Renewable energy access traction power supply system capacity optimization configuration system based on grey wolf algorithm

Abstract

The invention discloses a renewable energy access traction power supply system capacity optimization configuration system based on a grey wolf algorithm, comprising cost function and constraint condition construction of energy storage and renewable energy access traction power supply system, cost function construction, calculation of optimal capacity configuration by the grey wolf algorithm, and fusion of elite reverse learning mechanism module operation. The search process of the heuristic algorithm depends on the quality of an initial population, a basic grey wolf algorithm adopts a mode of randomly generating the initial population, but the mode is low in search efficiency and relatively weak in global search capability, and a method of combining logistic mapping and tent mapping is adopted to generate the initial population, so that the search efficiency is relatively high. The population can be more uniformly distributed in the search space, and the convergence speed of the algorithm can be improved compared with an original algorithm.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of electrical engineering and rail transit, in particular to a renewable energy access traction power supply system capacity optimization configuration system based on grey wolf algorithm. BACKGROUND

[0002] It has become a hot research topic to connect photovoltaic and other renewable energy systems to urban rail transit, however, the integration of energy storage and renewable energy with urban rail transit system still faces many challenges, on the one hand, renewable energy generation has intermittency and volatility, its power generation is unstable, and it is difficult to directly meet the strict requirements of rail transit system on power stability and reliability, and it is necessary to configure energy storage system to operate cooperatively, on the other hand, it is difficult to determine the optimal capacity configuration scheme of energy storage device and renewable energy generation system, resulting in high cost of energy storage and renewable energy access to traction power supply system, therefore, an algorithm capable of effectively configuring energy storage capacity and renewable energy installed capacity is urgently needed to maximize the benefits of energy storage and renewable energy access, at present, intelligent algorithms have been widely used in capacity configuration, such as ant colony algorithm, particle swarm algorithm, etc., which proves that intelligent algorithms have unique advantages in solving the nonlinear integer programming problem of capacity configuration.

[0003] Grey wolf algorithm (Grey Wolf Optimizer, GWO) is an optimization algorithm based on grey wolf swarm intelligence, including three behaviors of hierarchical system, surrounding prey and hunting behavior, the grey wolf swarm has strict hierarchical division, divided into alpha, beta and delta three levels, in the search space, the grey wolf individual approaches the prey by constantly updating its own position, each grey wolf adjusts its position according to the distance between itself and the current optimal solution (alpha wolf position), suboptimal solution (beta wolf position) and third optimal solution (delta wolf position), and moves to a better area, in the process of surrounding prey, the grey wolves will gradually reduce the surrounding circle and approach the position of the prey, through constant iteration and updating, the whole wolf pack gradually gathers around the optimal solution, realizing the search and approximation of the optimal solution, in the process of dealing with complex problems, the grey wolf algorithm is easy to fall into local optimal problem, an improved grey wolf algorithm is proposed, aiming at the problem of uneven population distribution in the initial stage of GWO, chaos mapping is added to initialize the population, so that the population is more evenly distributed in the search space, in order to improve the quality of the solution in the later stage of the algorithm, the elite reverse learning strategy is added, the results show that the proposed algorithm has certain advantages. SUMMARY

[0004] The present application provides a renewable energy access traction power supply system capacity optimization configuration system based on grey wolf algorithm, which can effectively solve the problems raised in the background art.

[0005] To achieve the above object, the application provides the following technical scheme: a renewable energy access traction power supply system capacity optimization configuration system based on a grey wolf algorithm, comprising a storage and renewable energy access traction power supply system cost function and constraint condition construction, cost function construction, grey wolf algorithm optimal capacity configuration calculation, and a fusion elite reverse learning mechanism module operation.

[0006] The cost function comprises initial investment costs, operation and maintenance costs, electric energy purchase and sale costs, and system operation losses of renewable energy and storage devices, and the constraint conditions cover system energy balance constraints, device capacity and power limitations, storage charging and discharging constraints, renewable energy volatility constraints, and traction load demand matching, etc.

[0007] The grey wolf optimization algorithm is used to optimize and calculate the system capacity, the elite reverse learning mechanism is fused to enhance the global search ability and convergence precision of the algorithm, and the capacity of the renewable energy and storage system in the traction power supply system is configured.

[0008] According to the above technical scheme, the storage and renewable energy access traction power supply system cost function and constraint condition construction comprises the establishment of a renewable energy mathematical model:

[0009] The photovoltaic output model can be calculated by (1.1):

[0010]

[0011] Wherein, f pv is a power derating factor for calculating the loss caused by dust and stains on the photovoltaic panel, generally taken as 0.9;

[0012] P ref is the rated output power of the photovoltaic cell under standard conditions (solar irradiance 1000W / m 2 , ambient temperature 25℃);

[0013] G ref is the light intensity under standard conditions, taken as 1000W / m 2 ;

[0014] k is the power temperature rise coefficient, generally taken as -0.47% / ℃;

[0015] Tref represents the rated temperature, generally 25℃.

[0016] According to the above technical scheme, the storage and renewable energy access traction power supply system cost function and constraint condition construction comprises a wind power output model that can be calculated by (1.2), which can be calculated by (1.2):

[0017]

[0018] where v ci is the cut-in wind speed, v co is the cut-out wind speed, v r is the rated wind speed;

[0019] The charging of the energy storage system in the energy storage system modeling can be calculated by (1.3):

[0020] E(t+1) = E(t) + ∫P ch (t)η ch dt (1.3)

[0021] where E(t) represents the energy storage capacity, P ch represents the energy storage charging power, and η ch represents the energy storage charging efficiency;

[0022] The discharging of the energy storage system in the model can be calculated by (1.4)

[0023]

[0024] where P dis represents the energy storage discharging power, and η dis represents the energy storage discharging efficiency.

[0025] According to the above technical solution, the constraint conditions of the cost function and the constraint conditions of the energy storage and the renewable energy access to the traction power supply system are as follows:

[0026] The maximum power of wind power and photovoltaic power

[0027] 0 < P PV < P PVN (1.5)

[0028] 0 < P PVN < P PV_max (1.6)

[0029] 0 < P wind < P windN (1.7)

[0030] 0 < P windN < P wind_max (1.8)

[0031] where P PVN represents the rated output of photovoltaic power, P PV_max represents the maximum value of the rated power of photovoltaic power, P windN represents the rated power of wind power, and P wind_max represents the maximum value of the rated power of wind power;

[0032] The energy storage power capacity limit is:

[0033] Energy storage power limit:

[0034] 0 < P < 1 ch P < 1 chN b bat (1.9)

[0035] 0 < P < 1 dis P < 1 disN (1-b bat ) (1.10)

[0036] where P chN represents the energy storage charging rated power, P disN represents the energy storage discharging rated power, b bat is a binary variable for controlling the state of energy storage charging and discharging, 1 represents charging, and 0 represents discharging;

[0037] Energy storage capacity limit:

[0038] Emax min (1.11)

[0039]

[0040] where E N represents the battery rated power, SOC min and SOC max represent the maximum and minimum values of the battery state of charge, E max and E min represent the maximum and minimum values of the energy storage capacity;

[0041] Power balance limit:

[0042] In actual working conditions, the catenary and the traction power supply station are both subject to power balance constraints:

[0043] p sub-grid +p bat,dis +p pv +p wind =p bat,ch +p T +p sub-grid,fed (1.14)

[0044] where p sub-grid is the power flowing from the power grid to the traction power supply station, p bat,dis is the power charged by the energy storage system, p pv is the photovoltaic power, p wind is the wind power, p bat,ch is the energy storage charging power, p T is the traction power, and p sub-grid,fed is the regenerative braking feedback power.

[0045] According to the technical solution, the initial investment cost of the cost function is mainly divided into four parts, namely, the isolation transformer, the converter, the photovoltaic and the energy storage device, and the investment cost can be calculated according to the following formula

[0046] Calculation: C invest = C tran + C con + C pv + C wind + C bat = λ tran p tran + λ con p con + λ pv p pv + λ wind p wind + λ bat p bat (2.1)

[0047] Wherein, C invest represents the total investment cost, C tran represents the traction transformer cost, C con represents the converter cost, C pv represents the photovoltaic investment cost, C wind represents the wind power investment cost, C bat represents the energy storage investment cost, λ tran , λ con , λ pv , λ wind , λ bat represents the investment cost per WM.

[0048] Therefore, the objective function of the grey wolf algorithm can be set as:

[0049] max R total = L (C orange -C optim )-C invest (2.2)

[0050] Wherein, R total represents the total income, L represents the life cycle of the system, C orange represents the daily average cost before renewable energy access, C optim represents the daily average cost after renewable energy access.

[0051] According to the technical solution, the grey wolf algorithm calculates the optimal capacity configuration including randomly generating a population:

[0052] The original grey wolf algorithm usually generates a group of grey wolf individuals randomly, each individual represents a potential solution to the problem, and its position is determined by the variables of the problem. Assuming that in a D-dimensional search space, the population size is N, the position of the i-th grey wolf can be represented as:

[0053] X i = i1 , i2 , …, iD i = 1, 2, …, N (3.1)

[0054] where X i represents the position set of the grey wolf, x i1 , x i2 , …, x iD represent the position of each grey wolf respectively;

[0055] L-T chaotic mapping strategy changes the initial position:

[0056]

[0057] where x n is the initial position of the grey wolf after n iterations, x n+1 is the position of the grey wolf after n+1 iterations, r is a control parameter, usually taking 3.57-4;

[0058] Calculate the fitness value and divide the level:

[0059] According to the objective function, calculate the fitness value of each grey wolf to evaluate the quality of its solution. The fitness function is determined according to the specific problem. According to the fitness value, sort the grey wolf population and divide it into three levels, δ, β, and α.

[0060] Alpha wolf is the individual with the best fitness value in the population, representing the current optimal solution; beta wolf is the suboptimal individual; delta wolf is the individual with poor fitness value, which mainly follows other better wolves in the search process;

[0061] Update the position:

[0062] In each iteration, update the position of other wolves according to the position information of the wolf, the update formula is as follows:

[0063] Calculate the distance:

[0064] D α = α - X(t) |

[0065] D β = β - X(t) |

[0066] D δ= |C3X δ - X(t) | (3.3)

[0067] Calculate the new position:

[0068] X1= X α - A1D α

[0069] X2= X β - A2D β

[0070] X3= X δ - A3D δ

[0071]

[0072] A i , C i Value mode:

[0073] A i = 2ar1-a

[0074] C i = 2r2

[0075]

[0076] Wherein, X(t) represents the current position of the wolf group, X α , X β , X δ respectively represent the positions of alpha, beta and delta wolves;

[0077] D α , D β , D δ respectively represent the distances of alpha, beta and delta wolves from the wolf group;

[0078] r1, r2 are random variables with [0,1] as the value range;

[0079] t represents the iteration number; T max represents the maximum iteration number.

[0080] According to the above technical scheme, in the D-dimensional space, the extreme point of the feasible solution in the gray wolf algorithm is taken as the elite individual of the population:

[0081]

[0082] The corresponding inverse solution is:

[0083]

[0084] ξ j = x max,j

[0085] ε j = x min,j (3.7)

[0086] wherein, is a position variable representing an elite individual; is the inverse solution of the elite individual i corresponding position; is a random variable with a value range of [0, 1];

[0087] ξ j , ε j respectively represent the upper and lower limits of the dynamic boundary, x max,j is the upper limit of the boundary; x min,j is the lower limit of the boundary; by using the dynamic boundary, the search experience of the grey wolf can be saved, thereby reducing the optimization time of the algorithm.

[0088] According to the above technical scheme, the system finally performs calculation verification, taking the cost function and the traction substation as the upper layer and the target function and the grey wolf solving algorithm as the lower layer, importing each capacity configuration solved by the grey wolf algorithm into the upper layer to obtain different target functions, and the target function can obtain different capacity configurations through the calculation of the grey wolf algorithm until the result converges.

[0089] Compared with the prior art, the beneficial effects of the present application are: the structure of the present application is scientific and reasonable, safe and convenient to use,

[0090] 1. The search process of the heuristic algorithm depends on the quality of the initial population, and the basic grey wolf algorithm adopts a random initial population generation method, but the search efficiency of this method is low and the global search ability is weak. The logistic mapping and tent mapping are combined to generate the initial population, which can make the population more uniformly distributed in the search space, and compared with the original algorithm, the convergence speed of the algorithm can be improved.

[0091] 2. The original grey wolf algorithm has slow convergence speed in the later period, low solution accuracy, and is prone to premature convergence. In order to effectively improve the quality of the algorithm solution and prevent the algorithm from entering premature, an elite inverse learning mechanism is added in each iteration process of the traditional grey wolf algorithm. The main principle of this mechanism is to calculate the inverse solution of the current solution, and select the optimal solution from the current solution and the inverse solution of the population as the next generation individual, which can improve the convergence speed of the original algorithm in the later period, improve the solution accuracy, and prevent the algorithm from appearing premature. BRIEF DESCRIPTION OF DRAWINGS

[0092] The accompanying drawings are included to provide a further understanding of the present application, and constitute a part of the specification, which together with the embodiments of the present application, serve to explain the present application, and do not constitute a limitation of the present application.

[0093] In the attached figure:

[0094] Fig. 1 This is a topological diagram of the renewable energy access traction power supply system of the present invention;

[0095] Fig. 2 It is a block diagram of the Gray Wolf Algorithm of the present invention;

[0096] Fig. 3 This is a comparison chart of the convergence curves of the grey wolf algorithm of the present invention, the particle swarm algorithm, and the sparrow algorithm. DETAILED DESCRIPTION

[0097] The preferred embodiments of the present invention are described below with reference to the accompanying drawings. It should be understood that the preferred embodiments described herein are only used to illustrate and explain the present invention, and are not used to limit the present invention.

[0098] Example: Figs. 1-3 As shown, the present invention provides a technical solution, a capacity optimization configuration system for renewable energy access to traction power supply system based on the gray wolf algorithm, including the construction of cost functions and constraints for energy storage and renewable energy access to traction power supply system, cost function construction, calculation of optimal capacity configuration by gray wolf algorithm, and operation of fusion elite reverse learning mechanism module;

[0099] The cost function includes the initial investment cost of renewable energy and energy storage equipment, operation and maintenance costs, electricity purchase and sales costs, and system operating losses. Constraints include system energy balance constraints, equipment capacity and power limitations, energy storage charging and discharging constraints, renewable energy volatility constraints, and traction load demand matching.

[0100] The Grey Wolf optimization algorithm is used to optimize the system capacity, and the elite reverse learning mechanism is integrated to enhance the algorithm's global search capability and convergence accuracy, thereby realizing the configuration of renewable energy and energy storage system capacity in the traction power supply system.

[0101] According to the above technical solution, the cost function and constraint conditions for integrating energy storage and renewable energy into the traction power supply system are constructed, including the establishment of a renewable energy mathematical model:

[0102] The photovoltaic output model can be calculated by (1.1):

[0103]

[0104] Among them, f pv The power derating factor is used to calculate the loss caused by dust and stains on the photovoltaic panel surface, and is generally taken as 0.9;

[0105] P ref Standard conditions (solar irradiance 1000W / m 2The rated output power of the photovoltaic cell under the ambient temperature of 25℃;

[0106] G ref is the standard condition light intensity, taking 1000W / m 2 ;

[0107] k is the power temperature rise coefficient, generally taking -0.47% / ℃;

[0108] Tref represents the rated temperature, generally 25℃.

[0109] According to the above technical scheme, the cost function and constraint condition of the energy storage and renewable energy access to the traction power supply system are constructed, including the wind power output model calculation, which can be calculated by (1.2):

[0110]

[0111] Wherein, v ci is the cut-in wind speed, v co is the cut-out wind speed, v r is the rated wind speed;

[0112] The energy storage system modeling can be calculated by (1.3) when the energy storage system is charged:

[0113] E(t+1)=E(t)+∫P ch (t)η ch dt (1.3)

[0114] Wherein E(t) represents the energy storage capacity, P ch represents the energy storage charging power, η ch represents the energy storage charging efficiency;

[0115] The energy storage system model can be calculated by (1.4) when the energy storage system is discharged

[0116]

[0117] Wherein, P dis represents the energy storage discharge power, η dis represents the energy storage discharge efficiency.

[0118] According to the above technical scheme, the constraint condition of the cost function and constraint condition of the energy storage and renewable energy access to the traction power supply system is as follows:

[0119] The maximum power of wind power photovoltaic

[0120] 0<P PV <P PVN (1.5)

[0121] 0<P PVN <P PV_max(1.6)

[0122] 0 < P wind < P windN (1.7)

[0123] 0 < P windN < P wind_max (1.8)

[0124] where P PVN represents the photovoltaic rated power, P PV_max represents the photovoltaic rated power maximum, P windN represents the wind power rated power, P wind_max represents the wind power rated power maximum;

[0125] Energy storage power capacity limit:

[0126] Energy storage power limit:

[0127] 0 < P ch < P chN b bat (1.9)

[0128] 0 < P dis < P disN (1 - b bat ) (1.10)

[0129] where P chN represents the energy storage charging rated power, P disN represents the energy storage discharging rated power, b bat is a binary variable to control the energy storage charging and discharging state, 1 represents charging and 0 represents discharging;

[0130] Energy storage capacity limit:

[0131] Emax min (1.11)

[0132] ENmin min (1.12)

[0133] ENmin min (1.13)

[0134] where E N represents the battery rated power, SOC min , SOC max represent the maximum and minimum values of the battery state of charge, respectively, E max , E min represent the maximum and minimum values of the energy storage capacity, respectively;

[0135] Power balance limit:

[0136] In actual working conditions, the catenary and traction power supply station are subject to power balance constraints:

[0137] p sub-grid +p bat,dis +p pv +p wind =p bat,ch +p T +p sub-grid,fed (1.14)

[0138] Wherein, p sub-grid is the power of the power grid flowing to the traction power supply station, p bat,dis is the power of the energy storage system charging, p pv is the photovoltaic power generation power, p wind is the wind power generation power, p bat,ch is the energy storage charging power, p T is the traction power, and p sub-grid,fed is the power of regenerative braking feedback.

[0139] According to the above technical scheme, the initial investment cost function is mainly divided into four parts: isolation transformer, converter, photovoltaic and energy storage device, and the investment cost can be calculated according to the following formula:

[0140] C invest =C tran +C con +C pv +C wind +C bat =λ tran p tran +λ con p con +λ pv p pv +λ wind p wind +λ bat p bat (2.1)

[0141] Wherein, C invest represents the total investment cost, C tran represents the cost of traction transformer, C con represents the cost of converter, C pv represents the investment cost of photovoltaic, C wind represents the investment cost of wind power, C bat represents the investment cost of energy storage, λ tran , λ con , λ pv , λ wind , λ bat represent the investment cost per WM;

[0142] Therefore, the objective function of the gray wolf algorithm can be set as:

[0143] max R total =L(C orange -C optim )-C invest (2.2)

[0144] Among them, R total represents the total benefit, L represents the life cycle of the system, C orange represents the average daily cost before renewable energy access, C optim Represents the average daily cost after renewable energy access.

[0145] According to the above technical solution, the Grey Wolf Algorithm calculates the optimal capacity configuration by randomly generating a population:

[0146] The original gray wolf algorithm usually randomly generates a group of gray wolf individuals. Each individual represents a potential solution to the problem, and its position is determined by the problem variables. Assuming that the population size is N in the D-dimensional search space, the position of the i-th gray wolf can be expressed as

[0147] X i =(x i1 ,x i2 ,…,x iD )i=1,2,...,N (3.1)

[0148] Among them, X i represents the location set of gray wolves, x i1 ,x i2 ,...,x iD Represents the position of each gray wolf;

[0149] LT chaos mapping strategy changes the initial position:

[0150]

[0151] Among them, x n is the initial position of the gray wolf after n iterations, x n+1 is the position of the gray wolf after n+1 iterations, r is the control parameter, usually 3.57-4;

[0152] Calculate the fitness value and divide it into levels:

[0153] The fitness value of each gray wolf is calculated according to the objective function to evaluate the quality of its solution. The fitness function is determined according to the specific problem. For example, in the minimization problem, the smaller the fitness value, the better the solution. Then, the gray wolf population is sorted according to the fitness value and divided into three levels: δ, β, and α.

[0154] The α wolf is the individual with the best fitness value in the population and represents the current optimal solution; the β wolf is the second-best individual; the δ wolf is the individual with the poorest fitness value, and they mainly follow other better wolves during the search process;

[0155] Update location:

[0156] In each iteration, the positions of other wolves are updated based on the wolf's position information. The update formula is as follows:

[0157] Calculate distance:

[0158] D α =|C1X α -X(t)|

[0159] D β =|C2X β -X(t)|

[0160] D δ =|C3X δ -X(t)| (3.3)

[0161] Calculate the new position:

[0162] X1=X α -A1D α

[0163] X2=X β -A2D β

[0164] X3=X δ -A3D δ

[0165]

[0166] A i 、C i Value method:

[0167] A i =2ar1-a

[0168]

[0169] Among them, X(t) represents the current position of the wolf pack, X α 、X β 、X δ denote the positions of α, β, and δ wolves respectively;

[0170] D α 、D β 、D δ They represent the distances of α, β, and δ wolves from the pack, respectively;

[0171] r1, r2 are random variables with [0, 1] as a value range;

[0172] t represents the number of iterations; T max represents the maximum number of iterations.

[0173] According to the above technical scheme, the elite reverse learning mechanism is added in each iteration process of the traditional grey wolf algorithm in order to effectively improve the quality of algorithm solution and prevent the algorithm from entering premature convergence, the main principle of the mechanism is to calculate the reverse solution according to the current solution, and select the optimal solution from the population of the current solution and the reverse solution as the next generation individual;

[0174] In D-dimensional space, the extreme point of the feasible solution in the grey wolf algorithm is taken as the elite individual of the population:

[0175]

[0176] The corresponding reverse solution is:

[0177]

[0178] ξ j = x max,j

[0179] ε j = x min,j (3.7)

[0180] Wherein, is a position variable representing the elite individual; is the reverse solution corresponding to the position of the elite individual i; That is, a random variable with a value range of [0, 1];

[0181] ξ j , ε j respectively represent the upper and lower limits of the dynamic boundary, x max,j is the upper limit of the boundary; x min,j is the lower limit of the boundary; by using the dynamic boundary, the search experience of the grey wolf can be saved, thereby reducing the optimization time of the algorithm.

[0182] According to the above technical scheme, the system finally performs calculation verification, taking the cost function and the traction substation as the upper layer, the objective function and the grey wolf solving algorithm as the lower layer, importing each capacity configuration solved by the grey wolf algorithm into the upper layer to obtain different objective functions, and the objective function can obtain different capacity configurations through the calculation of the grey wolf algorithm until the result converges.

[0183] The proposed scheme can improve the income by 35.5% compared with the traditional rule-based capacity configuration scheme, and can improve the income by 13.2% compared with the intelligent algorithm such as particle swarm.

[0184] Finally, it should be noted that: the above only for the preferred examples of the present application, and not for the limitation of the present application, although in the foregoing examples of the present application are described in detail, for those skilled in the art, it still can be modified, or part of the technical features of the equivalent replacement of the technical solutions described in the foregoing embodiments. Any modification, equivalent replacement, improvement, etc. within the spirit and principles of the present application, should be included in the protection scope of the present application.

Claims

1. A capacity optimization configuration system for renewable energy access to a traction power supply system based on a gray wolf algorithm, characterized by: This includes the construction of cost functions and constraints for integrating energy storage and renewable energy into the traction power supply system, cost function construction, calculation of optimal capacity configuration using the Grey Wolf algorithm, and operation of the fusion elite reverse learning mechanism module; The cost function includes the initial investment cost of renewable energy and energy storage equipment, operation and maintenance costs, electricity purchase and sales costs, and system operation losses. The constraints include system energy balance constraints, equipment capacity and power limitations, energy storage charging and discharging constraints, renewable energy volatility constraints, and traction load demand matching. The Grey Wolf optimization algorithm is used to optimize the system capacity, and the elite reverse learning mechanism is integrated to enhance the algorithm's global search capability and convergence accuracy, thereby realizing the configuration of renewable energy and energy storage system capacity in the traction power supply system.

2. The capacity optimization configuration system for renewable energy access to traction power supply system based on the grey wolf algorithm according to claim 1 is characterized in that: The cost function and constraint conditions for integrating energy storage and renewable energy into the traction power supply system include the establishment of a renewable energy mathematical model: The photovoltaic output model can be calculated by (1.1): Among them, f pv The power derating factor is used to calculate the loss caused by dust and stains on the photovoltaic panel surface, and is generally taken as 0.9; P ref is the rated output power of the photovoltaic cell under standard conditions; G ref The light intensity under standard conditions is 1000W / m 2 ; k is the power temperature rise coefficient, generally -0.47% / ℃; Tref represents the rated temperature, which is generally 25°C.

3. The capacity optimization configuration system for renewable energy access to traction power supply system based on the grey wolf algorithm according to claim 2 is characterized in that: The cost function and constraint conditions for integrating energy storage and renewable energy into the traction power supply system include the calculation of the wind power output model, which can be calculated through (1.2): Among them, v ci is the cut-in wind speed, v co is the cut-out wind speed, v r is the rated wind speed; In the energy storage system modeling, the energy storage system charging time can be calculated by (1.3): E(t+1)=E(t)+∫P ch (t)η ch dt (1.3) Where E(t) represents the energy storage capacity, P ch represents the energy storage charging power, η ch Indicates the energy storage charging efficiency; The model of energy storage system discharge can be calculated by (1.4) Among them, P dis Represents the energy storage discharge power, η dis Indicates the energy storage discharge efficiency.

4. The capacity optimization configuration system for renewable energy access to traction power supply system based on the grey wolf algorithm according to claim 3 is characterized in that: The cost function and constraints for integrating energy storage and renewable energy into the traction power supply system are constructed as follows: Maximum power of wind power and photovoltaic power 0<P PV <P PVN (1.5) 0<P PVN <P PV_max (1.6) 0<P wind <P windN (1.7) 0<P windN <P wind_max (1.8) Among them, P PVN Indicates the rated output of photovoltaic power, P PV_max Indicates the maximum rated power of photovoltaic power, P windN Indicates the rated power of wind power, P wind_max Indicates the maximum rated power of wind power; Energy storage power capacity limitations: Energy storage power limit: 0<P ch <P chN b bat (1.9) 0<P dis <P disN (1-b bat ) (1.10) Among them, P chN Indicates the energy storage charging rated power, P disN Indicates the rated power of energy storage discharge, b bat It is a binary variable that controls the charge and discharge status of energy storage, 1 means charging and 0 means discharging; Energy storage capacity limitations: Emax min (1.11) ENmin min (1.12) ENmin min (1.13) Among them, E N Indicates the battery rated power, SOC min , SOC max Respectively represent the maximum and minimum values ​​of the battery state of charge, E max 、E min Respectively represent the maximum and minimum values ​​of energy storage capacity; Power balancing limits: In actual working conditions, both the overhead line and traction power supply station are subject to power balance constraints: p sub-grid +p bat,dis +p pv +p wind =p bat,ch +p T +p sub-grid,fed (1.14) Among them, p sub-grid is the power flowing from the grid to the traction power station, p bat,dis is the charging power of the energy storage system, p pv is the photovoltaic power generation power, p wind is the wind power generation power, p bat,ch is the energy storage charging power, p T is the traction power, p sub-grid,fed is the power fed back by regenerative braking.

5. The capacity optimization configuration system for renewable energy access to traction power supply system based on grey wolf algorithm according to claim 1 is characterized in that: The initial investment in the cost function is mainly divided into four parts: isolation transformer, converter, photovoltaic and energy storage device. The investment cost can be calculated according to the following formula: C invest =C tran +C con +C pv +C wind +C bat =λ tran p tran +λ con p con +λ pv p pv +λ wind p wind +λ bat p bat (2.1) Among them, C invest Represents the total investment cost, C tran represents the cost of traction transformer, C con represents the converter cost, C pv represents the photovoltaic investment cost, C wind represents the investment cost of wind power generation, C bat represents the energy storage investment cost, λ tran ,λ con ,λ pv ,λ wind ,λ bat Indicates the investment cost per WM; Therefore, the objective function of the gray wolf algorithm can be set as: max R total =L(C orange -C optim )-C invest (2.2) Among them, R total represents the total benefit, L represents the life cycle of the system, C orange represents the average daily cost before renewable energy access, C optim Represents the average daily cost after renewable energy access.

6. The capacity optimization configuration system for renewable energy access to traction power supply system based on the grey wolf algorithm according to claim 1 is characterized in that: The gray wolf algorithm calculates the optimal capacity configuration by randomly generating a population: The original gray wolf algorithm usually randomly generates a group of gray wolf individuals. Each individual represents a potential solution to the problem, and its position is determined by the problem variables. Assuming that the population size is N in the D-dimensional search space, the position of the i-th gray wolf can be expressed as: X i =(x i1 ,x i2 ,…,x iD ) i=1,2,...,N (3.1) Among them, X i represents the location set of gray wolves, x i1 ,x i2 ,...,x iD Represents the position of each gray wolf; LT chaos mapping strategy changes the initial position: Among them, x n is the initial position of the gray wolf after n iterations, x n+1 is the position of the gray wolf after n+1 iterations, r is the control parameter, usually 3.57-4; Calculate the fitness value and divide it into levels: The fitness value of each gray wolf is calculated according to the objective function to evaluate the quality of its solution. The fitness function is determined according to the specific problem. The gray wolf population is sorted according to the fitness value and divided into three levels: δ, β, and α. The α wolf is the individual with the best fitness value in the population and represents the current optimal solution; the β wolf is the second-best individual; the δ wolf is the individual with the poorest fitness value, and they mainly follow other better wolves during the search process; Update location: In each iteration, the positions of other wolves are updated based on the wolf's position information. The update formula is as follows: Calculate distance: D α =|C1X α -X(t)| D β =|C2X β -X(t)| D δ =|C3X δ -X(t)| (3.3) Calculate the new position: X1=X α -A1D α X2=X β -A2D β X3=X δ -A3D δ A i 、C i Value method: A i =2ar1-a C i ＝2r2 Among them, X(t) represents the current position of the wolf pack, X α 、X β 、X δ denote the positions of α, β, and δ wolves respectively; D α 、D β 、D δ They represent the distances of α, β, and δ wolves from the pack, respectively; r1 and r2 are random variables with a value range of [0,1]; t represents the number of iterations; T max Indicates the maximum number of iterations.

7. The capacity optimization configuration system for renewable energy access to traction power supply system based on grey wolf algorithm according to claim 1 is characterized in that: In the D-dimensional space, the extreme point of the feasible solution in the gray wolf algorithm is used as the elite individual of the population: The corresponding inverse solution is: x j =x max,j e j =x min,j (3.7) in, is the position variable representing elite individuals; is the reverse solution of the corresponding position of elite individual i; That is, a random variable with a value range of [0,1]; ξ j , ε j Represent the upper and lower limits of the dynamic boundary, x max,j is the upper bound of the boundary; x min,j is the lower bound of the boundary; by utilizing the dynamic boundary, the search experience of the gray wolf can be saved, thereby reducing the optimization time of the algorithm.

8. The capacity optimization configuration system for renewable energy access to traction power supply system based on the grey wolf algorithm according to claim 1 is characterized in that: The system finally performs calculation verification, taking the cost function and traction substation as the upper layer, and the objective function and the Gray Wolf solution algorithm as the lower layer. Each capacity configuration obtained by the Gray Wolf algorithm is imported into the upper layer to obtain different objective functions. The objective function can be calculated by the Gray Wolf algorithm to obtain different capacity configurations until the results converge.

Images