Optimization Control Method for Suppressing Wind Power Fluctuations in Multi-Battery Energy Storage Systems

Abstract

The present invention relates to an optimal control method for suppressing wind power fluctuations in a multi-battery energy storage system, including: establishing a cost model, a rejection coefficient model, and an SoC recovery model for the energy storage power station. On this basis, taking the minimum of the total operating cost, the rejection coefficient, the SoC recovery factor, and the penalty term for wind power fluctuation exceeding the limit of the multi-battery energy storage system as the objective function; establishing model constraints; combining the objective function and the model constraints to obtain an optimal control model for suppressing wind power fluctuations in the multi-battery energy storage system; solving the optimal control model for suppressing wind power fluctuations in the multi-battery energy storage system according to the following steps: Step S1, convexifying the rejection coefficient calculation formula by using the McCormick method; Step S2, convexifying the objective function by using the McCormick method; Step S3, solving the optimal control model for suppressing wind power fluctuations in the multi-battery energy storage system; Step S4, determining whether the iteration termination condition is satisfied. If so, stop the iteration and output the optimal control result. Otherwise, update the limit values of the rejection coefficient and the energy storage cost, and then return to Step S2 to continue the iteration. This method is beneficial to reducing the operating cost of the energy storage power station and improving the charge and discharge capacity of the energy storage power station.

Description

Technical Field

[0001] The present invention belongs to the technical field of wind power grid connection, and particularly relates to an optimized control method for suppressing wind power fluctuations in a multi-battery energy storage system. Background Technique

[0002] Affected by the natural environment, the power fluctuations of high-proportion wind power grid connection are relatively large, seriously affecting the frequency and voltage stability of the power system, and bringing huge challenges to the safe and stable operation of the power system. Under this background, a battery energy storage system with the characteristic of storing energy can effectively solve the problem of power fluctuations in high-penetration wind power grid connection. Among them, compared with a single-battery energy storage system that needs to frequently switch the charge and discharge states, a multi-type battery energy storage system containing multiple units has the advantages of improving the output of energy storage devices and extending the service life of energy storage devices, and can participate in the process of suppressing wind power fluctuations more efficiently. However, the output plan of each battery energy storage device in the control system will affect the operation performance of the energy storage device. Therefore, when formulating a control strategy for suppressing wind power fluctuations in a multi-battery energy storage system, a model that can not only optimize the operation performance of the energy storage system but also be easy to solve needs to be established.

[0003] At present, for the optimized control method of suppressing wind power fluctuations in a multi-battery energy storage system, there are mainly a power dynamic proportional distribution control method and an optimal economic distribution control method. The power dynamic proportional distribution control method distributes power according to the proportion of the charge and discharge capacity of each energy storage device, without considering the differences in the operating costs of different energy storage devices, and does not have good economy. The optimal economic distribution control method only takes the operating cost of the energy storage system as the optimal goal, only considering the influence of the economy of the energy storage device, which may cause the energy storage power station with a higher operating cost to hardly output power, seriously hitting its output enthusiasm. At the same time, this control method fails to consider whether the state of charge (SoC) of the energy storage at the current moment can meet the output at the future moment, and cannot guarantee the sustainability of the output of the energy storage system.

[0004] In addition, the current research in the field of multi-battery energy storage system control for suppressing wind power fluctuations is more inclined to decentralized control. Often, a single battery energy storage power station containing multiple units is used as the control object, lacking the cooperative control technology for multiple different types of battery energy storage power stations in the regional power grid.

[0005] According to the above analysis, the current multi-battery energy storage control strategy for suppressing wind power fluctuations cannot comprehensively consider the operating costs and output enthusiasm of each energy storage power station while meeting the output of the energy storage power station at the future moment. Summary of the Invention

[0006] The purpose of the present invention is to provide an optimized control method for suppressing wind power fluctuations in a multi-battery energy storage system, which is beneficial to reducing the operating cost of the energy storage power station and at the same time improving the charge and discharge capacity of the energy storage power station.

[0007] To achieve the above object, the technical solution adopted by the present invention is: a method for optimizing the control of suppressing wind power fluctuations in a multi-battery energy storage system, including:

[0008] Establish an energy storage power station cost model, a rejection coefficient model, and an SoC recovery model. On this basis, take the total operating cost, rejection coefficient, SoC recovery factor, and wind power fluctuation over-limit penalty term of the multi-battery energy storage system as the objective function; establish model constraints; combine the objective function and model constraints to obtain an optimization control model for suppressing wind power fluctuations in the multi-battery energy storage system;

[0009] Solve the optimization control model for suppressing wind power fluctuations in the multi-battery energy storage system according to the following steps:

[0010] Step S1: Convexify the rejection coefficient calculation formula by using the McCormick method;

[0011] Step S2: Convexify the objective function by using the McCormick method;

[0012] Step S3: Solve the optimization control model for suppressing wind power fluctuations in the multi-battery energy storage system;

[0013] Step S4: Determine whether the iteration termination condition is satisfied. If so, stop the iteration and output the optimization control result. Otherwise, update the limit values of the rejection coefficient and energy storage cost, and then return to Step S2 to continue the iteration.

[0014] Furthermore, the energy storage power station cost model ignores the line losses caused by long-distance power transmission, and divides the cost into initial construction depreciation cost, energy consumption cost, and life loss cost. Its calculation formula is:

[0015]

[0016]

[0017]

[0018]

[0019] In the formula, C i (t) represents the total output cost of the i-th energy storage power station at time t, are respectively the initial construction depreciation cost, energy consumption cost, and life loss cost of the i-th energy storage power station at time t; c cap,i represents the unit capacity cost of the i-th energy storage power station; E rate,i is the rated capacity of the i-th energy storage power station; T float,i is the floating charge life of power station i; r is the discount rate; s is the maximum dispatching times in a day; c eis the feed-in tariff, without considering the impact of time-of-use electricity price on frequency regulation cost; Δt is the sampling interval between two sampling points; P char,i (t) and P dis,i (t) are the charging power and discharging power of the i-th energy storage power station at time t; η i c and η i d are the charging and discharging efficiencies of the power station; c FR,i and P rate,i are respectively the unit power cost and rated power of the i-th energy storage power station; N 0,i is the equivalent cycle number of energy storage power station i at 100% charge-discharge depth; k p is an empirical value, representing the value obtained by fitting the relationship between the energy storage cycle number and the charge-discharge depth based on the actual operation information of the energy storage device.

[0020] Furthermore, the resistance coefficient model promotes the participation of power stations with high costs and strong charge-discharge capabilities in suppressing fluctuations through the resistance coefficient REF i (t), and reduces the output of power stations with low costs and weak charge-discharge capabilities; the resistance coefficient represents the difference between the internal unit SoC of the energy storage power station at this moment and its upper and lower limit values, providing an evaluation or reference for the charge-discharge capabilities of the power station at the next moment; the resistance coefficient REF i (t) is divided into the charging resistance coefficient and the discharging resistance coefficient and its calculation formula is:

[0021]

[0022] REF i (t) = v(t)·REF i char (t) + u(t)·REF i dis (t) (6)

[0023] In the formula, J represents the number of energy storage units in each energy storage power station; SoC i,j (t) is the state of charge of the j-th unit of the i-th energy storage power station at time t, SoC max and SoC min are its maximum and minimum values.

[0024] Furthermore, the SoC recovery model quantifies the recovery degree of the energy storage unit SoC to the ideal state SoC ideal through the SoC recovery factor, and its calculation formula is:

[0025]

[0026] In the formula, It is the SoC recovery factor, representing the degree of SoC recovery of the j-th unit of the i-th energy storage power station.

[0027] Furthermore, based on the established cost model C i (t) of the energy storage power station, the resistance coefficient model REF i (t), and the SoC recovery model Comprehensively considering the operation economy, output enthusiasm, SoC recovery characteristics of each energy storage power station, and the impact of wind power fluctuation exceeding the limit, the objective function of the optimal control model for suppressing wind power fluctuation of the multi-battery energy storage system is constructed as follows:

[0028]

[0029] In the formula, N is the number of time sampling points, I is the number of energy storage power stations, ω is the penalty coefficient for wind power fluctuation exceeding the limit, and δ is the introduced slack variable.

[0030] Furthermore, the established model constraints are as follows:

[0031] Equations to are the constraints related to the SoC of the energy storage power station; Equation is the change range of the SoC of the j-th unit of the i-th energy storage power station; Equation is the calculation expression of the SoC value of a single energy storage unit at adjacent moments;

[0032] SoC min,i,j ≤SoC i,j (t)≤SoC max,i,j (9)

[0033]

[0034] In the formula, SoC i,j (t) is the state of charge of the j-th unit of the i-th energy storage power station at time t, SoC max,,i,j and SoC min,,i,j are its maximum and minimum values; P char,i,j (t) and P dis,i,j (t) are the charge and discharge powers of the j-th unit of the i-th energy storage power station at time t;

[0035] Equations - are the charge and discharge output constraints of the energy storage power station; Equation represents the output model of the energy storage unit; Equation represents the relationship between the output of the energy storage unit and the power station; Equation represents the relationship between the output of a single energy storage power station and the charge and discharge powers, where P ESS,i (t) represents the output of the i-th energy storage power station at time t, and its positive and negative values represent discharge and charge; Equation is the total output of all power stations at time t;

[0036]

[0037]

[0038] P ESS,i $(t)=P dis,i $(t)-P char,i $(t)(13)

[0039]

[0040] wherein, P rate,i,j is the rated power of the j-th unit of the i-th energy storage power station; introduce 0-1 variables v(t) and u(t) to represent the charge and discharge states of the j-th energy storage unit of the i-th energy storage power station at time t. When v(t)=1, the energy storage power station is charging. When u(t)=1, the energy storage power station is discharging. When both are equal to zero, the energy storage unit does not output power;

[0041] Equation - is the wind power grid connection constraint; Equation represents the power balance relationship, where P G $(t) is the wind power grid connection power at time t, and P WT $(t) is the original wind power output at time t; Equation represents that the grid connection power fluctuation does not exceed the value specified in the grid connection standard. In the equation, γ is the fluctuation limit value; considering that the maximum output of the energy storage cannot meet the requirements of the smoothing standard at some moments, that is, when Equation cannot hold at a certain moment, a slack variable δ is introduced in Equation (17) to ensure the feasibility of model solution; at the same time, a fluctuation over-limit penalty term with a coefficient greater than the set value is added to the objective function; when the constraint is satisfied, the penalty term is 0 to ensure that the slack tightens; when the fluctuation over-limit penalty term in the objective function is non-zero, it indicates that the wind power grid connection fluctuation standard cannot be met at this moment; Equation (18) represents the value range of the slack variable δ;

[0042] P G $(t)=P ESS $(t)+P WT $(t)(15)

[0043] -γ≤P G,i $(t)-P G,i $(t - 1)≤γ(16)

[0044] -(γ + δ)≤P G,i $(t)-P G,i $(t - 1)≤γ + δ(17)

[0045] 0≤δ≤0.4γ(18).

[0046] Furthermore, the specific method of convexifying the objective function using the McCormick method is as follows:

[0047] There are two continuous variables REF i $(t), C i $(t) multiplying the non-linear term, and the tightened McCormick method is used for convexification;

[0048] Assume that x and y are two continuous variables, and x ∈ [x l , x u , y ∈ [y l , y u , then the expression of the McCormick envelope of the product term of the two continuous variables is:

[0049]

[0050] In the formula, x l , x u and y l , y u are the lower and upper limits of x and y respectively;

[0051] For the objective function F, based on the McCormick method, equation (19) is used to linearize the product term in the objective function formula, and the specific expression is as follows:

[0052]

[0053] In the formula, W i (t) is the introduced auxiliary variable, which is used to replace the product term in the original formula; and are the upper and lower limits of REF i (t) respectively, and are the upper and lower limits of C i (t) respectively;

[0054] The McCormick method is that the four hyperplanes formed by the formula wrap the product term REF i (t)·C i (t) in a tetrahedron, and a certain value in the tetrahedron space is used to approximately replace the true value; the McCormick method is tightened through an iterative algorithm to improve the approximation accuracy; it includes the following steps:

[0055] 1) Initialization: Set the maximum number of iterations T soc , the iteration number index γ soc = 1, the error gap e soc , the strictly decreasing sequence {ξ k};

[0056] 2) Add the formula to the model constraints, solve the optimal control model for suppressing wind power fluctuations of the multi-battery energy storage system, and obtain the values of the resistance coefficient REF i and the energy storage cost C i ;

[0057] 3) Determine whether to stop iteration: If the differences between the adjacent resistance coefficients and the energy storage costs are both less than the error margin e soc , or the iteration reaches the maximum number of iterations T soc , exit the iteration process; otherwise, execute step 4);

[0058] 4) Tighten the constraint ranges of the resistance coefficient REF i and the energy storage cost C i , update the iteration number index γ soc , and return to step 2); The specific calculation formula is as follows:

[0059]

[0060]

[0061]

[0062]

[0063] After several iterations of the above-mentioned tightened McCormick method, the strictly decreasing sequence {ξ k} tends to 0. At the same time, the constraint ranges of the resistance coefficient REF i and the energy storage cost C i also approach 0, thereby tightening McCormick so that W i (t) = REF i (t)·C i (t), thus reducing the approximation error and improving the accuracy.

[0064] Furthermore, the specific method for convexifying the resistance coefficient calculation formula by using the McCormick method is as follows:

[0065] For the non-linear terms in the formula, that is, the product terms of the 0-1 variable v(t) and the continuous variable or the product terms of v(t) and REF(t), the McCormick method is used to achieve the linearization process;

[0066] Assume that x1 is a 0-1 variable, x2 is a continuous variable, and x2 ∈ [0, u]. The McCormick method is used to linearize the product term x1x2; The specific process is as follows:

[0067]

[0068] In the formula, α is the introduced auxiliary variable used to replace the non-linear term x1x2;

[0069] Using the above method to process the formula, for the following constraints are obtained:

[0070]

[0071] ξ(t) ≤ 1·v(t) (27)

[0072] ξ(t) ≤ REF i char (t) (28)

[0073] ξ(t) ≥ REF i char (t) - 1·(1 - v(t)) (29)

[0074] Wherein, ξ(t) is the introduced auxiliary variable, representing

[0075] Similarly,[[]]END]] It is also processed by the above method.

[0076] Compared with the prior art, the present invention has the following beneficial effects: It provides an optimized control method for suppressing wind power fluctuations in a multi - battery energy storage system. On the basis of effectively suppressing wind power fluctuations, this method comprehensively considers the impact of energy storage costs, the output enthusiasm of energy storage power stations, and the SoC recovery characteristics on the operation performance of the multi - battery energy storage system. First, on the basis of considering the safety operation constraints of the energy storage power station and the wind power grid - connection constraints, with the lowest total operation cost of multiple battery energy storage power stations with different output costs as the objective function, a multi - battery energy storage system model for suppressing wind power fluctuations is constructed, which reduces the operation cost of the energy storage power station while suppressing wind power fluctuations; Secondly, a resistance coefficient representing the charging (discharging) willingness of the energy storage power station is introduced into the objective function to promote the participation of higher - cost energy storage power stations in the suppression process and enhance the overall charge - discharge continuity of multiple energy storage power stations; Finally, to relieve the pressure of the energy storage power station to suppress wind power fluctuations at future moments, an SoC recovery factor is added to the objective function to further consider the impact of the SoC of each unit in the power station at the current moment on the output at future moments, improving the continuity and sustainability of energy storage output control; In addition, the tightened McCormick method is used to convexify the non - linear terms in the objective function of the multi - energy storage model, improving the solution accuracy and convergence reliability of the model. Brief Description of the Drawings

[0077] Figure 1 It is a block diagram of a multi - battery energy storage system for suppressing wind power fluctuations in an embodiment of the present invention;

[0078] Figure 2 It is a flowchart of the implementation of the optimized control method for suppressing wind power fluctuations in a multi - battery energy storage system in an embodiment of the present invention. Detailed Embodiment

[0079] The present invention will be further described below with reference to the drawings and embodiments.

[0080] It should be noted that the following detailed description is exemplary and is intended to provide further illustration of the present application. Unless otherwise specified, all technical and scientific terms used herein have the same meaning as commonly understood by those of ordinary skill in the technical field to which this application belongs.

[0081] It should be noted that the terms used herein are only for describing specific embodiments and are not intended to limit the exemplary embodiments according to the present application. As used herein, unless the context clearly indicates otherwise, the singular form is also intended to include the plural form. In addition, it should be understood that when the terms "comprising" and / or "including" are used in this specification, they specify the presence of features, steps, operations, devices, components, and / or combinations thereof.

[0082] As Figure 1-2 shown, for a multi-battery energy storage system for suppressing wind power fluctuations, the present embodiment provides an optimized control method for suppressing wind power fluctuations in a multi-battery energy storage system, including:

[0083] Model establishment: Establish a cost model of the energy storage power station, a rejection coefficient model, and an SoC recovery model. On this basis, take the total operating cost, rejection coefficient, SoC recovery factor, and wind power fluctuation over-limit penalty term of the multi-battery energy storage system as the objective function; establish model constraints; combine the objective function and model constraints to obtain an optimized control model for suppressing wind power fluctuations in the multi-battery energy storage system.

[0084] Model solution: Solve the optimized control model for suppressing wind power fluctuations in the multi-battery energy storage system according to the following steps:

[0085] Step S1: Use the McCormick method to convexify the rejection coefficient calculation formula;

[0086] Step S2: Use the McCormick method to convexify the objective function;

[0087] Step S3: Solve the optimized control model for suppressing wind power fluctuations in the multi-battery energy storage system;

[0088] Step S4: Determine whether the iteration termination condition is satisfied. If so, stop the iteration and output the optimized control result. Otherwise, update the limits of the rejection coefficient and energy storage cost, and then return to Step S2 to continue the iteration.

[0089] 1. Objective function

[0090] In the present embodiment, in order to simplify the cost model of the multi-type battery energy storage power station, the line loss caused by long-distance power transmission is ignored, and the cost is divided into initial construction depreciation cost, energy consumption cost, and life loss cost. The calculation formula of the cost model of the energy storage power station is:

[0091]

[0092]

[0093]

[0094]

[0095] In the formula, C i (t) represents the total output cost of the i-th energy storage power station at time t, are respectively the initial construction depreciation cost, energy consumption cost and life loss cost of the i-th energy storage power station at time t; c cap,i represents the unit capacity cost of the i-th energy storage power station; E rate,i is the rated capacity of the i-th energy storage power station; T float,i is the floating charge life of power station i, which is related to the battery model; r is the discount rate, taking 0.8; s is the maximum dispatching times in one day, taking 1440; c e is the on-grid electricity price, taking 520 yuan / MWh, without considering the impact of time-of-use electricity price on the frequency regulation cost; Δt is the sampling interval between two sampling points, taking 15 minutes; P char,i (t) and P dis,i (t) are the charging power and discharging power of the i-th energy storage power station at time t; η i c 、η i d are the charge and discharge efficiencies of the power station; c FR,i 、P rate,i are respectively the unit power cost and rated power of the i-th energy storage power station; N 0,i is the equivalent cycle number of energy storage power station i at 100% charge and discharge depth; k p is an empirical value, representing the value obtained by fitting the relationship between the energy storage cycle number and the charge and discharge depth based on the actual operation information of the energy storage device, with the value range of 0.8 - 2.1, taking 1 here.

[0096] If only considering the optimal economy of the energy storage, it may lead to the situation that the power stations with low energy storage operation cost and weak remaining charge (discharge) capacity of the energy storage continuously participate in the process of suppressing wind power fluctuations, while the power stations with high energy storage operation cost and strong remaining charge (discharge) capacity of the energy storage have a very small output proportion, or even do not output. This behavior cannot take into account the overall charge and discharge continuity of multiple energy storage power stations, greatly dampens the output enthusiasm of the energy storage power stations with high operation cost, and even leads to the situation that some power stations are idle for a long time, seriously affecting the output distribution among power stations. Based on this, the resistance coefficient model promotes the participation of power stations with high cost and strong charge and discharge capacity in suppression through the resistance coefficient REF i (t), and reduces the output of power stations with low cost and weak charge and discharge capacity.

[0097] The resistance coefficient represents the difference between the internal unit SoC of the energy storage power station at this moment and its upper (lower) limit value, providing an evaluation or reference for the charging (discharging) capacity of the power station at the next moment. The resistance coefficient REF i (t) is divided into the charging resistance coefficient and the discharging resistance coefficient The stronger the charging (discharging) capacity of the energy storage power station, the stronger its charging (discharging) willingness, and the greater the discharging (charging) resistance coefficient, and vice versa. The calculation formula of the resistance coefficient is:

[0098]

[0099] REF i (t) = v(t)·REF i char (t) + u(t)·REF i dis (t) (6)

[0100] In the formula, J represents the number of energy storage units in each energy storage power station; SoC i，j (t) is the state of charge of the jth unit of the ith energy storage power station at time t, SoC max and SoC min are its maximum and minimum values.

[0101] The SoC recovery model quantifies the recovery degree of the SoC of the energy storage unit to the ideal state SoC ideal through the SoC recovery factor, and adds the SoC recovery factor to the objective function. The calculation formula of the SoC recovery model is:

[0102]

[0103] In the formula, is the SoC recovery factor, indicating the SoC recovery degree of the jth unit of the ith energy storage power station.

[0104] Based on the established energy storage power station cost model C i (t), resistance coefficient model REF i (t) and SoC recovery model The present invention comprehensively considers the operation economy, output enthusiasm, SoC recovery characteristics of each energy storage power station and the influence of wind power fluctuation exceeding the limit, and constructs the objective function of the multi-battery energy storage system for suppressing wind power fluctuation and optimizing control model:

[0105]

[0106] In the formula, N is the number of time sampling points, I is the number of energy storage power stations, ω is the penalty coefficient for wind power fluctuation exceeding the limit, and δ is the introduced slack variable.

[0107] 2. Constraints

[0108] In this embodiment, the constraints of the established model are as follows.

[0109] Equations to are the constraints related to the SoC of the energy storage power station; Equation is the change range of the SoC of the j-th unit of the i-th energy storage power station; Equation is the calculation expression of the SoC value of a single energy storage unit at adjacent moments;

[0110] SoC min,i,j ≤SoC i,j (t) ≤ SoC max,i,j (9)

[0111]

[0112] In the formula, SoC i,j (t) is the state of charge of the j-th unit of the i-th energy storage power station at time t, SoC max,,i,j and SoC min,,i,j are its maximum and minimum values; P char,i,j (t) and P dis,i,j (t) are the charge and discharge powers of the j-th unit of the i-th energy storage power station at time t.

[0113] Equations - are the charge and discharge output constraints of the energy storage power station; Equation represents the output model of the energy storage unit; Equation represents the relationship between the output of the energy storage unit and the power station; Equation represents the relationship between the output of a single energy storage power station and the charge and discharge powers, where P ESS,i (t) represents the output of the i-th energy storage power station at time t, and its positive and negative values indicate discharge and charge; Equation is the total output of all power stations at time t;

[0114]

[0115]

[0116] P ESS,i (t) = P dis,i (t) - P char,i (t) (13)

[0117]

[0118] In the formula, P rate,i,j is the rated power of the j-th unit of the i-th energy storage power station; In the present invention, 0-1 variables v(t) and u(t) are introduced to represent the charge and discharge states of the j-th energy storage unit of the i-th energy storage power station at time t. When v(t) = 1, the energy storage power station is charging. When u(t) = 1, the energy storage power station is discharging. When both are equal to zero, the energy storage unit does not output power.

[0119] Equation - represents the constraints for wind power grid connection; the equation represents the power balance relationship, where P G (t) is the wind power grid - connected power at time t, and P WT (t) is the original output of wind power at time t; the equation represents that the grid - connected power fluctuation does not exceed the value specified in the grid - connection standard. In the equation, γ is the fluctuation limit value; considering that the maximum output of energy storage cannot meet the requirements of suppressing the standard at some moments, that is, when the equation cannot hold at a certain moment, a slack variable δ is introduced in Equation (17) to ensure the feasibility of model solution; at the same time, a penalty term for excessive fluctuation with a coefficient greater than the set value is added to the objective function; when the constraints are met, the penalty term is 0 to ensure the relaxation tightens; when the penalty term for excessive fluctuation in the objective function is non - zero, it indicates that the wind power grid - connection fluctuation standard cannot be met at this moment; Equation (18) represents the value range of the slack variable δ.

[0120] P G (t)=P ESS (t)+P WT (t)(15)

[0121] - γ≤P G,i (t)-P G,i (t - 1)≤γ(16)

[0122] -(γ + δ)≤P G,i (t)-P G,i (t - 1)≤γ + δ(17)

[0123] 0≤δ≤0.4γ(18)

[0124] 3. Model solution

[0125] The non - linearity in the optimal control model for suppressing wind power fluctuations by a multi - battery energy storage system comes from: the objective function and the calculation formula of the total resistance coefficient. To solve the problem of difficulty in solving non - linear equations, the present invention mainly uses the tightened McCormick method to achieve the linearization process.

[0126] (1) Method for linearizing the objective function

[0127] In this embodiment, the specific method of using the McCormick method to convexify the objective function is as follows:

[0128] There is a non - linear term in the equation where two continuous variables REF i (t), C i (t) are multiplied, and the tightened McCormick method is used for convexification.

[0129] Assume that x and y are two continuous variables, and x∈[x l ,x u and y∈[y l ,y u, the expression of the McCormick envelope for the product term of two continuous variables is as follows:

[0130]

[0131] In the formula, x l , x u and y l , y u are the lower and upper limits of x and y, respectively.

[0132] For the objective function F, based on the McCormick method, equation (19) is used to linearize the product term in the objective function formula. The specific expression is as follows:

[0133]

[0134] In the formula, W i (t) is the introduced auxiliary variable, which is used to replace the product term in the original formula; and are the upper and lower limits of REF i (t), and are the upper and lower limits of C i (t), respectively.

[0135] The McCormick method can be understood as that the four hyperplanes formed by the formula wrap the product term REF i (t)·C i (t) in a tetrahedron, and a certain value in the tetrahedron space is used to approximately replace the true value. However, due to the large difference between the upper and lower limits of REF i (t) and C i (t), the formed McCormick space is large, and the uncertainty of the product substitution value is large, resulting in the insufficient accuracy of the substituted result. Therefore, it is necessary to tighten McCormick to improve the approximation accuracy. The approximation accuracy depends on the tightening degree of McCormick, that is, narrowing the constraint ranges of the resistance coefficient and the cost. In this embodiment, the McCormick method is tightened by an iterative algorithm to improve the approximation accuracy. The detailed process is as follows:

[0136] 1) Initialization: Set the maximum number of iterations T soc , the iteration number index γ soc =1, the error gap e soc , and the strictly decreasing sequence {ξ k}}.

[0137] 2) Add the formula to the model constraints, solve the optimal control model for suppressing wind power fluctuations of the multi-battery energy storage system, and obtain the resistance coefficient REF i and the energy storage cost C iValue.

[0138] 3) Determine whether to stop iteration: If the differences between the adjacent two resistance coefficients and the energy storage costs are both less than the error gap e soc , or the iteration reaches the maximum number of iterations T soc , exit the iteration process; otherwise, execute step 4).

[0139] 4) Tighten the constraint ranges of the resistance coefficient REF i and the energy storage cost C i , update the iteration number index γ soc , and return to step 2); the specific calculation formula is as follows:

[0140]

[0141]

[0142]

[0143]

[0144] After several iterations of the above-mentioned tightened McCormick method, the strictly decreasing sequence {ξ k} tends to 0, and at the same time, the constraint ranges of the resistance coefficient REF i and the energy storage cost C i also approach 0, so as to tighten McCormick and make W i (t) = REF i (t)·C i (t), thereby reducing the approximation error and improving the accuracy.

[0145] (2) Linearization method for the resistance coefficient calculation formula

[0146] In this embodiment, the specific method for convexifying the resistance coefficient calculation formula by using the McCormick method is as follows:

[0147] For the non-linear terms in the formula, that is, the product terms of the 0-1 variable v(t) and the continuous variable (or the product terms of v(t) and ), the present invention uses the McCormick method to implement the linearization process.

[0148] Assume that x1 is a 0-1 variable, x2 is a continuous variable, and x2 ∈ [0, u], and use the McCormick method to linearize the product term x1x2; the specific process is as follows:

[0149]

[0150] Wherein, α is an introduced auxiliary variable used to replace the non-linear term x1x2.

[0151] Using the above method to process the formula, taking... as an example, the following constraints are obtained:

[0152]

[0153] ξ(t) ≤ 1·v(t) (27)

[0154] ξ(t) ≤ REF i char (t) (28)

[0155] ξ(t) ≥ REF i char (t) - 1·(1 - v(t)) (29)

[0156] Wherein, ξ(t) is an introduced auxiliary variable, representing

[0157] Similarly, is also processed by the above method.

[0158] In view of the differences in the dispatching costs and output enthusiasm of multiple different types of battery energy storage power stations during the process of suppressing wind power fluctuations, as well as the problems of too high (low) SOC of energy storage units, the present invention takes the minimum of the total operating cost, resistance coefficient, SoC recovery factor, and wind power fluctuation over-limit penalty term of the multi-battery energy storage system as the objective function, and at the same time considers the safety constraints of the energy storage power station and the wind power grid connection constraints to construct an optimized control model for suppressing wind power fluctuations of the multi-battery energy storage system. In addition, during the process of solving the optimized control model for suppressing wind power fluctuations of the multi-battery energy storage system, the tightened McCormick method is used to convexify the model to ensure that the model is easy to solve.

[0159] The above are only the preferred embodiments of the present invention, and are not limitations on the present invention in other forms. Any person skilled in the art may use the disclosed technical content to make changes or modifications into equivalent embodiments with equivalent changes. However, any simple modifications, equivalent changes, and modifications made to the above embodiments based on the technical essence of the present invention without departing from the technical solution content of the present invention still fall within the protection scope of the technical solution of the present invention.

Claims

1. An optimal control method for suppressing the fluctuations of wind power in a multi-battery energy storage system, characterized in that Including: Establish a cost model of the energy storage power station, a resistance coefficient model and an SoC recovery model. On this basis, take the minimum of the total operating cost, resistance coefficient, SoC recovery factor and wind power fluctuation over-limit penalty term of the multi-battery energy storage system as the objective function; establish model constraints; combine the objective function and model constraints to obtain an optimal control model for suppressing wind power fluctuations in the multi-battery energy storage system; Solve the optimal control model according to the following steps: Step S1: Use the McCormick method to convexify the resistance coefficient calculation formula; Step S2: Use the McCormick method to convexify the objective function; Step S3: Solve the optimal control model; Step S4: Determine whether the iteration termination condition is satisfied. If so, stop the iteration and output the optimal control result. Otherwise, update the limit values of the resistance coefficient and energy storage cost, and then return to Step S2 to continue the iteration; The cost model of the energy storage power station ignores the line losses caused by long-distance power transmission, and divides the cost into initial construction depreciation cost, energy consumption cost and life loss cost. Its calculation formula is: (1) (2) (3) (4) C i (t) represents the total output cost of the i-th energy storage power station at time t, , , are respectively the initial construction depreciation cost, energy consumption cost and life loss cost of the i-th energy storage power station at time t; c cap,i represents the unit capacity cost of the i-th energy storage power station; E rate,i is the rated capacity of the i-th energy storage power station; T float,i is the floating charge life of power station i; r is the discount rate; s is the maximum number of dispatches in one day; c e is the on-grid electricity price, without considering the impact of time-of-use electricity price on frequency regulation cost; Δt is the sampling interval between two sampling points; P char,i (t) and P dis,i (t) are the charging power and discharging power of the i-th energy storage power station at time t; η i c and η i d are the charging and discharging efficiencies of the power station; c FR,i and P rate,i are respectively the unit power cost and the rated power of the i-th energy storage power station; N 0,i is the equivalent cycle number of the energy storage power station i at 100% charge-discharge depth; k p is an empirical value, representing the value obtained by fitting the relationship between the energy storage cycle number and the charge-discharge depth based on the actual operation information of the energy storage device; The resistance coefficient model promotes the participation of power stations with high costs and strong charge-discharge capabilities in suppressing fluctuations through the resistance coefficient REF i (t), and reduces the output of power stations with low costs and weak charge-discharge capabilities; the resistance coefficient is the difference between the internal unit SoC of the energy storage power station at this moment and its upper and lower limits, providing an evaluation or reference for the charge-discharge capability of the power station at the next moment; the resistance coefficient REF i (t) is divided into the charging resistance coefficient and the discharging resistance coefficient , and its calculation formula is as follows: (5) (6) Wherein, J represents the number of energy storage units in each energy storage power station; SoC i,j (t) is the state of charge of the j-th unit of the i-th energy storage power station at time t, SoC max and SoC min are its maximum and minimum values; The SoC recovery model quantifies the recovery degree of the energy storage unit's SoC to the ideal state SoC through the SoC recovery factor, and its calculation formula is as follows: ideal The calculation formula is: (7) It is the SoC recovery factor, representing the SoC recovery degree of the j-th unit of the i-th energy storage power station.

Images