CN104485681B

CN104485681B - A kind of monitoring method of wind energy turbine set energy-storage system

Info

Publication number: CN104485681B
Application number: CN201510002860.3A
Authority: CN
Inventors: 董亮; 汪振东; 段卓华; 艾尼瓦·克然木; 陈晓云; 薛建德; 李冠龙; 赵彦文
Original assignee: Electric Power Research Institute of State Grid Xinjiang Electric Power Co Ltd
Current assignee: Electric Power Research Institute of State Grid Xinjiang Electric Power Co Ltd
Priority date: 2015-01-05
Filing date: 2015-01-05
Publication date: 2016-12-07
Anticipated expiration: 2035-01-05
Also published as: CN104485681A

Abstract

The invention provides a monitoring method for a wind farm energy storage system. The method can predict the generated power of the wind farm, predict the change of the load, detect the battery capacity of the battery module in real time and obtain the operation status of the distribution network in real time, formulate and Implement the most appropriate control strategy to ensure the stable output power of the wind farm and improve the safety and service life of the energy storage system.

Description

A monitoring method for a wind farm energy storage system

所属技术领域Technical field

本发明涉一种风电场储能系统的监控方法。The invention relates to a monitoring method for a wind farm energy storage system.

背景技术Background technique

近年来，风力发电凭借其绿色环保、资源丰富等优势，得到了世界各国的重视，成为非化石燃料发电的重要来源。但风能具有随机性和间歇性的特点，独立的风力发电系统难以提供稳定、连续的功率输出，波动性较大，直接并入电网必然会影响电力系统的安全稳定运行。因此，从电网安全角度考虑，为风电场引入储能装置来平抑其功率波动，建立风储联合发电系统是未来风力发电的必然趋势。In recent years, wind power has attracted the attention of countries all over the world due to its advantages of environmental protection and abundant resources, and has become an important source of non-fossil fuel power generation. However, wind energy has the characteristics of randomness and intermittentness. It is difficult for an independent wind power generation system to provide stable and continuous power output, and the fluctuation is large. Direct integration into the grid will inevitably affect the safe and stable operation of the power system. Therefore, from the perspective of power grid security, it is an inevitable trend for future wind power generation to introduce energy storage devices to wind farms to stabilize their power fluctuations and to establish wind-storage combined power generation systems.

风储系统是通过储能快速地吸收剩余能量或补充功率缺额来平抑风电场的功率波动的，所以在利用储能系统平抑风电场的功率波动的时候，无法保证对其进行有规律的充放电，容易出现过充过放，这不仅会影响其使用寿命，增加投入成本，而且在功率波动剧烈时可能会使其充放电能力不足，影响风电并网运行的安全。The wind storage system stabilizes the power fluctuation of the wind farm by quickly absorbing the remaining energy or supplementing the power shortage through energy storage. Therefore, when the energy storage system is used to stabilize the power fluctuation of the wind farm, it cannot be guaranteed to be charged and discharged regularly. , prone to overcharging and overdischarging, which will not only affect its service life and increase input costs, but also may make its charging and discharging capacity insufficient when the power fluctuates violently, affecting the safety of wind power grid-connected operation.

风储系统是通过储能快速地吸收剩余能量或补充功率缺额来平抑风电场的功率波动的，所以在利用储能系统平抑风电场的功率波动的时候，无法保证对其进行有规律的充放电，容易出现过充过放，这不仅会影响其使用寿命，增加投入成本，而且在功率波动剧烈时可能会使其充放电能力不足，影响风电并网运行的安全。如果在制定储能系统充放电策略时，加入对其SOC(State of Charge，荷电状态)的额外控制，就可以在平抑风电场功率波动的同时，避免储能系统的过充过放，使其能够长期平滑风电场的输出功率。The wind storage system stabilizes the power fluctuation of the wind farm by quickly absorbing the remaining energy or supplementing the power shortage through energy storage. Therefore, when the energy storage system is used to stabilize the power fluctuation of the wind farm, it cannot be guaranteed to be charged and discharged regularly. , prone to overcharging and overdischarging, which will not only affect its service life and increase input costs, but also may make its charging and discharging capacity insufficient when the power fluctuates violently, affecting the safety of wind power grid-connected operation. If additional control over the SOC (State of Charge) is added to the charge and discharge strategy of the energy storage system, the overcharge and overdischarge of the energy storage system can be avoided while stabilizing the power fluctuations of the wind farm. It can smooth the output power of wind farms in the long term.

发明内容Contents of the invention

本发明提供一种风电场储能系统的监控方法，该方法可以预测风电场的发电功率，预测负载的变化情况，实时检测的蓄电池模块电池容量和实时获取的配电网的运行情况，制定和实施最适宜的控制策略，保障风电场平稳输出功率，提升储能系统的安全性和使用寿命。The invention provides a monitoring method for a wind farm energy storage system. The method can predict the generated power of the wind farm, predict the change of the load, detect the battery capacity of the battery module in real time and obtain the operation status of the distribution network in real time, formulate and Implement the most appropriate control strategy to ensure the stable output power of the wind farm and improve the safety and service life of the energy storage system.

为了实现上述目的，本发明提供一种风电场储能系统的监控方法，该方法基于如下监控装置来实现，该监控装置包括：In order to achieve the above object, the present invention provides a monitoring method for a wind farm energy storage system, which is implemented based on the following monitoring device, which includes:

风电监控模块，用于实时监控风电模块，并对风电模块的发电功率进行预测；The wind power monitoring module is used to monitor the wind power module in real time and predict the power generation of the wind power module;

蓄电池监控模块，用于实时监控蓄电池模块；The battery monitoring module is used for real-time monitoring of the battery module;

负载监控模块，用于实时监控风电场储能系统中的负载，并对负载的功率变化情况进行预测；The load monitoring module is used to monitor the load in the wind farm energy storage system in real time and predict the power change of the load;

配电网联络模块，用于实时从配电网调控中心获知配电网的运行情况以及相关调度信息；The distribution network contact module is used to obtain the operation status of the distribution network and related dispatch information from the distribution network control center in real time;

并网运行监控模块，用于控制风电场储能系统连接或隔离配电网；The grid-connected operation monitoring module is used to control the connection or isolation of the wind farm energy storage system to the distribution network;

中控模块，用于确定风电场储能系统的运行策略，并向上述监控装置中的各模块发出指令，以执行该运行策略；The central control module is used to determine the operation strategy of the energy storage system of the wind farm, and issue instructions to each module in the above-mentioned monitoring device to execute the operation strategy;

总线模块，用于该监控装置的各个模块的通信联络；The bus module is used for the communication of each module of the monitoring device;

该监控方法包括如下步骤：The monitoring method includes the steps of:

(1)风电监控模块实时获取风电模块的运行数据，并存储数据，负载监控模块实时获取负载的负荷变化情况；(1) The wind power monitoring module obtains the operation data of the wind power module in real time and stores the data, and the load monitoring module obtains the load change of the load in real time;

(2)根据风电模块的运行数据，对未来预定时刻内的风电模块的输出功率进行预测，根据风电场负载的负荷变化情况，对负载的负荷需求进行预测；(2) According to the operation data of the wind power module, predict the output power of the wind power module within a predetermined time in the future, and predict the load demand of the load according to the load change of the wind farm load;

(3)实时检测获取蓄电池模块的SOC，实时获取配电网的参数和调度信息；(3) Real-time detection and acquisition of the SOC of the battery module, and real-time acquisition of parameters and scheduling information of the distribution network;

(4)以配电网的调度信息、当前蓄电池储能的SOC、未来风电模块输出功率、以及对未来负荷需求的变化作为约束条件，实现蓄电池模块SOC的优化控制。(4) With the dispatching information of the distribution network, the current SOC of the battery energy storage, the output power of the wind power module in the future, and the change of the future load demand as constraints, the optimal control of the SOC of the battery module is realized.

优选的，在步骤(2)采用如下方式预测风电模块的输出功率，所述风电模块包括风力发电机和SVG：Preferably, in step (2), the output power of the wind power module is predicted in the following manner, and the wind power module includes a wind power generator and an SVG:

(201)采集风电模块中当前各类电量实测值作为各类电量的预测值的初始值，预测值包括：风机有功预测值风机无功预测值风机机端电压预测值SVG无功预测值SVG机端电压预测值风电模块并网点(PCC)母线电压预测值 (201) Collect the current measured values of various electric quantities in the wind power module as the initial values of the predicted values of various electric quantities, and the predicted values include: predicted active power values of wind turbines Predicted value of fan reactive power Predicted value of fan terminal voltage SVG reactive power prediction value Predicted value of SVG terminal voltage Wind power module grid connection point (PCC) bus voltage prediction value

(202)根据所述预测值建立由优化目标函数和约束条件组成的MPC优化控制模型，并求解风电模块的有功和无功输出的预测值：(202) set up the MPC optimization control model that is made up of optimization objective function and constraint condition according to described predicted value, and solve the predicted value of the active and reactive power output of wind power module:

MPC优化控制模型的目标函数如式(1)所示：The objective function of the MPC optimal control model is shown in formula (1):

$\underset{{Q Q}_{WTG WTG}^{set set},, {V V}_{SVG SVG}^{set set}}{min min} (({Σ Σ}_{i i = = 00}^{N N - - 11} {Σ Σ}_{j j = = 00}^{M m - - 11} {ρ ρ}^{{t t}_{i i,, j j}} {F f}_{11},, {Σ Σ}_{i i = = 00}^{N N - - 11} {Σ Σ}_{j j = = 00}^{M m - - 11} {ρ ρ}^{{t t}_{i i,, j j}} {F f}_{22})) - - - - - - ((11))$

式(1)中和为优化变量，和含义分别为风机无功设定值和SVG电压设定值；N为时间窗覆盖控制周期的个数；M为单个控制周期下含预测点的个数；ρ为衰减系数，取值ρ＜1；时间变量ti,j＝(Mi+j)Δt意义为当前时刻起第i个控制周期内的第j个预测点，Δt为预测点间隔，Δt由风电模块功率预测时间间隔决定；In formula (1) and For the optimization variable, and The meanings are the reactive power setting value of the fan and the SVG voltage setting value; N is the number of control cycles covered by the time window; M is the number of prediction points under a single control cycle; ρ is the attenuation coefficient, and the value ρ<1 ; The time variable ti,j=(Mi+j)Δt means the j-th prediction point in the i-th control cycle from the current moment, Δt is the prediction point interval, and Δt is determined by the wind power module power prediction time interval;

F1为风电模块并网点母线电压与设定值的偏差水平，F1具体表达式如式(2)：F1 is the deviation level between the bus voltage of the grid-connected point of the wind power module and the set value, and the specific expression of F1 is as formula (2):

${F f}_{11} (({t t}_{i i,, j j})) = = {[[{V V}_{PCC PCC}^{pre pre} (({t t}_{i i,, j j})) - - {V V}_{PCC PCC}^{ref ref}]]}^{22} - - - - - - ((22))$

式(2)中表示PCC电压的参考值，从主站控制指令中提取后设定；In formula (2) Indicates the reference value of the PCC voltage, which is set after being extracted from the master station control command;

F2为SVG无功储备水平，F2具体表达式如式(3)：F2 is the reactive power reserve level of SVG, and the specific expression of F2 is as formula (3):

${F f}_{22} (({t t}_{i i,, j j})) = = {[[{Q Q}_{SVG SVG}^{pre pre} (({t t}_{i i,, j j})) - - {Q Q}_{SVG SVG}^{opr opr}]]}^{22} - - - - - - ((33))$

式(3)中为SVG无功最佳运行点；In formula (3) It is the optimal operating point of SVG reactive power;

MPC优化控制模型的约束条件，具体包括：The constraints of the MPC optimal control model include:

风机有功预测约束条件：Wind turbine active power prediction constraints:

${P P}_{WTG WTG}^{pre pre} (({t t}_{i i,, j j})) = = {Σ Σ}_{k k = = 11}^{{N N}_{a a}} {φ φ}_{k k} {P P}_{WTG WTG}^{pre pre} (({t t}_{i i,, j j - - k k})) + + {ϵ ϵ}_{WTG WTG}^{pre pre} (({t t}_{i i,, j j})) - - {Σ Σ}_{k k - - 11}^{{N N}_{m m}} {θ θ}_{k k} {ϵ ϵ}_{WTG WTG}^{pre pre} (({t t}_{i i,, j j - - k k})) - - - - - - ((44))$

式(4)中为风机有功预测误差；Na和Nm分别为AR和MA模型的阶数，φk和θk为相关权重，阶数与权重均根据风机有功历史值确定；ti,j-k为预测中参与计算数据(包括)对应时刻，下标k表征预测时刻前推kΔt时间，当ti,j-k≤0时，有功预测值应取对应时刻历史值；In formula (4) is the wind turbine active power prediction error; Na and Nm are the orders of the AR and MA models respectively, φk and θk are the relevant weights, and the order and weight are determined according to the historical value of the wind turbine active power; ti, jk are the calculation data involved in the prediction (including ) corresponds to the time, and the subscript k indicates that the prediction time is pushed forward by kΔt time. When ti, jk≤0, the active power prediction value should take the historical value at the corresponding time;

风机无功预测约束条件：Wind turbine reactive power prediction constraints:

风机无功在下次控制前达到设定值：The fan reactive power reaches the set value before the next control:

${Q Q}_{WTG WTG}^{pre pre} (({t t}_{i i,, 00})) = = {Q Q}_{WTG WTG}^{set set} (({t t}_{i i - - 1,0 1,0})) - - - - - - ((55))$

第i个控制周期内的各预测点，风机无功功率的变化过程以指数函数拟合：At each prediction point in the i-th control cycle, the change process of the reactive power of the fan is fitted with an exponential function:

${Q Q}_{WTG WTG}^{pre pre} (({t t}_{i i,, j j})) = = \frac{11 - - {e e}^{- - (({t t}_{i i,, j j} - - {t t}_{i i,, 00})) / / {T T}_{s the s}}}{11 - - {e e}^{- - MΔt MΔt / / {T T}_{s the s}}} {Q Q}_{WTG WTG}^{set set} (({t t}_{i i,, 00})) + + \frac{{e e}^{- - (({t t}_{i i,, j j} - - {t t}_{i i,, 00})) / / {T T}_{s the s}} - - {e e}^{- - MΔt MΔt / / {T T}_{s the s}}}{11 - - {e e}^{- - MΔt MΔt / / {T T}_{s the s}}} {Q Q}_{WTG WTG}^{pre pre} (({t t}_{i i,, 00})) - - - - - - ((66))$

式(6)中Ts为风机无功调节时间常数，可以根据风机无功调节测试试验获取。In formula (6), Ts is the reactive power adjustment time constant of the fan, which can be obtained according to the reactive power adjustment test of the fan.

SVG无功预测约束条件：SVG reactive power prediction constraints:

SVG无功参考值如式(7)所示：SVG reactive reference value As shown in formula (7):

$\begin{matrix} {Q Q}_{SVG SVG}^{ref ref} (({t t}_{i i,, j j})) = = {K K}_{P P} [[{V V}_{SVG SVG}^{pre pre} (({t t}_{i i,, j j})) - - {V V}_{SVG SVG}^{set set} (({t t}_{i i,, 00}))]] + + {K K}_{I I} Δt Δt {Σ Σ}_{k k = = 00}^{i i \times \times M m + + j j} [[{V V}_{SVG SVG}^{pre pre} (({t t}_{i i,, j j - - k k})) - - {V V}_{SVG SVG}^{set set} (({t t}_{i i,, - - k k}))]] \\ + + {Q Q}_{SVG SVG}^{pre pre} (({t t}_{0,0 0,0})) - - {K K}_{P P} [[{V V}_{SVG SVG}^{pre pre} (({t t}_{0,0 0,0})) - - {V V}_{SVG SVG}^{set set} (({t t}_{0,0 0,0}))]] \end{matrix} - - - - - - ((77))$

式(7)中KI和KP分别为比例环节和积分环节的系数；In formula (7), KI and KP are the coefficients of the proportional link and the integral link respectively;

SVG无功预测值如式(8)所示：The predicted reactive power value of SVG is shown in formula (8):

${Q Q}_{SVG SVG}^{pre pre} (({t t}_{i i,, j j})) = = {Q Q}_{SVG SVG}^{ref ref} (({t t}_{i i,, j j - - 11})) + + {[[{Q Q}_{SVG SVG}^{pre pre} (({t t}_{i i,, j j - - 11})) - - {Q Q}_{SVG SVG}^{ref ref} (({t t}_{i i,, j j - - 11}))]] e e}^{- - (({t t}_{i i,, j j} - - {t t}_{i i,, j j - - 11})) / / {T T}_{d d}} - - - - - - ((88))$

式(8)中时间常数Td为SVG电力电子装置动作延时；The time constant Td in formula (8) is the action delay of the SVG power electronic device;

电压预测约束条件：Voltage prediction constraints:

${V V}^{pre pre} (({t t}_{i i,, j j})) - - {V V}^{pre pre} (({t t}_{0,0 0,0})) = = S S [\begin{matrix} {P P}_{WTG WTG}^{pre pre} (({t t}_{i i,, j j})) - - {P P}_{WTG WTG}^{pre pre} (({t t}_{0,0 0,0})) \\ {Q Q}_{WTG WTG}^{pre pre} (({t t}_{i i,, j j})) - - {Q Q}_{WTG WTG}^{pre pre} (({t t}_{0,0 0,0})) \\ {Q Q}_{SVG SVG}^{pre pre} (({t t}_{i i,, j j})) - - {Q Q}_{SVG SVG}^{pre pre} (({t t}_{0,0 0,0})) \end{matrix}] - - - - - - ((99))$

式(9)中V^pre为风机机端、SVG机端、和PCC母线电压预测值构成的向量，S为灵敏度矩阵；In formula (9), V ^pre is a vector composed of the fan terminal, SVG terminal, and PCC bus voltage prediction value, and S is the sensitivity matrix;

系统电压、发电机运行和SVG运行的约束条件：Constraints for system voltage, generator operation, and SVG operation:

$\{\begin{matrix} {V V}^{min min} \leq \leq {V V}^{pre pre} (({t t}_{i i,, j j})) \leq \leq {V V}^{max max} \\ {Q Q}_{WTG WTG}^{min min} \leq \leq {Q Q}_{WTG WTG}^{pre pre} (({t t}_{i i,, j j})) \leq \leq {Q Q}_{WTG WTG}^{max max} \\ {Q Q}_{SVG SVG}^{min min} \leq \leq {Q Q}_{SVG SVG}^{pre pre} (({t t}_{i i,, j j})) \leq \leq {Q Q}_{SVG SVG}^{max max} \\ {ΔQ ΔQ}_{WTG WTG}^{min min} \leq \leq {Q Q}_{WTG WTG}^{pre pre} (({t t}_{i i,, 00})) - - {Q Q}_{WTG WTG}^{pre pre} (({t t}_{i i - - 1,0 1,0})) \leq \leq {ΔQ ΔQ}_{WTG WTG}^{max max} \\ {ΔQ ΔQ}_{SVG SVG}^{min min} \leq \leq {Q Q}_{SVG SVG}^{pre pre} (({t t}_{i i,, 00})) - - {Q Q}_{SVG SVG}^{pre pre} (({t t}_{i i - - 1,0 1,0})) {ΔQ ΔQ}_{SVG SVG}^{max max} \end{matrix} - - - - - - ((1010))$

式(11)中V^max和V^min分别为由PCC、风机和SVG电压预测值构成系统电压向量的上限和下限，其中PCC电压限值由配电网调度中心给出，而风机和SVG电压限值根据设备生产厂商给出的正常工作范围确定；和分别为风机无功运行上下限，和别为SVG无功运行上下限，皆根据设备生产厂商给出的正常工作范围确定；和分别为风机无功爬坡上下限，和分别为SVG无功爬坡上下限，皆需经过无功调速实验测试结果确定。In formula (11), V ^max and V ^min are the upper limit and lower limit of the system voltage vector composed of PCC, wind turbine and SVG voltage prediction values, respectively, where the PCC voltage limit is given by the distribution network dispatching center, and the wind turbine and SVG voltage limit The value is determined according to the normal working range given by the equipment manufacturer; and Respectively, the upper and lower limits of reactive power of the fan, and Not to mention the upper and lower limits of SVG reactive power operation, which are determined according to the normal working range given by the equipment manufacturer; and Respectively, the upper and lower limits of reactive power climbing of the fan, and Respectively, the upper and lower limits of SVG reactive power climbing, both of which need to be determined by the test results of reactive power speed regulation experiments.

优选的，步骤(4)中，上述蓄电池模块SOC的优化控制包括以下步骤：Preferably, in step (4), the optimal control of the battery module SOC includes the following steps:

(41)求解最优SOC范围，具体步骤为：(41) To solve the optimal SOC range, the specific steps are:

最优SOC范围优化模型的目标函数为：The objective function of the optimal SOC range optimization model is:

$\begin{matrix} min min F f = = {λ λ}_{11} {Σ Σ}_{i i = = 11}^{N N} {u u}_{optSOC optSOC min min} (({t t}_{i i})) Δt Δt + + {λ λ}_{22} {Σ Σ}_{i i = = 11}^{N N} {u u}_{optSOC optSOC max max} (({t t}_{i i})) Δt Δt \\ + + {λ λ}_{33} | | {SOC SOC}_{opt opt__min min} - - {SOC SOC}_{min min} | | + + {λ λ}_{44} | | {SOC SOC}_{opt opt__max max} - - {SOC SOC}_{max max} | | \end{matrix}$

约束条件为：The constraints are:

$\begin{matrix} \underset{j j = = 1,2 1,2,, . . . . . .,, {N N}_{k k}}{max max} {P P}_{out out} (({t t}_{i i - - j j})) - - \underset{j j = = 1,2 1,2,, . . . . . .,, {N N}_{k k}}{min min} {P P}_{out out} (({t t}_{i i - - j j})) \leq \leq {γ γ}_{k k},, k k = = 1,2 1,2,, . . . . . .,, K K \\ - - {P P}_{ch ch__max max} \leq \leq {P P}_{B B__ref ref} (({t t}_{i i})) \leq \leq {P P}_{disch disc__max max} \\ {SOC SOC}_{min min} \leq \leq {SOC SOC}_{opt opt__min min} \leq \leq {SOC SOC}_{max max} \\ {SOC SOC}_{min min} \leq \leq {SOC SOC}_{opt opt__max max} \leq \leq {SOC SOC}_{max max} \\ {SOC SOC}_{(({t t}_{i i}))} = = {SOC SOC}_{(({t t}_{i i - - 11}))} - - \frac{{P P}_{B B__ref ref} (({t t}_{i i})) Δt Δt}{{E E.}_{cap cap}} \\ {P P}_{out out} (({t t}_{i i})) = = {P P}_{B B__ref ref} (({t t}_{i i})) + + {P P}_{w w__pre pre} (({t t}_{i i})) \end{matrix} - - - - - - ((1212))$

其中，SOC_{opt_min}表示储能系统最优工作范围的下限，SOC_{opt_max}表示储能系统最优工作范围的上限，SOC_min表示储能系统正常工作范围的下限，SOC_max表示储能系统正常工作范围的上限，λ₁、λ₂、λ₃、λ₄分别为相应的权重系数，均为正数且权重系数和为1，SOC(t_i)和SOC(t_i-1)分别为t_i时刻和t_i-1时刻的储能系统荷电状态，P_{B_ref}(t_i)为储能系统在t_i时刻的设定功率，E_cap为储能系统的容量，P_out(t_i)为风电场经过储能系统平抑后的并网功率，u_optSOCmin(t_i)表示初始时刻荷电状态SOC(t₀)为SOC_{opt_min}时，t_i时刻储能系统是否出现过充过放；u_optSOCmax(t_i)表示初始时刻荷电状态SOC(t₀)为SOC_{opt_max}时，t_i时刻储能系统是否出现过充过放；P_{ch_max}为储能系统所允许的最大充电功率，P_{disch_max}为储能系统所允许的最大放电功率；N_k表示第k个波动控制时间范围内时间步长Δt的个数，K表示波动控制时间范围的数量，γ_k表示第k个波动控制时间范围内允许的功率最大变化量；Among them, SOC _{opt_min} represents the lower limit of the optimal operating range of the energy storage system, SOC _{opt_max} represents the upper limit of the optimal operating range of the energy storage system, SOC _min represents the lower limit of the normal operating range of the energy storage system, and SOC _max represents the upper limit of the normal operating range of the energy storage system upper limit, λ ₁ , λ ₂ , λ ₃ , and λ ₄ are the corresponding weight coefficients, all of which are positive numbers and the sum of the weight coefficients is 1. SOC(t _i ) and SOC(t _i-1 ) are the time t _i and The state of charge of the energy storage system at time t _i-1 , P _{B_ref} (t _i ) is the set power of the energy storage system at time t _i , E _cap is the capacity of the energy storage system, and P _out (t _i ) is the wind farm The grid-connected power stabilized by the energy storage system, u _optSOCmin (t _i ) indicates whether the energy storage system is overcharged or over-discharged at time t _i when the initial state of charge SOC(t ₀ ) is SOC _{opt_min} ; u _optSOCmax (t _i ) indicates whether the energy storage system is overcharged or overdischarged at time t _i when the state of charge SOC(t ₀ ) is SOC _{opt_max} at the initial moment; P _{ch_max} is the maximum charging power allowed by the energy storage system, and P _{disch_max} is the energy storage system The maximum discharge power allowed; N _k represents the number of time steps Δt in the k-th fluctuation control time range, K represents the number of fluctuation control time ranges, γ _k represents the maximum allowable power in the k-th fluctuation control time range Variation;

设定粒子的属性为SOC_{opt_min}、SOC_{opt_max}和P_{B_ref(ti)}，采用粒子群算法对最优SOC范围优化模型求解即可得到最优的SOCopt_min、SOCopt_max；Set the properties of the particles as SOC _{opt_min} , SOC _{opt_max} and P _{B_ref(ti)} , and use the particle swarm optimization algorithm to solve the optimal SOC range optimization model to obtain the optimal SOCopt_min and SOCopt_max;

(42)风储系统在运行过程中，根据实时的荷电状态偏移比例和储能系统的设定功率，周期性地调节滤波时间常数，每次调节的具体方法为：(42) During the operation of the wind storage system, the filter time constant is periodically adjusted according to the real-time charge state deviation ratio and the set power of the energy storage system. The specific method of each adjustment is as follows:

(421)偏移比例计算：(421) Offset ratio calculation:

根据得到的最优SOC范围(SOC_{opt_min},SOC_{opt_max})和储能系统的实时SOC计算荷电状态偏移比例pro_ΔSOC，计算公式为：According to the obtained optimal SOC range (SOC _{opt_min} , SOC _{opt_max} ) and the real-time SOC of the energy storage system, the state of charge offset ratio pro _{ΔSOC is} calculated, and the calculation formula is:

${pro pro}_{ΔSOC ΔSOC} = = \frac{SOC SOC - - \frac{11}{22} (({SOC SOC}_{opt opt__min min} + + {SOC SOC}_{opt opt__max max}))}{{SOC SOC}_{opt opt__max max} - - {SOC SOC}_{opt opt__min min}} - - - - - - ((1313))$

(422)将荷电状态偏移比例proΔSOC和储能系统当前时刻的设定功率P_{B_ref}作为输入，滤波时间常数T作为输出，根据预设的模糊控制规则，采用模糊控制策略得到滤波时间常数T；(422) The state of charge deviation ratio proΔSOC and the set power P _{B_ref} of the energy storage system at the current moment are taken as input, and the filtering time constant T is taken as output. According to the preset fuzzy control rule, the filtering time constant T is obtained by using the fuzzy control strategy ;

(423)本次调节周期内，根据步骤(422)得到的滤波时间常数T对风电场的实际输出功率P_w进行低通滤波，平抑后的期望并网功率记为P_{out_exp}，计算得到储能系统的目标设定功率并根据以下公式对目标设定功率进行限值处理，得到最终的设定功率P_{B_ref}，限制处理公式为：(423) In this adjustment period, according to the filtering time constant T obtained in step (422), the actual output power P _w of the wind farm is low-pass filtered, and the expected grid-connected power after smoothing is recorded as P _{out_exp} , and the energy storage is calculated System target setting power And according to the following formula, limit value processing is performed on the target set power to obtain the final set power P _{B_ref} , and the limit processing formula is:

$\begin{matrix} {\overset{~ ~}{P P}}_{B B__ref ref} \leq \leq \frac{((SOC SOC - - {SOC SOC}_{protect protect})) * * {E E.}_{cap cap}}{Δk Δk} \\ - - {P P}_{ch ch__max max} \leq \leq {\overset{~ ~}{P P}}_{B B__ref ref} \leq \leq {P P}_{disch disc__max max} \end{matrix} - - - - - - ((1414))$

其中，SOC_protect表示设定的荷电状态保护，Δk表示滤波时间常数调节的控制周期。Among them, SOC _protect represents the set state of charge protection, and Δk represents the control period of the filter time constant adjustment.

本发明的监控方法具有如下优点：(1)准确预测风电场的功率变化情况；(2)控制策略兼顾配电网调度要求、储能系统运行情况和负载的负荷需求，满足用户同时，兼顾了供电可靠性，保障储能系统的安全性，延长了系统储能系统的使用寿命。The monitoring method of the present invention has the following advantages: (1) accurately predict the power change of the wind farm; (2) the control strategy takes into account the dispatching requirements of the distribution network, the operation of the energy storage system and the load demand of the load, and satisfies users while taking into account The reliability of power supply ensures the safety of the energy storage system and prolongs the service life of the system energy storage system.

附图说明Description of drawings

图1示出了本发明方法所使用的一种风电场储能系统及其监控装置的框图；Fig. 1 shows the block diagram of a kind of wind farm energy storage system and monitoring device thereof used by the method of the present invention;

图2示出了本发明方法的流程图。Fig. 2 shows a flow chart of the method of the present invention.

具体实施方式detailed description

图1是示出了本发明的一种风电场储能系统监控装置11，该装置11包括：风电监控模块114，用于实时监控风电场储能系统10中的风电模块12，并对风电模块12的发电功率进行预测；蓄电池监控模块115，用于实时监控风电场储能系统10中的蓄电池模块13；负载监控模块116，用于实时监控风电场储能系统10中的负载17，并对负载17的功率变化情况进行预测；配电网联络模块112，用于实时从配电网20调控中心获知配电网20的运行情况以及相关调度信息；并网监控模块113，用于风电场储能系统10连接或隔离配电网20；中控模块117，用于确定风电场储能系统10的运行策略，并向上述各模块发出指令，以执行该供电策略；总线模块111，用于该监控装置11的各个模块的通信联络。Fig. 1 shows a wind farm energy storage system monitoring device 11 of the present invention, the device 11 includes: a wind power monitoring module 114 for real-time monitoring of the wind power module 12 in the wind farm energy storage system 10, and monitoring the wind power module 12 to predict the generated power; the battery monitoring module 115 is used for real-time monitoring of the battery module 13 in the wind farm energy storage system 10; the load monitoring module 116 is used for real-time monitoring of the load 17 in the wind farm energy storage system 10, and The power change of the load 17 is predicted; the distribution network contact module 112 is used to obtain the operation status of the distribution network 20 and related scheduling information from the distribution network 20 control center in real time; the grid-connected monitoring module 113 is used for wind farm storage The energy system 10 is connected to or isolated from the distribution network 20; the central control module 117 is used to determine the operation strategy of the wind farm energy storage system 10, and sends instructions to the above-mentioned modules to implement the power supply strategy; the bus module 111 is used for the The communication between each module of the monitoring device 11.

通信模块111，用于上述各个模块之间的通信，所述总线通信模块111通过冗余双CAN总线与其他模块相连。The communication module 111 is used for communication between the above modules, and the bus communication module 111 is connected to other modules through a redundant dual CAN bus.

风电模块包括多个风力发电机和SVG设备。风电监控模块114至少包括风力发电机定压、电流、频率检测设备、风速检测设备，以及SVG电压和电流检测设备。风力发电机的输出功率由风力发电机所在地点的风速、风向和自身特征所决定。The wind power module includes multiple wind turbines and SVG devices. The wind power monitoring module 114 at least includes wind generator constant voltage, current, frequency detection equipment, wind speed detection equipment, and SVG voltage and current detection equipment. The output power of the wind turbine is determined by the wind speed, wind direction and its own characteristics at the location of the wind turbine.

蓄电池监控模块116至少包括蓄电池端电压、电流、SOC检测设备以及温度检测设备。用于实时监控蓄电池模块的SOC。The battery monitoring module 116 at least includes battery terminal voltage, current, SOC detection equipment and temperature detection equipment. It is used to monitor the SOC of the battery module in real time.

中控模块117至少包括CPU单元、数据存储单元和显示单元。The central control module 117 at least includes a CPU unit, a data storage unit and a display unit.

配电网联络模块112至少包括无线通信设备。该无线通信设备可以为有线设备或无线设备。The distribution network communication module 112 includes at least a wireless communication device. The wireless communication device may be a wired device or a wireless device.

并网监控模块113至少包括用于检测配电网和风电场储能系统电压、电流和频率的检测设备、数据采集单元和数据处理单元。数据采集单元包含采集预处理和A/D转换模块，采集八路遥测信号量，包含电网侧A相电压、电流，风电场储能系统侧的三相电压、电流。遥测量可通过终端内的高精度电流和电压互感器将强交流电信号(5A/110V)不失真地转变为内部弱电信号，经滤波处理后进入A/D芯片进行模数转换，经转换后的数字信号经数据处理单元计算，获得风电场储能系统10侧的三相电压电流值和配电网20侧相电压电流值。本遥测信号量处理采用了高速高密度同步采样、频率自动跟踪技术还有改进的FFT算法，所以精度得到充分保证，能够完成风电场储能系统10侧有功、无功和电能从基波到高次谐波分量的测量和处理。The grid-connected monitoring module 113 includes at least detection equipment for detecting the voltage, current and frequency of the distribution network and wind farm energy storage system, a data acquisition unit and a data processing unit. The data acquisition unit includes an acquisition preprocessing and A/D conversion module, and collects eight channels of telemetry signals, including A-phase voltage and current on the grid side, and three-phase voltage and current on the wind farm energy storage system side. The remote measurement can convert the strong AC signal (5A/110V) into an internal weak current signal without distortion through the high-precision current and voltage transformers in the terminal, and enter the A/D chip for analog-to-digital conversion after filtering. The digital signal is calculated by the data processing unit to obtain the three-phase voltage and current values of the wind farm energy storage system 10 side and the phase voltage and current values of the distribution network 20 side. The telemetry signal quantity processing adopts high-speed high-density synchronous sampling, frequency automatic tracking technology and improved FFT algorithm, so the accuracy is fully guaranteed, and the wind farm energy storage system 10 side active power, reactive power and electric energy can be completed from fundamental to high Measurement and processing of subharmonic components.

参见附图2，本发明的方法包括如下步骤：Referring to accompanying drawing 2, method of the present invention comprises the steps:

S1.风电监控模块实时获取风电模块的运行数据，并存储数据，负载监控模块实时获取负载的负荷变化情况；S1. The wind power monitoring module obtains the operation data of the wind power module in real time and stores the data, and the load monitoring module obtains the load change of the load in real time;

S2.根据风电模块的运行数据，对未来预定时刻内的风电模块的输出功率进行预测，根据风电场负载的负荷变化情况，对负载的负荷需求进行预测；S2. According to the operation data of the wind power module, predict the output power of the wind power module within a predetermined time in the future, and predict the load demand of the load according to the load change of the wind farm load;

S3.实时检测获取蓄电池模块的SOC，实时获取配电网的参数和调度信息；S3. Real-time detection and acquisition of the SOC of the battery module, and real-time acquisition of parameters and scheduling information of the distribution network;

S4.以配电网的调度信息、当前蓄电池储能的SOC、未来风电模块输出功率、以及对未来负荷需求的变化作为约束条件，实现蓄电池模块SOC的优化控制。S4. Using the dispatching information of the distribution network, the current SOC of the battery energy storage, the output power of the wind power module in the future, and the change in the future load demand as constraints, the optimal control of the SOC of the battery module is realized.

优选的，在步骤S2.采用如下方式预测风电模块的输出功率，所述风电模块包括风力发电机和SVG：Preferably, in step S2. The output power of the wind power module is predicted in the following manner, and the wind power module includes a wind power generator and an SVG:

S201.采集风电模块中当前各类电量实测值作为各类电量的预测值的初始值，预测值包括：风机有功预测值风机无功预测值风机机端电压预测值SVG无功预测值SVG机端电压预测值风电模块并网点(PCC)母线电压预测值 S201. Collect the current measured values of various types of electricity in the wind power module as the initial values of the predicted values of various types of electricity. The predicted values include: the predicted value of active power of the wind turbine Predicted value of fan reactive power Predicted value of fan terminal voltage SVG reactive power prediction value Predicted value of SVG terminal voltage Wind power module grid connection point (PCC) bus voltage prediction value

S202.根据所述预测值建立由优化目标函数和约束条件组成的MPC优化控制模型，并求解风电模块的有功和无功输出的预测值：S202. Establish an MPC optimization control model composed of an optimization objective function and constraints according to the predicted value, and solve the predicted values of active and reactive power output of the wind power module:

SVG无功预测约束条件：SVG reactive power prediction constraints:

电压预测约束条件：Voltage prediction constraints:

优选的，步骤S4中，上述蓄电池模块SOC的优化控制包括以下步骤：Preferably, in step S4, the optimal control of the battery module SOC includes the following steps:

S41.求解最优SOC范围，具体步骤为：S41. Solving the optimal SOC range, the specific steps are:

约束条件为：The constraints are:

S42.风储系统在运行过程中，根据实时的荷电状态偏移比例和储能系统的设定功率，周期性地调节滤波时间常数，每次调节的具体方法为：S42. During the operation of the wind storage system, the filter time constant is periodically adjusted according to the real-time charge state deviation ratio and the set power of the energy storage system. The specific method of each adjustment is as follows:

S421.偏移比例计算：S421. Offset ratio calculation:

S422.将荷电状态偏移比例proΔSOC和储能系统当前时刻的设定功率P_{B_ref}作为输入，滤波时间常数T作为输出，根据预设的模糊控制规则，采用模糊控制策略得到滤波时间常数T；S422. Taking the state of charge offset ratio proΔSOC and the set power P _{B_ref} of the energy storage system at the current moment as input, and the filter time constant T as output, according to the preset fuzzy control rule, adopt the fuzzy control strategy to obtain the filter time constant T;

S423.本次调节周期内，根据步骤(422)得到的滤波时间常数T对风电场的实际输出功率P_w进行低通滤波，平抑后的期望并网功率记为P_{out_exp}，计算得到储能系统的目标设定功率并根据以下公式对目标设定功率进行限值处理，得到最终的设定功率P_{B_ref}，限制处理公式为：S423. In this adjustment period, according to the filter time constant T obtained in step (422), the actual output power P _w of the wind farm is low-pass filtered, and the expected grid-connected power after smoothing is recorded as P _{out_exp} , and the energy storage system is calculated target setting power And according to the following formula, limit value processing is performed on the target set power to obtain the final set power P _{B_ref} , and the limit processing formula is:

在S2中，采用神经网络模型预测负荷需求，具体步骤如下：In S2, the neural network model is used to predict the load demand, and the specific steps are as follows:

S211.每一天采集12组有功功率和无功功率，共连续采集8天，这样共有96组数据P(k)和Q(k)，k＝1,2,…,96。S211. Collect 12 sets of active power and reactive power every day for a total of 8 consecutive days, so there are 96 sets of data P(k) and Q(k), k=1,2,...,96.

S212.将96组数据P(k)和Q(k)进行归一化处理，使得 n＝1,2,…,96；首先将每一天的12个有功功率P(k)作为一组输入矢量R(m)，12个无功功率Q(k)作为一组输入矢量S(m)，m＝1,2,…,8，m表示神经网络的训练次数；同时预先假设第9天的12个有功功率P′(k)作为预测功率的输出矢量R′，第9天的12个无功功率Q′(k)作为预测功率的输出矢量S′；这样前8天的有功功率输入矢量就为S212. Normalize 96 sets of data P(k) and Q(k), so that n=1,2,...,96; first, 12 active powers P(k) of each day are taken as a set of input vectors R(m), and 12 reactive powers Q(k) are taken as a set of input vectors S(m ), m=1,2,...,8, m represents the training times of the neural network; at the same time, it is assumed that the 12 active powers P′(k) on the 9th day are used as the output vector R′ of the predicted power, and the 12 active powers on the 9th day Reactive power Q'(k) is used as the output vector S' of predicted power; so the input vector of active power in the first 8 days is

R(1),R(2),R(3),R(4),R(5),R(6),R(7),R(8)，第9天预测有功功率的输出矢量为R′；前8天的无功功率输入矢量就为R(1), R(2), R(3), R(4), R(5), R(6), R(7), R(8), the output vector of predicted active power on the 9th day is R′; the reactive power input vector in the first 8 days is

S(1),S(2),S(3),S(4),S(5),S(6),S(7),S(8)，第9天预测有功功率的输出矢量为S′。S(1), S(2), S(3), S(4), S(5), S(6), S(7), S(8), the output vector of predicted active power on the 9th day is S'.

S213.将8组输入矢量R(m)和S(m)作为神经网络的输入层，隐含层神经元的传递函数采用S型正切函数tansig，输出层神经元的传递函数采用S型对数函数logsig，如图2所示，这样经过8次神经网络训练后，就确定了神经网络中各连接权的权值。S213. Using 8 groups of input vectors R(m) and S(m) as the input layer of the neural network, the transfer function of the hidden layer neurons adopts the S-type tangent function tansig, and the transfer function of the output layer neurons adopts the S-type logarithm The function logsig, as shown in Figure 2, after eight times of neural network training, the weight of each connection weight in the neural network is determined.

S214.对于8个有功功率输入矢量R(m)，在隐含层神经元有a¹＝tansig(IW¹R+b¹)，其中a¹为隐含层神经元输出，IW¹为隐含层神经元的权值，b¹为隐含层神经元的阈值；在输出层神经元有a²＝log sig(LW²a¹+b²)，其中a²为输出层神经元输出，IW²为输出层神经元的权值，b²为输出层神经元的阈值。S214. For 8 active power input vectors R(m), there is a ¹ =tansig(IW ¹ R+b ¹ ) in the hidden layer neuron, where a ¹ is the output of the hidden layer neuron, and IW ¹ is the hidden The weight of the neurons in the layer, b ¹ is the threshold of the neurons in the hidden layer; the neurons in the output layer have a ² =log sig(LW ² a ¹ +b ² ), where a ² is the output of the neurons in the output layer, IW ² is the weight of neurons in the output layer, and b ² is the threshold of neurons in the output layer.

S215.对于8个有功功率输入矢量S(m)，在隐含层神经元有c¹＝tansig(IW¹S+b¹)，其中c¹为隐含层神经元输出，IW¹为隐含层神经元的权值，b¹为隐含层神经元的阈值；在输出层神经元有c²＝log sig(LW²c1+b²)，其中c²为输出层神经元输出，IW²为输出层神经元的权值，b²为输出层神经元的阈值。S215. For 8 active power input vectors S(m), there is c ¹ =tansig(IW ¹ S+b ¹ ) in hidden layer neurons, where c ¹ is the output of hidden layer neurons, and IW ¹ is hidden The weight of the neurons in the layer, b ¹ is the threshold of the neurons in the hidden layer; the neurons in the output layer have c ² =log sig(LW ² c1+b ² ), where c ² is the output of the neurons in the output layer, IW ² is the weight of neurons in the output layer, and b ² is the threshold of neurons in the output layer.

S216.将第8天的输入矢量R(8)和S(8)再次作为神经网络的输入层，此时神经网络中输出的预测功率的输出矢量R′和S′即为第九天的功率预测归一化值，再用反归一化算法，即k＝1,2,…,96，输出的矢量值R(9)和S(9)就是第九天预测功率的12个有功功率P′(k)和12个无功功率Q′(k)。这样以此类推，可以重复上面的步骤利用第二天到第九天的数据预测到第十天的功率，这样后面每一天的功率都可以被预测出来。S216. The input vectors R(8) and S(8) of the 8th day are used as the input layer of the neural network again, and the output vectors R' and S' of the predicted power output in the neural network are the power of the ninth day Predict the normalized value, and then use the denormalization algorithm, that is k=1,2,...,96, the output vector values R(9) and S(9) are the 12 active powers P′(k) and 12 reactive powers Q′(k) of the predicted power on the ninth day . By analogy, the above steps can be repeated to use the data from the second day to the ninth day to predict the power of the tenth day, so that the power of each subsequent day can be predicted.

在步骤S4中，风电场储能系统总功率Pg的约束为：In step S4, the constraint of the total power Pg of the wind farm energy storage system is:

在非响应调度时段1下，P_g,min≤P_g(l)≤P_g,max，P_g,min为风电场储能系统10能够从配电网20吸收的最大功率，P_g,max为风电场储能系统10能够向配电网20输送功率的最大功率；In the non-responsive scheduling period 1, P _g,min ≤P _g(l) ≤P _g,max , P _g,min is the maximum power that the wind farm energy storage system 10 can absorb from the distribution network 20, P _g,max is the maximum power that the wind farm energy storage system 10 can transmit power to the distribution network 20;

在响应调度时段2下，P_g(2)＝P_set，P_set为响应调度时段2下要求的联络线功率。In the response scheduling period 2, P _g(2) =P _set , where P _set is the tie line power required in the response scheduling period 2.

以上内容是结合具体的优选实施方式对本发明所作的进一步详细说明，不能认定本发明的具体实施只局限于这些说明。对于本发明所属技术领域的普通技术人员来说，在不脱离本发明构思的前提下，做出若干等同替代或明显变型，而且性能或用途相同，都应当视为属于本发明的保护范围。The above content is a further detailed description of the present invention in conjunction with specific preferred embodiments, and it cannot be assumed that the specific implementation of the present invention is limited to these descriptions. For those of ordinary skill in the technical field of the present invention, without departing from the concept of the present invention, making several equivalent substitutions or obvious modifications, and having the same performance or use, should be deemed to belong to the protection scope of the present invention.

Claims

1. A monitoring method for a wind farm energy storage system, the method is realized based on the following monitoring device, the monitoring device comprising:

The wind power monitoring module is used to monitor the wind power module in real time and predict the power generation of the wind power module;

The battery monitoring module is used for real-time monitoring of the battery module;

The load monitoring module is used to monitor the load in the wind farm energy storage system in real time and predict the power change of the load;

The distribution network contact module is used to obtain the operation status of the distribution network and related dispatch information from the distribution network control center in real time;

The grid-connected operation monitoring module is used to control the connection or isolation of the wind farm energy storage system to the distribution network;

The central control module is used to determine the operation strategy of the energy storage system of the wind farm, and issue instructions to each module in the above-mentioned monitoring device to execute the operation strategy;

The bus module is used for the communication of each module of the monitoring device;

The monitoring method includes the steps of:

(1) The wind power monitoring module obtains the operation data of the wind power module in real time and stores the data, and the load monitoring module obtains the load change of the load in real time;

(2) According to the operation data of the wind power module, predict the output power of the wind power module within a predetermined time in the future, and predict the load demand of the load according to the load change of the wind farm load;

(3) Real-time detection and acquisition of the SOC of the battery module, and real-time acquisition of parameters and scheduling information of the distribution network;

(4) Using the dispatching information of the distribution network, the current SOC of battery energy storage, the output power of future wind power modules, and changes in future load demand as constraints, the optimal control of battery module SOC is realized;

In step (2), the output power of the wind power module is predicted in the following manner, and the wind power module includes a wind generator and an SVG:

(201) Collect the current measured values of various electric quantities in the wind power module as the initial values of the predicted values of various electric quantities, and the predicted values include: predicted active power values of wind turbines Predicted value of fan reactive power Predicted value of fan terminal voltage SVG reactive power prediction value Predicted value of SVG terminal voltage Wind power module grid connection point (PCC) bus voltage prediction value

(202) set up the MPC optimization control model that is made up of optimization objective function and constraint condition according to described predicted value, and solve the predicted value of the active and reactive power output of wind power module:

The objective function of the MPC optimal control model is shown in formula (1):

In formula (1) and For the optimization variable, and The meanings are the reactive power setting value of the fan and the SVG voltage setting value; N is the number of control cycles covered by the time window; M is the number of prediction points under a single control cycle; ρ is the attenuation coefficient, and the value ρ<1 ; The time variable ti,j=(Mi+j)Δt means the j-th prediction point in the i-th control cycle from the current moment, Δt is the prediction point interval, and Δt is determined by the wind power module power prediction time interval;

F1 is the deviation level between the bus voltage of the grid-connected point of the wind power module and the set value, and the specific expression of F1 is as formula (2):

In formula (2) Indicates the reference value of the PCC voltage, which is set after being extracted from the master station control command;

F2 is the reactive power reserve level of SVG, and the specific expression of F2 is as formula (3):

In formula (3) It is the optimal operating point of SVG reactive power;

The constraints of the MPC optimal control model include:

Wind turbine active power prediction constraints:

In formula (4) is the wind turbine active power prediction error; Na and Nm are the orders of the AR and MA models respectively, φk and θk are the relevant weights, and the order and weight are determined according to the historical value of the wind turbine active power; ti, jk are the calculation data involved in the prediction (including ) corresponds to the time, and the subscript k indicates that the prediction time is pushed forward by kΔt time. When ti, jk≤0, the active power prediction value should take the historical value at the corresponding time;

Wind turbine reactive power prediction constraints:

The fan reactive power reaches the set value before the next control:

At each prediction point in the i-th control cycle, the change process of the reactive power of the fan is fitted with an exponential function:

In formula (6), Ts is the reactive power adjustment time constant of the fan, which can be obtained according to the reactive power adjustment test of the fan;

SVG reactive power prediction constraints:

SVG reactive reference value As shown in formula (7):

In formula (7), KI and KP are the coefficients of the proportional link and the integral link respectively;

The predicted reactive power value of SVG is shown in formula (8):

The time constant Td in formula (8) is the action delay of the SVG power electronic device;

Voltage prediction constraints:

In formula (9), V ^pre is a vector composed of the fan terminal, SVG terminal, and PCC bus voltage prediction value, and S is the sensitivity matrix;

Constraints for system voltage, generator operation, and SVG operation:

In formula (11), V ^max and V ^min are the upper limit and lower limit of the system voltage vector composed of PCC, wind turbine and SVG voltage prediction values, respectively, where the PCC voltage limit is given by the distribution network dispatching center, and the wind turbine and SVG voltage limit The value is determined according to the normal working range given by the equipment manufacturer; and Respectively, the upper and lower limits of reactive power of the fan, and Not to mention the upper and lower limits of SVG reactive power operation, which are all determined according to the normal working range given by the equipment manufacturer; and Respectively, the upper and lower limits of reactive power climbing of the fan, and Respectively, the upper and lower limits of SVG reactive power climbing, both of which need to be determined by the experimental test results of reactive power speed regulation;

In step (4), the optimal control of the battery module SOC includes the following steps:

(41) To solve the optimal SOC range, the specific steps are:

The objective function of the optimal SOC range optimization model is:

The constraints are:

Among them, SOC _{opt_min} represents the lower limit of the optimal operating range of the energy storage system, SOC _{opt_max} represents the upper limit of the optimal operating range of the energy storage system, SOC _min represents the lower limit of the normal operating range of the energy storage system, and SOC _max represents the upper limit of the normal operating range of the energy storage system upper limit, λ ₁ , λ ₂ , λ ₃ , and λ ₄ are the corresponding weight coefficients, all of which are positive numbers and the sum of the weight coefficients is 1. SOC(t _i ) and SOC(t _i-1 ) are the time t _i and The state of charge of the energy storage system at time t _i-1 , P _{B_ref} (t _i ) is the set power of the energy storage system at time t _i , E _cap is the capacity of the energy storage system, and P _out (t _i ) is the wind farm The grid-connected power stabilized by the energy storage system, u _optSOCmin (t _i ) indicates whether the energy storage system is overcharged or over-discharged at time t _i when the initial state of charge SOC(t ₀ ) is SOC _{opt_min} ; u _optSOCmax (t _i ) indicates whether the energy storage system is overcharged or overdischarged at time t _i when the state of charge SOC(t ₀ ) is SOC _{opt_max} at the initial moment; P _{ch_max} is the maximum charging power allowed by the energy storage system, and P _{disch_max} is the energy storage system The maximum discharge power allowed; N _k represents the number of time steps Δt in the k-th fluctuation control time range, K represents the number of fluctuation control time ranges, γ _k represents the maximum allowable power in the k-th fluctuation control time range Variation;

Set the properties of the particles as SOC _{opt_min} , SOC _{opt_max} and P _{B_ref(ti)} , and use the particle swarm optimization algorithm to solve the optimal SOC range optimization model to obtain the optimal SOCopt_min and SOCopt_max;

(42) During the operation of the wind storage system, the filter time constant is periodically adjusted according to the real-time charge state deviation ratio and the set power of the energy storage system. The specific method of each adjustment is as follows:

(421) Offset ratio calculation:

According to the obtained optimal SOC range (SOC _{opt_min} , SOC _{opt_max} ) and the real-time SOC of the energy storage system, the state of charge offset ratio pro _{ΔSOC is} calculated, and the calculation formula is:

(422) The state of charge deviation ratio proΔSOC and the set power P _{B_ref} of the energy storage system at the current moment are taken as input, and the filtering time constant T is taken as output. According to the preset fuzzy control rule, the filtering time constant T is obtained by using the fuzzy control strategy ;

(423) In this adjustment period, according to the filtering time constant T obtained in step (422), the actual output power P _w of the wind farm is low-pass filtered, and the expected grid-connected power after smoothing is recorded as P _{out_exp} , and the energy storage is calculated System target setting power And according to the following formula, limit value processing is performed on the target set power to obtain the final set power P _{B_ref} , and the limit processing formula is:

Among them, SOC _protect represents the set state of charge protection, and Δk represents the control period of the filter time constant adjustment.