Power system optimization scheduling method considering energy storage state dependent aging characteristics

Abstract

The invention discloses a power system optimization scheduling method considering an energy storage state dependent aging characteristic, and the method comprises the following steps: 1), depicting an energy storage state dependent aging discrete characteristic expression based on an Arrhenius empirical model; 2) setting parameters of the energy storage state dependent aging discrete characteristic expression, and constructing a discrete Arrhenius empirical model taking discharge power as a decision variable; 3) based on the discretized Arrhenius empirical model, constructing a power system optimization scheduling model considering energy storage parameter rolling update; and 4) solving the power system optimization scheduling model considering the rolling update of the energy storage parameters to obtain a power system optimization scheduling scheme. According to the power system optimization scheduling method considering the energy storage state dependent aging characteristic, the energy storage parameters can be dynamically updated according to the accumulated aging degree, and the energy storage full life cycle state dependent aging characteristic is described. And the stored energy is called in a more economical and reasonable manner within the energy storage operation period.

Description

Technical Field

[0001] The present invention relates to the field of power systems and automation thereof, and in particular to a method for optimizing and dispatching a power system taking into account the state-dependent aging characteristics of energy storage. Background Art

[0002] In recent years, energy storage technology represented by lithium-ion batteries has experienced rapid development, with unit costs continuously decreasing and installed capacity growing explosively. Energy storage with rapid response capabilities has been widely applied and promoted in all aspects of the power system. However, while accepting power system regulation instructions, the irreversible aging process also causes the loss of its own available capacity, reduces the efficiency and life of energy storage, and generates corresponding aging costs. Therefore, establishing an accurate energy storage aging model is of certain significance to the optimal dispatch of the power system.

[0003] At present, many scholars have studied the characterization of energy storage aging characteristics. Among them, the model with fixed energy storage aging parameters has a wide range of applications. This model allocates the energy storage aging loss to the release of each unit of electricity in the entire life cycle, and applies it as a fixed aging parameter in the power system optimization scheduling model. However, affected by time-varying operating conditions and usage frequency, the aging parameters of energy storage are changing dynamically, and there is a large error in calculating the energy storage aging loss. In order to consider the actual operating status of energy storage, aging models based on energy storage cycle depth have begun to attract attention. For the calculation of energy storage cycle depth, existing studies usually use the rain flow counting algorithm to achieve it. However, the rain flow counting algorithm cannot be expressed analytically, and it is necessary to initialize the energy storage operating conditions, calculate the cycle depth and perform iterative solutions. Therefore, it cannot be directly embedded in the power system optimization scheduling model, which increases the complexity of the solution process.

[0004] In addition, the above studies have all ignored the fact that the aging characteristics of energy storage are affected by changes in the life of energy storage. For the same type of energy storage, when receiving scheduling instructions to perform the same charging and discharging behavior at the beginning of its life and near retirement, the energy storage aging losses caused are not the same. The present invention refers to the phenomenon that the aging process of energy storage is affected by the coupling of operating status and life changes as "state-dependent aging characteristics." In the long-term operation simulation of energy storage, the entire life cycle of energy storage is covered, so the state-dependent aging characteristics of energy storage cannot be ignored.

[0005] In the field of electrochemical mechanism research of lithium-ion batteries, the Arrhenius empirical model is a classic model for describing chemical reaction rates and is widely recognized. Through this model, the energy storage aging process can be established as an energy storage aging model with discharge rate and cumulative discharge as parameters. However, the external characteristic parameters of energy storage in this model do not directly correspond to the modeling variables of the power system for energy storage, and can only achieve a post-estimation of the degree of energy storage aging. At the same time, the above-mentioned improved model is highly nonlinear and cannot be embedded in the power system optimization dispatch model.

[0006] In summary, there is an urgent need to study an optimal dispatching method for power systems that takes into account the state-dependent aging characteristics of energy storage in order to improve the economy of power system operation. Summary of the invention

[0007] The object of the present invention is to provide a method for optimizing the dispatching of a power system taking into account the state-dependent aging characteristics of energy storage, comprising the following steps:

[0008] 1) Based on the Arrhenius empirical model, the discrete characteristic expression of energy storage state-dependent aging is characterized;

[0009] 2) Adjust the parameters of the energy storage state-dependent aging discrete characteristic expression to construct a discharge power P dis is the discretized Arrhenius empirical model of the decision variables;

[0010] 3) Based on the discretized Arrhenius empirical model, a power system optimization dispatch model taking into account the rolling update of energy storage parameters is constructed;

[0011] 4) Solve the power system optimization dispatch model taking into account the rolling update of energy storage parameters and obtain the power system optimization dispatch plan.

[0012] Furthermore, the steps of describing the discrete characteristic expression of energy storage state-dependent aging include:

[0013] 1.1) Construct an energy storage aging model based on the Arrhenius empirical model;

[0014] 1.2) Process the energy storage aging model and construct the discrete characteristic expression of energy storage state-dependent aging.

[0015] Furthermore, the energy storage aging model based on the Arrhenius empirical model is as follows:

[0016]

[0017] Where: C loss is the accumulated aging loss of energy storage; B is the pre-exponential factor; E a is the reaction activation energy; R is a constant; T is the thermodynamic temperature of the energy storage operating environment; z is the power law factor;

[0018] Cumulative discharge capacity A h As shown below:

[0019] A h =n·DOD·C init (2)

[0020] Where: n is the number of cycles; DOD is the depth of cycle; C init is the initial available capacity of energy storage.

[0021] Furthermore, the steps of constructing the discrete characteristic expression of energy storage state-dependent aging include:

[0022] 1.2.1) Calculation Right now:

[0023]

[0024] Where: C loss is the accumulated aging loss of energy storage; B is the pre-exponential factor; E a is the reaction activation energy; R is a constant; T is the thermodynamic temperature of the energy storage operating environment; z is the power law factor;

[0025] 1.2.2) Transform formula (1) to obtain:

[0026]

[0027] 1.2.3) Substituting formula (4) into formula (3), we can obtain the discrete characteristic expression of energy storage state-dependent aging, namely:

[0028]

[0029] Where: ΔC loss,n ΔA is the corresponding aging loss when the energy storage receives the dispatch instruction for the nth time since it was put into use; h is the amount of electricity released when energy storage participates in scheduling; C loss,n-1 The accumulated aging loss of energy storage.

[0030] The amount of electricity released when energy storage participates in scheduling ΔA h As shown below:

[0031] ΔA h =P dis Δt (6)

[0032] Where: P dis is the discharge power of the energy storage; Δt is the discharge duration.

[0033] Further, construct a discharge power P dis The steps for discretizing the Arrhenius empirical model for the decision variables include:

[0034] 2.1) Reaction activation energy E a After tuning, we get:

[0035] E a =k 1 +k 2 C rate (7)

[0036] Where: C rateis the discharge rate; k 1 With k 2 To collect the performance of lithium iron phosphate batteries at different C using Arbin battery test system rate The coefficients obtained by fitting the cycle life data under ;

[0037] 2.2) Taking the logarithm of formula (1), we get:

[0038]

[0039] Where: C n is the available capacity of the energy storage when it runs in a complete cycle for the nth time;

[0040] 2.3) Taking the logarithm of formula (2), we get ln(A h );

[0041] 2.4) Determine the discharge rate and test environment temperature, and obtain the measured data set under standard test conditions, and determine the reaction activation energy E according to the measured data set a , R and T;

[0042] 2.5) Rearrange the terms in formula (8) to obtain:

[0043]

[0044] 2.6) Perform linear regression on formula (10), the slope in the linear regression fitting formula is the power law factor z, and the intercept is the logarithm of the pre-exponential factor corresponding to the current discharge rate ln(B);

[0045] 2.7) Update the discharge rate and the test environment temperature, return to step 2.4), and obtain the pre-exponential factor B under multiple discharge rates;

[0046] 2.8) Substitute the pre-exponential factor B and discharge rate C from step 2.7) rate Fitting, we get:

[0047] B=k 3 +k 4 C rate (11)

[0048] In the formula, k 3 With k 4 is the root fitting coefficient;

[0049] 2.9) Construct the discharge rate C rate The expression is:

[0050]

[0051] Where: P dis is the discharge power when energy storage participates in scheduling, Pnom is the energy storage rated power;

[0052] 2.10) Substituting formulas (6), (7), (11), and (12) into formula (5), we can obtain the discharge power P dis is the discretized Arrhenius empirical model of the decision variables, namely:

[0053]

[0054] The discretized Arrhenius empirical model is simplified as follows:

[0055] ΔC loss,n =λ n P dis (14)

[0056] Where: ΔC loss,n is the discrete aging loss of energy storage after linear regression; n The cumulative aging loss reaches C loss,n-1 The energy storage loss rate when energy storage participates in scheduling.

[0057] Furthermore, the objective function F of the power system optimization dispatch model taking into account the rolling update of energy storage parameters is as follows:

[0058]

[0059] Where: F is the total operating cost of the power system, including the operating cost of conventional units, wind and solar power abandonment penalties, energy storage operating costs, energy storage aging costs and load shedding penalties; a i Quote for unit i; is the power of unit i in period t; ρ w Penalty fee for unit abandoned wind volume, and is the maximum power that can be generated and the actual power of wind farm k in period t; is the maximum power that can be generated and the actual power of the photovoltaic power station r in time period t; and is the energy storage charging and discharging quotation for period t, and is the charging and discharging power of the energy storage station m in period t, E m is the available capacity of the energy storage station m, is the energy storage discrete aging loss corresponding to the operation of energy storage power station m during period t; Γ is the unit capacity construction cost of energy storage; π is the unit load shedding penalty of the power system, is the load shedding amount of node j in period t; N, N w 、N pv 、N B and N Lare the number of conventional units, wind farms, photovoltaic power stations, energy storage power stations and connected load nodes respectively; Δt is the time interval, and T is the number of time periods. ρ pv It is the penalty fee for the amount of abandoned light per unit.

[0060] Furthermore, the energy storage discrete aging loss corresponding to the energy storage station m when it is running in time period t is Calculated by discretizing the Arrhenius empirical model.

[0061] Furthermore, the constraints of the power system optimization dispatch model taking into account the rolling update of energy storage parameters include battery energy storage operation constraints, distributed renewable energy operation constraints, conventional unit operation constraints, and power system constraints;

[0062] Among them, battery energy storage operation constraints include charge and discharge power constraints, energy conservation constraints, and state of charge constraints;

[0063] Conventional unit operation constraints include power regulation constraints, ramp constraints, and start / stop time constraints;

[0064] Power system constraints include power balance constraints, flow constraints and load shedding constraints.

[0065] Furthermore, the upper and lower limits of charge and discharge power are as follows:

[0066]

[0067] In the formula, and is the upper limit of the charging and discharging power of the energy storage station m in period t; and is the charging and discharging power of the energy storage station m in period t;

[0068] The energy conservation constraints are as follows:

[0069]

[0070] Where: is the charge state of the energy storage station m in period t; η ch and η dis The charging and discharging efficiency of the energy storage power station; and is the charging and discharging power of the energy storage station m in period t, E m is the available capacity of energy storage power station m; Δt is the time interval; is the charge state of the energy storage station m at time t-1;

[0071] The state of charge constraints are as follows:

[0072]

[0073] Where: and are the lower and upper limits of the state of charge of the energy storage station m;

[0074] The operating constraints of distributed renewable energy are as follows:

[0075]

[0076] Where: and is the maximum power that can be generated and the actual power of wind farm k in period t; is the maximum power that can be generated and the actual power of the photovoltaic power station r in time period t;

[0077] The power regulation constraints are as follows:

[0078]

[0079] Where: and is the output P of unit i during period t i t The lower and upper limits of Indicates the operating status of unit i in period t, 1 indicates that the unit is running, and 0 indicates that the unit is shut down;

[0080] The hill climbing constraints are as follows:

[0081] R i,down ≤P i t -P i t-1 ≤R i,up (twenty three)

[0082] Where: R i,down and R i,up is the upper and lower limits of the climbing slope of unit i; P i t-1 is the output of unit i during period t-1; P i t is the output of unit i during period t;

[0083] The start and stop time constraints are as follows:

[0084]

[0085] Where: ts and to are the minimum start and stop time of the unit; Indicates the operating status of unit i in period k; Indicates the operating status of unit i in period t-1;

[0086] The power balance constraints are as follows:

[0087]

[0088] Where: is the load demand of node j in period t; N, N w 、N pv 、N B and N L They are the number of conventional units, wind farms, photovoltaic power stations, energy storage power stations and connected load nodes; Charge and discharge power for energy storage power stations; P i t Provide power for wind farms, photovoltaic power stations, and conventional units;

[0089] The power flow constraints are as follows:

[0090]

[0091] Where: G x-i represents the impact of the injected power of node i on line x; G x-j represents the impact of the injected power of node j on line x; are the maximum and minimum transmission capacities of branch l respectively;

[0092] The load shedding constraints are as follows:

[0093]

[0094] Where: is the load shedding amount of node j in period t.

[0095] Furthermore, in the process of solving the optimal dispatch model of the power system taking into account the rolling update of energy storage parameters, the available capacity of energy storage E m , cumulative aging loss C loss,n and energy storage loss rate λ n All are in a dynamic update state;

[0096] Energy storage available capacity E m , cumulative aging loss C loss,n and energy storage loss rate λ n The update process is as follows:

[0097] s1) Update the accumulated aging loss C loss,n ,Right now:

[0098] C loss,n =C loss,n-1 +ΔC loss,n (29)

[0099] s2) Update the accumulated aging loss C loss,nSubstituting the discretized Arrhenius empirical model, the accumulated aging loss is calculated as C in the theoretical discharge power range of energy storage. loss,n The corresponding energy storage discrete aging loss is

[0100] s3) Discrete aging loss of stored energy Substitute into formula (14) and update the energy storage loss rate λ n+1 ;

[0101] s4) Based on constraint (18), update the available energy storage capacity E m .

[0102] The technical effect of the present invention is unquestionable. The power system optimization dispatching method proposed by the present invention taking into account the state-dependent aging characteristics of energy storage can dynamically update the energy storage parameters according to the accumulated aging degree, characterize the state-dependent aging characteristics of energy storage throughout its life cycle, and call the energy storage in a more economical and reasonable way within the operating life of the energy storage. BRIEF DESCRIPTION OF THE DRAWINGS

[0103] Figure 1 Simulate input conditions and energy storage operating status for daily operation; Figure 1 (a) Changes in electricity prices; Figure 1 (b) Output of conventional units, wind power, photovoltaic power and energy storage; Figure 1 (c) is the change of energy storage SOC;

[0104] Figure 2 It is the annual aging loss and cumulative aging loss of energy storage. DETAILED DESCRIPTION

[0105] The present invention is further described below in conjunction with the embodiments, but it should not be understood that the above subject matter of the present invention is limited to the following embodiments. Without departing from the above technical ideas of the present invention, various substitutions and changes are made according to the common technical knowledge and customary means in the art, which should all be included in the protection scope of the present invention.

[0106] Embodiment 1:

[0107] A method for optimizing the dispatching of a power system considering the state-dependent aging characteristics of energy storage comprises the following steps:

[0108] 1) Based on the Arrhenius empirical model, the discrete characteristic expression of energy storage state-dependent aging is characterized;

[0109] 2) Adjust the parameters of the energy storage state-dependent aging discrete characteristic expression to construct a discharge power P dis is the discretized Arrhenius empirical model of the decision variables;

[0110] 3) Based on the discretized Arrhenius empirical model, a power system optimization dispatch model taking into account the rolling update of energy storage parameters is constructed;

[0111] 4) Solve the power system optimization dispatch model taking into account the rolling update of energy storage parameters and obtain the power system optimization dispatch plan.

[0112] The steps of describing the discrete characteristic expression of energy storage state-dependent aging include:

[0113] 1.1) Construct an energy storage aging model based on the Arrhenius empirical model;

[0114] 1.2) Process the energy storage aging model and construct the discrete characteristic expression of energy storage state-dependent aging.

[0115] The energy storage aging model based on the Arrhenius empirical model is as follows:

[0116]

[0117] Where: C loss is the accumulated aging loss of energy storage; B is the pre-exponential factor; E a is the reaction activation energy; R is a constant; T is the thermodynamic temperature of the energy storage operating environment; z is the power law factor;

[0118] Cumulative discharge capacity A h As shown below:

[0119] A h =n·DOD·C init (2)

[0120] Where: n is the number of cycles; DOD is the depth of cycle; C init is the initial available capacity of energy storage.

[0121] The steps of constructing the discrete characteristic expression of energy storage state-dependent aging include:

[0122] 1.2.1) Calculation Right now:

[0123]

[0124] Where: C loss is the accumulated aging loss of energy storage; B is the pre-exponential factor; E a is the reaction activation energy; R is a constant; T is the thermodynamic temperature of the energy storage operating environment; z is the power law factor;

[0125] 1.2.2) Transform formula (1) to obtain:

[0126]

[0127] 1.2.3) Substituting formula (4) into formula (3), we can obtain the discrete characteristic expression of energy storage state-dependent aging, namely:

[0128]

[0129] Where: ΔC loss,n ΔA is the corresponding aging loss when the energy storage receives the dispatch instruction for the nth time since it was put into use; h is the amount of electricity released when energy storage participates in scheduling; C loss,n-1 The accumulated aging loss of energy storage.

[0130] The amount of electricity released when energy storage participates in scheduling ΔA h As shown below:

[0131] ΔA h =P dis Δt (6)

[0132] Where: P dis is the discharge power of the energy storage; Δt is the discharge duration.

[0133] Constructed with discharge power P dis The steps for discretizing the Arrhenius empirical model for the decision variables include:

[0134] 2.1) Reaction activation energy E a After tuning, we get:

[0135] E a =k 1 +k 2 C rate (7)

[0136] Where: C rate is the discharge rate; k 1 With k 2 To analyze the measured data set of Sandia National Laboratory (using Arbin battery test system to collect data on lithium iron phosphate batteries at different C rate The coefficients obtained by fitting the cycle life data under

[0137] 2.2) Taking the logarithm of formula (1), we get:

[0138]

[0139] Where: C n is the available capacity of the energy storage when it runs in a complete cycle for the nth time;

[0140] 2.3) Taking the logarithm of formula (2), we get ln(A h );

[0141] 2.4) Determine the discharge rate and test environment temperature, and obtain the measured data set under standard test conditions, and determine the reaction activation energy E according to the measured data set a , R and T;

[0142] 2.5) Rearrange the terms in formula (8) to obtain:

[0143]

[0144] 2.6) Perform linear regression on formula (10), the slope in the linear regression fitting formula is the power law factor z, and the intercept is the logarithm of the pre-exponential factor corresponding to the current discharge rate ln(B);

[0145] 2.7) Update the discharge rate and the test environment temperature, return to step 2.4), and obtain the pre-exponential factor B under multiple discharge rates;

[0146] 2.8) Substitute the pre-exponential factor B and discharge rate C from step 2.7) rate Fitting, we get:

[0147] B=k 3 +k 4 C rate (11)

[0148] In the formula, k 3 With k 4 is the root fitting coefficient;

[0149] 2.9) Construct the discharge rate C rate The expression is:

[0150]

[0151] Where: P dis is the discharge power when energy storage participates in scheduling, P nom is the energy storage rated power;

[0152] 2.10) Substituting formulas (6), (7), (11), and (12) into formula (5), we can obtain the discharge power P dis is the discretized Arrhenius empirical model of the decision variables, namely:

[0153]

[0154] The discretized Arrhenius empirical model is simplified as follows:

[0155] ΔC loss,n =λ n P dis (14)

[0156] Where: ΔC loss,nis the discrete aging loss of energy storage after linear regression; n The cumulative aging loss reaches C loss,n-1 The energy storage loss rate when energy storage participates in scheduling.

[0157] The objective function F of the power system optimal dispatch model taking into account the rolling update of energy storage parameters is as follows:

[0158]

[0159] Where: F is the total operating cost of the power system, including the operating cost of conventional units, wind and solar power abandonment penalties, energy storage operating costs, energy storage aging costs and load shedding penalties; a i Quote for unit i; is the power of unit i in period t; ρ w Penalty fee for unit abandoned wind volume, and is the maximum power that can be generated and the actual power of wind farm k in period t; is the maximum power that can be generated and the actual power of the photovoltaic power station r in time period t; and is the energy storage charging and discharging quotation for period t, and is the charging and discharging power of the energy storage station m in period t, E m is the available capacity of the energy storage station m, is the energy storage discrete aging loss corresponding to the operation of energy storage power station m during period t; Γ is the unit capacity construction cost of energy storage; π is the unit load shedding penalty of the power system, is the load shedding amount of node j in period t; N, N w 、N pv 、N B and N L are the number of conventional units, wind farms, photovoltaic power stations, energy storage power stations and connected load nodes respectively; Δt is the time interval, and T is the number of time periods. ρ pv It is the penalty fee for the amount of abandoned light per unit.

[0160] The energy storage discrete aging loss corresponding to the energy storage station m when it is running in time period t Calculated by discretizing the Arrhenius empirical model.

[0161] The constraints of the power system optimization dispatch model taking into account the rolling update of energy storage parameters include battery energy storage operation constraints, distributed renewable energy operation constraints, conventional unit operation constraints, and power system constraints.

[0162] Among them, battery energy storage operation constraints include charge and discharge power constraints, energy conservation constraints, and state of charge constraints;

[0163] Conventional unit operation constraints include power regulation constraints, ramp constraints, and start / stop time constraints;

[0164] Power system constraints include power balance constraints, flow constraints and load shedding constraints.

[0165] The upper and lower limits of charge and discharge power are as follows:

[0166]

[0167] In the formula, and is the upper limit of the charging and discharging power of the energy storage station m in period t; and is the charging and discharging power of the energy storage station m in period t;

[0168] The energy conservation constraints are as follows:

[0169]

[0170] Where: is the charge state of the energy storage station m in period t; η ch and η dis The charging and discharging efficiency of the energy storage power station; and is the charging and discharging power of the energy storage station m in period t, E m is the available capacity of energy storage power station m; Δt is the time interval; is the charge state of the energy storage station m at time t-1;

[0171] The state of charge constraints are as follows:

[0172]

[0173] Where: and are the lower and upper limits of the state of charge of the energy storage station m;

[0174] The operating constraints of distributed renewable energy are as follows:

[0175]

[0176] Where: and is the maximum power that can be generated and the actual power of wind farm k in period t; is the maximum power that can be generated and the actual power of the photovoltaic power station r in time period t;

[0177] The power regulation constraints are as follows:

[0178]

[0179] Where: and is the output P of unit i during period t i t The lower and upper limits of Indicates the operating status of unit i in period t, 1 indicates that the unit is running, and 0 indicates that the unit is shut down;

[0180] The hill climbing constraints are as follows:

[0181] R i,down ≤P i t -P i t-1 ≤R i,up (twenty three)

[0182] Where: R i,down and R i,up is the upper and lower limits of the climbing slope of unit i; P i t-1 is the output of unit i during period t-1; P i t is the output of unit i during period t;

[0183] The start and stop time constraints are as follows:

[0184]

[0185] Where: ts and to are the minimum start and stop time of the unit; Indicates the operating status of unit i in period k; Indicates the operating status of unit i in period t-1;

[0186] The power balance constraints are as follows:

[0187]

[0188] Where: is the load demand of node j in period t; N, N w 、N pv 、N B and N L They are the number of conventional units, wind farms, photovoltaic power stations, energy storage power stations and connected load nodes; Charge and discharge power for energy storage power stations; P i t Provide power for wind farms, photovoltaic power stations, and conventional units;

[0189] The power flow constraints are as follows:

[0190]

[0191] Where: G x-i represents the impact of the injected power of node i on line x; G x-j represents the impact of the injected power of node j on line x; are the maximum and minimum transmission capacities of branch l respectively;

[0192] The load shedding constraints are as follows:

[0193]

[0194] Where: is the load shedding amount of node j in period t.

[0195] In the process of solving the optimal dispatch model of the power system taking into account the rolling update of energy storage parameters, the available capacity E of energy storage m , cumulative aging loss C loss,n and energy storage loss rate λ n All are in a dynamic update state;

[0196] Energy storage available capacity E m , cumulative aging loss C loss,n and energy storage loss rate λ n The update process is as follows:

[0197] s1) Update the accumulated aging loss C loss,n ,Right now:

[0198] C loss,n =C loss,n-1 +ΔC loss,n (29)

[0199] s2) Update the accumulated aging loss C loss,n Substituting the discretized Arrhenius empirical model, the accumulated aging loss is calculated as C in the theoretical discharge power range of energy storage. loss,n The corresponding energy storage discrete aging loss is

[0200] s3) Discrete aging loss of stored energy Substitute into formula (14) and update the energy storage loss rate λ n+1 ;

[0201] s4) Based on constraint (18), update the available energy storage capacity E m .

[0202] Embodiment 2:

[0203] See also Figure 1 to Figure 2 , a method for optimizing the dispatching of a power system considering the state-dependent aging characteristics of energy storage, comprising the following steps:

[0204] 1) Based on the Arrhenius empirical model, the discrete characteristic expression of energy storage state-dependent aging is characterized;

[0205] 2) Adjust the parameters of the energy storage state-dependent aging discrete characteristic expression to construct a discharge power P dis is the discretized Arrhenius empirical model of the decision variables;

[0206] 3) Based on the discretized Arrhenius empirical model, a power system optimization dispatch model taking into account the rolling update of energy storage parameters is constructed;

[0207] 4) Solve the power system optimization dispatch model taking into account the rolling update of energy storage parameters and obtain the power system optimization dispatch plan.

[0208] Embodiment 3:

[0209] A method for optimizing the dispatch of a power system considering the energy storage state dependent aging characteristics, the technical content of which is the same as any one of Embodiments 1-2, and further, the step of describing the energy storage state dependent aging discrete characteristic expression comprises:

[0210] 1.1) Construct an energy storage aging model based on the Arrhenius empirical model;

[0211] 1.2) Process the energy storage aging model and construct the discrete characteristic expression of energy storage state-dependent aging.

[0212] Embodiment 4:

[0213] A method for optimizing the dispatching of a power system considering the dependent aging characteristics of energy storage state, the technical content of which is the same as any one of Embodiments 1-3, and further, an energy storage aging model based on the Arrhenius empirical model is as follows:

[0214]

[0215] Where: C loss is the accumulated aging loss of energy storage; B is the pre-exponential factor; E a is the reaction activation energy; R is a constant; T is the thermodynamic temperature of the energy storage operating environment; z is the power law factor;

[0216] Cumulative discharge capacity A h As shown below:

[0217] A h =n·DOD·C init (2)

[0218] Where: n is the number of cycles; DOD is the depth of cycle; C init is the initial available capacity of energy storage.

[0219] Embodiment 5:

[0220] A method for optimizing the dispatch of a power system considering the energy storage state-dependent aging characteristics, the technical content of which is the same as any one of Embodiments 1-4, and further, the step of constructing a discrete characteristic expression of the energy storage state-dependent aging characteristics comprises:

[0221] 1.2.1) Calculation Right now:

[0222]

[0223] Where: C loss is the accumulated aging loss of energy storage; B is the pre-exponential factor; E a is the reaction activation energy; R is a constant; T is the thermodynamic temperature of the energy storage operating environment; z is the power law factor;

[0224] 1.2.2) Transform formula (1) to obtain:

[0225]

[0226] 1.2.3) Substituting formula (4) into formula (3), we can obtain the discrete characteristic expression of energy storage state-dependent aging, namely:

[0227]

[0228] Where: ΔC loss,n ΔA is the corresponding aging loss when the energy storage receives the dispatch instruction for the nth time since it was put into use; h is the amount of electricity released when energy storage participates in scheduling; C loss,n-1 The accumulated aging loss of energy storage.

[0229] The amount of electricity released when energy storage participates in scheduling ΔA h As shown below:

[0230] ΔA h =P dis Δt (6)

[0231] Where: P dis is the discharge power of the energy storage; Δt is the discharge duration.

[0232] Embodiment 6:

[0233] A method for optimizing the dispatching of a power system considering the energy storage state-dependent aging characteristics, the technical content of which is the same as any one of Embodiments 1-5, and further, a method for optimizing the dispatching of a power system based on the discharge power P dis The steps for discretizing the Arrhenius empirical model for the decision variables include:

[0234] 2.1) Reaction activation energy E a After tuning, we get:

[0235] E a =k 1 +k 2 C rate (7)

[0236] Where: C rate is the discharge rate; k 1 With k 2 is the coefficient obtained by fitting the measured data set;

[0237] 2.2) Taking the logarithm of formula (1), we get:

[0238]

[0239] Where: C n is the available capacity of the energy storage when it runs in a complete cycle for the nth time;

[0240] 2.3) Taking the logarithm of formula (2), we get ln(A h );

[0241] 2.4) Determine the discharge rate and test environment temperature, and obtain the measured data set under standard test conditions, and determine the reaction activation energy E according to the measured data set a , R and T;

[0242] 2.5) Rearrange the terms in formula (8) to obtain:

[0243]

[0244] 2.6) Perform linear regression on formula (10), the slope in the linear regression fitting formula is the power law factor z, and the intercept is the logarithm of the pre-exponential factor corresponding to the current discharge rate ln(B);

[0245] 2.7) Update the discharge rate and the test environment temperature, return to step 2.4), and obtain the pre-exponential factor B under multiple discharge rates;

[0246] 2.8) Substitute the pre-exponential factor B and discharge rate C from step 2.7) rate Fitting, we get:

[0247] B=k 3 +k 4 C rate (11)

[0248] In the formula, k 3 With k 4 is the root fitting coefficient;

[0249] 2.9) Construct the discharge rate C rate The expression is:

[0250]

[0251] Where: P dis is the discharge power when energy storage participates in scheduling, P nom is the energy storage rated power;

[0252] 2.10) Substituting formulas (6), (7), (11), and (12) into formula (5), we can obtain the discharge power P dis is the discretized Arrhenius empirical model of the decision variables, namely:

[0253]

[0254] The discretized Arrhenius empirical model is simplified as follows:

[0255] ΔC loss,n =λ n P dis (14)

[0256] Where: ΔC loss,n is the discrete aging loss of energy storage after linear regression; n The cumulative aging loss reaches C loss,n-1 The energy storage loss rate when energy storage participates in scheduling.

[0257] Embodiment 7:

[0258] A method for optimizing and dispatching a power system considering the dependent aging characteristics of energy storage states, the technical content of which is the same as any one of Embodiments 1-6, and further, the objective function F of the power system optimizing and dispatching model taking into account the rolling update of energy storage parameters is as follows:

[0259]

[0260] Where: F is the total operating cost of the power system, including the operating cost of conventional units, wind and solar power abandonment penalties, energy storage operating costs, energy storage aging costs and load shedding penalties; a i Quote for unit i; is the power of unit i in period t; ρ w Penalty fee for unit abandoned wind volume, and is the maximum power that can be generated and the actual power of wind farm k in period t; is the maximum power that can be generated and the actual power of the photovoltaic power station r in time period t; and is the energy storage charging and discharging quotation for period t, and is the charging and discharging power of the energy storage station m in period t, E mis the available capacity of the energy storage station m, is the energy storage discrete aging loss corresponding to the operation of energy storage power station m during period t; Γ is the unit capacity construction cost of energy storage; π is the unit load shedding penalty of the power system, is the load shedding amount of node j in period t; N, N w 、N pv 、N B and N L are the number of conventional units, wind farms, photovoltaic power stations, energy storage power stations and connected load nodes respectively; Δt is the time interval, and T is the number of time periods.

[0261] Embodiment 8:

[0262] A method for optimizing the dispatching of a power system considering the dependent aging characteristics of energy storage states, the technical content of which is the same as any one of Embodiments 1-7, and further, the energy storage discrete aging loss corresponding to the energy storage power station m when it is running in time period t Calculated by discretizing the Arrhenius empirical model.

[0263] Embodiment 9:

[0264] A method for optimizing and dispatching a power system considering the dependent aging characteristics of energy storage states, the technical content of which is the same as any one of embodiments 1-8, and further, the constraints of the power system optimization and dispatching model taking into account the rolling update of energy storage parameters include battery energy storage operation constraints, distributed renewable energy operation constraints, conventional unit operation constraints, and power system constraints;

[0265] Among them, battery energy storage operation constraints include charge and discharge power constraints, energy conservation constraints, and state of charge constraints;

[0266] Conventional unit operation constraints include power regulation constraints, ramp constraints, and start / stop time constraints;

[0267] Power system constraints include power balance constraints, flow constraints and load shedding constraints.

[0268] Embodiment 10:

[0269] A method for optimizing and dispatching a power system considering the dependent aging characteristics of energy storage states, the technical content of which is the same as any one of Embodiments 1-9, and further, the upper and lower limit constraints of the charge and discharge power are as follows:

[0270]

[0271] In the formula, and is the upper limit of the charging and discharging power of the energy storage station m in period t; and is the charging and discharging power of the energy storage station m in period t;

[0272] The energy conservation constraints are as follows:

[0273]

[0274] Where: is the charge state of the energy storage station m in period t; η ch and η dis The charging and discharging efficiency of the energy storage power station; and is the charging and discharging power of the energy storage station m in period t, E m is the available capacity of energy storage power station m; Δt is the time interval; is the charge state of the energy storage station m at time t-1;

[0275] The state of charge constraints are as follows:

[0276]

[0277] Where: and are the lower and upper limits of the state of charge of the energy storage station m;

[0278] The operating constraints of distributed renewable energy are as follows:

[0279]

[0280] Where: and is the maximum power that can be generated and the actual power of wind farm k in period t; is the maximum power that can be generated and the actual power of the photovoltaic power station r in time period t;

[0281] The power regulation constraints are as follows:

[0282]

[0283] Where: and is the output P of unit i during period t i t The lower and upper limits of Indicates the operating status of unit i in period t, 1 indicates that the unit is running, and 0 indicates that the unit is shut down;

[0284] The hill climbing constraints are as follows:

[0285] R i,down ≤P i t -P i t-1 ≤R i,up (twenty three)

[0286] Where: R i,down and R i,up is the upper and lower limits of the climbing slope of unit i; P i t-1 is the output of unit i during period t-1;

[0287] The start and stop time constraints are as follows:

[0288]

[0289] Where: ts and to are the minimum start and stop time of the unit; Indicates the operating status of unit i in period k; Indicates the operating status of unit i in period t-1;

[0290] The power balance constraints are as follows:

[0291]

[0292] Where: is the load demand of node j in period t; N, N w 、N pv 、N B and N L They are the number of conventional units, wind farms, photovoltaic power stations, energy storage power stations and connected load nodes;

[0293] The power flow constraints are as follows:

[0294]

[0295] Where: G x-i represents the impact of the injected power of node i on line x; G x-j represents the impact of the injected power of node j on line x; are the maximum and minimum transmission capacities of branch l respectively;

[0296] The load shedding constraints are as follows:

[0297]

[0298] Where: is the load shedding amount of node j in period t.

[0299] Embodiment 11:

[0300] A method for optimizing and dispatching a power system considering the state-dependent aging characteristics of energy storage, the technical content of which is the same as any one of Embodiments 1-10, and further, in the process of solving the power system optimization dispatching model taking into account the rolling update of energy storage parameters, the energy storage available capacity E m , cumulative aging loss Closs,n and energy storage loss rate λ n All are in a dynamic update state;

[0301] Energy storage available capacity E m , cumulative aging loss C loss,n and energy storage loss rate λ n The update process is as follows:

[0302] s1) Update the accumulated aging loss C loss,n ,Right now:

[0303] C loss,n =C loss,n-1 +ΔC loss,n (29)

[0304] s2) Update the accumulated aging loss C loss,n Substituting the discretized Arrhenius empirical model, the accumulated aging loss is calculated as C in the theoretical discharge power range of energy storage. loss,n The corresponding energy storage discrete aging loss is

[0305] s3) Discrete aging loss of stored energy Substitute into formula (14) and update the energy storage loss rate λ n+1 ;

[0306] s4) Based on constraint (18), update the available energy storage capacity E m .

[0307] Embodiment 12:

[0308] A method for optimizing the dispatching of a power system considering the state-dependent aging characteristics of energy storage, comprising:

[0309] 1. Discrete characterization method of energy storage state-dependent aging characteristics

[0310] The specific form of the energy storage aging model based on the Arrhenius empirical model is as follows:

[0311]

[0312] Where: C loss is the accumulated aging loss of energy storage; B is the pre-exponential factor, which is related to the discharge rate; E a is the reaction activation energy, which is related to the type of energy storage and the discharge rate; R is a constant, which remains unchanged under normal operating conditions; T is the thermodynamic temperature of the energy storage operating environment; z is the power law factor, which is obtained by fitting the measured data set; A h The cumulative discharge amount can be calculated by the following formula

[0313] A h =n·DOD·Cinit (2)

[0314] Where: n is the number of cycles; DOD is the depth of cycle; C init is the initial available capacity of energy storage.

[0315] The discretization characterization method of energy storage state-dependent aging characteristics based on the Arrhenius empirical model is mainly divided into the following steps. First, A is converted into h Partial derivative calculation We get formula (3).

[0316]

[0317] Then, transform equation (1) into the following form:

[0318]

[0319] Finally, substituting equation (4) into equation (3) yields the discretized Arrhenius empirical model.

[0320]

[0321] Where: ΔC loss,n ΔA is the corresponding aging loss when the energy storage receives the dispatch instruction for the nth time since it was put into use; h is the amount of electricity released when energy storage participates in scheduling; C loss,n-1 ΔA is the accumulated aging loss of energy storage. h It can be calculated by the following formula

[0322] ΔA h =P dis Δt (6)

[0323] Where: P dis is the discharge power of the energy storage; Δt is the discharge duration.

[0324] 2. Parameter setting method

[0325] Pre-exponential factor B, reaction activation energy E a The adjustment of the power law factor z is discussed further. a It is related to the type of energy storage and the discharge rate of energy storage. a The setting is usually in the form of

[0326] E a =k 1 +k 2 C rate (7)

[0327] Where: C rate is the discharge rate; k 1 With k2 It needs to be fitted according to the measured data set.

[0328] Regarding the setting of the pre-exponential factor B and the power law factor z, by taking the logarithm of equation (1), we can obtain the following equation:

[0329]

[0330] ln(C loss ) can be calculated by the following formula

[0331]

[0332] Where: C n It is the available capacity when the energy storage runs in a complete cycle for the nth time.

[0333] Then take the logarithm of equation (2) to obtain ln(A h ), the reaction activation energy E can be determined based on the discharge rate and test environment temperature of the standard test conditions of the measured data set a , R and T, we can get the following equation by transposing the terms in equation (8)

[0334]

[0335] Perform linear regression on the above formula. The slope in the linear regression fitting formula is the power law factor z, and the intercept is the logarithm of the pre-exponential factor corresponding to the current discharge rate, ln(B). Perform the above linear regression on the measured data set under different discharge rates under standard test conditions to obtain a series of pre-exponential factors B under different discharge rates. rate Fitting to the following relationship

[0336] B=k 3 +k 4 C rate (11)

[0337] In the above parameter setting process, the reaction activation energy E a and pre-exponential factor B are established for the discharge rate C rate However, the external characteristics of energy storage participating in power system dispatch are usually expressed as discharge power, so the above parameters can be converted by the following formula:

[0338]

[0339] Where: P dis is the discharge power when energy storage participates in scheduling, P nom is the rated power of energy storage. Substituting equations (6), (7), (11), (12) into equation (5), we get equation (13) with discharge power P dis The discretized Arrhenius empirical model of the decision variable

[0340]

[0341] At different stages of cumulative aging loss, as the energy storage discharge power increases, the aggravation process of energy storage aging loss is approximately linear, and as the cumulative aging loss increases, at the same discharge power, the energy storage aging loss gradually decreases. In order to reduce the complexity of solving the optimization scheduling model, equation (13) is simplified to the following form based on the linear regression results.

[0342] ΔC loss,n =λ n P dis (14)

[0343] Where: ΔC loss,n is the discrete aging loss of energy storage after linear regression; n The cumulative aging loss reaches C loss,n-1 The energy storage loss rate when the energy storage participates in the dispatch is λ, which specifically means that the aging loss caused by each MWh discharge when the energy storage participates in the dispatch in the current period is n .

[0344] 3. Power system optimization dispatch method considering rolling update of energy storage parameters

[0345] The power system optimization dispatch model taking into account the rolling update of energy storage parameters takes minimizing the power system operation cost as the objective function F, that is,

[0346]

[0347] Where: F is the total operating cost of the power system, including the operating cost of conventional units, wind and solar power abandonment penalties, energy storage operating costs, energy storage aging costs and load shedding penalties. i Quote for unit i, is the power of unit i in period t; ρ w Penalty fee for unit abandoned wind volume, and is the maximum power that can be generated and the actual power of wind farm k in period t; is the maximum power that can be generated and the actual power of the photovoltaic power station r in time period t; and is the energy storage charging and discharging quotation for period t, and is the charging and discharging power of the energy storage station m in period t, E m is the available capacity of the energy storage station m, is the aging loss caused by the operation of energy storage station m during period t, which is calculated by the discretized Arrhenius empirical model. Γ is the construction cost of energy storage unit capacity; π is the unit load shedding penalty of the power system. is the load shedding amount of node j in period t; N, N w 、N pv 、N B and N L are the number of conventional units, wind farms, photovoltaic power stations, energy storage power stations and connected load nodes respectively; Δt is the time interval, and T is the number of time periods.

[0348] The constraints are as follows:

[0349] 1) Battery energy storage operation constraints include charging and discharging power constraints, energy conservation constraints and state of charge constraints, which are described in detail as follows.

[0350] Charge and discharge power upper and lower limits:

[0351]

[0352] In the formula, and is the upper limit of the charging and discharging power of the energy storage station m during period t.

[0353] Energy conservation constraints:

[0354]

[0355] Where: is the charge state of the energy storage station m in period t; η ch and η dis It is the charging and discharging efficiency of the energy storage power station.

[0356] In order to avoid overcharging or over-discharging of energy storage, upper and lower limit constraints on energy storage charge state are introduced.

[0357]

[0358] Where: and are the lower and upper limits of the state of charge of the energy storage station m.

[0359] 2) Distributed renewable energy such as wind power and photovoltaic power has the following operating constraints:

[0360]

[0361] 3) The operating constraints of conventional units mainly include power regulation constraints, ramp constraints and start-stop time constraints, which are specifically expressed as follows.

[0362] Power regulation constraints:

[0363]

[0364] Where: and are the lower and upper limits of the output of unit i during period t; It indicates the operating status of unit i in period t, 1 indicates that the unit is running, and 0 indicates that the unit is shut down.

[0365] Hill climbing constraints:

[0366] R i,down ≤P i t -P i t-1 ≤R i,up (twenty three)

[0367] Where: R i,down and R i,up are the upper and lower limits of the climbing slope of unit i.

[0368] Start and stop time constraints:

[0369]

[0370] Where: ts and to are the minimum start and stop time of the unit.

[0371] 4) Power system constraints mainly include power balance constraints, line capacity constraints, spinning reserve constraints and load shedding constraints.

[0372] Power balance constraints:

[0373]

[0374] Where: is the load demand of node j in period t.

[0375] The power flow constraint adopts the DC power flow model and describes the power flow of each branch by introducing the generator output power transfer distribution factor matrix G:

[0376]

[0377] Where: G x-i Describes the impact of the injected power of node i on line x; G x-j Describe the impact of the injected power at node j on line x; are the maximum and minimum transmission capacities of branch l respectively.

[0378] Load shedding constraints:

[0379]

[0380] Where: is the load shedding amount of node j in period t.

[0381] In the above power system optimization dispatch model, the available energy storage capacity Em , cumulative aging loss C loss,n and energy storage loss rate λ n All are in the process of dynamic updating. Calculate the energy storage daily operation aging loss ΔC according to the energy storage daily operation curve loss,n , and the accumulated aging loss C loss,n-1 sum

[0382] C loss,n =C loss,n-1 +ΔC loss,n (29)

[0383] Get the updated cumulative aging loss C loss,n , then, the updated cumulative aging loss C loss,n Substituting the discretized Arrhenius empirical model, the accumulated aging loss is calculated as C in the theoretical discharge power range of energy storage. loss,n The energy storage discrete aging loss at time , the energy storage loss rate in formula (14) is updated to λ n+1 , and then update the available energy storage capacity in the optimal dispatch model constraint (18).

[0384] Embodiment 13:

[0385] The verification of a power system optimal dispatch method considering the state-dependent aging characteristics of energy storage is as follows:

[0386] This embodiment is tested on the actual power system of a certain province in China, and the power system is simulated for intraday operation with electricity price information, power system load curve, distributed renewable energy output, etc. as boundary conditions. The available energy storage capacity configured in the power system operation simulation model is set to 10MWh, the construction cost per MWh is 1.2 million yuan, and the total construction cost is 12 million yuan. The electricity price information adopts the actual node marginal electricity price of SPP (Southern Power Pool), and the distributed renewable energy output and load adopt the actual operation data of a province in Northwest my country in 2021.

[0387] In order to illustrate the rationality and effectiveness of the design method of the present invention, the fixed parameter model method is used as a comparative method. The energy storage loss rate in the fixed parameter model method is set to 4×10-6, which specifically means that the energy storage will lose 0.004% of the rated capacity for each complete charge and discharge behavior of the energy storage. The calendar aging process is added to both the fixed parameter model method and the method of this article, and the calendar aging loss rate is set to 2.5×10-5, which specifically means that after one natural day, the energy storage will lose 0.0025% of the rated capacity.

[0388] The operation of energy storage is analyzed based on the single-day simulation results. The electricity price of the SPP power market on the 327th day of 2018 is selected as the input, and its changes are as follows: Figure 1 (a) shows the charging and discharging and SOC changes of energy storage participating in power system dispatch under the guidance of electricity price signals. Figure 1 As shown in (b) and (c).

[0389] According to the simulation results of long-term operation of energy storage, the aging loss of energy storage participating in power system dispatch under the method of the present invention and the fixed parameter model method is calculated as follows: Figure 2 As shown. The fixed parameter model method ignores the changes in the energy storage loss rate at different stages of energy storage, so it still frequently participates in scheduling when the aging cost is high, resulting in annual aging losses as high as 2.38%. According to the aging loss of 20% as the retirement threshold, the service life of the energy storage is 8 years. For the method of the present invention, the state-dependent aging characteristics of energy storage are taken into account, and the power system scheduling is participated in in a more economical way. In the first year of long-term operation simulation, the annual aging loss of energy storage is 2.26%, and the annual aging loss before retirement is 1.52%, and the service life of energy storage is 11 years.

[0390] In summary, the power system optimization dispatching method proposed in the present invention that considers the state-dependent aging characteristics of energy storage can dynamically update energy storage parameters according to the accumulated aging degree, characterize the state-dependent aging characteristics of energy storage throughout its life cycle, and call energy storage in a more economical and reasonable manner within the operating life of energy storage.

Claims

1. A method for optimizing the dispatch of a power system considering the state-dependent aging characteristics of energy storage, characterized in that: The following steps are involved: 1) Based on the Arrhenius empirical model, the discrete characteristic expression of energy storage state-dependent aging is characterized. 2) Adjust the parameters of the energy storage state-dependent aging discrete characteristic expression to construct a discharge power P dis is the discretized Arrhenius empirical model of the decision variables; 3) Based on the discretized Arrhenius empirical model, a power system optimization dispatch model taking into account the rolling update of energy storage parameters is constructed; 4) Solve the power system optimization dispatch model taking into account the rolling update of energy storage parameters and obtain the power system optimization dispatch plan.

2. A method for optimizing and dispatching a power system considering the energy storage state-dependent aging characteristics according to claim 1, characterized in that: The steps of describing the discrete characteristic expression of energy storage state-dependent aging include: 1) Construct an energy storage aging model based on the Arrhenius empirical model; 2) Process the energy storage aging model and construct a discrete characteristic expression of energy storage state-dependent aging.

3. A method for optimizing and dispatching a power system considering the energy storage state-dependent aging characteristics according to claim 2, characterized in that: The energy storage aging model based on the Arrhenius empirical model is as follows: Where: C loss is the accumulated aging loss of energy storage; B is the pre-exponential factor; E a is the reaction activation energy; R is a constant; T is the thermodynamic temperature of the energy storage operating environment; z is the power law factor; Cumulative discharge capacity A h As shown below: A h =n·DOD·C in i t (2) Where: n is the number of cycles; DOD is the depth of cycle; C init is the initial available capacity of energy storage.

4. The method for optimizing the dispatching of a power system considering the energy storage state-dependent aging characteristics according to claim 1, characterized in that: The steps of constructing the discrete characteristic expression of energy storage state-dependent aging include: 1) Calculation Right now: Where: C loss is the accumulated aging loss of energy storage; B is the pre-exponential factor; E a is the reaction activation energy; R is a constant; T is the thermodynamic temperature of the energy storage operating environment; z is the power law factor; A h is the cumulative discharge amount; 2) Transform formula (1) to obtain: 3) Substituting formula (4) into formula (3), we can obtain the discrete characteristic expression of energy storage state-dependent aging, namely: Where: ΔC loss,n ΔA is the corresponding aging loss when the energy storage receives the dispatch instruction for the nth time since it was put into use; h is the amount of electricity released when energy storage participates in scheduling; C loss,n-1 The accumulated aging loss of energy storage. The amount of electricity released when energy storage participates in scheduling ΔA h As shown below: ΔA h =P dis Δt (6) Where: P dis is the discharge power of the energy storage; Δt is the discharge duration.

5. The method for optimizing and dispatching a power system considering the energy storage state-dependent aging characteristics according to claim 3, characterized in that: Constructed with discharge power P dis The steps for discretizing the Arrhenius empirical model for the decision variables include: 1) Reaction activation energy E a After tuning, we get: HAVE BEEN a =k1+k2C rate (7) Where: C rate is the discharge rate; k1 and k2 are the data collected by Arbin battery test system for lithium iron phosphate batteries at different C rate The coefficients obtained by fitting the cycle life data under ; 2) Taking the logarithm of formula (1), we get: Where: C n is the available capacity of the energy storage when it runs in a complete cycle for the nth time; 3) Take the logarithm of formula (2) and get ln(A h ); 4) Determine the discharge rate and test environment temperature, and obtain the measured data set under standard test conditions, and determine the reaction activation energy E according to the measured data set a , R and T; 5) By shifting the terms in formula (8), we get: 6) Perform linear regression on formula (10), the slope in the linear regression fitting formula is the power law factor z, and the intercept is the logarithm of the pre-exponential factor corresponding to the current discharge rate ln(B); 7) updating the discharge rate and the test environment temperature, returning to step 4), and obtaining the pre-exponential factor B at multiple discharge rates; 8) The pre-exponential factor B and discharge rate C of step 7) are rate Fitting, we get: B=k3+k4C rate (11) In the formula, k3 and k4 are root fitting coefficients; 9) Construct discharge rate C rate The expression is: Where: P dis is the discharge power when energy storage participates in scheduling, P nom is the energy storage rated power; 10) Substituting formulas (6), (7), (11), and (12) into formula (5), we can obtain the discharge power P dis is the discretized Arrhenius empirical model of the decision variables, namely: The discretized Arrhenius empirical model is simplified as follows: ΔC loss,n =λ n P dis (14) Where: ΔC loss,n is the discrete aging loss of energy storage after linear regression; n The cumulative aging loss reaches C loss,n-1 The energy storage loss rate when energy storage participates in scheduling.

6. The method for optimizing and dispatching a power system considering the energy storage state-dependent aging characteristics according to claim 1, characterized in that: The objective function F of the power system optimal dispatch model taking into account the rolling update of energy storage parameters is as follows: Where: F is the total operating cost of the power system, including the operating cost of conventional units, wind and solar power abandonment penalties, energy storage operating costs, energy storage aging costs and load shedding penalties; a i Quote for unit i; is the power of unit i in period t; ρ w Penalty fee for unit abandoned wind volume, and is the maximum power that can be generated and the actual power of wind farm k in period t; is the maximum power that can be generated and the actual power of the photovoltaic power station r in time period t; and is the energy storage charging and discharging quotation for period t, and is the charging and discharging power of the energy storage station m in period t, E m is the available capacity of the energy storage station m, is the energy storage discrete aging loss corresponding to the energy storage station m when it is running in period t; Γ is the unit capacity construction cost of energy storage; π is the unit load shedding penalty of the power system, is the load shedding amount of node j in period t; N, N w 、N pv 、N B and N L are the number of conventional units, wind farms, photovoltaic power stations, energy storage power stations and connected load nodes respectively; Δt is the time interval, T is the number of time periods; ρ pv It is the penalty fee for the amount of abandoned light per unit.

7. A method for optimizing and dispatching a power system considering the energy storage state-dependent aging characteristics according to claim 6, characterized in that: The energy storage discrete aging loss corresponding to the energy storage station m when it is running in time period t Calculated by discretizing the Arrhenius empirical model.

8. The method for optimizing and dispatching a power system considering the energy storage state-dependent aging characteristics according to claim 1, characterized in that: The constraints of the power system optimization dispatch model taking into account the rolling update of energy storage parameters include battery energy storage operation constraints, distributed renewable energy operation constraints, conventional unit operation constraints, and power system constraints. Among them, battery energy storage operation constraints include charge and discharge power constraints, energy conservation constraints, and state of charge constraints; Conventional unit operation constraints include power regulation constraints, ramp constraints, and start / stop time constraints; Power system constraints include power balance constraints, flow constraints and load shedding constraints.

9. A method for optimizing and dispatching a power system considering the energy storage state-dependent aging characteristics according to claim 8, characterized in that: The upper and lower limits of charge and discharge power are as follows: In the formula, and is the upper limit of the charging and discharging power of the energy storage station m in period t; and is the charging and discharging power of the energy storage station m in period t; and is the charging and discharging power of the energy storage station m in period t; The energy conservation constraints are as follows: Where: is the charge state of the energy storage station m in period t; η ch and η dis The charging and discharging efficiency of the energy storage power station; and is the charging and discharging power of the energy storage station m in period t, E m is the available capacity of energy storage power station m; Δt is the time interval; is the charge state of the energy storage station m at time t-1; The state of charge constraints are as follows: Where: and are the lower and upper limits of the state of charge of the energy storage station m; The operating constraints of distributed renewable energy are as follows: Where: and is the maximum power that can be generated and the actual power of wind farm k in period t; is the maximum power that can be generated and the actual power of the photovoltaic power station r in time period t; The power regulation constraints are as follows: Where: and is the output P of unit i during period t i t The lower and upper limits of Indicates the operating status of unit i in period t, 1 indicates that the unit is running, and 0 indicates that the unit is shut down; The hill climbing constraints are as follows: R i,down ≤P i t -P i t-1 ≤R i,up (23) Where: R i,down and r i,up is the upper and lower limits of the climbing slope of unit i; P i t-1 is the output of unit i during period t-1; P i t is the output of unit i during period t; The start and stop time constraints are as follows: Where: ts and to are the minimum start and stop time of the unit; Indicates the operating status of unit i in period k; Indicates the operating status of unit i in period t-1; The power balance constraints are as follows: Where: is the load demand of node j in period t; N, N w 、N pv 、N B and N L They are the number of conventional units, wind farms, photovoltaic power stations, energy storage power stations and connected load nodes; Charge and discharge power for energy storage power stations; P i t Provide power for wind farms, photovoltaic power stations, and conventional units; The power flow constraints are as follows: Where: G x-i represents the impact of the injected power of node i on line x; G x-j represents the impact of the injected power of node j on line x; are the maximum and minimum transmission capacities of branch l respectively; The load shedding constraints are as follows: Where: is the load shedding amount of node j in period t.

10. The method for optimizing and dispatching a power system considering the energy storage state-dependent aging characteristics according to claim 1, characterized in that: In the process of solving the optimal dispatch model of the power system taking into account the rolling update of energy storage parameters, the available capacity E of energy storage m , cumulative aging loss C loss,n and energy storage loss rate λ n All are in a dynamic update state; Energy storage available capacity E m , cumulative aging loss C loss,n and energy storage loss rate λ n The update process is as follows: 1) Update the accumulated aging loss C loss,n ,Right now: C loss,n =C loss,n-1 +ΔC loss,n (29) 2) Update the accumulated aging loss C loss,n Substituting the discretized Arrhenius empirical model, the accumulated aging loss is calculated as C in the theoretical discharge power range of energy storage. loss,n The corresponding energy storage discrete aging loss is 3) Discrete energy storage into aging losses Substitute into formula (14) and update the energy storage loss rate λ n+1 ; 4) Based on constraint (18), update the available energy storage capacity E m .

Images