A method for operating a battery swap station based on photovoltaic power generation

Abstract

The present invention relates to the technical field of battery swap station operation, and in particular to a method for operating a battery swap station based on photovoltaic power generation. The present invention adopts a method based on BP neural network to predict photovoltaic power generation power, so that the data is more accurate. The present invention adopts immune genetic algorithm and shortest path algorithm model to plan the layout of battery swap stations, which can quickly find the optimal path between charging demand points and the optimal charging points of battery swap stations in the entire map network. The present invention constructs a direct operation model of photovoltaic battery swap stations to reduce the charging costs of battery swap stations. The present invention constructs an optimized operation model of photovoltaic battery swap stations under the participation of the electricity market to increase the additional income of battery swap stations.

Description

Technical Field

[0001] The present invention relates to the technical field of battery swap station operation, and in particular to a battery swap station operation method based on photovoltaic power generation. Background Art

[0002] With the development of economy and people's pursuit of higher quality of life, over-exploitation of resources has become an inevitable trend. Rapid development has caused huge environmental pollution problems, seriously affecting the development of the country and the health of the people. Under the national green development policy, vigorously promoting the popularization of electric vehicles has become an important measure to solve greenhouse gas emissions; in addition, with the advancement of electric vehicles and related technologies, people's recognition of electric vehicles has continued to increase.

[0003] Due to some characteristics of electric energy, electric vehicles are currently unable to achieve instant energy replenishment. For some specific groups of people, fast charging is still not very convenient. Battery swap charging uses replacing electric vehicle batteries as the main "charging" method, and the battery swap operation can be completed within 3-5 minutes, which greatly solves the problem of long charging time caused by plug-in charging, and can achieve true "fast" charging.

[0004] In recent years, the State Grid Corporation of China has built a large number of public transportation battery swap stations across the country, which has greatly promoted the development of the battery swap model. In the battery swap model, the batteries can be managed uniformly in the battery swap station, which can not only improve the service life of the battery and reduce the waste of resources; it can also alleviate the overload problem of the distribution system caused by the disorderly charging of electric vehicles. In addition to the "fast charging" function, the large number of power batteries contained in the battery swap station can also be used as energy storage to better participate in grid regulation, play a role in peak load regulation, and accept more renewable energy generation.

[0005] At present, although a certain number of battery swap stations have been put into use, there is a lack of relevant technical research on the analysis of taxi battery swap demand and the optimal capacity planning of battery swap stations, which makes it impossible to reasonably allocate resources in the construction and operation of battery swap stations. Blindly expanding the scale of battery swap stations will increase investment while causing problems such as waste of resources; on the contrary, the battery inventory is too small to meet the necessary battery swap demand, which restricts the development of the battery swap model.

[0006] In view of this, some battery swap station operation strategies are needed to help promote the construction and development of battery swap stations. Summary of the invention

[0007] In order to solve the above problems, the present invention provides a method for operating a battery swap station based on photovoltaic power generation. The specific technical solution is as follows:

[0008] A method for operating a battery swap station based on photovoltaic power generation, comprising the following steps:

[0009] Step S1, the photovoltaic power generation device is connected to the battery swap station, and the photovoltaic power generation power is predicted by combining environmental parameters and using a BP neural network model;

[0010] Step S2, using immune genetic algorithm combined with shortest path algorithm to plan the layout of the battery swap station;

[0011] Step S3, based on the photovoltaic power generation data, according to the planning and layout of the battery swap station, with the minimum daily charging cost of the battery swap station as the goal, establish a photovoltaic-battery swap station direct operation model, and calculate the power supply of the battery swap station;

[0012] Step S4, based on the photovoltaic power generation data, according to the planning and layout of the battery swap station, with the minimum daily charging cost of the battery swap station as the goal, establish a photovoltaic-battery swap station optimization operation model, and calculate the power available for the battery swap station; in the photovoltaic-battery swap station optimization operation model, the battery swap station sells part of the electricity while meeting its own battery swap needs, and the electricity price when the battery swap station sells electricity is higher than the market price.

[0013] Preferably, the step S1 specifically includes the following steps:

[0014] S11: Normalize the historical data of environmental parameters and convert them into data in the [0.1] interval. The calculation formula is as follows:

[0015]

[0016] The historical data of environmental parameters include solar radiation intensity, ambient temperature, wind speed, humidity and visibility. In the formula, p(t) represents the environmental parameter data at time t; p max (t) and p min (t) represents the maximum and minimum values ​​of the environmental parameter data at time t; p * (t) is the normalized data;

[0017] S12: Train the BP model. The sample input matrix is ​​X=(X1, X2, …X n ), where the row vector represents the input sample and the column vector X1,X2,…X n They represent a set of sample values ​​respectively, and the number of input samples is relatively large; each set of model input variables is x = [x1, x2, x3, x4, x5] T , where x1, x2, x3, x4, and x5 represent solar radiation intensity, ambient temperature, wind speed, humidity, and visibility respectively; the output variable is the photovoltaic power generation corresponding to each set of sample data;

[0018] Step S13, establish a BP neural network model, and during the modeling process, determine the number of hidden layer nodes m represents the number of input layer nodes;

[0019] Step S14, using the Levenberg-Marquardt optimization method of the BP network to train the BP neural network model, the weight and threshold update formula is:

[0020] X k+1 =X k -(J T J+uI) -1 J T e;

[0021] Among them, X k is the kth iteration data, X k+1 is the k+1th iteration data, J is the Jacobian matrix of the error differential with respect to the weight, e is the error vector, and u is a scalar;

[0022] The training function with the Levenberg-Marquardt optimization method in the matlab toolbox is trainlm.m, and its training parameters include: training times epochs, training step length show, error function indicator goal, training time time, minimum gradient min-grad, memory reduction coefficient men-reduc, and the initial value, increase coefficient, decrease coefficient and maximum value of u; step S15, obtaining the environmental parameters of the prediction day, normalizing the environmental parameters and inputting them into the trained BP neural network model for prediction, so as to obtain the photovoltaic power generation power of the prediction day.

[0023] Preferably, the step S2 specifically includes the following steps:

[0024] Step S21, constructing a cost objective function of a battery swap station, where the cost of constructing a battery swap station includes the construction investment cost of the battery swap station, the network loss cost of the battery swap station, and the user queuing time cost of the battery swap station;

[0025] Step S22, constructing the constraints of the cost objective function,

[0026] Step S23, using the Stra Dijk algorithm to calculate the optimal path between the charging demand point and the optimal charging point of the battery swap station in the entire map network;

[0027] Step S24, using a genetic immune algorithm to optimize the layout planning of the battery swap station.

[0028] Preferably, the construction investment cost of the battery swap station in step S21 is as follows:

[0029]

[0030] Among them, C Ji is the purchase cost of the ith battery swap station, C Wi is the operation and maintenance cost of the i-th battery swap station; I is the number of battery swap stations; the purchase cost of the i-th battery swap station is calculated as follows:

[0031]

[0032] Where: λ0 represents the rate of return of the battery swap station; A i represents the building area of ​​the i-th battery swap station; D i represents the land price of the i-th battery swap station; E i represents the distribution transformer capacity of the i-th battery swap station; D Ei represents the price of the distribution transformer of the i-th battery swap station; n1 represents the number of charging piles purchased; D Ci represents the purchase price of the charging pile of the i-th battery swap station; D qi represents other expenses of the ith battery swap station; f represents the number of years the battery swap station can operate;

[0033] The calculation method of the operation and maintenance cost of the i-th battery swap station is as follows:

[0034] C Wi =(a+b)n i C Ci m i ;

[0035] Where a represents the power loss coefficient; b represents the operation and maintenance conversion rate; n i represents the number of times electric vehicles are used at the i-th battery swap station; C Ci represents the electricity cost of charging once at the i-th battery swap station; m i represents the annual usage of electric vehicles at the i-th battery swap station;

[0036] The network loss cost of the i-th battery swap station is calculated as follows:

[0037] W2=365t i C Pi E L ;

[0038] Among them, t i Indicates the available time of the representative day; C Pi Indicates the charging electricity price; E L Indicates network loss;

[0039] Network loss E L The calculation method is as follows:

[0040]

[0041] Among them, M represents the total number of nodes; Gxy Represents the total conductance of the line between nodes x and y; cosφ xy Represents the cosine value of the power angle difference between nodes x and y; U x , U y are the voltages at nodes x and y respectively;

[0042] The calculation of the user queuing time cost at the battery swap station is as follows:

[0043] W3=365C W (T c +T P );

[0044] Among them, T c is the total arrival time at the battery swap station, T P is the total queuing time; C W is the operation and maintenance cost of the battery swap station, including

[0045]

[0046]

[0047]

[0048] Where, j represents the jth location where the electric vehicle needs to be charged; P ij Indicates that the selection probability factor takes values ​​of 0 and 1; L ij H represents the distance required to charge at point i; i represents the expected value of waiting time; V represents the average speed of the vehicle;

[0049] The cost objective function is:

[0050] W min =ω1W1+ω2W2+ω3W3; where ω1, ω2, ω3 are the weight values ​​of the corresponding cost functions, and the sum of the three is 1.

[0051] Preferably, the constraints of the cost objective function are:

[0052] Constraint 1: Total number of sites:

[0053]

[0054] P i Whether to build a battery swap station is 1 or 0, and the total number N is obtained;

[0055] Constraint 2: Charging distance:

[0056] L ij <L m ;

[0057] The charging distance is subject to the maximum service radius L m constraint;

[0058] Constraint 3: Demand quantity constraint:

[0059] The formula for building only one battery swap station within the constraint point is as follows:

[0060]

[0061] Constraint 4: Power flow calculation:

[0062]

[0063] Among them, G xy , B xy ,θ xy are the conductance, susceptance, and voltage phase angle of the x and y nodes respectively; P x Inject active power into the node, Q x Inject reactive power into the node, M is the number of nodes, U x , U y are the voltages at nodes x and y respectively;

[0064] Constraint 5: Voltage range:

[0065] U xmin <U x <U xmax ;

[0066] Among them, U xmin , U xmax are the voltage lower limit and voltage upper limit of node x respectively;

[0067] Load:

[0068] P x <P xmax ;

[0069] The load should be at the maximum load P of the x node xmax The value is within the range.

[0070] Preferably, the step S3 specifically includes:

[0071] S31: Taking the minimum daily charging cost of the battery swap station as the goal:

[0072]

[0073] Among them, P BSS.t is the charging power of the battery swap station at time t, P PV.t is the output power of photovoltaic power generation at time t, t is the charging duration, u tis the time-of-use electricity price; D is the length of an operation cycle;

[0074]

[0075] P charge,h (n,N) is the charging power of the battery swap station at time h; n is the number of battery stocks in the battery swap station; N is the number of battery swap electric vehicles served by the battery swap station;

[0076] S32: Constraints on charging power at battery swap stations:

[0077] 0≤P BSS.t ≤P BSS.max n;

[0078] P BSS.max is the maximum charging power of a single battery in the battery swap station; n is the battery inventory in the battery swap station;

[0079] S33: Photovoltaic power generation output constraints:

[0080] 0≤P PV.t ≤P PV.max ;

[0081] P PV.max It is the maximum power of photovoltaic power generation, and its value depends on the installed capacity of photovoltaic power generation and the weather conditions of the day.

[0082] Preferably, the step S4 specifically includes:

[0083] S41: Targeting the minimum daily charging cost at the battery swap station

[0084]

[0085] P BSS.t and P PV.t are the charging power of the battery swap station and the output power of photovoltaic power generation at time t; t in is the charging duration; t out u is the duration of electricity sales at the battery swap station; t is the time-of-use electricity purchase price; v t P is the time-of-use electricity price; sell.t is the electricity sales power of the battery swap station at time t;

[0086] S42: In addition to meeting the charging power constraint conditions and photovoltaic power generation power constraint conditions, the battery swap station also needs to meet the power sales constraint conditions of the battery swap station:

[0087] 0≤P sell.t ≤P sell,max ·P t n ;

[0088] Psell,max P is the maximum power sold by a single battery in the battery swap station; t n is the battery surplus of the battery swap station at time t;

[0089]

[0090] L h is the battery surplus of the battery swap station at time h;

[0091]

[0092] g h-7 is the number of batteries that have just been charged at the swap station at time h, is the sum of the battery swapping demands h-1 moments before the battery swapping station, d h is the battery replacement demand at time h;

[0093] S43: When a battery swap station participates in the electricity market to sell part of its energy, it must first meet the basic battery swap needs of the battery swap station:

[0094]

[0095] P in.t is the total charging power of the battery swap station at time t; Q battery,S0 is the initial power capacity of the battery swap station; The amount of electricity required to meet the battery replacement demand at time t+1;

[0096] The overall charging power P of the battery swap station in.t :

[0097] P in.t =P BSS.t -P PV.t ;

[0098] The initial power capacity of the battery swap station is Q battery,S0 It is expressed as:

[0099] Q battery.S0 =P t 0 C;

[0100] P t 0 is the initial battery surplus of the electric vehicle battery swap station, and C is the capacity of a single battery in the battery swap station;

[0101] The amount of electricity required for battery replacement at time t+1

[0102]

[0103] is the battery replacement demand of taxis at time t+1.

[0104] Compared with the prior art, the present invention has the following beneficial effects:

[0105] The present invention adopts a method based on BP neural network to predict photovoltaic power generation, making the data more accurate. The present invention adopts immune genetic algorithm and shortest path algorithm model to plan the layout of battery swap stations, which can quickly find the optimal path between charging demand points and the optimal charging points of battery swap stations in the entire map network. The present invention constructs a photovoltaic-battery swap station direct operation model to reduce the charging cost of the battery swap station. The present invention constructs an optimized operation model of photovoltaic-battery swap stations under the participation of the electricity market to increase the additional income of the battery swap station. BRIEF DESCRIPTION OF THE DRAWINGS

[0106] In order to more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the following is a brief introduction to the drawings required for the specific embodiments or the description of the prior art. In all the drawings, similar elements or parts are generally identified by similar reference numerals. In the drawings, the elements or parts are not necessarily drawn according to the actual scale.

[0107] Figure 1 It is a flow chart of the operation strategy method of the battery swap station based on photovoltaic power generation of the present invention;

[0108] Figure 2 It is a graph of the input and output variables of the BP neural network model;

[0109] Figure 3 A flow chart for planning the layout of battery swap stations. DETAILED DESCRIPTION

[0110] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0111] It should be understood that when used in this specification and the appended claims, the terms "include" and "comprises" indicate the presence of described features, integers, steps, operations, elements and / or components, but do not exclude the presence or addition of one or more other features, integers, steps, operations, elements, components and / or combinations thereof.

[0112] It should also be understood that the terms used in the present specification are only for the purpose of describing specific embodiments and are not intended to limit the present invention. As used in the present specification and the appended claims, unless the context clearly indicates otherwise, the singular forms "a", "an" and "the" are intended to include plural forms.

[0113] It should be further understood that the term "and / or" used in the present description and the appended claims refers to and includes any and all possible combinations of one or more of the associated listed items.

[0114] like Figure 1 As shown, an embodiment of the present invention provides a method for operating a battery swap station based on photovoltaic power generation, comprising the following steps:

[0115] Step S1, the photovoltaic power generation device is connected to the battery swap station, and the photovoltaic power generation power is predicted by combining environmental parameters and using a BP neural network model;

[0116] Step S2, using immune genetic algorithm combined with shortest path algorithm to plan the layout of the battery swap station;

[0117] Step S3, based on the photovoltaic power generation data, according to the planning and layout of the battery swap station, with the minimum daily charging cost of the battery swap station as the goal, establish a photovoltaic-battery swap station direct operation model, and calculate the power supply of the battery swap station; wherein the battery swap station can obtain the power provided by photovoltaic power generation, and can also purchase insufficient power from the external power grid;

[0118] Step S4, based on the photovoltaic power generation data, according to the planning and layout of the battery swap station, with the minimum daily charging cost of the battery swap station as the goal, establish a photovoltaic-battery swap station optimization operation model, and calculate the power available for the battery swap station; in the photovoltaic-battery swap station optimization operation model, the battery swap station sells part of the electricity while meeting its own battery swap needs, and the electricity price when the battery swap station sells electricity is higher than the market price.

[0119] Step S1 specifically includes the following steps:

[0120] S11: Normalize the historical data of environmental parameters and convert them into data in the [0.1] interval. The calculation formula is as follows:

[0121]

[0122] The historical data of environmental parameters include solar radiation intensity, ambient temperature, wind speed, humidity and visibility. In the formula, p(t) represents the environmental parameter data at time t; p max (t) and p min (t) represents the maximum and minimum values ​​of the environmental parameter data at time t; p* (t) is the normalized data;

[0123] S12: Train the BP model. The sample input matrix is ​​X=(X1, X2, …X n ), where the row vector represents the input sample and the column vector X1,X2,…X n They represent a set of sample values ​​respectively, and the number of input samples is relatively large; each set of model input variables is x = [x1, x2, x3, x4, x5] T , where x1, x2, x3, x4, and x5 represent solar radiation intensity, ambient temperature, wind speed, humidity, and visibility respectively; the output variable is the photovoltaic power generation corresponding to each set of sample data;

[0124] Step S13, establish a BP neural network model. During the modeling process, determine the number of hidden layer nodes n1=log2 m , m represents the number of input layer nodes;

[0125] Step S14, using the Levenberg-Marquardt optimization method of the BP network to train the BP neural network model, the weight and threshold update formula is:

[0126] X k+1 =X k -(J T J+uI) -1 J T e;

[0127] Among them, X k is the kth iteration data, X k+1 is the k+1th iteration data, J is the Jacobian matrix of the error differential with respect to the weight, e is the error vector, and u is a scalar;

[0128] The training function with the Levenberg-Marquardt optimization method in the matlab toolbox is trainlm.m, and its training parameters include: training times epochs, training step length show, error function indicator goal, training time time, minimum gradient min-grad, memory reduction coefficient men-reduc, and the initial value, increase coefficient, decrease coefficient and maximum value of u; step S15, obtaining the environmental parameters of the prediction day, normalizing the environmental parameters and inputting them into the trained BP neural network model for prediction, so as to obtain the photovoltaic power generation power of the prediction day.

[0129] Step S2 specifically includes the following steps:

[0130] Step S21, construct a cost objective function for the battery swap station. The cost of constructing the battery swap station includes the construction investment cost of the battery swap station, the network loss cost of the battery swap station, and the user waiting time cost of the battery swap station. The construction investment of the battery swap station includes not only the equipment purchase and manpower installation costs, but also the subsequent maintenance costs.

[0131] The construction cost of the battery swap station is as follows:

[0132]

[0133] Among them, C Ji is the purchase cost of the ith battery swap station, C Wi is the operation and maintenance cost of the i-th battery swap station; I is the number of battery swap stations; the purchase cost of the i-th battery swap station is calculated as follows:

[0134]

[0135] Where: λ0 represents the rate of return of the battery swap station; A i represents the building area of ​​the i-th battery swap station; D i represents the land price of the i-th battery swap station; E i represents the distribution transformer capacity of the i-th battery swap station; D Ei represents the price of the distribution transformer of the i-th battery swap station; n1 represents the number of charging piles purchased; D Ci represents the purchase price of the charging pile of the i-th battery swap station; D qi represents other expenses of the ith battery swap station; f represents the number of years the battery swap station can operate;

[0136] Transformer capacity E i Select the impact formula:

[0137]

[0138] where k i represents the simultaneity rate; Indicates power factor; D Ci Indicates the purchase price of the charging pile; P i represents charging power; L represents load rate; α represents charging load ratio; η represents charging efficiency;

[0139] The calculation method of the operation and maintenance cost of the i-th battery swap station is as follows:

[0140] C Wi =(a+b)n i C Ci m i ;

[0141] Where a represents the power loss coefficient; b represents the operation and maintenance conversion rate; n irepresents the number of times electric vehicles are used at the i-th battery swap station; C Ci represents the electricity cost of charging once at the i-th battery swap station; m i represents the annual usage of electric vehicles at the i-th battery swap station;

[0142] The network loss cost of the i-th battery swap station is calculated as follows:

[0143] W2=365t i C Pi E L ;

[0144] Among them, t i Indicates the available time of the representative day; C Pi Indicates the charging electricity price; E L Indicates network loss;

[0145] The power network loss for electric vehicles is also an important factor we consider when planning charging piles. Our goal is to choose a location with the smallest network loss. L The calculation method is as follows:

[0146]

[0147] Among them, M represents the total number of nodes; G xy Represents the total conductance of the line between nodes x and y; cosφ xy Represents the cosine value of the power angle difference between nodes x and y; U x , U y are the voltages at nodes x and y respectively;

[0148] The calculation of the user queuing time cost at the battery swap station is as follows:

[0149] W3=365C W (T c +T P );

[0150] Among them, T c is the total arrival time at the battery swap station, T P is the total queuing time; C W is the operation and maintenance cost of the battery swap station, including

[0151]

[0152]

[0153]

[0154] Where, j represents the jth location where the electric vehicle needs to be charged; P ij Indicates that the selection probability factor takes values ​​of 0 and 1; Lij H represents the distance required to charge at point i; i represents the expected value of waiting time; V represents the average speed of the vehicle;

[0155] The mathematical model of charging queue time cost is constructed based on the M / M / K model.

[0156] The average queue length formula is:

[0157]

[0158] The average waiting time formula is:

[0159]

[0160] Where μ represents the probability of the user entering.

[0161] The probability Pn of no charging pile working in the battery swap station is as follows:

[0162]

[0163] Site work intensity formula:

[0164]

[0165] Where ε represents the probability of charging pile usage, and n1 represents the number of charging piles.

[0166] The cost objective function is:

[0167] W min =ω1W1+ω2W2+ω3W3; where ω1, ω2, ω3 are the weight values ​​of the corresponding cost functions, and the sum of the three is 1.

[0168] Step S22, constructing the constraint conditions of the cost objective function, the constraint conditions of the cost objective function are:

[0169] Constraint 1: Total number of sites:

[0170]

[0171] P i Whether to build a battery swap station is 1 or 0, and the total number N is obtained;

[0172] Constraint 2: Charging distance:

[0173] L ij <L m ;

[0174] The charging distance is subject to the maximum service radius L m constraint;

[0175] Constraint 3: Demand quantity constraint:

[0176] The formula for building only one battery swap station within the constraint point is as follows:

[0177]

[0178] Constraint 4: Power flow calculation:

[0179]

[0180] Among them, G xy , B xy ,θ xy are the conductance, susceptance, and voltage phase angle of the x and y nodes respectively; P x Inject active power into the node, Q x Inject reactive power into the node, M is the number of nodes, U x , U y are the voltages at nodes x and y respectively;

[0181] Constraint 5: Voltage range:

[0182] U xmin <U x <U xmax ;

[0183] Among them, U xmin , U xmax are the voltage lower limit and voltage upper limit of node x respectively;

[0184] Load:

[0185] P x <P xmax ;

[0186] The load should be at the maximum load P of the x node xmax The value is within the range.

[0187] Step S23, using the Stra Dijk algorithm to calculate the optimal path between the charging demand point and the optimal charging point of the battery swap station in the entire map network.

[0188] For the overall planning model of the battery swap station, the waiting time for users in queues is: that is, the time to drive to the battery swap station and the waiting time in queues. When calculating the time taken to drive to the battery swap station for the shortest path, the Stra Dijk algorithm can be used to find the optimal path between the charging demand point and the optimal charging point of the battery swap station in the entire map network. The overall idea of ​​the Stra Dijk algorithm is to first initially mark the adjacent points around the demand point, then calculate the shortest distance for the initial marked points, find the closest distance point and mark it as the final mark. When the final mark is spread over all the points in the set, the algorithm process ends. The details are as follows:

[0189] (1) First, set two sets, set A is the initial mark, set B is the final mark, L(a,b) is the distance between point a and point b, and s is the change value of the initial starting point;

[0190] (2) Its distance λ x The number of the point before the shortest path between the target position f and x, initially 0, a≠x, d a =+∞

[0191] (3) Compare the distances of the adjacent positions r of s, and determine the distance d from the starting point to r. r The distance d greater than the initial point s s When the distance between s and r is L(s,r), then d r =d s +L(s,r)λ x =s. The distance from the starting point to r is d r Less than or equal to the distance d from the initial point s s When the sum of the distances L(s,r) between s and r is calculated, no change is made and all adjacent points are traversed.

[0192] (4) Determine the point in A with the shortest distance. Add it to set B and mark it as a point.

[0193] (5) Repeat steps 3 and 4 until set A becomes an empty set, so that the calculation is terminated and the minimum path value of s that can reach all points is obtained.

[0194] Step S24, using a genetic immune algorithm to optimize the layout planning of the battery swap station.

[0195] Genetic Algorithm (GA) is a computational model that simulates biological evolution, natural selection, and inheritance. First, it creates individuals by coding. After a certain number of individuals are generated, an initial population is formed. Then, genetic operations are performed on the population, including selection, crossover, and mutation. Then, a new population is generated and the optimal solution is found through evolution. The evolution of the genetic algorithm is determined by genetic operations. Different genetic operators are used in the entire genetic calculation, and the results and calculation effects are different.

[0196] (1) Select operator:

[0197] Establishing the objective function as an evaluation indicator for individuals corresponds to Darwin's theory of evolution's selection of the environment and individual evolution. Select high-quality individuals that are more adaptable to the environment in the population for genetic operations. The following formula represents the probability of an individual being selected:

[0198]

[0199] i represents an individual in the population, with a total of N individuals, f i represents the fitness of the ith individual. (2) Cross

[0200] The crossover operation is to select any two individuals in the population with a certain probability, so that part of the information of the individuals is exchanged. The crossover point of the exchange is randomly generated. The two new individuals formed in this way are then judged to see whether they are better than the previous generation of individuals. In binary coding operators, there are often single-point crossover and multi-point crossover, and in digital computing, there is arithmetic crossover.

[0201] Arithmetic crossover expression:

[0202]

[0203] It can be seen from the above formula that the entire operation is a linear combination process.

[0204] Mutations:

[0205] In order to prevent the overall algorithm from local convergence during the genetic operation, which leads to a decrease in accuracy, mutation operators are often used to mutate it. The method often used in binary is: select the mutation point in the individual, if it is 1, then the inverse value is 0, if it is 0, then the inverse value is 1.

[0206] The genetic immune algorithm is based on the original genetic algorithm and integrates the immune algorithm. This immune ability can memorize and store excellent individuals. Extended to genetics, the antigen corresponds to the objective function of the immune genetic algorithm, and the antibody corresponds to the optimal solution. By comparing the affinity between the antibody and the antigen and the fitness value, the antibody concentration is obtained, and the excellent quality individuals are memorized and stored, which improves the local search performance of the original genetic algorithm as a whole. The antibody concentration C represents the ratio of the total number of individuals whose similarity to the antibody is less than a similarity threshold to the total population.

[0207] The similarity calculation formula between real-coded antibodies and binary-coded antibodies is:

[0208]

[0209]

[0210]

[0211] Here, n represents the position number of the antibody and N is the total length of the gene.

[0212] S24: Genetic immune algorithm process:

[0213] (1) Set the objective function for the population to be optimized and identify the antigen;

[0214] (2) Setting constraint parameters to form a population;

[0215] (3) Detection of affinity with antigen;

[0216] (4) Memory stores antibodies with higher fitness, forming the memory bank of the next generation population;

[0217] (5) Calculate the antibody concentration C and select antibodies with high reproduction probability;

[0218] (6) Conduct genetic evolution to generate new antibodies;

[0219] (7) replacing the generated new antibodies by comparing their affinity to form a new population;

[0220] (8) When the termination index is reached, the calculation is stopped; otherwise, the antibody usage value is recalculated and the cycle returns to the third step.

[0221] Process such as Figure 3 As shown, based on the above analysis, this time we use the immune genetic algorithm and the shortest path algorithm model to conduct layout planning analysis on the battery swap station.

[0222] Step S3 specifically includes:

[0223] S31: Taking the minimum daily charging cost of the battery swap station as the goal:

[0224]

[0225] Among them, P BSS.t is the charging power of the battery swap station at time t, P PV.t is the output power of photovoltaic power generation at time t, t is the charging duration, u t is the time-of-use electricity price; since the prediction accuracy of photovoltaic power generation is 1 hour, in this model, the resolution of time period t is 1 hour. D is the length of an operation cycle, that is, 24 hours.

[0226]

[0227] P charge,h (n,N) is the charging power of the battery swap station at time h; n is the number of battery stocks in the battery swap station; N is the number of battery swap electric vehicles served by the battery swap station;

[0228] S32: Constraints on charging power at battery swap stations:

[0229] 0≤P BSS.t ≤P BSS.max n;

[0230] P BSS.max is the maximum charging power of a single battery in the battery swap station; n is the battery inventory in the battery swap station;

[0231] S33: Photovoltaic power generation output constraints:

[0232] 0≤P PV.t ≤P PV.max ;

[0233] P PV.max It is the maximum power of photovoltaic power generation, and its value depends on the installed capacity of photovoltaic power generation and the weather conditions of the day.

[0234] Preferably, the step S4 specifically includes:

[0235] S41: Targeting the minimum daily charging cost at the battery swap station

[0236]

[0237] P BSS.t and P PV.t are the charging power of the battery swap station and the output power of photovoltaic power generation at time t; t in is the charging duration; t out u is the duration of electricity sales at the battery swap station; t is the time-of-use electricity purchase price; v t P is the time-of-use electricity price; sell.t is the electricity sales power of the battery swap station at time t; the resolution of time period t is also 1 hour.

[0238] S42: In addition to meeting the charging power constraint conditions and photovoltaic power generation power constraint conditions, the battery swap station also needs to meet the power sales constraint conditions of the battery swap station:

[0239] 0≤P sell.t ≤P sell,max ·P t n ;

[0240] P sell,max P is the maximum power sold by a single battery in the battery swap station; t n is the battery surplus of the battery swap station at time t;

[0241]

[0242] L h is the battery surplus of the battery swap station at time h;

[0243]

[0244] g h-7 is the number of batteries that have just been charged at the swap station at time h, is the sum of the battery swapping demands h-1 moments before the battery swapping station, d h is the battery replacement demand at time h;

[0245] S43: When a battery swap station participates in the electricity market to sell part of its energy, it must first meet the basic battery swap needs of the battery swap station:

[0246]

[0247] P in.t is the total charging power of the battery swap station at time t; Q battery,S0 is the initial power capacity of the battery swap station; The amount of electricity required to meet the battery replacement demand at time t+1;

[0248] The overall charging power P of the battery swap station in.t :

[0249] P in.t =P BSS.t -P PV.t ;

[0250] The initial power capacity of the battery swap station is Q battery,S0 It is expressed as:

[0251] Q battery.S0 =P t 0 C;

[0252] P t 0 is the initial battery surplus of the electric vehicle battery swap station, and C is the capacity of a single battery in the battery swap station;

[0253] The amount of electricity required for battery replacement at time t+1

[0254]

[0255] is the battery replacement demand of taxis at time t+1.

[0256] A direct operation model of a photovoltaic-battery swap station was constructed and compared with the battery swap station model without considering photovoltaic power generation. It was found that the charging cost of the battery swap station can be reduced.

[0257] By constructing an optimal operation model of photovoltaic-battery swap stations under the electricity market and comparing it with the battery swap station model without considering photovoltaic power generation, it is found that the additional revenue of the battery swap station can be increased.

[0258] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or replace some or all of the technical features therein by equivalents. These modifications or replacements do not make the essence of the corresponding technical solutions deviate from the scope of the technical solutions of the embodiments of the present invention, and they should all be included in the scope of the claims and specification of the present invention.

Claims

1. A method for operating a battery swap station based on photovoltaic power generation, characterized in that: Includes the following steps: Step S1, the photovoltaic power generation device is connected to the battery swap station, and the photovoltaic power generation power is predicted by combining environmental parameters and using a BP neural network model; Step S2, using immune genetic algorithm combined with shortest path algorithm to plan the layout of the battery swap station; Step S3, based on the photovoltaic power generation data, according to the planning and layout of the battery swap station, with the minimum daily charging cost of the battery swap station as the goal, establish a photovoltaic-battery swap station direct operation model, and calculate the power supply of the battery swap station; Step S4, based on the photovoltaic power generation data, according to the planning and layout of the battery swap station, with the minimum daily charging cost of the battery swap station as the goal, establish a photovoltaic-battery swap station optimization operation model, and calculate the power supply of the battery swap station; in the photovoltaic-battery swap station optimization operation model, the battery swap station sells part of the electricity when meeting its own battery swap needs, and the electricity price when the battery swap station sells the electricity is higher than the market price; The step S3 specifically includes: S31: Taking the minimum daily charging cost of the battery swap station as the goal: Among them, P BSS.t is the charging power of the battery swap station at time t, P PV.t is the output power of photovoltaic power generation at time t, t is the charging duration, u t is the time-of-use electricity price; D is the length of an operation cycle; P charge,h (n', N') is the charging power of the battery swap station at time h; n is the number of battery stocks in the battery swap station; N is the number of battery swap electric vehicles served by the battery swap station; S32: Constraints on charging power at battery swap stations: 0≤P BSS.t ≤P BSS.max ·n; P BSS.max is the maximum charging power of a single battery in the battery swap station; n is the battery inventory in the battery swap station; S33: Photovoltaic power generation output constraints: 0≤P PV.t ≤P PV.max ; P PV.max The maximum power of photovoltaic power generation, its value depends on the installed capacity of photovoltaic power generation and the weather conditions of the day; The step S4 specifically includes: S41: Targeting the minimum daily charging cost at the battery swap station P BSS.t and P PV.t are the charging power of the battery swap station and the output power of photovoltaic power generation at time t; t in is the charging duration; t out v is the duration of electricity sales at the battery swap station; t P is the time-of-use electricity price; sell.t is the electricity sales power of the battery swap station at time t; S42: In addition to meeting the charging power constraint conditions and photovoltaic power generation power constraint conditions, the battery swap station also needs to meet the power sales constraint conditions of the battery swap station: 0≤P sell.t ≤P sell,max ·P t n ; P sell,max P is the maximum power sold by a single battery in the battery swap station; t n is the battery surplus of the battery swap station at time t; L h is the battery surplus of the battery swap station at time h; g h-7 is the number of batteries that have just been charged at the swap station at time h, is the sum of the battery swapping demands h-1 moments before the battery swapping station, d h is the battery replacement demand at time h; S43: When a battery swap station participates in the electricity market to sell part of its energy, it must first meet the basic battery swap needs of the battery swap station: P in.t is the total charging power of the battery swap station at time t; Q battery,S0 is the initial power capacity of the battery swap station; The amount of electricity required to meet the battery replacement demand at time t+1; The overall charging power P of the battery swap station in.t : P in.t =P BSS.t -P PV.t ; The initial power capacity of the battery swap station is Q battery,S0 It is expressed as: Q battery.S0 =P t 0 ·C; P t 0 is the initial battery surplus of the electric vehicle battery swap station, and C is the capacity of a single battery in the battery swap station; The amount of electricity required for battery replacement at time t+1 is the battery replacement demand of taxis at time t+1.

2. A method for operating a battery swap station based on photovoltaic power generation according to claim 1, characterized in that: The step S1 specifically includes the following steps: S11: Normalize the historical data of environmental parameters and convert them into data in the [0.1] interval. The calculation formula is as follows: The historical data of environmental parameters include solar radiation intensity, ambient temperature, wind speed, humidity and visibility. In the formula, p(t) represents the environmental parameter data at time t; p max (t) and p min (t) represents the maximum and minimum values ​​of the environmental parameter data at time t; p * (t) is the normalized data; S12: Train the BP model. The sample input matrix is ​​X=(X1, X2, …X n ), where the row vector represents the input sample and the column vector X1,X2,…X n Represent a set of sample values ​​respectively, and each set of model input variables is x=[x1,x2,x3,x4,x5] T , where x1, x2, x3, x4, and x5 represent solar radiation intensity, ambient temperature, wind speed, humidity, and visibility respectively; the output variable is the photovoltaic power generation corresponding to each set of sample data; Step S13, establish a BP neural network model, and during the modeling process, determine the number of hidden layer nodes m represents the number of input layer nodes; Step S14, using the Levenberg-Marquardt optimization method of the BP network to train the BP neural network model, the weight and threshold update formula is: X k+1 =X k -(J T J+uI) -1 J T yes; Among them, X k is the kth iteration data, X k+1 is the k+1th iteration data, J is the Jacobian matrix of the error differential with respect to the weight, e is the error vector, and u is a scalar; The training function with Levenberg-Marquardt optimization method in MATLAB toolbox is trainlm.m, and its training parameters are: training times epochs, training step length show, error function indicator goal, training time time, minimum gradient min-grad, memory reduction coefficient men-reduc, and initial value of u, increase coefficient, decrease coefficient and maximum value; Step S15, obtaining environmental parameters on the forecast day, normalizing the environmental parameters and inputting them into the trained BP neural network model for forecasting, thereby obtaining the photovoltaic power generation on the forecast day.

3. The method for operating a battery swap station based on photovoltaic power generation according to claim 1, characterized in that: The step S2 specifically includes the following steps: Step S21, constructing a cost objective function of a battery swap station, where the cost of constructing a battery swap station includes the construction investment cost of the battery swap station, the network loss cost of the battery swap station, and the user queuing time cost of the battery swap station; Step S22, constructing the constraints of the cost objective function, Step S23, using the Stra Dijk algorithm to calculate the optimal path between the charging demand point and the optimal charging point of the battery swap station in the entire map network; Step S24, using a genetic immune algorithm to optimize the layout planning of the battery swap station.

4. The method for operating a battery swap station based on photovoltaic power generation according to claim 3, characterized in that: The construction investment cost of the battery swap station in step S21 is as follows: Among them, C Ji is the purchase cost of the ith battery swap station, C Wi is the operation and maintenance cost of the i-th battery swap station; I is the number of battery swap stations; the purchase cost of the i-th battery swap station is calculated as follows: Where: λ0 represents the rate of return of the battery swap station; A i represents the building area of ​​the i-th battery swap station; D i represents the land price of the i-th battery swap station; E i represents the distribution transformer capacity of the i-th battery swap station; D Ei represents the price of the distribution transformer of the i-th battery swap station; n1 represents the number of charging piles purchased; D Ci represents the purchase price of the charging pile of the i-th battery swap station; D qi represents other expenses of the ith battery swap station; f represents the number of years the battery swap station can operate; The calculation method of the operation and maintenance cost of the i-th battery swap station is as follows: C Wi =(a+b)n i C Ci m i ; Where a represents the power loss coefficient; b represents the operation and maintenance conversion rate; n i represents the number of times electric vehicles are used at the i-th battery swap station; C Ci represents the electricity cost of charging once at the i-th battery swap station; m i represents the annual usage of electric vehicles at the i-th battery swap station; The network loss cost of the i-th battery swap station is calculated as follows: W2=365t i C Pi E L ; Among them, t i Indicates the available time of the representative day; C Pi Indicates the charging electricity price; E L Indicates network loss; Network loss E L The calculation method is as follows: Among them, M represents the total number of nodes; G xy Represents the total conductance of the line between nodes x and y; cosφ xy Represents the cosine value of the power angle difference between nodes x and y; U x , U y are the voltages at nodes x and y respectively; The calculation of the user queuing time cost at the battery swap station is as follows: W3=365C W (T c +T P ); Among them, T c is the total arrival time at the battery swap station, T P is the total queuing time; C W is the operation and maintenance cost of the battery swap station, including Where, j represents the jth location where the electric vehicle needs to be charged; P ij Indicates that the selection probability factor takes values ​​of 0 and 1; L ij H represents the distance required to charge at point i; i represents the expected value of waiting time; V represents the average speed of the vehicle; The cost objective function is: W min =ω1W1+ω2W2+ω3W3; where ω1, ω2, ω3 are the weight values ​​of the corresponding cost functions, and the sum of the three is 1.

5. A method for operating a battery swap station based on photovoltaic power generation according to claim 4, characterized in that: The constraints of the cost objective function are: Constraint 1: Total number of sites: P i Whether to build a battery swap station is 1 or 0, and the total number N is obtained; Constraint 2: Charging distance: L ij <L m ; The charging distance is subject to the maximum service radius L m constraint; Constraint 3: Demand quantity constraint: The formula for building only one battery swap station within the constraint point is as follows: Constraint 4: Power flow calculation: Among them, G xy , B xy ,θ xy are the conductance, susceptance, and voltage phase angle of the x and y nodes respectively; P x Inject active power into the node, Q x Inject reactive power into the node, M is the number of nodes, U x , U y are the voltages at nodes x and y respectively; Constraint 5: Voltage range: IN xmin <In x <In xmax ; Among them, U xmin , U xmax are the voltage lower limit and voltage upper limit of node x respectively; Load: P x <P xmax ; The load should be at the maximum load P of the x node xmax The value is within the range.

Images