029] In the above method, the energy balance constraint is calculated using the following formula:

[0030]

[0031]

[0032]

[0033] E K (t) represents the amount of electricity required by the car during period t; E PV (t) represents the power generation of the photovoltaic system during period t; E XS (t) represents the charge capacity of the energy storage system during period t; E SX (t) represents the discharge amount of the energy storage system during period t.

[0034] In the above method, the power storage constraint is calculated using the following formula:

[0035] MinE S ≤E S (t)≤MaxE S .

[0036] In the above method, the energy flow constraint is calculated using the following formula:

[0037] [S PVS (t)+S GS (t)]×S SW (t) = 0

[0038] [S PVS (t)+S GS (t)]×S SK (t) = 0

[0039] S PVS (t)×[S SG (t)+S SK (t)]=0

[0040] S GS (t)×[S SG (t)+S SK (t)]=0

[0041] S PVS (t) represents the charging state of the photovoltaic energy storage system, S GS (t) represents the charging status of the energy storage system from the public grid; S SW (t) represents the discharge state of the energy storage system to the public grid; S SK (t) represents the discharge state of the energy storage system to the charging pile. PVS (t), S GS (t), S SW (t), S SK (t) The value range is 0 or 1, 1 means the current state exists, and 0 means it is not in the current state.

[0042] In the above method, the solution process specifically includes

[0043] Step 1: Initialize the population Np, the maximum number of iterations Tmax, and the initial data of the integrated energy station optimization scheduling model;

[0044] Step 2: Select the population N according to the number of iterations P Four individuals (P1, P2, P3, P4) are selected from population N0, where population N0 is the dominant population composed of individuals in the first layer after non-dominated sorting.

[0045] 1) When the number of iterations is 1, P1, P2, P3, and P4 all come from population N. P elect;

[0046] 2) When the number of iterations is not 1, P1 and P2 are selected from population N0, and P3 and P4 are selected from population N P Selected from.

[0047] Step 3: Generate three random numbers (r1, r2, r3) and select different individual update methods based on the values ​​of the random numbers.

[0048] When r1<0.7&&r2<0.7, the individual updates its position using formula (1)

[0049]

[0050] N S (i) is the population N p The new individual after the position of the i-th individual is updated, k is a parameter with a value of 0 or 1.

[0051] When r1<0.7&&r2≥0.7, the individual updates its position using formula (2)

[0052]

[0053] When r1≥0.7&&r3<0.9, the individual uses formula (3) to update the position

[0054] N S (i) = λ × N P (i)×e α×δ ×cos(δ×1.5π)+P1 (3)

[0055] N P (i) represents the population N P λ, α, and δ are the parameters in the position update equation, which can be expressed by formulas (4)-(7).

[0056]

[0057]

[0058]

[0059]

[0060] rand is a random number between 0 and 1; t is the current iteration number.

[0061] When r1≥0.7&&r3≥0.9, the individual updates its position using formula (8)

[0062] N S (i) = N P (i)×Levy(D)×e α×δ +N P (i) (8)

[0063] Levy stands for levy flight, and D is the dimension of decision variables.

[0064] Step 4: Put the individuals updated by equations (1)-(8) into the offspring population N S , and N pop =N P ∪NS .

[0065] Through the following three principles, the population N pop All individuals are transformed into E GK (t), E SK (t), E GS (t), E SG (t), E PVG (t), E PVS (t). Substitute the above six values ​​into the objective function and find the corresponding objective function value.

[0066] 1) When the energy storage system is in a charging state, the power generated by the photovoltaic system during the current period is preferentially transmitted to the energy storage system. When the photovoltaic power generation is greater than the energy storage system's charge capacity, the photovoltaic system transmits the excess energy to the grid. When the photovoltaic power generation is less than the energy storage system's charge capacity, the grid charges the energy storage system. At this time, the power demand of the electric vehicle is completely provided by the grid;

[0067] 2) When the energy storage system is discharging, the energy storage system prioritizes the electric vehicle's power needs. If the electric vehicle's power needs exceed the energy storage system's maximum capacity for the current period, the remaining power will be provided by the grid. If the electric vehicle's power needs are less than the energy storage system's maximum capacity for the current period, the excess power provided by the energy storage system will be transferred to the grid. At this point, the photovoltaic power generation is fully transferred to the grid.

[0068] 3) When the energy storage system is neither charging nor discharging, the photovoltaic power generation is fully transmitted to the grid.

[0069] The electricity demanded by electric vehicles is completely provided by the power grid.

[0070] Substitute the 6 transformed values ​​of each individual into all objective functions, solve the results, and perform non-dominated sorting on all individuals (individuals will be divided into F1, F2, ... layers). All individuals in the F1 layer are placed in the dominant population N0.

[0071] Step 5: Starting from the first layer, put the individuals in each layer into the population S t until the population S t The number of |S t |≥N, the current accumulated number of layers is recorded as F l . And according to the population S t The final number of individuals is selected and placed into the new parent population

[0072] If |S t |=N, then F1,F2,…,F l All individuals in the new parent population are placed in the new parent population.

[0073] If |St |≥N, then F l The individuals in each layer are normalized using formulas (9)-(11) and the reference points are generated.

[0074]

[0075]

[0076]

[0077] where a i represents the intercept between the axis of the i-th target and the linear hyperplane; M represents the number of objective functions; Represents the population S t The minimum value of the i-th objective function in the solution set; w i =10 -6 ;f i ′ represents the target of the i-th transformation.

[0078] Step 6: Calculate F l The vertical distance between all individuals in the layer and each reference point, and F l All individuals in the layer are connected to their corresponding reference points, where the vertical distance can be calculated using formula (12).

[0079]

[0080] Z r Represents the set of generated reference points

[0081] Step 7: Calculate the number of individuals related to the reference point j, denoted as ρ j . Calculate the number of individuals to be selected (expressed as K) and added to the new parent population. Where K = N P -|N S |.

[0082] Step 8: Randomly select a non-repeated reference point. If the ρ of the reference point j ≥1, select the individual with the smallest vertical distance to the selected reference point. If the ρ of all reference points j ≥1, but k<K, then choose the reference point ρ j Individuals with ≥2 are placed into the new parent population. The specific individual selection steps are shown in Algorithm 1.

[0083] Step 9: Repeat steps 2-8 until the exit condition is met.

[0084]

[0085] Therefore, the present invention has the following advantages: The present invention introduces photovoltaic systems and energy storage systems into integrated energy stations, establishes a multi-objective optimization and scheduling model for integrated photovoltaic and energy storage stations, reduces electricity purchase costs, and improves the economic benefits within the station. Compared with traditional car charging stations, the station's revenue is increased, and the electricity purchase cost during periods of high grid load is reduced; and an improved NSGA-III algorithm (BML-NSGA-III) is proposed by combining the BWOA position update strategy, the MOA spiral search strategy, and the Levy flight strategy. The global search capability of the algorithm is enhanced, and the distribution of settlement results within the feasible domain is improved. In addition, the improved algorithm significantly speeds up the calculation speed while ensuring calculation accuracy. BRIEF DESCRIPTION OF THE DRAWINGS

[0086] Figure 1 This is the structural diagram of the photovoltaic and energy storage integrated energy station;

[0087] Figure 2 It is the energy flow diagram of the integrated energy station;

[0088] Figure 3 It is a schematic diagram of electricity prices in different time periods of the public power grid;

[0089] Figure 4 It is a schematic diagram of the load situation of the integrated energy station in each period;

[0090] Figure 5 This is a schematic diagram of photovoltaic power generation in each period of the integrated energy station;

[0091] Figure 6 Schematic diagram of car charging prices at different time periods;

[0092] Figure 7 It is a schematic diagram of the electricity price that the station sells to the grid in each period;

[0093] Figure 8 This is a schematic diagram comparing the amount of electricity purchased by the photovoltaic and energy storage integrated energy station and the traditional charging station in different time periods;

[0094] Figure 9 It is a schematic diagram of the charging and discharging status of the energy storage system in each period;

[0095] Figure 10 This is a diagram comparing the benefits of nine optimization schemes in a solar-storage integrated energy station with those of a traditional charging station.

[0096] Figure 11 This is a comparison chart between BML-NSGA-Ⅲ and seven other algorithms. DETAILED DESCRIPTION

[0097] The technical solution of the present invention will be further specifically described below through embodiments and in conjunction with the accompanying drawings.

[0098] Example:

[0099] The present invention introduces photovoltaic system and energy storage system into the integrated energy station, and its structure is as follows: Figure 1 The energy flow configuration diagram of the photovoltaic integrated energy station is shown in Figure 2 As shown in the figure, the core component of the photovoltaic system (photovoltaic panels) will be installed on the roof of the integrated energy station building. Three types of inverters are used: DC-DC inverter, DC-AC inverter, and AC-DC inverter. The DC-DC inverter is used to supply photovoltaic power to the station's energy storage system or to the charging station. The DC-AC inverter transmits photovoltaic power to the public grid. The AC-DC inverter is used to charge the energy storage system or power the charging station.

[0100] 1.1 Photovoltaic system model

[0101] The output power of solar photovoltaic panels is as follows:

[0102]

[0103] P pv (t) represents the photovoltaic power at time t. G(t) represents the solar radiation at time t. sr is the rated solar radiation. pvr Indicates the rated power of the photovoltaic panel. η pv is the power generation efficiency. T is the temperature coefficient. T(t) represents the temperature of the environment at time t. CT Indicates the temperature of the photovoltaic panel under normal conditions. T crT Indicates the reference temperature of the photovoltaic panel.

[0104] 1.2 Energy Storage System Model

[0105] The energy storage system of the photovoltaic and storage integrated energy station can be charged by the public power grid, and can also supply power to the charging pile or discharge to the public power grid. Figure 2 As shown in Figure 2, the photovoltaic integrated energy station has six energy flow directions. The energy of the energy storage system comes from the public power grid and the photovoltaic system. Therefore, the current energy state of the energy storage system can be expressed by formula (2).

[0106]

[0107] E S (t+1) represents the remaining energy of the energy storage system at t+1; E GS (t) represents the amount of charge from the power grid to the energy storage system at time t; E PVS (t) represents the amount of charge from the photovoltaic system to the energy storage system at time t; E SG (t) represents the amount of electricity fed back to the grid by the energy storage system at time t; E SK(t) represents the amount of electricity that the energy storage system charges the car at time t; P GS (t) represents the charging power of the power grid to the energy storage system at time t; P PVS (t) represents the charging power of the photovoltaic system to the energy storage system at time t; P SG (t) represents the discharge power of the energy storage system to the grid at time t; P SK (t) represents the charging power of the energy storage system to the car at time t. SX (t) represents the discharge power of the energy storage system at time t; η c Indicates charging efficiency; η d Indicates discharge efficiency; Δt=1h.

[0108] 1.3 Energy Storage System Loss Model

[0109] Q c =(α×SOCi+β)exp Η ×S Z (15)

[0110]

[0111] Qc is the battery capacity loss rate; SOC is the initial state of charge of the battery; S represents the cumulative battery throughput; α, β, E a 、R g 、R g , Z, and Η are related parameters.

[0112] 1.4 Objective Function

[0113] The solar-energy storage integrated energy station is converted into a mathematical model with three objective functions. The remaining power of the energy storage system at each time period will be used as a decision variable.

[0114] Objective function 1: Maximize daily profit within the site

[0115]

[0116] R is the daily profit of the integrated energy station; E SK (t) is the amount of electricity provided by the energy storage system to the car during period t; E GK (t) is the amount of electricity provided by the public power grid to the car during period t; E PVG (t) is the amount of electricity that the photovoltaic system transmits to the public grid during period t; E SG (t) is the amount of electricity delivered from the energy storage system to the public grid during period t; E GS (t) is the amount of electricity from the grid to the charging pile during period t; Pe XK (t) is the charging price of the car in period t; Pe GX (t) is the grid electricity price during period t; Pe XG(t) is the price at which the station sells electricity to the grid during period t.

[0117] Objective function 2: Minimize the daily electricity purchase within the station

[0118]

[0119] E G It is the daily electricity purchase amount of the integrated energy station, which mainly comes from the charging amount of the energy storage system from the public power grid and the power supply of the public power grid to the charging piles.

[0120] Objective function 3: Minimize daily energy storage system losses within the station

[0121]

[0122] Q represents the total loss rate of the energy storage system in one day. C (t) is the loss rate of the energy storage system during period t. The loss rate of the energy storage system can be calculated using equations (3)-(4).

[0123] 1.5 Constraints

[0124] (1) Energy balance constraints

[0125] The amount of energy generated in a photovoltaic and energy storage integrated energy station should be equal to the amount of energy consumed, which can be expressed by equations (8)-(9).

[0126]

[0127]

[0128]

[0129] E K (t) represents the amount of electricity required by the car during period t; E PV (t) represents the power generation of the photovoltaic system during period t; E XS (t) represents the charge capacity of the energy storage system during period t; E SX (t) represents the discharge amount of the energy storage system during period t.

[0130] (2) Inequality constraints

[0131] The amount of electricity stored in the energy storage system must meet the constraints, as shown in formula (11).

[0132]

[0133] (3) Energy flow constraints

[0134] The energy storage system cannot have charging and discharging behaviors at the same time. Therefore, the following constraints must be met, as shown in equations (12)-(15).

[0135] [S PVS (t)+S GS (t)]×S SW (t)=0 (24)

[0136] [S PVS (t)+S GS (t)]×S SK (t)=0 (25)

[0137] S PVS (t)×[S SG (t)+S SK (t)]=0 (26)

[0138] S GS (t)×[S SG (t)+S SK (t)]=0 (27)

[0139] S PVS (t) represents the charging state of the photovoltaic energy storage system, S GS (t) represents the charging status of the energy storage system from the public grid; S SW (t) represents the discharge state of the energy storage system to the public grid; S SK (t) represents the discharge state of the energy storage system to the charging pile. PVS (t), S GS (t), S SW (t), S SK (t) The value range is 0 or 1, 1 means the current state exists, and 0 means it is not in the current state.

[0140] 1.6 An Improved NSGA-III Algorithm

[0141] Step 1: Initialize the population Np, the maximum number of iterations Tmax, and the initial data of the integrated energy station optimization scheduling model;

[0142] Step 2: Select the population N according to the number of iterations P Four individuals (P1, P2, P3, P4) are selected from population N0, where population N0 is the dominant population composed of individuals in the first layer after non-dominated sorting.

[0143] 1) When the number of iterations is 1, P1, P2, P3, and P4 all come from population N. P elect;

[0144] 2) When the number of iterations is not 1, P1 and P2 are selected from population N0, and P3 and P4 are selected from population N P Selected.

[0145] Step 3: Generate three random numbers (r1, r2, r3) and select different individual update methods based on the values ​​of the random numbers.

[0146] When r1<0.7&&r2<0.7, the individual updates its position using formula (16)

[0147]

[0148] N S (i) is the population N p The new individual after the position of the i-th individual is updated, k is a parameter with a value of 0 or 1.

[0149] When r1<0.7&&r2≥0.7, the individual updates its position using formula (17)

[0150]

[0151] When r1≥0.7&&r3<0.9, the individual updates its position using formula (18)

[0152] N S (i) = λ × N P (i)×e α×δ ×cos(δ×1.5π)+P1 (30)

[0153] N P (i) represents the population N P λ, α, and δ are the parameters in the position update equation, which can be expressed by formulas (19)-(22).

[0154]

[0155]

[0156]

[0157]

[0158] rand is a random number between 0 and 1; t is the current iteration number.

[0159] When r1≥0.7&&r3≥0.9, the individual updates its position using formula (23)

[0160] N S (i) = N P (i)×Levy(D)×e α×δ +N P (i) (35)

[0161] Levy stands for levy flight, and D is the dimension of decision variables.

[0162] Step 4: Put the individuals updated by equations (16)-(23) into the offspring population N S , and N p o p =N P ∪N S .

[0163] Through the following three principles, the population N pop All individuals are transformed into E GK (t), E SK (t), E GS (t), E SG (t), E PVG (t), E PVS (t). Substitute the above six values ​​into the objective function and find the corresponding objective function value.

[0164] 1) When the energy storage system is in a charging state, the power generated by the photovoltaic system during the current period is preferentially transmitted to the energy storage system. When the photovoltaic power generation is greater than the energy storage system's charge capacity, the photovoltaic system transmits the excess energy to the grid. When the photovoltaic power generation is less than the energy storage system's charge capacity, the grid charges the energy storage system. At this time, the power demand of the electric vehicle is completely provided by the grid;

[0165] 2) When the energy storage system is discharging, the energy storage system prioritizes the electric vehicle's power needs. If the electric vehicle's power needs exceed the energy storage system's maximum capacity for the current period, the remaining power will be provided by the grid. If the electric vehicle's power needs are less than the energy storage system's maximum capacity for the current period, the excess power provided by the energy storage system will be transferred to the grid. At this point, the photovoltaic power generation is fully transferred to the grid.

[0166] 3) When the energy storage system is neither charging nor discharging, the photovoltaic power generation is fully transmitted to the power grid, and the power demand of electric vehicles is fully provided by the power grid.

[0167] Substitute the 6 transformed values ​​of each individual into all objective functions, solve the results, and perform non-dominated sorting on all individuals (individuals will be divided into F1, F2, ... layers). All individuals in the F1 layer are placed in the dominant population N0.

[0168] Step 5: Starting from the first layer, put the individuals in each layer into the population S t until the population S t The number of |S t |≥N, the current accumulated number of layers is recorded as F l . And according to the population S tThe final number of individuals is selected and placed into the new parent population

[0169] If |S t |=N, then F1,F2,…,F l All individuals in the new parent population are placed in the new parent population.

[0170] If |S t |≥N, then F l The individuals in each layer are normalized using formulas (24)-(26) and the reference points are generated.

[0171]

[0172]

[0173]

[0174] where a i represents the intercept between the axis of the i-th target and the linear hyperplane; M represents the number of objective functions; Represents the population S t The minimum value of the i-th objective function in the solution set; w i =10 -6 ;f i ′ represents the target of the i-th transformation.

[0175] Step 6: Calculate F l The vertical distance between all individuals in the layer and each reference point, and F l All individuals in the layer are connected to their corresponding reference points, where the vertical distance can be calculated using formula (27).

[0176]

[0177] Z r Represents the set of generated reference points

[0178] Step 7: Calculate the number of individuals related to the reference point j, denoted as ρ j . Calculate the number of individuals to be selected (expressed as K) and added to the new parent population. Where K = N P -|N S |.

[0179] Step 8: Randomly select a non-repeated reference point. If the ρ of the reference point j ≥1, select the individual with the smallest vertical distance to the selected reference point. If the ρ of all reference points j ≥1, but k<K, then choose the reference point ρ j Individuals with ≥2 are placed into the new parent population. The specific individual selection steps are shown in Algorithm 1.

[0180] Step 9: Repeat steps 2-8 until the exit condition is met.

[0181]

[0182] 1.7 Analysis of calculation results of integrated energy station optimization model

[0183] 1.7.1 Basic Data

[0184] In the present invention, the time-of-use electricity price of the power grid ( Figure 3 ), the traffic flow in the station at each time period ( Figure 4 ), photovoltaic power generation in each period ( Figure 5 ), car charging prices at different time periods ( Figure 6 ), the electricity price of the station selling electricity to the grid in each period ( Figure 7 ) is the basic input data. The photovoltaic power generation in each period can be calculated using formula (1).

[0185] 1.7.2 Comparison between Solar-Storage Integrated Energy Stations and Traditional Charging Stations

[0186] from Figure 8 As can be seen, during periods of high grid load (10:00-15:00 and 18:00-21:00), the solar-plus-storage integrated energy station purchases less electricity than traditional vehicle charging stations. However, during periods of low grid load, the solar-plus-storage integrated energy station purchases more electricity than traditional vehicle charging stations. This approach can both increase the station's daily revenue and reduce energy waste.

[0187] Figure 9 This is a set of optimization solutions in a centralized solution. The figure shows the energy storage system's charging and discharging quantities for each time period. It can be seen that the energy storage system's charging period is concentrated during the period of low grid load (0:00-7:00). The purpose is to store the electricity during this period so that the energy storage system can power electric vehicles during periods of higher electricity prices. The energy storage system's discharging period is concentrated during periods of high grid load (10:00-14:00 and 18:00-21:00). This is to save station electricity purchase costs and reduce grid load. The energy storage system also charges during the 16:00-18:00 period, which is the grid's normal electricity price. Its purpose is to meet the demand of electric vehicles in the next period and reduce the amount of electricity purchased from the grid by the station during the period of high grid load (18:00-22:00).

[0188] Figure 10This figure compares the daily revenue of a photovoltaic-storage integrated energy station with that of a traditional vehicle charging station. The first nine scenarios represent nine data sets selected arithmetic progression from the non-dominated solution set, while the remaining scenario shows the daily revenue of a traditional vehicle charging station. It can be seen that the daily revenue of the photovoltaic-storage integrated energy station is higher than that of a traditional vehicle charging station. This is because photovoltaic-storage integrated energy stations can store energy when electricity prices are low; when prices are high, the energy storage system charges vehicles, reducing the amount of electricity purchased from the grid during this period. Furthermore, the photovoltaic system can also provide electricity during the day, reducing the amount of electricity purchased from the grid by the station and saving on electricity purchase costs.

[0189] 1.8 Comparison of BML-NSGA-III and other 7 algorithms

[0190] To demonstrate the effectiveness of the BML-NSGA-III algorithm, eight algorithms (ANSGA-III, BIGE, θ-DEA, KNEA, NSGA-III, RPEA, SPEAR, and BML-NSGA-III) were used to calculate the optimal scheduling model for a solar-energy storage integrated energy station. All algorithms had a population size Np = 200 and a maximum number of iterations Tmax = 200. Each algorithm was run 10 times. Figure 11 is the calculation result.

[0191] Table 1 Comparison of eight algorithms in terms of Spread when solving the photovoltaic and energy storage integrated energy station model

[0192]

[0193]

[0194] Table 2 Comparison of eight algorithms in terms of HV when solving the photovoltaic and energy storage integrated energy station model

[0195]

[0196] Table 3 Comparison of the computational time of eight algorithms in solving the photovoltaic and energy storage integrated energy station model (seconds)

[0197]

[0198] Tables 1-3 show that BM-NSGA-III outperforms the other seven algorithms in terms of speard, HV, and computation time. Table 4 shows that the non-dominated solution sets calculated by the BML-NSGA-III algorithm are more evenly distributed within the feasible region than those of the other seven algorithms. Table 2 shows that the non-dominated solution sets calculated by the BML-NSGA-III algorithm are more widely distributed within the feasible region than those of the other seven algorithms. Table 3 shows that the computation time of the BML-NSGA-III algorithm is shorter than that of the other seven algorithms, indicating that the BML-NSGA-III algorithm has a lower computational complexity.

[0199] The specific embodiments described herein are merely illustrative of the spirit of the present invention. Persons skilled in the art may make various modifications, additions, or substitutions to the described specific embodiments without departing from the spirit of the present invention or exceeding the scope of the appended claims.

Claims

1. A method for optimizing energy dispatching of integrated energy stations based on high-dimensional multi-objectives, characterized in that: Collecting integrated energy station system data; The integrated energy station system data is input into the integrated energy station energy optimization model, and the improved NSGA-Ⅲ algorithm is used to solve it and output the optimized control parameters. The integrated energy station includes a photovoltaic system model, an energy storage system model, and an energy storage system loss model, among which: The output power of the solar photovoltaic panel in the photovoltaic system model is expressed by the following formula: represents the photovoltaic power at time t; represents the solar radiation at time t; is the rated solar radiation; Indicates the rated power of the photovoltaic panel; is the power generation efficiency; is the temperature coefficient; represents the temperature of the environment at time t; Indicates the temperature of the photovoltaic panel under normal conditions; Indicates the reference temperature of the photovoltaic panel; The energy state of the energy storage system model is expressed by the following formula: Represents the remaining energy of the energy storage system at t+1; represents the amount of charge from the power grid to the energy storage system at time t; represents the amount of charge from the photovoltaic system to the energy storage system at time t; represents the amount of electricity fed back to the grid by the energy storage system at time t; represents the amount of electricity charged to the car by the energy storage system at time t; represents the charging power of the grid to the energy storage system at time t; represents the charging power of the photovoltaic system to the energy storage system at time t; represents the discharge power of the energy storage system to the grid at time t; represents the charging power of the energy storage system to charge the car at time t; represents the discharge power of the energy storage system at time t; Indicates charging efficiency; Indicates discharge efficiency; ; The energy storage system loss model is expressed by the following formula: is the battery capacity loss rate; It is the initial state of battery charging; Indicates the cumulative battery throughput; 、 、 、 、 、 、 are relevant parameters; The energy optimization model of the integrated energy station includes three objective functions: maximizing daily profit within the station, minimizing daily power purchase within the station, and minimizing daily loss of the energy storage system within the station, as well as three constraints: energy balance constraint, power storage constraint, and energy flow constraint. The maximum daily profit on the site is calculated using the following formula: is the daily profit of the integrated energy station; is the amount of electricity the energy storage system provides to the car during period t; is the amount of electricity provided by the public grid to the car during period t; is the amount of electricity delivered from the photovoltaic system to the public grid during period t; is the amount of electricity delivered from the energy storage system to the public grid during period t; is the amount of electricity from the grid to the charging pile during period t; is the charging price of the car during period t; is the grid electricity price during period t; is the price at which the station sells electricity to the grid during period t; The minimum daily electricity purchase within the station is calculated using the following formula: The daily electricity purchase amount of the integrated energy station is mainly derived from the amount of electricity the energy storage system charges from the public grid and the amount of power supplied by the public grid to the charging piles. Calculated using the following formula: Indicates the total loss rate of the energy storage system in one day; is the loss rate of the energy storage system during period t.

2. The energy optimization scheduling method for an integrated energy station based on high-dimensional multi-objectives according to claim 1 is characterized in that: The energy balance constraint is calculated using the following formula: It represents the power demand of the car during period t; represents the power generation of the photovoltaic system during period t; represents the charging capacity of the energy storage system during period t; Indicates the discharge amount of the energy storage system during period t.

3. The energy optimization scheduling method for an integrated energy station based on high-dimensional multi-objectives according to claim 1 is characterized in that: The energy storage constraint is calculated using the following formula: 。 4. The energy optimization scheduling method for an integrated energy station based on high-dimensional multi-objectives according to claim 1 is characterized in that: The energy flow constraint is calculated using the following formula: Indicates the charging status of photovoltaic energy storage system. Indicates the charging status of the energy storage system from the public grid; Indicates the discharge status of the energy storage system to the public grid; Indicates the discharge status of the energy storage system to the charging pile; 、 、 、 The value range is 0 or 1, 1 means the current state exists, and 0 means it is not in the current state.

5. The energy optimization scheduling method for an integrated energy station based on high-dimensional multi-objectives according to claim 1 is characterized in that: The solution process specifically includes Step 1: Initialize the population Np, the maximum number of iterations Tmax, and the initial data parameters of the integrated energy station optimization scheduling model; Step 2: Select the population according to the number of iterations and populations Four individuals were selected ( 、 、 、 ); where population It is the dominant population composed of individuals in the first layer after non-dominated sorting; 1) When the number of iterations is 1, 、 、 、 All from the population elect; 2) When the number of iterations is not 1, 、 From the population Selected from 、 From the population selected from Step 3: Generate 3 random numbers ( , , ), select different individual update methods according to the random number value; when <0.7 && <0.7, the individual uses formula (1) to update the position (1) It is a population The new individual after the position of the i-th individual is updated, It is a parameter, the value is 0 or 1; when <0.7 && ≥0.7, the individual uses formula (2) to update the position (2) when ≥0.7 && <0.9, the individual uses formula (3) to update the position (3) Represents population The i-th individual in ; 、 、 is the parameter in the position update equation, expressed by formulas (4)-(7); (4) (5) (6) (7) is a random number between 0 and 1; is the current iteration number; when ≥0.7 && ≥0.9, the individual uses formula (8) to update the position (8) On behalf of Levy Flight, is the dimension of the decision variables; Step 4: Put the individuals updated by equations (1)-(8) into the offspring population ,and ; The population is divided into three categories: All individuals are transformed into 、 、 、 、 、 ;Substitute the above six values ​​into the objective function and find the corresponding objective function value; 1) When the energy storage system is in a charging state, the power generated by the photovoltaic system during the current period is preferentially transmitted to the energy storage system. When the photovoltaic power generation is greater than the energy storage system's charge capacity, the photovoltaic system transmits the excess energy to the grid. When the photovoltaic power generation is less than the energy storage system's charge capacity, the grid charges the energy storage system. At this time, the power demand of the electric vehicle is fully provided by the grid. 2) When the energy storage system is in a discharging state, the power demanded by the electric vehicle is first provided by the energy storage system. If the power demanded by the electric vehicle is greater than the maximum power provided by the energy storage system during the current period, the remaining power demanded by the grid is provided. If the power demanded by the electric vehicle is less than the maximum power provided by the energy storage system during the current period, the excess power provided by the energy storage system is transmitted to the grid. At this point, the photovoltaic power generation is fully transmitted to the grid. 2) When the energy storage system is neither charging nor discharging, the photovoltaic power generation is fully transmitted to the grid, and the power demand of electric vehicles is fully provided by the grid; Substitute the 6 transformed values ​​of each individual into all objective functions, solve the respective results, and sort all individuals into non-dominated order. The individuals will be divided into layer; and All individuals in the layer are placed into the dominant population middle; Step 5: Starting from the first layer, place the individuals in each layer into the population until the population The number of , the current accumulated number of layers is recorded as ; and according to population The final number of individuals is selected and placed into the new parent population if , then All individuals in the layer are placed into the new parent population; if , then The individuals in each layer are normalized using formulas (9)-(11) and the reference points are generated. (9) (10) (11) in represents the intercept between the axis of the i-th target and the linear hyperplane; Indicates the number of objective functions; Represents population The minimum value of the i-th objective function in the solution set; ; represents the target of the i-th transformation; Step 6: Calculation The vertical distance between all individuals in the layer and each reference point, and All individuals in the layer are connected to their corresponding reference points, where the vertical distance is calculated by formula (12); (12) Represents the set of generated reference points Step 7: Calculation and reference points The number of related individuals is recorded as ; Calculate the number of individuals to be selected, expressed as , joins the new parent population; ; Step 8: Randomly select a non-repeated reference point; if the reference point , select the individual with the smallest vertical distance to the selected reference point; if all the reference points ,but , then select the reference point Individuals are placed into the new parent population; Step 9: Repeat steps 2-8 until the exit condition is met.

Images