Optimization method and device for grid-connected photovoltaic battery system

Abstract

The present disclosure provides an optimization method and device for a grid-connected photovoltaic battery system. The method includes: establishing an output power model of the grid-connected photovoltaic battery system; establishing an economic model of the grid-connected photovoltaic battery system according to the output power model of the grid-connected photovoltaic battery system; establishing an economic scale optimization model of the grid-connected photovoltaic battery system according to the economic model of the grid-connected photovoltaic battery system; optimizing the economic scale optimization model of the grid-connected photovoltaic battery system based on the coyote optimization algorithm. The coyote optimization algorithm solves the scale optimization problem of the grid-connected photovoltaic battery system considering reliability and greenhouse gas emissions. Compared with the non-renewable energy grid-connected system, the NPC and COE of the system are reduced.

Description

Technical Field

[0001] The present invention belongs to the technical field of power grid frequency regulation and control, and specifically relates to an optimization method and device for a grid-connected photovoltaic battery system. Background Art

[0002] With the expansion of technology and industrial tools and the rising demand for energy supply in different industries, the use of renewable energy has become inevitable. Among all renewable resources, solar energy has a special importance. The total capacity of photovoltaic (PV) systems in the world has increased from 7 GW in 2006 to about 300 GW at the end of 2016.

[0003] On the other hand, solar energy is less reliable than fossil fuels due to the uncertainty of power generation, such as the unevenness of solar radiation, climatic conditions, and the mismatch between power generation and consumption time. Therefore, it is necessary to use energy storage systems. In addition, it is expected that the demand for electricity will increase significantly in the future, which will lead to challenges in the operation of the power grid, especially during the peak period of power demand. Energy storage systems installed in residential areas, optimized scheduling of household equipment, and joint scheduling of household appliances and batteries can solve these challenges to a certain extent by shaving peaks and filling valleys.

[0004] Since the cost of purchasing and installing photovoltaic battery systems is high, the exact size of photovoltaic panels and battery storage systems needs to be determined to avoid oversizing and supply energy needs at the lowest possible price. Today, using computer software to determine the size of renewable energy systems is the most common method due to its accuracy and simplicity. The HOMER software designed and developed by the National Renewable Energy Laboratory in the United States is one of the most popular software. In addition, artificial intelligence techniques are becoming increasingly popular as an alternative to traditional methods, such as particle swarm optimization techniques, genetic algorithms, harmony search algorithms, ant colony algorithms, cuckoo search, and hybrid algorithms.

[0005] In recent studies, different energy resources have been chosen, such as photovoltaic panels, diesel generators, wind turbines, hydro turbines, etc. There are also different energy storage systems, such as batteries, fuel cells, water pumps, flywheels, etc., used to store the energy produced. Due to the complex systems used in these studies, the initial cost of the components is too high. Time-of-use pricing is the differentiation of electricity prices based on the time of electricity consumption. Utility providers try to reduce demand during peak hours by implementing a price structure that charges higher rates during typical peak consumption hours. Although in previous studies, time-of-use pricing and grid reliability have been considered in the optimization process, there have been no reports on the use of renewable systems to compensate for power outages and cover grid peaks. Summary of the invention

[0006] The present disclosure aims to solve at least one of the technical problems existing in the prior art and to provide an optimization method and device for a grid-connected photovoltaic battery system.

[0007] In one aspect of the present disclosure, a method for optimizing a grid-connected photovoltaic battery system is provided, the method comprising:

[0008] Establish output power model of grid-connected photovoltaic battery system;

[0009] Establishing an economic model of the grid-connected photovoltaic battery system according to the output power model of the grid-connected photovoltaic battery system;

[0010] According to the economic model of the grid-connected photovoltaic battery system, an economic scale optimization model of the grid-connected photovoltaic battery system is established;

[0011] Based on the coyote optimization algorithm, the economic scale optimization model of the grid-connected photovoltaic battery system is optimized.

[0012] In some embodiments, establishing an output power model of a grid-connected photovoltaic battery system includes:

[0013] The output power of the photovoltaic cell is obtained by using an accurate model of a single diode photovoltaic cell with a series resistor; the following formula (1) shows the maximum power generation of the photovoltaic cell:

[0014] P max.c (t) = V OC.c (t).I SC.c (t).FF(t) (1)

[0015] Among them, P max.c (t) is the maximum power generation of the photovoltaic cell, I SC.c (t) is the short-circuit current of the photovoltaic cell, V OC.c (t) is the open circuit voltage of the photovoltaic cell; I SC.c (t) and V OC.c The calculation formula of (t) is as follows: Formula (2) and Formula (3):

[0016]

[0017] V OC.c (t) = V OC.c.stc +K v (T C (t)-25) (3)

[0018] Among them, I SC.c.stc is the nominal short-circuit current of the photovoltaic cell, V OC.c.stc is the nominal open circuit voltage of the photovoltaic cell, G(t) is the global solar radiation intensity at time t, K i and K vare the temperature coefficients of current and voltage respectively; T C (t) is the photovoltaic cell temperature at time t, calculated according to the following formula (4):

[0019]

[0020] Among them, T a (t) is the ambient temperature at time t, NOCT is the rated operating temperature of the PV cell;

[0021] FF(t) is the fill factor of the photovoltaic cell at time t, calculated by the following formula (5):

[0022]

[0023] Among them, r s (t) and v oc (t) is the normalized series resistance and open circuit voltage of the photovoltaic cell, calculated by the following equations (6) and (7), respectively:

[0024]

[0025]

[0026] Among them, FF stc is the nominal value of the fill factor, FF 0stc is the nominal value of the fill factor at time zero;

[0027]

[0028] In the above formula, n is the inertia coefficient of the diode, K B is the Boltzmann constant, T a.k (t) is the ambient temperature at time t; In addition, formulas (9), (10), (11), (12), and (13) show the steps for obtaining the output power of the photovoltaic array:

[0029]

[0030] It should be noted that generally available data is for PV modules and not for PV cells; therefore:

[0031]

[0032]

[0033] I SC.c.stc (t) = I SC.m.stc (t) (12)

[0034] P av.PV (t) = N S.M ·NP.M N C.M P max.C (13)

[0035] Among them, P max.c.stc (t) is the nominal maximum power of the photovoltaic cell, P max.m.stc (t) is the nominal maximum power of the photovoltaic module, V OC.m.stc (t) is the nominal open circuit voltage of the photovoltaic module, I SC.m.stc (t) is the nominal short-circuit current of the photovoltaic module, P max.C is the maximum power of the photovoltaic cell, N C.M is the number of photovoltaic cells in the photovoltaic module, N S.M and N P.M is the number of PV modules connected in series and in parallel, P av.PV (t) is the output power of the entire photovoltaic array;

[0036] Battery packs are widely used as renewable energy generators. Photovoltaic power is stored in the battery pack. If the battery pack is fully charged, the photovoltaic power is directly fed into the residential part. In the case of insufficient power, the battery discharges to supply the demand. The power of the battery pack in the charging state is obtained by the following formula (14):

[0037]

[0038] Among them, P BAT (t) is the battery charge state at time t, P BAT (t-1) represents the charging state at time t-1; σ is the self-discharge rate of the battery, η INV and η BAT are the inverter and battery charging efficiencies, which are considered to be 0.95 and 0.9, respectively; P Load (t) is the load demand, N PV is the number of photovoltaic panels;

[0039] If the photovoltaic power generation is less than the load consumption, the battery starts to discharge; the discharge efficiency is equal to the charging efficiency; P BAT In the discharge state, it is expressed by the following formula (15):

[0040]

[0041] In some embodiments, establishing an economic model of the grid-connected photovoltaic battery system according to the output power model of the grid-connected photovoltaic battery system includes:

[0042] The NPC of a system is the present value of all costs during the life cycle of the system minus the present value of all revenues during its life cycle. Since the reliability coefficient must be converted into cost, the power supply loss LPS is multiplied by an arbitrary reliability penalty coefficient RPF to form a single target cost function. The formula for reliability cost CRE is as follows (16):

[0043] CRE=LPS*RPF (16)

[0044] The calculation formula of emission cost CE is as follows (17):

[0045] CE==(Ccd*Pcd+Ccm*Pcm+Cuh*Puh+Cpm*Ppm+Csd*Psd+Cno*Pno) / 10 6 (17)

[0046] Among them, C i is the cost, P i is the pollution generated by different pollutants, i = {cd, cm, uh, pm, sd, no}, refers to different pollutants; the net present value cost NPC of the system will be obtained from the total cost of components, replacement, operation and maintenance, reliability and emission impact, as shown in the following formula (18):

[0047] NPC=CT+CE+CRE (18)

[0048] Where CT is the total cost of the system; its value is calculated from the following formula (19):

[0049]

[0050] Among them, CI is the initial cost, CR is the replacement cost, CM is the operation and maintenance cost, C PV is the cost of photovoltaic panels, N PV is the number of photovoltaic panels, C BAT is the battery unit cost, N BAT is the number of battery cells; CPV is the cumulative present value, C PV,R is the replacement cost of photovoltaics, C BAT,R is the replacement cost of the battery, C PV,M is the maintenance cost of photovoltaics, C BAT,M is the maintenance cost of the battery; the cost of purchasing electricity from the grid is considered to be C Grid ; f is the annual inflation rate (%), i is the annual interest rate (%); In order to obtain the COE parameter, the total annualized cost must be calculated first, as shown in the following formula (20):

[0051] C A =NPC*CRF (i,LT) (20)

[0052] Among them, C A is the annualized total cost (USD / year), CRF (i,LT) is the capital recovery coefficient, which is calculated as follows (21):

[0053]

[0054] COE is the average cost per kilowatt-hour of effective electricity. It is one of the most important parameters for comparing the cost-effectiveness of hybrid energy systems and finding the optimal size of the system. The calculation formula is as follows (22):

[0055]

[0056] Among them, E served It is the energy provided in one year.

[0057] In some embodiments, establishing an economic scale optimization model of the grid-connected photovoltaic battery system according to the economic model of the grid-connected photovoltaic battery system includes:

[0058] The objective function of the optimization problem is to minimize NPC and COE under two different conditions; the objective function is defined as follows (23):

[0059]

[0060] There are some constraints in the optimization process; these constraints are as follows (24):

[0061]

[0062] Among them, PV max is the maximum number of photovoltaic panels, Battery max is the maximum available storage capacity; and is the minimum and maximum charge capacity of the battery, DOD is the depth of discharge of the battery, S Bat is the rated capacity of the battery.

[0063] In some embodiments, the optimizing the economic scale optimization model of the grid-connected photovoltaic battery system based on the coyote optimization algorithm includes:

[0064] (1) Initialization and random grouping:

[0065] Set parameters such as the number of coyote groups N p 、N number of coyotes in the group c and the maximum number of iterations N gen,Since COA is a random algorithm, the initial social state factors of each coyote are randomly set as follows (25); the social adaptability of each coyote is calculated and randomly grouped;

[0066] soc j =lb j +r j *(ub j -lb j ) (25)

[0067] Among them, lb j andub j denote the lower and upper bounds of the j-th state factor of the coyote, j = 1, 2, ..., D, r j is a random number uniformly distributed within [0,1];

[0068] (2) Growth of coyotes within the group:

[0069] Determine the optimal wolf alpha in the group, calculate the group cultural trend, and randomly select two coyotes. The four factors affect the growth of coyotes. The calculation of the group cultural trend is as follows (26):

[0070] cult j =median(A j ) (26)

[0071] Where A is an N c A matrix with D rows and D columns, representing N c A solution vector, A j represents the j-th column of matrix A, and median represents the median;

[0072] During the growth of the coyote, the difference δ between the best coyote alpha in the group and a randomly selected coyote in the group is first calculated. 1 , the difference between the group cultural trend and another random coyote in the group δ 2 , then the coyotes in the group are in δ 1 and δ 2 It grows under the influence of, as shown in the following equations (27) and (28):

[0073] δ 1 =Lbest-soc r1 ,δ 2 =cult-soc r2 (27)

[0074] new_soc c =soc c +s 1 *δ 1 +s 2 *δ 2(28)

[0075] Among them, r 1 and r 2 Represents two different random coyote numbers, L best Represents the best coyote alpha wolf in the group;s 1 and 2 They are δ 1 and δ 2 The random weight of s 1 and 2 is a random number uniformly distributed within [0,1];

[0076] After each coyote in the group grows up, its social adaptability is calculated and greedy selection is adopted, as shown in the following formula (29). By retaining high-quality coyotes to participate in the growth of other coyotes in the group, the convergence speed of the algorithm is accelerated:

[0077]

[0078] (3) Life and Death of Coyotes:

[0079] In COA, the age of coyotes is measured in years, and after each group of coyotes grows up, a newborn coyote is produced;

[0080] The birth of new coyotes is influenced by the social conditions and social environment of two randomly selected parent coyotes; the way new coyotes are produced is shown in the following formula (30):

[0081]

[0082] Among them, cr 1 and cr 2 are two random different coyotes in group p, j 1 and j 2 are two random dimensions of the newborn coyote; P s is the probability of dispersion, P a is the association probability, as shown below (31); the magnitude of the dispersion and association probability affects the diversity of newborn coyotes; R j is a random number within the value range of the j-th dimension of the decision variable; rnd j is a random number uniformly distributed in [0,1]:

[0083]

[0084] where ω represents a coyote pack that is less adaptable than newborn wolves, is the number of coyotes in the group, and the age of newborn coyotes is 0;

[0085] (4) Coyotes are driven away and accepted

[0086] Initially, coyotes were randomly assigned to groups, but sometimes they left and joined other groups; the probability of a coyote being excluded or accepted by a group is expressed as P. e This mechanism helps the exchange of information between COA groups and promotes mutual influence between coyotes in the population, as shown in the following formula (32);

[0087]

[0088] After initialization and random grouping, the three steps of coyote growth, coyote life and death, and coyote expulsion and acceptance are carried out in sequence; if the iteration termination condition is met, the optimal coyote is output; otherwise, jump to step (2);

[0089] Finally, the optimal size of the hybrid power system, i.e. the optimal size of the PV cells / inverter, is obtained through the Coyote optimization algorithm.

[0090] Another aspect of the present disclosure provides an optimization device for a grid-connected photovoltaic battery system, the device comprising:

[0091] The first building module is used to build an output power model of a grid-connected photovoltaic battery system;

[0092] A second establishing module is used to establish an economic model of the grid-connected photovoltaic battery system according to the output power model of the grid-connected photovoltaic battery system;

[0093] A third establishing module is used to establish an economic scale optimization model of the grid-connected photovoltaic battery system according to the economic model of the grid-connected photovoltaic battery system;

[0094] The optimization module is used to optimize the economic scale optimization model of the grid-connected photovoltaic battery system based on the coyote optimization algorithm.

[0095] In some implementations, the first establishing module is further configured to:

[0096] The output power of the photovoltaic cell is obtained by using an accurate model of a single diode photovoltaic cell with a series resistor; the following formula (1) shows the maximum power generation of the photovoltaic cell:

[0097] P max.c (t) = V OC.c (t).I SC.c (t).FF(t) (1)

[0098] Among them, P max.c (t) is the maximum power generation of the photovoltaic cell, I SC.c (t) is the short-circuit current of the photovoltaic cell, V OC.c (t) is the open circuit voltage of the photovoltaic cell; I SC.c (t) and VOC.c The calculation formula of (t) is as follows: Formula (2) and Formula (3):

[0099]

[0100] V OC.c (t) = V OC.c.stc +K v (T C (t)-25) (3)

[0101] Among them, I SC.c.stc is the nominal short-circuit current of the photovoltaic cell, V OC.c.stc is the nominal open circuit voltage of the photovoltaic cell, G(t) is the global solar radiation intensity at time t, K i and K v are the temperature coefficients of current and voltage respectively; T C (t) is the photovoltaic cell temperature at time t, calculated according to the following formula (4):

[0102]

[0103] Among them, T a (t) is the ambient temperature at time t, NOCT is the rated operating temperature of the PV cell;

[0104] FF(t) is the fill factor of the photovoltaic cell at time t, calculated by the following formula (5):

[0105]

[0106] Among them, r s (t) and v oc (t) is the normalized series resistance and open circuit voltage of the photovoltaic cell, calculated by the following equations (6) and (7), respectively:

[0107]

[0108]

[0109] Among them, FF stc is the nominal value of the fill factor, FF 0stc is the nominal value of the fill factor at time zero;

[0110]

[0111] In the above formula, n is the inertia coefficient of the diode, K B is the Boltzmann constant, T a.k (t) is the ambient temperature at time t; In addition, formulas (9), (10), (11), (12), and (13) show the steps for obtaining the output power of the photovoltaic array:

[0112]

[0113] It should be noted that generally available data is for PV modules and not for PV cells; therefore:

[0114]

[0115]

[0116] I SC.c.stc (t) = I SC.m.stc (t) (12)

[0117] P av.PV (t) = N S.M ·N P.M N C.M P max.C (13)

[0118] Among them, P max.c.stc (t) is the nominal maximum power of the photovoltaic cell, P max.m.stc (t) is the nominal maximum power of the photovoltaic module, V OC.m.stc (t) is the nominal open circuit voltage of the photovoltaic module, I SC.m.stc (t) is the nominal short-circuit current of the photovoltaic module, P max.C is the maximum power of the photovoltaic cell, N C.M is the number of photovoltaic cells in the photovoltaic module, N S.M and N P.M is the number of PV modules connected in series and in parallel, P av.PV (t) is the output power of the entire photovoltaic array;

[0119] Battery packs are widely used as renewable energy generators. Photovoltaic power is stored in the battery pack. If the battery pack is fully charged, the photovoltaic power is directly fed into the residential part. In the case of insufficient power, the battery discharges to supply the demand. The power of the battery pack in the charging state is obtained by the following formula (14):

[0120]

[0121] Among them, P BAT (t) is the battery charge state at time t, P BAT (t-1) represents the charging state at time t-1; σ is the self-discharge rate of the battery, η INV and η BAT are the inverter and battery charging efficiencies, which are considered to be 0.95 and 0.9, respectively; P Load (t) is the load demand, N PV is the number of photovoltaic panels;

[0122] If the photovoltaic power generation is less than the load consumption, the battery starts to discharge; the discharge efficiency is equal to the charging efficiency; P BAT In the discharge state, it is expressed by the following formula (15):

[0123]

[0124] In some implementations, the second establishing module is further configured to:

[0125] The NPC of a system is the present value of all costs during the life cycle of the system minus the present value of all revenues during its life cycle. Since the reliability coefficient must be converted into cost, the power supply loss LPS is multiplied by an arbitrary reliability penalty coefficient RPF to form a single target cost function. The formula for reliability cost CRE is as follows (16):

[0126] CRE=LPS*RPF (16)

[0127] The calculation formula of emission cost CE is as follows (17):

[0128] CE==(Ccd*Pcd+Ccm*Pcm+Cuh*Puh+Cpm*Ppm+Csd*Psd+Cno*Pno) / 10 6 (17)

[0129] Among them, C i is the cost, P i is the pollution generated by different pollutants, i = {cd, cm, uh, pm, sd, no}, refers to different pollutants; the net present value cost NPC of the system will be obtained from the total cost of components, replacement, operation and maintenance, reliability and emission impact, as shown in the following formula (18):

[0130] NPC=CT+CE+CRE (18)

[0131] Where CT is the total cost of the system; its value is calculated from the following formula (19):

[0132]

[0133] Among them, CI is the initial cost, CR is the replacement cost, CM is the operation and maintenance cost, C PV is the cost of photovoltaic panels, N PV is the number of photovoltaic panels, C BAT is the battery unit cost, N BAT is the number of battery cells; CPV is the cumulative present value, C PV,R is the replacement cost of photovoltaics, C BAT,R is the replacement cost of the battery, C PV,Mis the maintenance cost of photovoltaics, C BAT,M is the maintenance cost of the battery; the cost of purchasing electricity from the grid is considered to be C Grid ; f is the annual inflation rate (%), i is the annual interest rate (%); In order to obtain the COE parameter, the total annualized cost must be calculated first, as shown in the following formula (20):

[0134] C A =NPC*CRF (i,LT) (20)

[0135] Among them, C A is the annualized total cost (USD / year), CRF (i,LT) is the capital recovery coefficient, which is calculated as follows (21):

[0136]

[0137] COE is the average cost per kilowatt-hour of effective electricity. It is one of the most important parameters for comparing the cost-effectiveness of hybrid energy systems and finding the optimal size of the system. The calculation formula is as follows (22):

[0138]

[0139] Among them, E served It is the energy provided in one year.

[0140] In some implementations, the third establishing module is further configured to:

[0141] The objective function of the optimization problem is to minimize NPC and COE under two different conditions; the objective function is defined as follows (23):

[0142]

[0143] There are some constraints in the optimization process; these constraints are as follows (24):

[0144]

[0145] Among them, PV max is the maximum number of photovoltaic panels, Battery max is the maximum available storage capacity; and is the minimum and maximum charge capacity of the battery, DOD is the depth of discharge of the battery, S Bat is the rated capacity of the battery.

[0146] In some implementations, the optimization module is further configured to:

[0147] (1) Initialization and random grouping:

[0148] Set parameters such as the number of coyote groups N p 、N number of coyotes in the group c and the maximum number of iterations N gen ,Since COA is a random algorithm, the initial social state factors of each coyote are randomly set as follows (25); the social adaptability of each coyote is calculated and randomly grouped;

[0149] soc j =lb j +r j *(ub j -lb j ) (25)

[0150] Among them, lb j andub j denote the lower and upper bounds of the j-th state factor of the coyote, j = 1, 2, ..., D, r j is a random number uniformly distributed within [0,1];

[0151] (2) Growth of coyotes within the group:

[0152] Determine the optimal wolf alpha in the group, calculate the group cultural trend, and randomly select two coyotes. The four factors affect the growth of coyotes. The calculation of the group cultural trend is as follows (26):

[0153] cult j =median(A j ) (26)

[0154] Where A is an N c A matrix with D rows and D columns, representing N c A solution vector, A j represents the j-th column of matrix A, and median represents the median;

[0155] During the growth of the coyote, the difference δ between the best coyote alpha in the group and a randomly selected coyote in the group is first calculated. 1 , the difference between the group cultural trend and another random coyote in the group δ 2 , then the coyotes in the group are in δ 1 and δ 2 It grows under the influence of, as shown in the following equations (27) and (28):

[0156] δ 1 =Lbest-soc r1 ,δ 2 =cult-soc r2 (27)

[0157] new_soc c=soc c +s 1 *δ 1 +s 2 *δ 2 (28)

[0158] Among them, r 1 and r 2 Represents two different random coyote numbers, L best Represents the best coyote alpha wolf in the group;s 1 and 2 They are δ 1 and δ 2 The random weight of s 1 and 2 is a random number uniformly distributed within [0,1];

[0159] After each coyote in the group grows up, its social adaptability is calculated and greedy selection is adopted, as shown in the following formula (29). By retaining high-quality coyotes to participate in the growth of other coyotes in the group, the convergence speed of the algorithm is accelerated:

[0160]

[0161] (3) Life and Death of Coyotes:

[0162] In COA, the age of coyotes is measured in years, and after each group of coyotes grows up, a newborn coyote is produced;

[0163] The birth of new coyotes is influenced by the social conditions and social environment of two randomly selected parent coyotes; the way new coyotes are produced is shown in the following formula (30):

[0164]

[0165] Among them, cr 1 and cr 2 are two random different coyotes in group p, j 1 and j 2 are two random dimensions of the newborn coyote; P s is the probability of dispersion, P a is the association probability, as shown below (31); the magnitude of the dispersion and association probability affects the diversity of newborn coyotes; R j is a random number within the value range of the j-th dimension of the decision variable; rnd j is a random number uniformly distributed in [0,1]:

[0166]

[0167] where ω represents a coyote pack that is less adaptable than newborn wolves, is the number of coyotes in the group, and the age of newborn coyotes is 0;

[0168] (4) Coyotes are driven away and accepted

[0169] Initially, coyotes were randomly assigned to groups, but sometimes they left and joined other groups; the probability of a coyote being excluded or accepted by a group is expressed as P. e This mechanism helps the exchange of information between COA groups and promotes mutual influence between coyotes in the population, as shown in the following formula (32);

[0170]

[0171] After initialization and random grouping, the three steps of coyote growth, coyote life and death, and coyote expulsion and acceptance are carried out in sequence; if the iteration termination condition is met, the optimal coyote is output; otherwise, jump to step (2);

[0172] Finally, the optimal size of the hybrid power system, i.e. the optimal size of the PV cells / inverter, is obtained through the Coyote optimization algorithm.

[0173] The optimization method and device of the grid-connected photovoltaic battery system disclosed in the present invention apply the coyote optimization algorithm to solve the scale optimization problem of the grid-connected photovoltaic battery system considering reliability and greenhouse gas emissions. Compared with the non-renewable energy grid-connected system, the NPC and COE of the system are reduced. BRIEF DESCRIPTION OF THE DRAWINGS

[0174] Figure 1 A flow chart of an optimization method for a grid-connected photovoltaic battery system according to an embodiment of the present disclosure;

[0175] Figure 2 It is a structural schematic diagram of an optimization device for a grid-connected photovoltaic battery system according to another embodiment of the present disclosure. DETAILED DESCRIPTION

[0176] In order to enable those skilled in the art to better understand the technical solution of the present disclosure, the present disclosure is further described in detail below in conjunction with the accompanying drawings and specific implementation methods.

[0177] like Figure 1 It can be seen that a method S100 for optimizing a grid-connected photovoltaic battery system includes:

[0178] S110. Establish an output power model of a grid-connected photovoltaic battery system.

[0179] Specifically, in this step, an accurate model of a single diode photovoltaic cell with a series resistor is used to obtain the output power of the photovoltaic cell; the following formula (1) shows the maximum power generation of the photovoltaic cell:

[0180] Pmax.c (t) = V OC.c (t).I SC.c (t).FF(t) (1)

[0181] Among them, P max.c (t) is the maximum power generation of the photovoltaic cell, I SC.c (t) is the short-circuit current of the photovoltaic cell, V OC.c (t) is the open circuit voltage of the photovoltaic cell; I SC.c (t) and V OC.c The calculation formula of (t) is as follows: Formula (2) and Formula (3):

[0182]

[0183] V OC.c (t) = V OC.c.stc +K v (T C (t)-25) (3)

[0184] Among them, I SC.c.stc is the nominal short-circuit current of the photovoltaic cell, V OC.c.stc is the nominal open circuit voltage of the photovoltaic cell, G(t) is the global solar radiation intensity at time t, K i and K v are the temperature coefficients of current and voltage respectively; T C (t) is the photovoltaic cell temperature at time t, calculated according to the following formula (4):

[0185]

[0186] Among them, T a (t) is the ambient temperature at time t, NOCT is the rated operating temperature of the PV cell;

[0187] FF(t) is the fill factor of the photovoltaic cell at time t, calculated by the following formula (5):

[0188]

[0189] Among them, r s (t) and v oc (t) is the normalized series resistance and open circuit voltage of the photovoltaic cell, calculated by the following equations (6) and (7), respectively:

[0190]

[0191]

[0192] Among them, FF stc is the nominal value of the fill factor, FF 0stcis the nominal value of the fill factor at time zero;

[0193]

[0194] In the above formula, n is the inertia coefficient of the diode, K B is the Boltzmann constant, T a.k (t) is the ambient temperature at time t; In addition, formulas (9), (10), (11), (12), and (13) show the steps for obtaining the output power of the photovoltaic array:

[0195]

[0196] It should be noted that generally available data is for PV modules and not for PV cells; therefore:

[0197]

[0198]

[0199] I SC.c.stc (t) = I SC.m.stc (t) (12)

[0200] P av.PV (t) = N S.M ·N P.M N C.M P max.C (13)

[0201] Among them, P max.c.stc (t) is the nominal maximum power of the photovoltaic cell, P max.m.stc (t) is the nominal maximum power of the photovoltaic module, V OC.m.stc (t) is the nominal open circuit voltage of the photovoltaic module, I SC.m.stc (t) is the nominal short-circuit current of the photovoltaic module, P max.C is the maximum power of the photovoltaic cell, N C.M is the number of photovoltaic cells in the photovoltaic module, N S.M and N P.M is the number of PV modules connected in series and in parallel, P av.PV (t) is the output power of the entire photovoltaic array;

[0202] Battery packs are widely used as renewable energy generators. Photovoltaic power is stored in the battery pack. If the battery pack is fully charged, the photovoltaic power is directly fed into the residential part. In the case of insufficient power, the battery discharges to supply the demand. The power of the battery pack in the charging state is obtained by the following formula (14):

[0203]

[0204] Among them, P BAT (t) is the battery charge state at time t, P BAT (t-1) represents the charging state at time t-1; σ is the self-discharge rate of the battery, η INV and η BAT are the inverter and battery charging efficiencies, which are considered to be 0.95 and 0.9, respectively; P Load (t) is the load demand, N PV is the number of photovoltaic panels;

[0205] If the photovoltaic power generation is less than the load consumption, the battery starts to discharge; the discharge efficiency is equal to the charging efficiency; P BAT In the discharge state, it is expressed by the following formula (15):

[0206]

[0207] S120: Establishing an economic model of the grid-connected photovoltaic battery system according to the output power model of the grid-connected photovoltaic battery system.

[0208] Specifically, in this step, the NPC of the system is the present value of all costs during the life cycle of the system minus the present value of all revenues during its life cycle; since the reliability coefficient must be converted into cost, the power supply loss LPS is multiplied by an arbitrary reliability penalty coefficient RPF to form a single target cost function; the formula of reliability cost CRE is as follows (16):

[0209] CRE=LPS*RPF (16)

[0210] The calculation formula of emission cost CE is as follows (17):

[0211] CE==(Ccd*Pcd+Ccm*Pcm+Cuh*Puh+Cpm*Ppm+Csd*Psd+Cno*Pno) / 10 6 (17)

[0212] Among them, C i is the cost, P i Pollution caused by different pollutants C i is the cost of different pollutants, P i The pollution generated by different pollutants, i = {cd, cm, uh, pm, sd, no}, refers to different pollutants, cd: carbon dioxide, cm: carbon monoxide, uh: unburned hydrocarbons, pm: special substances, sd: sulfur dioxide, no: nitrogen oxides; the net present value cost NPC of the system will be obtained from the total cost of components, replacement, operation and maintenance, reliability and emission impact, as shown in the following formula (18):

[0213] NPC=CT+CE+CRE (18)

[0214] Where CT is the total cost of the system; its value is calculated from the following formula (19):

[0215]

[0216] Among them, CI is the initial cost, CR is the replacement cost, CM is the operation and maintenance cost, C PV is the cost of photovoltaic panels, N PV is the number of photovoltaic panels, C BAT is the battery unit cost, N BAT is the number of battery cells; CPV is the cumulative present value, C PV,R is the replacement cost of photovoltaics, C BAT,R is the replacement cost of the battery, C PV,M is the maintenance cost of photovoltaics, C BAT,M is the maintenance cost of the battery; the cost of purchasing electricity from the grid is considered to be C Grid ; f is the annual inflation rate (%), i is the annual interest rate (%); In order to obtain the COE parameter, the total annualized cost must be calculated first, as shown in the following formula (20):

[0217] C A =NPC*CRF (i,LT) (20)

[0218] Among them, C A is the annualized total cost (USD / year), CRF (i,LT) is the capital recovery coefficient, which is calculated as follows (21):

[0219]

[0220] COE is the average cost per kilowatt-hour of effective electricity. It is one of the most important parameters for comparing the cost-effectiveness of hybrid energy systems and finding the optimal size of the system. The calculation formula is as follows (22):

[0221]

[0222] Among them, E served It is the energy provided in one year.

[0223] S130: Establishing an economic scale optimization model for the grid-connected photovoltaic battery system according to the economic model of the grid-connected photovoltaic battery system.

[0224] Specifically, in this step, the objective function of the optimization problem is to minimize NPC and COE under two different conditions; the objective function is defined as follows (23):

[0225]

[0226] There are some constraints in the optimization process; these constraints are as follows (24):

[0227]

[0228] Among them, PV max is the maximum number of photovoltaic panels, Battery max is the maximum available storage capacity; and is the minimum and maximum charge capacity of the battery, DOD is the depth of discharge of the battery, S Bat is the rated capacity of the battery.

[0229] S140, optimizing the economic scale optimization model of the grid-connected photovoltaic battery system based on the coyote optimization algorithm.

[0230] Specifically, this step includes the following steps:

[0231] (1) Initialization and random grouping:

[0232] Set parameters such as the number of coyote groups N p 、N number of coyotes in the group c and the maximum number of iterations N gen ,Since COA is a random algorithm, the initial social state factors of each coyote are randomly set as follows (25); the social adaptability of each coyote is calculated and randomly grouped;

[0233] soc j =lb j +r j *(ub j -lb j ) (25)

[0234] Among them, lb j andub j denote the lower and upper bounds of the j-th state factor of the coyote, j = 1, 2, ..., D, r j is a random number uniformly distributed within [0,1];

[0235] (2) Growth of coyotes within the group:

[0236] Determine the optimal wolf alpha in the group, calculate the group cultural trend, and randomly select two coyotes. The four factors affect the growth of coyotes. The calculation of the group cultural trend is as follows (26):

[0237] cult j =median(A j ) (26)

[0238] Where A is an N c A matrix with D rows and D columns, representing N c A solution vector, A j represents the j-th column of matrix A, and median represents the median;

[0239] During the growth of the coyote, the difference δ between the best coyote alpha in the group and a randomly selected coyote in the group is first calculated. 1 , the difference between the group cultural trend and another random coyote in the group δ 2 , then the coyotes in the group are in δ 1 and δ 2 It grows under the influence of, as shown in the following equations (27) and (28):

[0240] δ 1 =Lbest-soc r1 ,δ 2 =cult-soc r2 (27)

[0241] new_soc c =soc c +s 1 *δ 1 +s 2 *δ 2 (28)

[0242] Among them, r 1 and r 2 Represents two different random coyote numbers, L best Represents the best coyote alpha wolf in the group;s 1 and 2 They are δ 1 and δ 2 The random weight of s 1 and 2 is a random number uniformly distributed within [0,1];

[0243] After each coyote in the group grows up, its social adaptability is calculated and greedy selection is adopted, as shown in the following formula (29). By retaining high-quality coyotes to participate in the growth of other coyotes in the group, the convergence speed of the algorithm is accelerated:

[0244]

[0245] (3) Life and Death of Coyotes:

[0246] In COA, the age of coyotes is measured in years, and after each group of coyotes grows up, a newborn coyote is produced;

[0247] The birth of new coyotes is influenced by the social conditions and social environment of two randomly selected parent coyotes; the way new coyotes are produced is shown in the following formula (30):

[0248]

[0249] Among them, cr 1 and cr 2 are two random different coyotes in group p, j 1 and j 2 are two random dimensions of the newborn coyote; P s is the probability of dispersion, P a is the association probability, as shown below (31); the magnitude of the dispersion and association probability affects the diversity of newborn coyotes; R j is a random number within the value range of the j-th dimension of the decision variable; rnd j is a random number uniformly distributed in [0,1]:

[0250]

[0251] where ω represents a coyote pack that is less adaptable than newborn wolves, is the number of coyotes in the group, and the age of newborn coyotes is 0;

[0252] (4) Coyotes are driven away and accepted

[0253] Initially, coyotes were randomly assigned to groups, but sometimes they left and joined other groups; the probability of a coyote being excluded or accepted by a group is expressed as P. e This mechanism helps the exchange of information between COA groups and promotes mutual influence between coyotes in the population, as shown in the following formula (32);

[0254]

[0255] After initialization and random grouping, the three steps of coyote growth, coyote life and death, and coyote expulsion and acceptance are carried out in sequence; if the iteration termination condition is met, the optimal coyote is output; otherwise, jump to step (2);

[0256] Finally, the optimal size of the hybrid power system, i.e. the optimal size of the PV cells / inverter, is obtained through the Coyote optimization algorithm.

[0257] The optimization method for the grid-connected photovoltaic battery system of this embodiment uses the coyote optimization algorithm to solve the scale optimization problem of the grid-connected photovoltaic battery system considering reliability and greenhouse gas emissions. Compared with the non-renewable energy grid-connected system, the NPC and COE of the system are reduced.

[0258] Another aspect of the present disclosure is Figure 2As shown, an optimization device 100 for a grid-connected photovoltaic battery system is provided. The device can be applied to the method described above. The specific technical content and corresponding beneficial effects can be referred to the relevant records above and will not be elaborated here.

[0259] The device 100 comprises:

[0260] A first establishing module 110 is used to establish an output power model of a grid-connected photovoltaic battery system;

[0261] A second establishing module 120, configured to establish an economic model of the grid-connected photovoltaic battery system according to an output power model of the grid-connected photovoltaic battery system;

[0262] A third establishing module 130 is used to establish an economic scale optimization model of the grid-connected photovoltaic battery system according to the economic model of the grid-connected photovoltaic battery system;

[0263] The optimization module 140 is used to optimize the economic scale optimization model of the grid-connected photovoltaic battery system based on the coyote optimization algorithm.

[0264] It is to be understood that the above embodiments are merely exemplary embodiments used to illustrate the principles of the present disclosure, but the present disclosure is not limited thereto. For those of ordinary skill in the art, various modifications and improvements can be made without departing from the spirit and substance of the present disclosure, and these modifications and improvements are also considered to be within the scope of protection of the present disclosure.

Claims

1. A method for optimizing a grid-connected photovoltaic battery system, characterized in that: The method comprises: Establish output power model of grid-connected photovoltaic battery system; Establishing an economic model of the grid-connected photovoltaic battery system according to the output power model of the grid-connected photovoltaic battery system; According to the economic model of the grid-connected photovoltaic battery system, an economic scale optimization model of the grid-connected photovoltaic battery system is established; Based on the Coyote optimization algorithm, optimizing the economic scale optimization model of the grid-connected photovoltaic battery system; The step of establishing an output power model of a grid-connected photovoltaic battery system comprises: The output power of the photovoltaic cell is obtained by using an accurate model of a single diode photovoltaic cell with a series resistor; the following formula (1) shows the maximum power generation of the photovoltaic cell: P max.c (t)=V OC.c (t).I SC.c (t).FF(t) (1) Among them, P max.c (t) is the maximum power generation of the photovoltaic cell, I SC.c (t) is the short-circuit current of the photovoltaic cell, V OC.c (t) is the open circuit voltage of the photovoltaic cell; I SC.c (t) and V OC.c The calculation formula of (t) is as follows: Formula (2) and Formula (3): V OC.c (t)=V OC.c.stc +K v (T C (t)-25) (3) Among them, I SC.c.stc is the nominal short-circuit current of the photovoltaic cell, V OC.c.stc is the nominal open circuit voltage of the photovoltaic cell, G(t) is the global solar radiation intensity at time t, K i and K v are the temperature coefficients of current and voltage respectively; T C (t) is the photovoltaic cell temperature at time t, calculated according to the following formula (4): Among them, T a (t) is the ambient temperature at time t, NOCT is the rated operating temperature of the photovoltaic cell; FF(t) is the fill factor of the photovoltaic cell at time t, which is calculated by the following formula (5): Among them, r s (t) and v oc (t) is the normalized series resistance and open circuit voltage of the photovoltaic cell, which are calculated by the following equations (6) and (7), respectively: Among them, FF stc is the nominal value of the fill factor, FF 0stc is the nominal value of the fill factor at time zero; In the above formula, n is the inertia coefficient of the diode, K B is the Boltzmann constant, T a.k (t) is the ambient temperature at time t; In addition, formulas (9), (10), (11), (12), and (13) show the steps for obtaining the output power of the photovoltaic array: It should be noted that generally available data is for PV modules and not for PV cells; therefore: I SC.c.stc (t)=I SC.m.stc (t) (12) P av.PV (t)=N S.M ·N P.M N C.M P max.C (13) Among them, P max.c.stc (t) is the nominal maximum power of the photovoltaic cell, P max.m.stc (t) is the nominal maximum power of the photovoltaic module, V OC.m.stc (t) is the nominal open circuit voltage of the photovoltaic module, I SC.m.stc (t) is the nominal short-circuit current of the photovoltaic module, P max.C is the maximum power of the photovoltaic cell, N C.M is the number of photovoltaic cells in the photovoltaic module, N S.M and N P.M is the number of PV modules connected in series and in parallel, P av.PV (t) is the output power of the entire photovoltaic array; Battery packs are widely used as renewable energy generators. Photovoltaic power is stored in the battery pack. If the battery pack is fully charged, the photovoltaic power is directly fed into the residential part. In the case of insufficient power, the battery discharges to supply the demand. The power of the battery pack in the charging state is obtained by the following formula (14): Among them, P BAT (t) is the battery charging state at time t, P BAT (t-1) represents the charging state at time t-1; σ is the self-discharge rate of the battery, η INV and η BAT are the inverter and battery charging efficiencies, which are considered to be 0.95 and 0.9, respectively; P Load (t) is the load demand, N PV is the number of photovoltaic panels; If the photovoltaic power generation is less than the load consumption, the battery starts to discharge; the discharge efficiency is equal to the charging efficiency; PBAT in the discharge state is expressed by the following formula (15): The step of establishing an economic model of the grid-connected photovoltaic battery system according to the output power model of the grid-connected photovoltaic battery system comprises: The NPC of a system is the present value of all costs during the life cycle of the system minus the present value of all revenues during its life cycle. Since the reliability coefficient must be converted into cost, the power supply loss LPS is multiplied by an arbitrary reliability penalty coefficient RPF to form a single target cost function. The formula for reliability cost CRE is as follows (16): CRE=LPS*RPF (16) The calculation formula of emission cost CE is as follows (17): CE==(Ccd*Pcd+Ccm*Pcm+Cuh*Puh+Cpm*Ppm+Csd*Psd+Cno*Pno) / 10 6 (17) Among them, C i is the cost of different pollutants, P i The pollution generated by different pollutants, i = {cd, cm, uh, pm, sd, no}, refers to different pollutants; The net present value cost NPC of the system will be obtained from the total cost of components, replacement, operation and maintenance, reliability and emission impact as follows (18): NPC=CT+CE+CRE (18) Where CT is the total cost of the system; its value is calculated from the following formula (19): Among them, CI is the initial cost, CR is the replacement cost, CM is the operation and maintenance cost, C PV is the cost of photovoltaic panels, N PV is the number of photovoltaic panels, C BAT is the battery unit cost, N BAT is the number of battery cells; CPV is the cumulative present value, C PV,R is the replacement cost of photovoltaic, C BAT,R is the replacement cost of the battery, C PV,M is the maintenance cost of photovoltaic, C BAT,M is the maintenance cost of the battery; the cost of purchasing electricity from the grid is considered to be C Grid ; f is the annual inflation rate (%), i is the annual interest rate (%); In order to obtain the COE parameter, the total annualized cost must be calculated first, as shown in the following formula (20): C A =NPC*CRF (i,LT) (20) Among them, C A is the annualized total cost (US dollars / year), CRF (i,LT) is the capital recovery coefficient, and its calculation formula is as follows (21): COE is the average cost per kilowatt-hour of effective electricity. It is one of the most important parameters for comparing the cost-effectiveness of hybrid energy systems and finding the optimal scale of the system. The calculation formula is as follows (22): Among them, E served The energy provided in a year; The step of establishing an economic scale optimization model of the grid-connected photovoltaic battery system according to the economic model of the grid-connected photovoltaic battery system comprises: The objective function of the optimization problem is to minimize NPC and COE under two different conditions; the objective function is defined as follows (23): There are some constraints in the optimization process; these constraints are as follows (24): Among them, PV max is the maximum number of photovoltaic panels, Battery max is the maximum available storage capacity; and is the minimum and maximum charge capacity of the battery, DOD is the depth of discharge of the battery, S Bat is the rated capacity of the battery.

2. The method according to claim 1, characterized in that The economic scale optimization model of the grid-connected photovoltaic battery system is optimized based on the coyote optimization algorithm, including: (1) Initialization and random grouping: Set parameters such as the number of coyote groups N p 、N number of coyotes in the group c and the maximum number of iterations N gen ,Since COA is a random algorithm, the initial social state factors of each coyote are randomly set as follows (25); the social adaptability of each coyote is calculated and randomly grouped; soc j =lb j +r j *(ub j -lb j ) (25) Among them, lb j andub j denote the lower and upper bounds of the j-th state factor of the coyote, j = 1, 2, ..., D, r j is a random number uniformly distributed within [0,1]; (2) Growth of coyotes within the group: Determine the optimal wolf alpha in the group, calculate the group cultural trend, and randomly select two coyotes. The four factors affect the growth of coyotes. The calculation of the group cultural trend is as follows (26): cult j =median(A j )(26) Where A is an N c A matrix with D rows and D columns, representing N c A solution vector, A j represents the j-th column of matrix A, and median represents the median; In the process of coyote growth, the difference δ1 between the optimal coyote alpha in the group and a randomly selected coyote in the group and the difference δ2 between the group cultural trend and another random coyote in the group are calculated first, and then the coyotes in the group grow under the influence of δ1 and δ2, as shown in the following equations (27) and (28): δ1=Lbest-soc r1 ,δ2=cult-soc r2 (27) new-soc c =s o c c +s1*δ1+s2*δ2(28) Among them, r1 and r2 represent two different random coyote numbers, L best represents the best alpha wolf in the group; s1 and s2 are random weights of δ1 and δ2 respectively; s1 and s2 are random numbers uniformly distributed in [0,1]; After each coyote in the group grows up, its social adaptability is calculated and greedy selection is adopted, as shown in the following formula (29). By retaining high-quality coyotes to participate in the growth of other coyotes in the group, the convergence speed of the algorithm is accelerated: (3) Life and Death of Coyotes: In COA, the age of coyotes is measured in years, and after each group of coyotes grows up, a newborn coyote is produced; The birth of new coyotes is influenced by the social conditions and social environment of two randomly selected parent coyotes; the way new coyotes are produced is shown in the following formula (30): where cr1 and cr2 are two random different coyotes in group p, j1 and j2 are two random dimensions of the newborn coyotes; P s is the probability of dispersion, P a is the association probability, as shown below (31); the magnitude of the dispersion and association probability affects the diversity of newborn coyotes; R j is a random number within the value range of the j-th dimension of the decision variable; rnd j is a random number uniformly distributed in [0,1]: where ω represents a coyote pack that is less adaptable than newborn wolves, is the number of coyotes in the group, and the age of newborn coyotes is 0; (4) Coyotes are driven away and accepted Initially, coyotes were randomly assigned to groups, but sometimes they left and joined other groups; the probability of a coyote being excluded or accepted by a group is expressed as P. e This mechanism helps the exchange of information between COA groups and promotes mutual influence between coyotes in the population, as shown in the following formula (32); After initialization and random grouping, the three steps of coyote growth, coyote life and death, and coyote expulsion and acceptance are carried out in sequence; if the iteration termination condition is met, the optimal coyote is output; otherwise, jump to step (2); Finally, the optimal size of the hybrid power system, i.e. the optimal size of the PV cells / inverter, is obtained through the Coyote optimization algorithm.

3. An optimization device for a grid-connected photovoltaic battery system, characterized in that: The device comprises: The first building module is used to build an output power model of a grid-connected photovoltaic battery system; A second establishing module is used to establish an economic model of the grid-connected photovoltaic battery system according to the output power model of the grid-connected photovoltaic battery system; A third establishing module is used to establish an economic scale optimization model of the grid-connected photovoltaic battery system according to the economic model of the grid-connected photovoltaic battery system; An optimization module, used for optimizing the economic scale optimization model of the grid-connected photovoltaic battery system based on a coyote optimization algorithm; The first establishing module is further specifically used for: The output power of the photovoltaic cell is obtained by using an accurate model of a single diode photovoltaic cell with a series resistor; the following formula (1) shows the maximum power generation of the photovoltaic cell: P max.c (t)=V OC.c (t).I SC.c (t).FF(t) (1) Among them, P max.c (t) is the maximum power generation of the photovoltaic cell, I SC.c (t) is the short-circuit current of the photovoltaic cell, V OC.c (t) is the open circuit voltage of the photovoltaic cell; I SC.c (t) and V OC.c The calculation formula of (t) is as follows: Formula (2) and Formula (3): V OC.c (t)=V OC.c.stc +K v (T C (t)-25) (3) Among them, I SC.c.stc is the nominal short-circuit current of the photovoltaic cell, V OC.c.stc is the nominal open circuit voltage of the photovoltaic cell, G(t) is the global solar radiation intensity at time t, K i and K v are the temperature coefficients of current and voltage respectively; T C (t) is the photovoltaic cell temperature at time t, calculated according to the following formula (4): Among them, T a (t) is the ambient temperature at time t, NOCT is the rated operating temperature of the photovoltaic cell; FF(t) is the fill factor of the photovoltaic cell at time t, which is calculated by the following formula (5): Among them, r s (t) and v oc (t) is the normalized series resistance and open circuit voltage of the photovoltaic cell, which are calculated by the following equations (6) and (7), respectively: Among them, FF stc is the nominal value of the fill factor, FF 0stc is the nominal value of the fill factor at time zero; In the above formula, n is the inertia coefficient of the diode, K B is the Boltzmann constant, T a.k (t) is the ambient temperature at time t; In addition, formulas (9), (10), (11), (12), and (13) show the steps for obtaining the output power of the photovoltaic array: It should be noted that generally available data is for PV modules and not for PV cells; therefore: I SC.c.stc (t)=I SC.m.stc (t) (12) P av.PV (t)=N S.M ·N P.M N C.M P max.C (13) Among them, Pma x .c. stc (t) is the nominal maximum power of the photovoltaic cell, Pma x .m. stc (t) is the nominal maximum power of the photovoltaic module, V OC .m. stc (t) is the nominal open circuit voltage of the photovoltaic module, I SC .m. stc (t) is the nominal short-circuit current of the photovoltaic module, Pma x.C is the maximum power of the photovoltaic cell, N C.M is the number of photovoltaic cells in the photovoltaic module, N S.M and N P.M is the number of PV modules connected in series and in parallel, Pa v.PV (t) is the output power of the entire photovoltaic array; Battery packs are widely used as renewable energy generators. Photovoltaic power is stored in the battery pack. If the battery pack is fully charged, the photovoltaic power is directly fed into the residential part. In the case of insufficient power, the battery discharges to supply the demand. The power of the battery pack in the charging state is obtained by the following formula (14): Among them, P BAT (t) is the battery charging state at time t, P BAT (t-1) represents the charging state at time t-1; σ is the self-discharge rate of the battery, η INV and η BAT are the inverter and battery charging efficiencies, which are considered to be 0.95 and 0.9, respectively; P Load (t) is the load demand, N PV is the number of photovoltaic panels; If the photovoltaic power generation is less than the load consumption, the battery starts to discharge; the discharge efficiency is equal to the charging efficiency; P B AT is expressed by the following formula (15) in the discharge state: The second establishing module is further specifically used for: The NPC of a system is the present value of all costs during the life cycle of the system minus the present value of all revenues during its life cycle. Since the reliability coefficient must be converted into cost, the power supply loss LPS is multiplied by an arbitrary reliability penalty coefficient RPF to form a single target cost function. The formula for reliability cost CRE is as follows (16): CRE=LPS*RPF (16) The calculation formula of emission cost CE is as follows (17): CE==(Ccd*Pcd+Ccm*Pcm+Cuh*Puh+Cpm*Ppm+Csd*Psd+Cno*Pno) / 10 6 (17) Among them, C i is the cost, P i is the pollution produced by different pollutants, i = {cd, cm, uh, pm, sd, no}, referring to different pollutants; The net present value cost NPC of the system will be obtained from the total cost of components, replacement, operation and maintenance, reliability and emission impact as follows (18): NPC=CT+CE+CRE (18) Where CT is the total cost of the system; its value is calculated from the following formula (19): Among them, CI is the initial cost, CR is the replacement cost, CM is the operation and maintenance cost, C PV is the cost of photovoltaic panels, N PV is the number of photovoltaic panels, C BAT is the battery unit cost, N BAT is the number of battery cells; CPV is the cumulative present value, C PV,R is the replacement cost of photovoltaic, C BAT,R is the replacement cost of the battery, C PV,M is the maintenance cost of photovoltaic power generation, C BAT,M is the maintenance cost of the battery; the cost of purchasing electricity from the grid is considered to be C Grid ; f is the annual inflation rate (%), i is the annual interest rate (%); In order to obtain the COE parameter, the total annualized cost must be calculated first, as shown in the following formula (20): C A =NPC*CRF (i,LT) (20) Among them, C A is the annualized total cost (US dollars / year), CRF (i,LT) is the capital recovery coefficient, and its calculation formula is as follows (21): COE is the average cost per kilowatt-hour of effective electricity. It is one of the most important parameters for comparing the cost-effectiveness of hybrid energy systems and finding the optimal scale of the system. The calculation formula is as follows (22): Among them, E served The energy provided in a year; The third establishing module is further specifically used for: The objective function of the optimization problem is to minimize NPC and COE under two different conditions; the objective function is defined as follows (23): There are some constraints in the optimization process; these constraints are as follows (24): Among them, PV max is the maximum number of photovoltaic panels, Battery max is the maximum available storage capacity; and is the minimum and maximum charge capacity of the battery, DOD is the depth of discharge of the battery, S Bat is the rated capacity of the battery.

4. The device according to claim 3, characterized in that The optimization module is further specifically used for: (1) Initialization and random grouping: Set parameters such as the number of coyote groups N p 、N number of coyotes in the group c and the maximum number of iterations N gen ,Since COA is a random algorithm, the initial social state factors of each coyote are randomly set as follows (25); the social adaptability of each coyote is calculated and randomly grouped; soc j =lb j +r j *(ub j -lb j ) (25) Among them, lb j andub j denote the lower and upper bounds of the j-th state factor of the coyote, j = 1, 2, ..., D, r j is a random number uniformly distributed within [0,1]; (2) Growth of coyotes within the group: Determine the optimal wolf alpha in the group, calculate the group cultural trend, and randomly select two coyotes. The four factors affect the growth of coyotes. The calculation of the group cultural trend is as follows (26): cult j =median(A j ) (26) Where A is an N c A matrix with D rows and D columns, representing N c A solution vector, A j represents the j-th column of matrix A, and median represents the median; In the process of coyote growth, the difference δ1 between the optimal coyote alpha in the group and a randomly selected coyote in the group and the difference δ2 between the group cultural trend and another random coyote in the group are calculated first, and then the coyotes in the group grow under the influence of δ1 and δ2, as shown in the following equations (27) and (28): δ1=Lbest-soc r1 ,δ2=cult-soc r2 (27) new-soc c =s o c c +s1*δ1+s2*δ2 (28) Among them, r1 and r2 represent two different random coyote numbers, L best represents the best alpha wolf in the group; s1 and s2 are random weights of δ1 and δ2 respectively; s1 and s2 are random numbers uniformly distributed in [0,1]; After each coyote in the group grows up, its social adaptability is calculated and greedy selection is adopted, as shown in the following formula (29). By retaining high-quality coyotes to participate in the growth of other coyotes in the group, the convergence speed of the algorithm is accelerated: (3) Life and Death of Coyotes: In COA, the age of coyotes is measured in years, and after each group of coyotes grows up, a newborn coyote is produced; The birth of new coyotes is influenced by the social conditions and social environment of two randomly selected parent coyotes; the way new coyotes are produced is shown in the following formula (30): where cr1 and cr2 are two random different coyotes in group p, j1 and j2 are two random dimensions of the newborn coyotes; P s is the probability of dispersion, P a is the association probability, as shown below (31); the magnitude of the dispersion and association probability affects the diversity of newborn coyotes; R j is a random number within the value range of the j-th dimension of the decision variable; rnd j is a random number uniformly distributed in [0,1]: where ω represents a coyote pack that is less adaptable than newborn wolves, is the number of coyotes in the group, and the age of newborn coyotes is 0; (4) Coyotes are driven away and accepted Initially, coyotes were randomly assigned to groups, but sometimes they left and joined other groups; the probability of a coyote being excluded or accepted by a group is expressed as P. e This mechanism helps the exchange of information between COA groups and promotes mutual influence between coyotes in the population, as shown in the following formula (32); After initialization and random grouping, the three steps of coyote growth, coyote life and death, and coyote expulsion and acceptance are carried out in sequence; if the iteration termination condition is met, the optimal coyote is output; otherwise, jump to step (2); Finally, the optimal size of the hybrid power system, i.e. the optimal size of the PV cells / inverter, is obtained through the Coyote optimization algorithm.

Images