A real-time efficiency optimal energy management method for battery energy storage system cluster
Through real-time load demand management and energy consumption characteristic modeling, the power distribution of the lithium-ion battery energy storage system cluster is optimized, and the energy consumption inconsistency caused by performance differences between energy storage systems is solved, and the energy consumption efficiency and service life of the cluster is improved.
Patent Information
- Application Number
- CN202410495914.3
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2024-04-24
- Publication Date
- 2024-12-03
- Estimated Expiration
- 2044-04-24
AI Technical Summary
Lithium-ion batteries have inconsistent energy consumption efficiency and service life due to performance differences in energy storage system clusters, which affects the economy and durability of the energy storage system clusters.
A method for optimum energy management of real-time efficiency of battery energy storage system clusters is proposed. Through system control parameter initialization and energy consumption characteristic modeling, power distribution is performed based on real-time load requirements, and the Taylor series expansion and cascade converter characteristic curve is constructed. The energy consumption characteristic model performs local secondary order reduction and iterative optimization to ensure that the relative error of power distribution between energy storage systems is within a reasonable range.
It improves the real-time efficiency of the energy storage system cluster, balances the consistency of SOCs between energy storage systems, and improves energy consumption economy and overall system performance.
Smart Images

Figure CN118589552B_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the field of energy storage system performance optimization control, and in particular to a real-time efficiency optimal energy management method for a battery energy storage system cluster. Background Art
[0002] With the continuous development of the economy, electricity consumption is also increasing year by year. Relying solely on traditional fossil fuels for power generation will significantly increase the power generation cost of the power system. It is urgent to get rid of its dependence on traditional fossil fuels. Wind power generation and photovoltaic power generation have been unanimously recognized as the medium for the power system to achieve energy structure transformation. However, technologies such as wind power generation and photovoltaic power generation are affected by factors such as climate conditions, surrounding environment and geographical location. Their energy conversion process is volatile and uncertain, which brings certain challenges to the continuous and stable operation of the power grid, thus restricting the large-scale grid-connected operation of such technologies. Energy storage technology can effectively solve the mismatch between new energy power generation and the power grid in time and space, thereby effectively improving the stability and continuity of power grid operation.
[0003] As a representative of energy storage batteries, lithium-ion batteries are favored by the energy storage field due to their excellent dynamic response speed when facing the rapidly changing power demand in actual power grid conditions, as well as longer cycle life and higher energy density compared to other types of energy storage batteries, thus driving the continuous increase in demand for lithium-ion batteries. However, the energy consumption characteristics of lithium-ion batteries are easily affected by many internal and external factors, such as temperature, fluctuations in charging and discharging conditions, efficiency of cascade converters, and inconsistency in charge state between battery packs. Therefore, how to improve the energy efficiency of lithium batteries as much as possible while controlling interference factors has become one of the research focuses in the field of energy storage power station systems.
[0004] In the field of energy storage technology, large-scale energy storage power station systems are composed of multiple energy storage systems, each of which is composed of hundreds to thousands of batteries. This results in extremely large charging and discharging powers of energy storage system clusters. The extremely large charging and discharging powers make the energy storage system clusters have high requirements for the consistency of performance between energy storage systems. However, due to the differences in performance of the single batteries that make up each energy storage system, there are bound to be inconsistencies between energy storage systems, which has a huge impact on the energy efficiency and service life of the energy storage system cluster. Therefore, formulating a reasonable energy optimization management strategy is of great significance to improving the inconsistency of performance between energy storage systems and thus improving the economy and durability of energy storage system clusters. Summary of the invention
[0005] In order to solve the above problems, the present invention proposes a real-time efficiency optimal energy management method for a battery energy storage system cluster based on the changes in load demand during the actual operation of the power grid, with the goal of optimizing real-time efficiency. The method optimizes the control and management of the energy storage system cluster, has a reasonable design, solves the shortcomings of the existing technology, and has good results.
[0006] In order to achieve the above object, the present invention adopts the following technical scheme:
[0007] A method for real-time efficiency optimal energy management of a battery energy storage system cluster comprises the following steps:
[0008] S1. System control parameter initialization and energy consumption characteristic modeling, including initialization of energy storage system cluster power allocation results, update of energy storage system cluster real-time power limit sequence, and modeling of energy storage battery energy consumption characteristic model and cascade converter characteristic curve;
[0009] S2. Power allocation of the system is performed based on real-time load demand, which is divided into two cases: charging power allocation and discharging power allocation. The specific steps include local quadratic reduction of the energy consumption characteristic model of the energy storage system, calculation of the power allocation coefficient matrix and power switching sequence of the energy storage system cluster, and power allocation according to the size of the load demand at different times;
[0010] S3. Optimizing and approximating the real-time power allocation structure of the energy storage system cluster based on iteration. The specific steps include calculating the relative error of the power allocation result and determining whether it is necessary to iterate the power allocation result.
[0011] Furthermore, the step S1 includes the following steps:
[0012] S101. For initialization of the power allocation result between energy storage system clusters, an average power allocation method is used to divide the load demand in a single operating cycle by the number of energy storage systems in the energy storage system cluster, and the load demand is evenly allocated to each energy storage system as the initial power allocation result of the energy storage system cluster, as shown in formula (1):
[0013]
[0014] Among them, P d,t represents the load demand of the energy storage system cluster at time t, Indicates the load power that each energy storage system should bear for the load demand at time t;
[0015] S102: According to the SOC and RUL of each energy storage system at time t, the power limit sequence of the energy storage system cluster is updated, as shown in formulas (2) and (3):
[0016]
[0017]
[0018] in, and They represent the maximum power limit sequence and the minimum power limit sequence of the energy storage system cluster at time t respectively; Q t , and They represent the open circuit voltage, available capacity, and equivalent internal resistance of the energy storage system cluster at time t, which are related to the current SOC and RUL of each energy storage system; CR max Represents the maximum charge and discharge rate of the energy storage system cluster, which is a constant;
[0019] S103. First, an energy storage battery equivalent model is established based on the Rint equivalent circuit, and the energy consumption characteristic expression of the relationship between the energy storage battery output power and the net output power is obtained through Taylor series expansion, as shown in formula (4):
[0020]
[0021] Among them, P gross P is the output power of the energy storage battery without considering the loss; bat is the net output power of the energy storage battery after deducting the loss, which is also the input power of the cascade converter; c α The equivalent model of the energy storage battery is represented by P bat =0 is the αth coefficient of the Mth Taylor expansion of the Taylor expansion point, which is related to the current SOC, RUL and charge and discharge state of the energy storage battery; the specific formula is as follows:
[0022]
[0023] Secondly, the Buck-Boost bidirectional converter is selected as the cascade converter of the energy storage system, and according to the characteristic curve of the cascade converter, the high-order nonlinear fitting expression of the input power and output power of the cascade converter is obtained, as shown in formula (6):
[0024]
[0025] Among them, d β The βth coefficient of the Nth order fitting polynomial of the energy storage battery converter model is related to the current SOC and charge and discharge state of the energy storage battery; P net is the net output power of the cascaded converter;
[0026] Finally, the expression of the energy storage battery and the cascade converter is combined to form the expression describing the output power P of a single energy storage system. gross With the net output power Pnet The energy consumption characteristic model of the relationship between is shown in formula (7):
[0027]
[0028] Among them, e γ Represents the γth coefficient of the M*Nth energy consumption characteristic function expression of a single energy storage system; e γ With formula (4) c α , d in formula (6) β The relationship is as follows:
[0029]
[0030] Furthermore, the step S2 comprises the following steps:
[0031] S201. Based on the energy consumption characteristic model, local quadratic reduction is performed to construct approximate local quadratic energy consumption charging and discharging functions of a single energy storage system at time t, as shown in formula (9):
[0032]
[0033] in, and are the quadratic term coefficient, linear term coefficient and constant term coefficient of the discharge energy consumption function model of the i-th energy storage system related to SOC at time t; and are the quadratic term coefficient, linear term coefficient and constant term coefficient of the charging energy consumption function model of the i-th energy storage system related to SOC at time t;
[0034] The above coefficients will vary according to the Taylor expansion points selected at different times, which can be divided into the following two cases:
[0035] Case 1: If it is at the initial time t = 1, the charging and discharging model takes the initialization power average distribution result at time t = 1 as the Taylor expansion point, and adopts different energy consumption functions according to the positive and negative power demand at this time. At this time, the expressions of the coefficients are shown in formulas (10) and (11):
[0036]
[0037]
[0038] Case 2: If it is not at the initial time t=1, the power allocation result at time t-1 is used as the Taylor expansion point of the charge-discharge model, and different energy consumption functions are used according to the positive or negative power demand at that time. At this time, the expressions of the coefficients are shown in formulas (12) and (13):
[0039]
[0040]
[0041] Among them, f, f' and f" are the single energy storage system P during discharge. gross With P net The relationship between the function, first-order derivative and second-order derivative; g, g' and g" are the single energy storage system P during charging gross With P net The relationship between functions, first-order derivatives and second-order derivatives; is the average distribution result of the initialization power at time t=1, as shown in formula (14):
[0042]
[0043] S202. Calculate the power allocation coefficient matrix within the energy storage system cluster, as shown in formula (15):
[0044]
[0045] Among them, K DDS and L DDS They represent the linear coefficient and constant coefficient matrix of the energy storage system cluster power allocation related to the power upper limit respectively; K DES and L DES They represent the linear coefficient and constant coefficient matrix of the energy storage system cluster power allocation related to the power lower limit respectively; k ρ,s and l ρ,s They respectively represent the linear coefficient and constant coefficient of the ρth energy storage system involved in the power allocation expression of the energy storage system cluster; the expressions of each coefficient are shown in formula (16):
[0046]
[0047] Among them, s represents the number of energy storage systems involved in regulation; m represents the number of single energy storage systems in the energy storage system cluster; ω ρ A represents the penalty coefficient for adjusting the SOC consistency of each energy storage system involved in the regulation; ρ and q ρ They respectively represent the reciprocal of the quadratic coefficient and the linear coefficient of the quadratic function form of the charging and discharging equivalent model of each single energy storage system involved in the regulation; ω ρ , A ρ and q ρ The expressions of are shown in equations (17) and (18):
[0048]
[0049]
[0050] Where μ represents the constant for adjusting the consistency speed between energy storage systems; SOC HL and SOC LL They represent the upper and lower limits of the state of charge set by the energy storage system cluster; SOC ρ represents the state of charge of the ρth energy storage system involved in regulation;
[0051] Calculate the power switching sequence of the energy storage system cluster. First, determine the order in which each energy storage system reaches the maximum and minimum power limits according to the performance of each energy storage system, as shown in formula (19):
[0052]
[0053] Then, the calculated m and m Arrange in ascending and descending order respectively to obtain the maximum power limit sequence of each single energy storage system and minimum power limit sequence The relationship between h and j is h = m-j+1;
[0054] Finally, according to the obtained maximum power limit sequence and minimum power limit sequence The maximum power limit of the corresponding energy storage system cluster and minimum power limit After sorting, the load demand of the system cluster corresponding to each single energy storage system reaching the maximum power limit and dropping to the minimum power limit is calculated, that is, the power limit power switching point, as shown in formula (20):
[0055]
[0056] Get the maximum power limit sequence and minimum power limit sequence
[0057] S203. According to the current load demand, the power distribution of the energy storage system cluster is divided into the following five cases:
[0058] In the first case, when the load demand of the energy storage system cluster At this time, all single energy storage systems have not escaped from their own minimum output power constraints and do not participate in power regulation. The net output power of each single energy storage system at this moment is shown in formula (21):
[0059]
[0060] In the second case, when the load demand of the energy storage system cluster At this time, some energy storage systems still have not escaped their own minimum output power limit, and these energy storage systems do not participate in power regulation. At this moment, the power allocation of the energy storage system cluster satisfies the relationship shown in formula (22):
[0061]
[0062] Among them, K h' and L h' They represent the power allocation coefficient matrix of the energy storage system involved in the regulation with respect to the load demand at that moment; the specific forms are shown in formulas (23) and (24):
[0063] K h' =[k 1,h' k 2,h' … k h'-1,h' k h',h' ] (twenty three)
[0064] L h' =[l 1,h' l 2,h' … l h'-1,h' l h',h' ] (twenty four)
[0065] At this time, the power demand allocated to each set of power batteries participating in the regulation is shown in formula (25):
[0066] P net,i' =k i',h' P d +l i',h' ,i=1,2,…,h (25)
[0067] In the third case, when the load demand of the energy storage system cluster At this time, all single energy storage systems are out of their own minimum output power limit and have not reached their own maximum output power limit, and all participate in power regulation. At this moment, the power allocation of the energy storage system cluster satisfies the relationship shown in formula (26):
[0068]
[0069] Among them, K m and L m They represent the power allocation coefficient matrix of the energy storage system involved in the regulation with respect to the load demand at that moment; the specific forms are shown in formulas (27) and (28):
[0070] K m =[k 1,m k 2,m … k m-1,m km,m ] (27)
[0071] L m =[l 1,m l 2,m … l m-1,m l m,m ] (28)
[0072] At this time, the power demand allocated to each set of power batteries participating in the regulation is shown in formula (29):
[0073] P net,i =k i,m P d +l i,m ,i=1,2,…,m (29)
[0074] In the fourth case, when the load demand of the energy storage system cluster At this time, only part of the energy storage system reaches its maximum output power limit, and this part of the energy storage system does not participate in power regulation. At this moment, the power allocation of the energy storage system cluster satisfies the relationship shown in formula (30):
[0075]
[0076] Among them, K j' and L j' They respectively represent the power allocation coefficient matrix of the energy storage system involved in the regulation with respect to the load demand at that moment; the specific forms are shown in formulas (31) and (32):
[0077] K j' =[k j',m k j'+1,m … k m-1,m k m,m ] (31)
[0078] L j '=[l j',m l j'+1,m … l m-1,m l m,m ] (32)
[0079] At this time, the power demand allocated to each set of power batteries participating in the regulation is shown in formula (33):
[0080] P net,i” =k i”,m P d +l i”,m ,i=j,j+1,…,m (33)
[0081] The fifth case is when the load demand of the energy storage system cluster At this time, all single energy storage systems have reached their maximum output power limits and do not participate in power regulation. The net output power of each single energy storage system at this moment is shown in formula (34):
[0082]
[0083] Furthermore, step S3 includes the following steps:
[0084] S301. Depending on whether time t is the initial time, there are two cases:
[0085] Case 1: If time t is the initial time, the relative error between the updated power allocation result at time t and the average power allocation result at time t is calculated, as shown in formula (35):
[0086]
[0087] in, Represents the relative error obtained by dividing the difference between the average power allocation result of the i-th energy storage system at time t and the updated power allocation result at time t by the absolute value of the average power allocation result at time t;
[0088] Case 2: If time t is not the initial time, the relative error between the updated power allocation result at time t and the final power allocation result at time t-1 is calculated, as shown in formula (36):
[0089]
[0090] in, Relative error is the difference between the final power allocation result of the i-th energy storage system at time t-1 and the updated power allocation result at time t divided by the final power allocation result at time t-1;
[0091] S302, determining whether the power allocation result updated at time t meets the set conditions;
[0092] If the current number of iterations n does not reach the set upper limit n max And the relative error of the energy storage system cluster power allocation in this iteration is There is a relative error ε of some energy storage systems in the sequence t If the relative error ε0 of each energy storage system is greater than the set value ε0 (ε0 = 0.1%), the updated power allocation result at time t replaces the final power allocation result at time t-1 or the average power allocation result at time t, and returns to step S201 to re-iterate until the relative error ε0 of each energy storage system is less than 0.1%. t are not greater than ε0 or the number of iterations at this moment has reached the upper limit of the number of iterations n max ;
[0093] If the current number of iterations n does not reach the upper limit of the number of iterations n max , the relative error ε of each energy storage system t Already satisfied ε t ≤ε0, the updated power allocation result obtained in this iteration is used as the final power allocation result at time t; if the current number of iterations n has reached the upper limit of the number of iterations n max , then the updated power allocation result obtained in the last iteration is taken as the final power allocation result at time t;
[0094] Finally, if the energy storage system cluster stops running, the output power of all energy storage systems is zero, that is, P=0.
[0095] Beneficial technical effects brought by the present invention:
[0096] Compared with the existing energy management strategies for energy storage system clusters, the present invention comprehensively considers the impact of performance differences between energy storage systems on the energy consumption characteristics of the energy storage system cluster, allocates real-time grid load demands to the energy storage system cluster in a more reasonable manner, optimizes the real-time efficiency of the energy storage system cluster, and balances the consistency of SOC between the energy storage systems, thereby achieving improved energy consumption economy of the energy storage system cluster. BRIEF DESCRIPTION OF THE DRAWINGS
[0097] Figure 1 An overall flow chart of the real-time efficiency optimal energy management method for a battery energy storage system cluster provided by the present invention;
[0098] Figure 2 A detailed flow chart of the initialization process provided by the present invention;
[0099] Figure 3 A detailed flow chart of the power distribution of the system based on real-time load demand provided by the present invention;
[0100] Figure 4 A detailed flow chart of power allocation according to the comparison between power demand and power switching point provided by the present invention;
[0101] Figure 5 A detailed flow chart of optimizing and approximating the real-time power allocation results of the energy storage system cluster based on iteration provided by the present invention;
[0102] Figure 6 A graph showing the distribution power changing with the total power demand of the system during the discharge process provided in Example 1;
[0103] Figure 7 A graph showing the distribution power during charging as a function of the total system power demand provided in Example 1; DETAILED DESCRIPTION
[0104] The specific implementation of the present invention is further described below in conjunction with the accompanying drawings and specific embodiments:
[0105] A real-time efficiency optimal energy management method for a battery energy storage system cluster, such as Figure 1 As shown, the following steps are included:
[0106] S1. System control parameter initialization and energy consumption characteristic modeling, including initialization of energy storage system cluster power allocation results, update of energy storage system cluster real-time power limit sequence, and modeling of energy storage battery energy consumption characteristic model and cascade converter characteristic curve;
[0107] like Figure 2 As shown, step S1 includes the following steps:
[0108] S101. For initialization of the power allocation result between energy storage system clusters, an average power allocation method is used to divide the load demand in a single operating cycle by the number of energy storage systems in the energy storage system cluster, and the load demand is evenly allocated to each energy storage system as the initial power allocation result of the energy storage system cluster, as shown in formula (1):
[0109]
[0110] Among them, P d,t represents the load demand of the energy storage system cluster at time t, Indicates the load power that each energy storage system should bear for the load demand at time t;
[0111] S102: According to the SOC and RUL of each energy storage system at time t, the power limit sequence of the energy storage system cluster is updated, as shown in formulas (2) and (3):
[0112]
[0113]
[0114] in, and They represent the maximum power limit sequence and the minimum power limit sequence of the energy storage system cluster at time t respectively; Q t , and They represent the open circuit voltage, available capacity, and equivalent internal resistance of the energy storage system cluster at time t, which are related to the current SOC and RUL of each energy storage system; CR max Represents the maximum charge and discharge rate of the energy storage system cluster, which is a constant.
[0115] S103. First, an energy storage battery equivalent model is established based on the Rint equivalent circuit, and the energy consumption characteristic expression of the relationship between the energy storage battery output power and the net output power is obtained through Taylor series expansion, as shown in formula (4):
[0116]
[0117] Among them, P gross P is the output power of the energy storage battery without considering the loss; bat is the net output power of the energy storage battery after deducting the loss, which is also the input power of the cascade converter; c α The equivalent model of the energy storage battery is represented by P bat =0 is the αth coefficient of the Mth Taylor expansion of the Taylor expansion point, which is related to the current SOC, RUL and charge and discharge state of the energy storage battery; the specific formula is as follows:
[0118]
[0119] Secondly, the Buck-Boost bidirectional converter is selected as the cascade converter of the energy storage system, and according to the characteristic curve of the cascade converter, the high-order nonlinear fitting expression of the input power and output power of the cascade converter is obtained, as shown in formula (6):
[0120]
[0121] Among them, d β The βth coefficient of the Nth order fitting polynomial of the energy storage battery converter model is related to the current SOC and charge and discharge state of the energy storage battery; P net is the net output power of the cascaded converter;
[0122] Finally, the expression of the energy storage battery and the cascade converter is combined to form the expression describing the output power P of a single energy storage system. gross With the net output power P net The energy consumption characteristic model of the relationship between is shown in formula (7):
[0123]
[0124] Among them, e γ Represents the γth coefficient of the M*Nth energy consumption characteristic function expression of a single energy storage system; e γ With formula (4) c α , d in formula (6) β The relationship is as follows:
[0125]
[0126] S2. Power allocation of the system is performed based on real-time load demand, which is divided into two cases: charging power allocation and discharging power allocation. The specific steps include local quadratic reduction of the energy consumption characteristic function of the energy storage system, calculation of the power allocation coefficient matrix and power switching sequence of the energy storage system cluster, and power allocation according to the size of the load demand at different times;
[0127] like Figure 3 As shown, step S2 includes the following steps:
[0128] S201. Since the energy consumption characteristic model of the single energy storage system constructed in step S1 is complex in structure and has the characteristics of high nonlinearity and strong coupling, local quadratic reduction is performed on the basis of the original model to construct the approximate local quadratic energy consumption charging and discharging functions of the single energy storage system at time t, as shown in formula (9):
[0129]
[0130] in, and are the quadratic term coefficient, linear term coefficient and constant term coefficient of the discharge energy consumption function model of the i-th energy storage system related to SOC at time t; and are the quadratic term coefficient, linear term coefficient and constant term coefficient of the charging energy consumption function model of the i-th energy storage system related to SOC at time t;
[0131] The above coefficients will vary according to the Taylor expansion points selected at different times, which can be divided into the following two cases:
[0132] Case 1: If it is at the initial time t = 1, the charging and discharging model takes the initialization power average distribution result at time t = 1 as the Taylor expansion point, and adopts different energy consumption functions according to the positive and negative power demand at this time. At this time, the expressions of the coefficients are shown in formulas (10) and (11):
[0133]
[0134]
[0135] Case 2: If it is not at the initial time t=1, the power allocation result at time t-1 is used as the Taylor expansion point of the charge-discharge model, and different energy consumption functions are used according to the positive or negative power demand at that time. At this time, the expressions of the coefficients are shown in formulas (12) and (13):
[0136]
[0137]
[0138] Among them, f, f' and f" are the single energy storage system P during discharge. gross With P net The relationship between the function, first-order derivative and second-order derivative; g, g' and g" are the single energy storage system P during charging gross With P net The relationship between functions, first-order derivatives and second-order derivatives; is the average distribution result of the initialization power at time t=1, as shown in formula (14):
[0139]
[0140] S202. Calculate the power allocation coefficient matrix within the energy storage system cluster, as shown in formula (15):
[0141]
[0142] Among them, K DDS and L DDS They represent the linear coefficient and constant coefficient matrix of the energy storage system cluster power allocation related to the power upper limit respectively; K DES and L DES They represent the linear coefficient and constant coefficient matrix of the energy storage system cluster power allocation related to the power lower limit respectively; k ρ,s and l ρ,s They respectively represent the linear coefficient and constant coefficient of the ρth energy storage system involved in the power allocation expression of the energy storage system cluster; the expressions of each coefficient are shown in formula (16):
[0143]
[0144] Among them, s represents the number of energy storage systems involved in regulation; m represents the number of single energy storage systems in the energy storage system cluster; ω ρ A represents the penalty coefficient for adjusting the SOC consistency of each energy storage system involved in the regulation; ρ and q ρ They respectively represent the reciprocal of the quadratic coefficient and the linear coefficient of the quadratic function form of the charging and discharging equivalent model of each single energy storage system involved in the regulation; ω ρ , A ρ and q ρ The expressions of are shown in equations (17) and (18):
[0145]
[0146]
[0147] Where μ represents the constant for adjusting the consistency speed between energy storage systems; SOC HL and SOC LLThey represent the upper and lower limits of the state of charge set by the energy storage system cluster; SOC ρ represents the state of charge of the ρth energy storage system involved in regulation;
[0148] Calculate the power switching sequence of the energy storage system cluster. First, determine the order in which each energy storage system reaches the maximum and minimum power limits according to the performance of each energy storage system, as shown in formula (19):
[0149]
[0150] Then, the calculated m and m Arrange in ascending and descending order respectively to obtain the maximum power limit sequence of each single energy storage system and minimum power limit sequence The relationship between h and j is h = m-j+1;
[0151] Finally, according to the obtained maximum power limit sequence and minimum power limit sequence The maximum power limit of the corresponding energy storage system cluster and minimum power limit After sorting, the load demand of the system cluster corresponding to each single energy storage system reaching the maximum power limit and dropping to the minimum power limit is calculated, that is, the power limit power switching point, as shown in formula (20):
[0152]
[0153] Get the maximum power limit sequence and minimum power limit sequence
[0154] S203. According to the current load demand, the power distribution of the energy storage system cluster is divided into the following five cases:
[0155] In the first case, when the load demand of the energy storage system cluster At this time, all single energy storage systems have not escaped from their own minimum output power constraints and do not participate in power regulation. The net output power of each single energy storage system at this moment is shown in formula (21):
[0156]
[0157] In the second case, when the load demand of the energy storage system cluster At this time, some energy storage systems still have not escaped their own minimum output power limit, and these energy storage systems do not participate in power regulation. At this moment, the power allocation of the energy storage system cluster satisfies the relationship shown in formula (22):
[0158]
[0159] Among them, K h' and L h' They represent the power allocation coefficient matrix of the energy storage system involved in the regulation with respect to the load demand at that moment; the specific forms are shown in formulas (23) and (24):
[0160] K h' =[k 1,h' k 2,h' … k h'-1,h' k h',h' ] (twenty three)
[0161] L h' =[l 1,h' l 2,h' … l h'-1,h' l h',h' ] (twenty four)
[0162] At this time, the power demand allocated to each set of power batteries participating in the regulation is shown in formula (25):
[0163] P net,i' =k i',h' P d +l i',h' ,i=1,2,…,h (25)
[0164] In the third case, when the load demand of the energy storage system cluster At this time, all single energy storage systems are out of their own minimum output power limit and have not reached their own maximum output power limit, and all participate in power regulation. At this moment, the power allocation of the energy storage system cluster satisfies the relationship shown in formula (26):
[0165]
[0166] Among them, K m and L m They represent the power allocation coefficient matrix of the energy storage system involved in the regulation with respect to the load demand at that moment; the specific forms are shown in formulas (27) and (28):
[0167] K m =[k 1,m k 2,m … k m-1,m k m,m ] (27)
[0168] L m =[l 1,m l 2,m … l m-1,m l m,m ] (28)
[0169] At this time, the power demand allocated to each set of power batteries participating in the regulation is shown in formula (29):
[0170] P net,i =k i,m P d +l i,m ,i=1,2,…,m (29)
[0171] In the fourth case, when the load demand of the energy storage system cluster At this time, only part of the energy storage system reaches its maximum output power limit, and this part of the energy storage system does not participate in power regulation. At this moment, the power allocation of the energy storage system cluster satisfies the relationship shown in formula (30):
[0172]
[0173] Among them, K j' and L j' They respectively represent the power allocation coefficient matrix of the energy storage system involved in the regulation with respect to the load demand at that moment; the specific forms are shown in formulas (31) and (32):
[0174] K j' =[k j',m k j'+1,m … k m-1,m k m,m ] (31)
[0175] L j' =[l j',m l j'+1,m … l m-1,m l m,m ] (32)
[0176] At this time, the power demand allocated to each set of power batteries participating in the regulation is shown in formula (33):
[0177] P net,i” =k i”,m P d +l i”,m ,i=j,j+1,…,m (33)
[0178] The fifth case is when the load demand of the energy storage system cluster At this time, all single energy storage systems have reached their maximum output power limits and do not participate in power regulation. The net output power of each single energy storage system at this moment is shown in formula (34):
[0179]
[0180] Figure 4 A detailed flow chart of power allocation based on the comparison between power demand and power switching point is shown. The real-time power demand is compared with the power switching point, i.e., the formula in the diamond on the left side of the figure, and then the interval in which it is located is determined, and then the power allocation formula is determined, i.e., the formula in the corresponding rectangle on the right side of the figure, so as to obtain the power allocation result at that moment.
[0181] S3. Optimizing and approximating the real-time power allocation result of the energy storage system cluster based on iteration, including calculating the relative error of the power allocation result, and determining whether it is necessary to iterate the power allocation result;
[0182] like Figure 5 As shown, step S3 includes the following steps:
[0183] S301. Depending on whether time t is the initial time, there are two cases:
[0184] Case 1: If time t is the initial time, the relative error between the updated power allocation result at time t and the average power allocation result at time t is calculated, as shown in formula (35):
[0185]
[0186] in, Represents the relative error obtained by dividing the difference between the average power allocation result of the i-th energy storage system at time t and the updated power allocation result at time t by the absolute value of the average power allocation result at time t;
[0187] Case 2: If time t is not the initial time, the relative error between the updated power allocation result at time t and the final power allocation result at time t-1 is calculated, as shown in formula (36):
[0188]
[0189] in, Relative error is the difference between the final power allocation result of the i-th energy storage system at time t-1 and the updated power allocation result at time t divided by the final power allocation result at time t-1;
[0190] S302, determining whether the power allocation result updated at time t meets the set conditions;
[0191] If the current number of iterations n does not reach the set upper limit n max And the relative error of the energy storage system cluster power allocation in this iteration is There is a relative error ε of some energy storage systems in the sequence t If the relative error ε0 of each energy storage system is greater than the set value ε0 (ε0 = 0.1%), the updated power allocation result at time t replaces the final power allocation result at time t-1 or the average power allocation result at time t, and returns to step S201 to re-iterate until the relative error ε0 of each energy storage system is less than 0.1%. t are not greater than ε0 or the number of iterations at this moment has reached the upper limit of the number of iterations n max ;
[0192] If the current number of iterations n does not reach the upper limit of the number of iterations n max , the relative error ε of each energy storage system t Already satisfied ε t ≤ε0, the updated power allocation result obtained in this iteration is used as the final power allocation result at time t; if the current number of iterations n has reached the upper limit of the number of iterations n max , then the updated power allocation result obtained in the last iteration is taken as the final power allocation result at time t;
[0193] Finally, if the energy storage system cluster stops running, the output power of all energy storage systems is zero, that is, P=0.
[0194] Example 1
[0195] This embodiment is an application of collecting a set of real data in the energy storage system and using the energy management strategy proposed in the present invention to optimize the efficiency of the energy storage system cluster. In this embodiment, the number of single energy storage systems in the energy storage system cluster is 4.
[0196] Among them, the internal parameters of each single energy storage system at the initial moment are shown in Table 1:
[0197] Table 1 Internal parameters of each single power battery system at the initial moment
[0198] Single energy storage system parameters j=1 j=2 j=3 j=4 <![CDATA[SOC j ]]> 0.5 0.5 0.5 0.5 <![CDATA[RUL j ]]> 0.98 0.96 0.94 0.92 <![CDATA[Q j ]]> 58.80 57.60 56.40 55.20
[0199] Step S1, initializing system control parameters and modeling energy consumption characteristics;
[0200] S101, initializing the power allocation result of the energy storage system cluster according to formula (1) so as to update the power allocation result at the initial moment;
[0201] The number of single energy storage systems in formula (1) is m=4;
[0202] S102: Calculate the real-time power limit sequence of the energy storage system cluster according to formula (2)-(3). The open circuit voltage E in formula (2)-(3) is m , equivalent internal resistance R chg and R dis , available capacity Q and the corresponding power limit sequence and It will change with the SOC and RUL at different times;
[0203] The maximum charge and discharge rate CR in formula (2)-(3) max =4;
[0204] S103, first, according to formula (4), calculate the energy consumption characteristic expression of the energy storage battery as the SOC and RUL change at the current moment;
[0205] Then, the characteristic curve fitting polynomial of the cascade converter changing with the current SOC is calculated according to formula (6);
[0206] Finally, the expressions obtained by combining formulas (4)-(6) in formula (7) are used to obtain the energy consumption characteristic expression of a single energy storage system.
[0207] Step S2: Allocate power to the system based on real-time load demand;
[0208] S201. Calculate the approximate local secondary energy consumption charge and discharge function of each single energy storage system according to formula (9), and select different Taylor expansion points according to the current moment.
[0209] S202, according to formulas (15)-(18), calculate the linear coefficient matrix K of the power allocation expression related to the energy storage system cluster and the upper and lower power limits DDS and K DES And the constant term matrix L DDS and L DES ;
[0210] In formula (18), the upper and lower limits of the state of charge of the energy storage system cluster are taken as 0.8 and 0.2 respectively, and the constant μ for adjusting the consistency speed between energy storage systems is taken as 0.5830;
[0211] The maximum power limit of each single energy storage system at the initial moment obtained by formula (2)-(3) is: [7.1094×10 4 6.9766×10 4 6.8093×10 4 6.613×10 4 ];
[0212] The minimum power limit is: [-13.4624×10 4-13.179×10 4 -12.9199×10 4 12.6815×10 4 ];
[0213] According to formula (19), the corresponding The value is [2.7642×10 5 2.7503×10 5 2.7439×10 5 2.7444×10 5 ];
[0214] The calculated In ascending order: [2.7439 × 10 5 2.7444×10 5 2.7503×10 5 2.7642×10 5 ];
[0215] The corresponding energy storage system serial numbers are: [3 4 2 1];
[0216] According to formula (19), the corresponding The value is [-5.2395×10 5 -5.1979×10 5 -5.2048×10 5 -5.2559×10 5 ];
[0217] The calculated Sort in descending order: [-5.1979×10 5 -5.2048×10 5 -5.2395×10 5 -5.2559×10 5 ];
[0218] The corresponding energy storage system serial numbers are: [2 3 1 4];
[0219] According to formula (20), the corresponding The value is [2.7439×10 5 2.7443×10 5 2.7473×10 5 2.7508×10 5 ];
[0220] Corresponding The value is [-5.1979×10 5 -5.2031×10 5-5.2203×10 5 -5.2243×10 5 ];
[0221] S203, power allocation of energy storage system cluster;
[0222] Calculate the real-time power allocation results of each single energy storage system according to formulas (21)-(34);
[0223] The power allocated in real time during the discharge process of each single energy storage system changes with the required power. Figure 6 As shown; the power allocated in real time during the charging process changes with the required power. Figure 7 As shown in the figure, it can be seen that the energy storage system with better performance bears more power demand, and the energy storage system with poorer performance bears less power demand;
[0224] Step S3, optimizing and approximating the real-time power allocation result of the energy storage system cluster based on iteration;
[0225] S301, calculating the relative error of the power allocation result of each single energy storage system according to formula (35);
[0226] S302, first compare the number of iterations n of the current loop with the upper limit of iteration n max In comparison, if it is less than n max Then the relative error calculated by formula (35) is compared with the set value ε0 (ε0 = 0.1%) to determine whether the next iteration is needed;
[0227] The iteration upper limit n used in the present invention is max =10; setting value ε0=0.1%.
[0228] Of course, the above description is not a limitation of the present invention, and the present invention is not limited to the above examples. Changes, modifications, additions or substitutions made by technicians in this technical field within the essential scope of the present invention should also fall within the protection scope of the present invention.
Claims
1. A real-time efficiency optimal energy management method for a battery energy storage system cluster, characterized in that: The following steps are involved: S1. System control parameter initialization and energy consumption characteristic modeling, including initialization of energy storage system cluster power allocation results, update of energy storage system cluster real-time power limit sequence, and modeling of energy storage battery energy consumption characteristic model and cascade converter characteristic curve; S2. Power allocation of the system is performed based on real-time load demand, which is divided into two cases: charging power allocation and discharging power allocation. The specific steps include local quadratic reduction of the energy consumption characteristic model of the energy storage system, calculation of the power allocation coefficient matrix and power switching sequence of the energy storage system cluster, and power allocation according to the size of the load demand at different times; S3, optimizing and approximating the real-time power allocation structure of the energy storage system cluster based on iteration, the specific steps include calculating the relative error of the power allocation result, and determining whether it is necessary to iterate the power allocation result; The step S1 comprises the following steps: S101. For initialization of the power allocation result between energy storage system clusters, an average power allocation method is used to divide the load demand in a single operating cycle by the number of energy storage systems in the energy storage system cluster, and the load demand is evenly allocated to each energy storage system as the initial power allocation result of the energy storage system cluster, as shown in formula (1): Among them, P d,t represents the load demand of the energy storage system cluster at time t, Indicates the load power that each energy storage system should bear for the load demand at time t; S102: According to the SOC and RUL of each energy storage system at time t, the power limit sequence of the energy storage system cluster is updated, as shown in formulas (2) and (3): in, and They represent the maximum power limit sequence and the minimum power limit sequence of the energy storage system cluster at time t respectively; Q t , and They represent the open circuit voltage, available capacity, and equivalent internal resistance of the energy storage system cluster at time t, which are related to the current SOC and RUL of each energy storage system; CR max Represents the maximum charge and discharge rate of the energy storage system cluster, which is a constant; S103. First, an energy storage battery equivalent model is established based on the Rint equivalent circuit, and the energy consumption characteristic expression of the relationship between the energy storage battery output power and the net output power is obtained through Taylor series expansion, as shown in formula (4): Among them, P gross P is the output power of the energy storage battery without considering the loss; bat is the net output power of the energy storage battery after deducting the loss, which is also the input power of the cascade converter; c α The equivalent model of the energy storage battery is represented by P bat =0 is the αth coefficient of the Mth Taylor expansion of the Taylor expansion point, which is related to the current SOC, RUL and charge and discharge state of the energy storage battery; the specific formula is as follows: Secondly, the Buck-Boost bidirectional converter is selected as the cascade converter of the energy storage system, and according to the characteristic curve of the cascade converter, the high-order nonlinear fitting expression of the input power and output power of the cascade converter is obtained, as shown in formula (6): Among them, d β The βth coefficient of the Nth order fitting polynomial of the energy storage battery converter model is related to the current SOC and charge and discharge state of the energy storage battery; P net is the net output power of the cascaded converter; Finally, the expression of the energy storage battery and the cascade converter is combined to form the expression describing the output power P of a single energy storage system. gross With the net output power P net The energy consumption characteristic model of the relationship between is shown in formula (7): Among them, e γ Represents the γth coefficient of the M*Nth energy consumption characteristic function expression of a single energy storage system; e γ With formula (4) c α , d in formula (6) β The relationship is as follows:
2. A method for real-time efficiency optimal energy management of a battery energy storage system cluster according to claim 1, characterized in that: The step S2 comprises the following steps: S201. Based on the energy consumption characteristic model, local quadratic reduction is performed to construct approximate local quadratic energy consumption charging and discharging functions of a single energy storage system at time t, as shown in formula (9): in, and are the quadratic term coefficient, linear term coefficient and constant term coefficient of the discharge energy consumption function model of the i-th energy storage system related to SOC at time t; and are the quadratic term coefficient, linear term coefficient and constant term coefficient of the charging energy consumption function model of the i-th energy storage system related to SOC at time t; The above coefficients will vary according to the Taylor expansion points selected at different times, which can be divided into the following two cases: Case 1: If it is at the initial time t = 1, the charging and discharging model takes the initialization power average distribution result at time t = 1 as the Taylor expansion point, and adopts different energy consumption functions according to the positive and negative power demand at this time. At this time, the expressions of the coefficients are shown in formulas (10) and (11): Case 2: If it is not at the initial time t=1, the power allocation result at time t-1 is used as the Taylor expansion point of the charge-discharge model, and different energy consumption functions are used according to the positive or negative power demand at that time. At this time, the expressions of the coefficients are shown in formulas (12) and (13): Among them, f, f' and f" are the single energy storage system P during discharge. gross With P net The relationship between the function, first-order derivative and second-order derivative; g, g' and g" are the single energy storage system P during charging gross With P net The relationship between functions, first-order derivatives and second-order derivatives; is the average distribution result of the initialization power at time t=1, as shown in formula (14): S202. Calculate the power allocation coefficient matrix within the energy storage system cluster, as shown in formula (15): Among them, K DDS and L DDS They represent the linear coefficient and constant coefficient matrix of the energy storage system cluster power allocation related to the power upper limit respectively; K DES and L DES They represent the linear coefficient and constant coefficient matrix of the energy storage system cluster power allocation related to the power lower limit respectively; k ρ,s and l ρ,s They respectively represent the linear coefficient and constant coefficient of the ρth energy storage system involved in the power allocation expression of the energy storage system cluster; the expressions of each coefficient are shown in formula (16): Among them, s represents the number of energy storage systems involved in regulation; m represents the number of single energy storage systems in the energy storage system cluster; ω ρ A represents the penalty coefficient for adjusting the SOC consistency of each energy storage system involved in the regulation; ρ and q ρ They respectively represent the reciprocal of the quadratic coefficient and the linear coefficient of the quadratic function form of the charging and discharging equivalent model of each single energy storage system involved in the regulation; ω ρ , A ρ and q ρ The expressions of are shown in equations (17) and (18): Where μ represents the constant for adjusting the consistency speed between energy storage systems; SOC HL and SOC LL They represent the upper and lower limits of the state of charge set by the energy storage system cluster; SOC ρ represents the state of charge of the ρth energy storage system involved in regulation; Calculate the power switching sequence of the energy storage system cluster. First, determine the order in which each energy storage system reaches the maximum and minimum power limits according to the performance of each energy storage system, as shown in formula (19): Then, the calculated m and m Arrange in ascending and descending order respectively to obtain the maximum power limit sequence of each single energy storage system and minimum power limit sequence The relationship between h and j is h = m-j+1; Finally, according to the obtained maximum power limit sequence and minimum power limit sequence The maximum power limit of the corresponding energy storage system cluster and minimum power limit After sorting, the load demand of the system cluster corresponding to each single energy storage system reaching the maximum power limit and dropping to the minimum power limit is calculated, that is, the power limit power switching point, as shown in formula (20): Get the maximum power limit sequence and minimum power limit sequence S203. According to the current load demand, the power distribution of the energy storage system cluster is divided into the following five cases: In the first case, when the load demand of the energy storage system cluster At this time, all single energy storage systems have not escaped from their own minimum output power constraints and do not participate in power regulation. The net output power of each single energy storage system at this moment is shown in formula (21): In the second case, when the load demand of the energy storage system cluster At this time, some energy storage systems still have not escaped their own minimum output power limit, and these energy storage systems do not participate in power regulation. At this moment, the power allocation of the energy storage system cluster satisfies the relationship shown in formula (22): Among them, K h' and L h' They represent the power allocation coefficient matrix of the energy storage system involved in the regulation with respect to the load demand at that moment; the specific forms are shown in formulas (23) and (24): At this time, the power demand allocated to each set of power batteries participating in the regulation is shown in formula (25): P net,i' =k i',h' P d +l i',h' ,i'=1,2,…,h′ (25) In the third case, when the load demand of the energy storage system cluster At this time, all single energy storage systems are out of their own minimum output power limit and have not reached their own maximum output power limit, and all participate in power regulation. At this moment, the power allocation of the energy storage system cluster satisfies the relationship shown in formula (26): Among them, K m and L m They represent the power allocation coefficient matrix of the energy storage system involved in the regulation with respect to the load demand at that moment; the specific forms are shown in formulas (27) and (28): K m =[k 1,m k 2,m … k m-1,m k m,m ] (27) L m =[l 1,m l 2,m … l m-1,m l m,m ] (28) At this time, the power demand allocated to each set of power batteries participating in the regulation is shown in formula (29): P net,i =k i,m P d +l i,m ,i=1,2,…,m (29) In the fourth case, when the load demand of the energy storage system cluster At this time, only part of the energy storage system reaches its maximum output power limit, and this part of the energy storage system does not participate in power regulation. At this moment, the power allocation of the energy storage system cluster satisfies the relationship shown in formula (30): Among them, K j' and L j' They respectively represent the power allocation coefficient matrix of the energy storage system involved in the regulation with respect to the load demand at that moment; the specific forms are shown in formulas (31) and (32): K j' =[k j',m k j'+1,m … k m-1,m k m,m ] (31) L j' =[l j',m l j'+1,m … l m-1,m l m,m ] (32) At this time, the power demand allocated to each set of power batteries participating in the regulation is shown in formula (33): P net,i” =k i”,m P d +l i”,m ,i”=j',j'+1,…,m (33) The fifth case is when the load demand of the energy storage system cluster At this time, all single energy storage systems have reached their maximum output power limits and do not participate in power regulation. The net output power of each single energy storage system at this moment is shown in formula (34):
3. A method for real-time efficiency optimal energy management of a battery energy storage system cluster according to claim 2, characterized in that: The step S3 comprises the following steps: S301. Depending on whether time t is the initial time, there are two cases: Case 1: If time t is the initial time, the relative error between the updated power allocation result at time t and the average power allocation result at time t is calculated, as shown in formula (35): in, Represents the relative error obtained by dividing the difference between the average power allocation result of the i-th energy storage system at the initial moment and the updated power allocation result at the initial moment by the absolute value of the average power allocation result at the initial moment; Case 2: If time t is not the initial time, the relative error between the updated power allocation result at time t and the final power allocation result at time t-1 is calculated, as shown in formula (36): in, Relative error is the difference between the final power allocation result of the i-th energy storage system at time t-1 and the updated power allocation result at time t divided by the final power allocation result at time t-1; S302, determining whether the power allocation result updated at time t meets the set conditions; If the current number of iterations n does not reach the set upper limit n max And the relative error of the energy storage system cluster power allocation in this iteration is There is a relative error ε of some energy storage systems in the sequence t If the relative error ε0 of each energy storage system is greater than the set value ε0=0.1%, the updated power allocation result at time t replaces the final power allocation result at time t-1 or the average power allocation result at time t, and returns to step S201 to re-iterate until the relative error ε0 of each energy storage system is less than 0.1%. t are not greater than ε0 or the number of iterations at this moment has reached the upper limit of the number of iterations n max ; If the current number of iterations n does not reach the upper limit of the number of iterations n max , the relative error ε of each energy storage system t Already satisfied ε t ≤ε0, the updated power allocation result obtained in this iteration is used as the final power allocation result at time t; if the current number of iterations n has reached the upper limit of the number of iterations n max , then the updated power allocation result obtained in the last iteration is taken as the final power allocation result at time t; Finally, if the energy storage system cluster stops running, the output power of all energy storage systems is zero, that is, P=0.
Citation Information
Patent Citations
Two-layer real-time efficiency optimization method for heavy-load fuel cell hybrid power system
CN114335632A
Multi-scene shared energy storage planning method and system of multi-data center alliance
CN115425663A
Optimization method for power distribution strategy model of hybrid electric vehicle
CN116467791A
Fuel cell vehicle power distribution method and device based on experimental data
CN117261695A
Method for operating a hybrid energy storage system
DE102017201061A1
