WO2017161787A1 - 基于未来信息的光伏功率波动的动态平抑方法 - Google Patents

基于未来信息的光伏功率波动的动态平抑方法 Download PDF

Info

Publication number
WO2017161787A1
WO2017161787A1 PCT/CN2016/091982 CN2016091982W WO2017161787A1 WO 2017161787 A1 WO2017161787 A1 WO 2017161787A1 CN 2016091982 W CN2016091982 W CN 2016091982W WO 2017161787 A1 WO2017161787 A1 WO 2017161787A1
Authority
WO
WIPO (PCT)
Prior art keywords
lib
vrb
soc
max
charge
Prior art date
Application number
PCT/CN2016/091982
Other languages
English (en)
French (fr)
Inventor
杨立滨
Original Assignee
严利容
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 严利容 filed Critical 严利容
Publication of WO2017161787A1 publication Critical patent/WO2017161787A1/zh

Links

Images

Classifications

    • HELECTRICITY
    • H02GENERATION; CONVERSION OR DISTRIBUTION OF ELECTRIC POWER
    • H02JCIRCUIT ARRANGEMENTS OR SYSTEMS FOR SUPPLYING OR DISTRIBUTING ELECTRIC POWER; SYSTEMS FOR STORING ELECTRIC ENERGY
    • H02J3/00Circuit arrangements for ac mains or ac distribution networks
    • H02J3/24Arrangements for preventing or reducing oscillations of power in networks
    • H02J3/383
    • YGENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y02TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
    • Y02EREDUCTION OF GREENHOUSE GAS [GHG] EMISSIONS, RELATED TO ENERGY GENERATION, TRANSMISSION OR DISTRIBUTION
    • Y02E10/00Energy generation through renewable energy sources
    • Y02E10/50Photovoltaic [PV] energy
    • Y02E10/56Power conversion systems, e.g. maximum power point trackers

Definitions

  • the optimal parameter combination of the hybrid energy storage system is obtained;
  • M n, M n-1 division ratio of the fitness value of particles present and the previous cycle of calculation n is the current number of cycles; c 1, c 2 is the particle weight coefficient; W is the inertia weight; r 1, r 2 The random number is uniformly distributed in (0, 1); x i and v i are the position and velocity of the i-th particle; g is the constraint factor.
  • the energy storage system is a hybrid energy storage system of an all vanadium redox flow battery VRB and a lithium battery LiB.
  • the main action of VRB, LiB is started under limited conditions.
  • the single charge or discharge interval energy in the progressive control step is lower than the limit E min .
  • the VRB takes precedence and does not exceed the limit of the SOC.
  • the charge and discharge power is within the limit range. Under the premise of independent suppression of fluctuations, when the charge and discharge power exceeds the limit, the LiB action assists the cooperation, and does not operate when the time limit is not exceeded;
  • the charge and discharge model in the step S3, "the main action of the VRB, and the activation of LiB under the limited conditions" corresponds to the specific operation mode:
  • the SOC VRB (t) is the instantaneous value of the SOC of the VRB
  • the SOC max-VRB and SOC min-VRB are the upper and lower limits of the SOC operation of the VRB respectively; when the P max-discha-VRB ⁇ P(t) ⁇ P is satisfied
  • P max-cha-VRB and P max-discha-VRB are the maximum charge and discharge power of VRB , respectively.
  • VRB independently completes power stabilization; otherwise, LiB assists start coordination and stabilizes.
  • the charging and discharging strategy in the step S3, "LiB preferentially starts, and the VRB operates under the met condition" is:
  • SOC max-LiB and SOC min-LiB are the upper and lower limits of the SOC operation of LiB , respectively;
  • P max-cha-LiB and P max-discha-LiB are the maximum charge and discharge powers of LiB , respectively;
  • ⁇ P max-cha-LiB and ⁇ P max-discha-LiB are the maximum charge and discharge power change rates of LiB;
  • SOC LiB (t) is the instantaneous value of SOC of LiB.
  • the upper and lower limit values SOC max-LiB and SOC min-LiB of the SOC operation of LiB are 0.9 and 0.2, respectively.
  • the optimal operating SOC OLiB and SOC OVRB of LiB and VRB are 0.6 and 0.5, respectively.
  • the invention introduces ultra-short-term photovoltaic power prediction into the control process of the energy storage power station, and constructs a dynamic stabilization method of photovoltaic power fluctuation based on future information.
  • the method fully exploits the advantages of the hybrid energy storage system, realizes the dynamic economic control of the energy storage power station by optimizing the progressive interval control, and constructs the optimal control model with the minimum state of the state of charge state as the objective function, and considers the actual constraints.
  • the implementation flow and solution method based on particle swarm optimization are given.
  • the actual PV power plant operation data is used for verification. According to the analysis results of multiple evaluation indexes such as the power offset after the translation and the SOC operation interval, the present invention can significantly improve the number of charge and discharge conversions and the suppression effect of the energy storage power station. The effect has certain theoretical value and practical application value.
  • FIG. 1 is a schematic flow chart of a dynamic smoothing method for photovoltaic power fluctuation based on future information according to the present invention.
  • FIG. 2 is a schematic diagram of the leveling offset of the energy storage power station of the present invention during a certain day.
  • Embodiment 3 is a graph showing optimized charge and discharge power of each medium in Embodiment 1 of the present invention.
  • Fig. 5 is a graph showing an optimized charge and discharge power of each medium in the second embodiment of the present invention.
  • Figure 6 is a graph showing an optimized SOC of each medium in Embodiment 2 of the present invention.
  • a dynamic stabilization method for photovoltaic power fluctuation based on future information of the present invention includes the following steps:
  • M n, M n-1 division ratio of the fitness value of particles present and the previous cycle of calculation n is the current number of cycles; c 1, c 2 is the particle weight coefficient; W is the inertia weight; r 1, r 2 The random number is uniformly distributed in (0, 1); x i and v i are the position and velocity of the i-th particle; g is the constraint factor.
  • the invention uses a vanadium redox batty (VRB) and a lithium battery (LiB) as an example to construct a hybrid energy storage system.
  • Power-type energy storage VRB has frequent charge-discharge switching response capability, high charge and discharge times, and is suitable for fluctuations of random components exhibiting frequent and fast-changing characteristics.
  • Energy-type energy storage devices represented by lead-acid batteries and LiB have energy storage. The advantages of high energy density and long energy storage time have the disadvantages of short cycle life, and the number of times of charge and discharge state conversion should be strictly limited.
  • the operation of the hybrid energy storage power station should fully utilize the characteristics of VRB that can be frequently charged and discharged while defining the state of charge (SOC) interval, which plays an important role in the small-amplitude energy charging and discharging interval;
  • SOC state of charge
  • the operation needs to be in the limited SOC interval, and avoid the frequent charge-discharge switching while appropriately increasing its relative capacity to work in the charging and discharging interval of large-value energy;
  • the charging and discharging strategy of the energy storage power station will be constructed based on the above-mentioned medium characteristics to form coordination Complementary effective charge and discharge mode.
  • the selection of the future control step size of the hybrid energy storage power station is related to the accuracy of the PV power prediction, and the PV power prediction accuracy is closely related to the prediction scale. Considering that the light energy distribution has significant time periodicity, the present invention determines the future information interval by the probability distribution statistics of the annual operational data. The annual operating data of the actual photovoltaic power plant is selected, and the sampling step is 5 min. The power fluctuations required for the energy storage power station are determined according to the flattening target, as shown in Fig. 2.
  • the corresponding probability of ⁇ t in the sampling point interval (1, 7) is 85%, which is the main aggregation region of ⁇ t, and thus the two charging and discharging intervals are selected as the present invention.
  • Progressive control step size further statistical analysis of the sampling point distribution of a single charging and discharging interval, and obtaining the progressive control step size with [5min, 35min] as the main gathering interval.
  • the specific charge and discharge model is:
  • the main action of VRB, LiB starts under limited conditions.
  • the single charge or discharge interval energy within the progressive control step is lower than the limit E min , and the VRB takes precedence and independently stabilizes fluctuations under the premise that the SOC is not limited and the charge and discharge power is within the limit range.
  • the LiB action assists the cooperation; if it does not exceed the limit, it does not operate.
  • the charging and discharging model corresponds to a specific operation mode:
  • the charging and discharging energy interval; SOC VRB (t) is the instantaneous value of SOC of VRB
  • SOC max-VRB and SOC min-VRB are the upper and lower limits of SOC operation of VRB , respectively.
  • the control principle of this model is that for the progressive interval that satisfies the qualification condition, the VRB priority action even completes the pacing task of the interval independently, and when the VRB power is insufficient, it is started by LiB and assists the cooperation. As a result, the startup frequency of LiB is relatively low, and the number of charge and discharge switching is reduced.
  • the core goal is to take advantage of VRB's wide range of SOC changes and strictly control the number of LiB charge and discharge switching.
  • the determination of the charge and discharge energy limits E min-discha and E min-cha is the key to measuring whether VRB can independently undertake the reconciliation task.
  • VRB LiB is activated first, and VRB operates under certain conditions.
  • non-VRB priority actions In this case, the advantage of high LiB energy density is exerted, and it is initiated and assumed the main leveling task.
  • start of VRB it depends on the rate of change of LiB charging and discharging power and its SOC. The goal is to assist Lib to stabilize the target or adjust its own SOC to be Better running status.
  • the VRB auxiliary start condition is:
  • SOC max-LiB and SOC min-LiB are the upper and lower limits of the SOC operation of LiB, respectively, and 0.9 and 0.2 are respectively used; P max-cha-LiB and P max-discha-LiB are the maximum charge and discharge of LiB , respectively.
  • SOC LiB (t) is the instantaneous value of SOC of LiB. When LiB and VRB are simultaneously activated, when the respective SOC or charging/discharging power is out of synchronization, the situation of insufficient light and insufficient power will appear respectively.
  • an optimal control model aiming at the optimal SOC operating state of the energy storage power station is constructed. It is known that the initial SOC int-LiB and SOC int-VRB of each medium in the progressive control step interval are based on the charge and discharge strategy, so that the variance of the optimum SOC of each medium in the interval is minimized.
  • the objective function is as shown in equation (3).
  • the objective function mainly solves the problem of coordinated distribution of charge and discharge energy between media in the progressive control step interval.
  • SOC OLiB and SOC OVRB are the optimal operating SOC respectively, and 0.6 and 0.5 are respectively taken in the present invention
  • SOC LiB (t) and SOC VRB (t) are the real-time SOC values of the media in the interval, respectively, and the values are based on charge and discharge.
  • Strategy and PV power output are determined.
  • the same charge and discharge energy has different degrees of SOC change for media of different rated capacities.
  • the objective function is mainly for the energy allocation problem when the mixed medium is started at the same time.
  • the LiB capacity in the energy storage power station is relatively larger than the VRB. Therefore, under the objective function, the VRB priority operation mode LiB only partially limits the energy for charging and discharging power.
  • LiB main operation mode when LiB starts and its charging and discharging power and its rate of change and SOC satisfy the flattening condition, it will act as the main body of the suppression energy.
  • the constraints in the present invention mainly include charge and discharge power constraints, SOC constraints, specifically:
  • the charging and discharging process uses the VRB to independently undertake the charging and discharging tasks of the weak energy interval, so the number of startups of LiB is significantly reduced, compared with the reduction of 76.7. %; Suppression effect, due to the coordination of LiB and VRB, effectively avoiding the SOC or the charging and discharging power of a certain medium, so that the variance of the power offset after the suppression is reduced by 43.2%, ensuring the smoothing effect; the objective function value M Reduced by 45.1%.
  • Embodiment 2 Extracting the operating data of a photovoltaic power plant in October, the calculation results are shown in Table 2; selecting a certain time section interval P LiB (t), P VRB (t), SOC LiB (t), SOC VRB ( t) As shown in Figures 5 and 6, respectively.
  • the dynamic charge and discharge strategy of the energy storage power station considering power prediction proposed by the present invention can effectively optimize the operation of the progressive control interval, and can effectively reduce the LiB under the premise of ensuring the SOC operating state and the smoothing effect of each medium.
  • the number of charge and discharge conversions, giving full play to the dielectric properties of VRB, has a good superiority.

Landscapes

  • Engineering & Computer Science (AREA)
  • Power Engineering (AREA)
  • Charge And Discharge Circuits For Batteries Or The Like (AREA)
  • Supply And Distribution Of Alternating Current (AREA)

Abstract

本发明公开了一种基于未来信息的光伏功率波动的动态平抑方法,包括:S1、根据递进协调控制算法和步长数据确定优化目标函数;S2、设置粒子群维数D,最大迭代次数Mmax,收敛精度σthresh,同时初始化粒子群位置x和速度v,并给定初始SOCint-LiB、SOCint-VRB数值;S3、根据充放电策略和目标函数计算各粒子适应度值M;S4、将各粒子适应度值与自身粒子极值及全局粒子极值比较,若适应度值较小,则更新各粒子个体极值ebest及全局例子适应度极值gbest;S5、判断当前计算是否满足收敛条件,若是则提取当前PLiB、PVRB即为最优充放电功率;。本发明可保证各介质SOC运行状态及平抑效果的前提下有效减小LiB的充放电转换次数,充分发挥VRB的介质特性。

Description

基于未来信息的光伏功率波动的动态平抑方法 技术领域
本发明涉及储能系统技术领域,特别是涉及一种基于未来信息的光伏功率波动的动态平抑方法。
背景技术
光伏功率具有间歇和随机性的固有属性,随着光伏发电渗透率的持续提升,增加了负荷的波动性,同时对配电网的稳定运行带来了一定的挑战。储能电站能够实现能量的存储和释放,对电网的削峰填谷、改善负荷特性、提高电能质量等方面具有重要作用。因此配置一定容量的储能系统成为平滑光伏功率输出,提升大规模光电可调控能力的重要途径。
随着对储能系统的研究,许多学者指出单一储能介质配置特性固定,应对复杂光伏功率波动具有局限性,提出优势特性互补的混合储能是储能技术未来重要的发展方向。国内外学者对于混合储能系统能量优化和控制问题展开了相关研究并取得了一系列成果。
现有技术中公开了以下技术方案:
利用模糊控制理论将在混合储能介质中分配功率平抑任务,优先由超级电容电池来独立平抑功率波动,以此减少蓄电池的充放电次数;
考虑储能系统特性参数与平滑效果间的关系,并基于储能系统参数-平滑度、成本特性利用自学习的神经网络建立长期数学模型,获取混合储能系统的最佳参数组合;
结合平抑效果、剩余容量等因素,对功率型储能和能量型储能进行平抑任务的分配,实现更好的储能容量优化;采用滤波器进行混合储能的功率分配,由超级电容和锂电池(Li-ion Battery,LiB)分别承担短时间尺度和长时间尺度的波动平抑;
由蓄电池和超级电容构成的混合储能系统的优化能量管理方案。
但由于光电未来出力相对风电功率输出,具有更强的不确定性,上述既定模式的控制策略在光伏电站中将可能无法应对复杂的光伏功率波动。随着超短期光伏功率预测水平的提升,为储能电站基于未来信息构建动态优化控制提供了可能。
发明内容
为克服现有技术的不足,本发明的目的在于提供一种基于未来信息的光伏功率波动的动态平抑方法。
为了实现上述目的,本发明实施例提供的技术方案如下:
一种基于未来信息的光伏功率波动的动态平抑方法,所述方法包括:
S1、根据递进协调控制算法和步长数据确定优化目标函数,
Figure PCTCN2016091982-appb-000001
S2、设置粒子群维数D,最大迭代次数Mmax,收敛精度σthresh,同时初始化粒子群位置x和速度v,并给定初始SOCint-LiB、SOCint-VRB数值;
S3、根据充放电策略和目标函数计算各粒子适应度值M;
S4、将各粒子适应度值与自身粒子极值及全局粒子极值比较,若适应度值较小,则更新各粒子个体极值ebest及全局例子适应度极值gbest
S5、判断当前计算是否满足收敛条件|Mn-Mn-1|<σthresh,若是则提取当前PLiB、PVRB即为最优充放电功率;若否则更新各粒子位置x及速度v,并重复步骤S3-S5.
Figure PCTCN2016091982-appb-000002
Figure PCTCN2016091982-appb-000003
其中,Mn、Mn-1分比为当前和前一次循环计算的粒子适应度值,n为当前循环次数;c1、c2为粒子权重系数;w为惯性权重;r1、r2为(0,1)内均匀分布随机数;xi、vi为第i维粒子的位置与速度;g为约束因子。
作为本发明的进一步改进,所述方法中,储能系统为全钒液流电池VRB和锂电池LiB的混合储能系统。
作为本发明的进一步改进,所述步骤S3中的充放电策略具体为:
VRB主要动作,LiB在限定条件下启动,递进控制步长内的单次充或放区间能量低于限值Emin,VRB优先动作并在其SOC不越限、充放功率在限值范围的前提下独立平抑波动,当充放电功率越限时,LiB动作辅助协作,不越限时则不动作;
LiB优先启动,VRB在满足限定条件下动作,对于非VRB优先动作状况,由LiB启动并承担平抑任务,对于VRB的启动,取决于LiB充放电功率的变化速率及其SOC,辅助LiB平抑目标或调整自身SOC以处于较优运行状态。
作为本发明的进一步改进,所述步骤S3中的充放电策略的约束条件包括充放电功率约束、SOC约束,其中,
充放电功率约束为:
Pmax-discha-LiB<P(t)<Pmax-cha-LiB
Pmax-discha-VRB<P(t)<Pmax-cha-VRB
P(t)=PVRB(t)+PLiB(t);
SOC约束为:
SOCmin-LiB<SOCLiB(t)<SOCmax-LiB
SOCmin-VRB<SOCVRB(t)<SOCmax-VRB
作为本发明的进一步改进,所述步骤S3中的充放电策略,“VRB主要动作,LiB在限定条件下启动”时的充放电模型对应具体运行模式为:
Figure PCTCN2016091982-appb-000004
式中,Ei(i=1,2,3,4)为递进控制步长对应的两个充放区间的能量,且
Figure PCTCN2016091982-appb-000005
[ts-i,te-i]为各充放区间的始末时刻;Ei>0则代表HESS吸收能量,否则为释放能量;[Emin-discha,Emin-cha]分别为VRB优先启动对应的充放能量区间;SOCVRB(t)为VRB的SOC瞬时值,SOCmax-VRB、SOCmin-VRB分别为VRB的SOC运行上下限值;当满足Pmax-discha-VRB<P(t)<Pmax-cha-VRB时,其中Pmax-cha-VRB、Pmax-discha-VRB分别为VRB的最大充、放电功率,本控制步长内VRB独立完成功率平抑;否则,LiB辅助启动协调平抑。
作为本发明的进一步改进,所述步骤S3中的充放电策略,“LiB优先启动,VRB在满足限定条件下动作”中VRB辅助启动条件为:
Figure PCTCN2016091982-appb-000006
式中,SOCmax-LiB、SOCmin-LiB分别为LiB的SOC运行上下限值;Pmax-cha-LiB、Pmax-discha-LiB分别为LiB的最大充、放电功率;ΔP(t)为充放电功率的变化率,且ΔP(t)=P(t)-P(t-1);ΔPmax-cha-LiB、ΔPmax-discha-LiB分别为LiB的最大充放电功率变化率;SOCLiB(t)为LiB的SOC瞬时值。
作为本发明的进一步改进,所述VRB辅助启动条件中,LiB的SOC运行上下限值SOCmax-LiB、SOCmin-LiB分别为0.9和0.2。
作为本发明的进一步改进,所述步骤S1目标函数中,LiB和VRB的最佳运行SOCOLiB和SOCOVRB分别为0.6和0.5。
本发明具有以下有益效果:
本发明将超短期光伏功率预测引入到储能电站的控制过程,构建了基于未来信息的光伏功率波动的动态平抑方法。该方法充分发挥混合储能系统的优势,通过优化递进区间控制来实现储能电站的动态经济控制,以荷电状态转移量为最小为目标函数构建优化控制模型,并考虑实际约束条件,同时给出基于粒子群算法的实现流程和求解方法。利用实际光伏电站运行数据进行验证,根据平移后的功率偏移量、SOC运行区间等多个评价指标的分析结果,表明本发明可对储能电站的充放电转换次数及平抑效果等均具有显著效果,具有一定的理论价值和实际应用价值。
附图说明
图1为本发明基于未来信息的光伏功率波动的动态平抑方法的流程示意图。
图2为本发明储能电站某天时间内的平抑偏移量示意图。
图3为本发明实施例1中各介质的优化充放电功率曲线图。
图4为本发明实施例1中各介质的优化SOC曲线图。
图5为本发明实施例2中各介质的优化充放电功率曲线图。
图6为本发明实施例2中各介质的优化SOC曲线图。
具体实施方式
为了使本技术领域的人员更好地理解本发明中的技术方案,下面将结合本发明实施例中的附图,对本发明实施例中的技术方案进行清楚、完整地描述,显然,所描述的实施例仅仅是本发明一部分实施例,而不是全部的实施例。基于本发明中的实施例,本领域普通技术人员在没有做出创造性劳动前提下所获得的所有其他实施例,都应当属于本发明保护的范围。
本发明采用粒子群优化算法(Particle Swarm Optimization,PSO)对算例进行求解,并对其进行适度改进,以克服动态边界问题,同时递进优化的区间计算量相对较小,利于发挥PSO搜索精度高和收敛效果好的优势。参图1所示,本发明的一种基于未来信息的光伏功率波动的动态平抑方法,包括以下步骤:
S1、根据递进协调控制算法和步长数据确定优化目标函数,
Figure PCTCN2016091982-appb-000007
S2、设置粒子群维数D,最大迭代次数Mmax,收敛精度σthresh,同时初始化粒子群位置x和速度v,并给定初始SOCint-LiB、SOCint-VRB数值;
S3、根据充放电策略和目标函数计算各粒子适应度值M;
S4、将各粒子适应度值与自身粒子极值及全局粒子极值比较,若适应度值较小,则更新各粒子个体极值ebest及全局例子适应度极值gbest
S5、判断当前计算是否满足收敛条件|Mn-Mn-1|<σthresh,若是则提取当前PLiB、PVRB即为最优充放电功率;若否则更新各粒子位置x及速度v,并重复步骤S3-S5.
Figure PCTCN2016091982-appb-000008
Figure PCTCN2016091982-appb-000009
其中,Mn、Mn-1分比为当前和前一次循环计算的粒子适应度值,n为当前循环次数;c1、c2为粒子权重系数;w为惯性权重;r1、r2为(0,1)内均匀分布 随机数;xi、vi为第i维粒子的位置与速度;g为约束因子。
本发明以全钒液流电池(vanadium redox batty,VRB)和锂电池(Li-ion Battery,LiB)为例构建混合储能系统。功率型储能VRB具备频繁充放电切换响应能力,可充放电次数高,适用于呈现频繁快速变化特性的随机分量的波动平抑;以铅酸蓄电池和LiB为代表的能量型储能设备具有储能能量密度大、储能时间长的优势,同时存在循环寿命短的缺点,应严格限制其充放电状态转换次数。由此可见,混合储能电站的运行应充分发挥VRB可频繁充放电的特性同时限定荷电状态(state of charge,SOC)区间,使其在小幅度能量充放区间中发挥重要作用;而LiB运行需在限定SOC区间,且避免频繁充放电切换的同时适当提升其相对容量,使其在大幅值能量的充放电区间中工作;储能电站的充放电策略将基于上述介质特性构建,形成协调互补的有效充放电模式。
充放电过程模型:
混合储能电站未来控制步长选取与光伏功率预测精度相互关联,同时光伏功率预测精度与预测尺度关系紧密。考虑到光能分布具有明显的时间周期性,为此,本发明以年度运行数据的概率分布统计来确定未来信息区间。选取实际光伏电站年度运行数据,采样步长为5min,根据平抑目标确定储能电站所需平抑的功率波动如图2所示。以充放各一次的时长Δt为统计量,可得Δt在采样点区间(1,7]的对应概率和达到85%,为Δt主要聚集区域,由此本发明选定两个充放区间作为递进控制步长,进一步统计分析单次充放区间的采样点数分布,得到递进控制步长将以[5min,35min]为主要聚集区间。由此确定本发明充放电动态控制的步长,同时发现该步长与超短期光伏功率预测时间尺度完全符合。
根据储能电站介质的运行特性构建有序控制策略,具体充放电模型为:
1)VRB主要动作,LiB在限定条件下启动。递进控制步长内的单次充或放 区间能量低于限值Emin,VRB优先动作并在其SOC不越限、充放功率在限值范围的前提下独立平抑波动。当充放电功率越限时,LiB动作辅助协作;不越限时则不动作。该充放电模型对应具体运行模式为:
Figure PCTCN2016091982-appb-000010
式中Ei(i=1,2,3,4)为递进控制步长对应的两个充放区间的能量,且
Figure PCTCN2016091982-appb-000011
[ts-i,te-i]为各充放区间的始末时刻;同理Ei>0则代表HESS吸收能量,反之为释放能量;[Emin-discha,Emin-cha]分别为VRB优先启动对应的充放能量区间;SOCVRB(t)为VRB的SOC瞬时值,SOCmax-VRB、SOCmin-VRB分别为VRB的SOC运行上下限值。当满足Pmax-discha-VRB<P(t)<Pmax-cha-VRB时,其中Pmax-cha-VRB、Pmax-discha-VRB分别为VRB的最大充、放电功率,本控制步长内VRB独立完成功率平抑;反之,则LiB辅助启动协调平抑。
此模型控制原则为对于满足限定条件的递进区间,由VRB优先动作甚至独立完成本区间的平抑任务,当VRB功率不足时由LiB启动并辅助协作。由此使得LiB启动频次相对较低,减少其充放电切换次数。核心目标在于发挥VRB的SOC大范围变化特性,严格控制LiB充放电切换次数。充放能量限值Emin-discha、Emin-cha的确定成为衡量VRB能否独立承担平抑任务的关键。为此,利用光能分布的年度周期性,统计青海某地区光伏电站全年各充放区间能量概率分布,可得单次充放能量在[-0.90,0.75](MWh)内的分布概率达73%,具有较强参考性和代表性。因此,实际储能电站运行协调中Emin-discha、Emin-cha可由年度充放电区间能量概率分布和VRB额定容量的对比关系共同确定。
2)LiB优先启动,VRB在满足限定条件下动作。对于非VRB优先动作状 况,发挥LiB能量密度高的优势,由其启动并承担主要平抑任务;对于VRB的启动,取决于LiB充放电功率的变化速率及其SOC,其目标在于辅助Lib平抑目标或调整自身SOC以处于较优运行状态。VRB辅助启动条件为:
Figure PCTCN2016091982-appb-000012
式中,SOCmax-LiB、SOCmin-LiB分别为LiB的SOC运行上下限值,文中分别取0.9和0.2;Pmax-cha-LiB、Pmax-discha-LiB分别为LiB的最大充、放电功率;ΔP(t)为充放电功率的变化率,且ΔP(t)=P(t)-P(t-1);ΔPmax-cha-LiB、ΔPmax-discha-LiB分别为LiB的最大充放电功率变化率;SOCLiB(t)为LiB的SOC瞬时值。当LiB、VRB同时启动时,当各自SOC或充放电功率同步越限时,将分别出现弃光和平抑功率不足的状况。
基于上述充放模型,构建以储能电站SOC运行状态最优为目标的优化控制模型。已知递进控制步长区间的各介质初始SOCint-LiB、SOCint-VRB,基于充放电策略,使得本区间内各介质偏移最佳SOC的方差和最小。目标函数如式(3)所示。
Figure PCTCN2016091982-appb-000013
该目标函数主要解决本递进控制步长区间内充放能量在各介质间的协调分配问题。其中,SOCOLiB、SOCOVRB分别为最佳运行SOC,本发明中分别取0.6和0.5;SOCLiB(t)、SOCVRB(t)分别为本区间各介质的实时SOC数值,其数值基于充放电策略和光伏功率输出确定。
同等的充放电能量对于不同额定容量的介质,其SOC变化程度不等。该目标函数主要针对混合介质同时启动情况下的能量分配问题,一般而言,考虑到成本投入问题,储能电站中LiB容量相对VRB较大,因此,在该目标函数下,VRB优先动作模式中,LiB仅针对充放电功率越限部分能量;而LiB主要动作模式下,在LiB启动且其充放电功率及其变化率和SOC均满足平抑条件时,将作为平抑能量主体。
本发明中的约束条件主要包括充放电功率约束、SOC约束,具体地:
充放电功率约束条件为:
Pmax-discha-LiB<P(t)<Pmax-cha-LiB                  (4)
Pmax-discha-VRB<P(t)<Pmax-cha-VRB                  (5)
P(t)=PVRB(t)+PLiB(t)                        (6)
SOC约束条件为:
SOCmin-LiB<SOCLiB(t)<SOCmax-LiB                 (7)
SOCmin-VRB<SOCVRB(t)<SOCmax-VRB                   (8)
在本发明的一具体实施例中,光伏电站装机容量为15MW,储能电站中LiB额定容量配置为3MWh,VRB为1.5MWh,各采样点间隔为5min。由平抑后的功率偏移量方差γ、LiB充放电切换次数N、SOC运行区间及过程曲线构建控制效果评价指标系统,并与常规方法对比分析验证本发明的有效性和优越性。运行参数中,LiB的SOC运行允许限值为[0.1,0.9],而VRB的SOC运行限值为[0,1.0];LiB的充放电功率限值均为4MW,而VRB的充放电功率均为1.5MW;经统计该光场功率特征,Emin-discha取值为-1.1MWh,而Emin-cha取值为1.0MWh。
1)实施例1:提取该光伏电站某年度5月份运行数据,基于本发明所提方法,计算结果如表1所示:
表1 计算结果
Figure PCTCN2016091982-appb-000014
如表1所示,本发明所提方法在相关评价指标上均有大幅改变,其中充放电过程因采用VRB独立承担弱能量区间的充放电任务,因此LiB的启动次数显著降低,相比减少76.7%;平抑效果方面,由于LiB和VRB的协调配合,有效避免某一介质SOC或充放电功率的越限,使得平抑后的功率偏移量方差降低43.2%,保证了平抑效果;目标函数数值M降低45.1%。
为进一步考察利用本发明的方法,各介质的充放电功率和SOC的变化,选取一定时间截面区间PLiB(t)、PVRB(t)显示如图3所示,SOCLiB(t)、SOCVRB(t)如图4所示。充放电功率方面,可以看出,PLiB(t)和PVRB(t)的协调使得各自功率越限次数明显降低,同时由于PVBR(t)独立承担弱能量区间的充放电,使PLiB(t)有效减少充放电启动,而在两者同时充放启动的情况下,PLiB(t)可承担更多的平抑任务;结合图4中SOC可以看出,其SOCLiB变化较SOCVRB要小,LiB适合于浅充浅放,而VRB则可发挥其SOC可大范围变化的优势,既可实现其额定容量的充分利用,也可将其控制于限值范围内。
2)实施例2:提取光伏电站某年度10月份运行数据,计算结果如表2所示;选取一定时间截面区间PLiB(t)、PVRB(t)、SOCLiB(t)、SOCVRB(t)分别如图5、图6所示。
表2 计算结果
Figure PCTCN2016091982-appb-000015
Figure PCTCN2016091982-appb-000016
表2中相关评价指标同样均有较大幅度优化,其中LiB的启动次数相比减少76.2%,而平抑后的功率偏移量方差降低44.1%;同时本发明目标函数数值M降低47.9%。算例中,当放电平抑任务较重,需较大容量放电容量时,此时两者同时启动,LiB承担较大放电容量,但若其SOC接近下限,此时VRB则承担起了剩余放电功率。
综上所述,本发明提出的考虑功率预测的储能电站动态充放电策略可有效实现了递进控制区间的运行最优化,可保证各介质SOC运行状态及平抑效果的前提下有效减小LiB的充放电转换次数,充分发挥VRB的介质特性,具有较好的优越性。
由以上技术方案可以看出,本发明具有以下有益效果:
本发明将超短期光伏功率预测引入到储能电站的控制过程,构建了基于未来信息的光伏功率波动的动态平抑方法。该方法充分发挥混合储能系统的优势,通过优化递进区间控制来实现储能电站的动态经济控制,以荷电状态转移量为最小为目标函数构建优化控制模型,并考虑实际约束条件,同时给出基于粒子群算法的实现流程和求解方法。利用实际光伏电站运行数据进行验证,根据平移后的功率偏移量、SOC运行区间等多个评价指标的分析结果,表明本发明可对储能电站的充放电转换次数及平抑效果等均具有显著效果,具有一定的理论价值和实际应用价值。
对于本领域技术人员而言,显然本发明不限于上述示范性实施例的细节,而且在不背离本发明的精神或基本特征的情况下,能够以其他的具体形式实现本发明。因此,无论从哪一点来看,均应将实施例看作是示范性的,而且是非 限制性的,本发明的范围由所附权利要求而不是上述说明限定,因此旨在将落在权利要求的等同要件的含义和范围内的所有变化囊括在本发明内。不应将权利要求中的任何附图标记视为限制所涉及的权利要求。
此外,应当理解,虽然本说明书按照实施方式加以描述,但并非每个实施方式仅包含一个独立的技术方案,说明书的这种叙述方式仅仅是为清楚起见,本领域技术人员应当将说明书作为一个整体,各实施例中的技术方案也可以经适当组合,形成本领域技术人员可以理解的其他实施方式。

Claims (8)

  1. 一种基于未来信息的光伏功率波动的动态平抑方法,其中,所述方法包括:
    S1、根据递进协调控制算法和步长数据确定优化目标函数,
    Figure PCTCN2016091982-appb-100001
    S2、设置粒子群维数D,最大迭代次数Mmax,收敛精度σthresh,同时初始化粒子群位置x和速度v,并给定初始SOCint-LiB、SOCint-VRB数值;
    S3、根据充放电策略和目标函数计算各粒子适应度值M;
    S4、将各粒子适应度值与自身粒子极值及全局粒子极值比较,若适应度值较小,则更新各粒子个体极值ebest及全局例子适应度极值gbest
    S5、判断当前计算是否满足收敛条件|Mn-Mn-1|<σthresh,若是则提取当前PLiB、PVRB即为最优充放电功率;若否则更新各粒子位置x及速度v,并重复步骤S3-S5.
    Figure PCTCN2016091982-appb-100002
    Figure PCTCN2016091982-appb-100003
    其中,Mn、Mn-1分比为当前和前一次循环计算的粒子适应度值,n为当前循环次数;c1、c2为粒子权重系数;w为惯性权重;r1、r2为(0,1)内均匀分布随机数;xi、vi为第i维粒子的位置与速度;g为约束因子。
  2. 根据权利要求1所述的基于未来信息的光伏功率波动的动态平抑方法,其中,所述方法中,储能系统为全钒液流电池VRB和锂电池LiB的混合储能系统。
  3. 根据权利要求2所述的基于未来信息的光伏功率波动的动态平抑方法,其中,所述步骤S3中的充放电策略具体为:
    VRB主要动作,LiB在限定条件下启动,递进控制步长内的单次充或放区间能量低于限值Emin,VRB优先动作并在其SOC不越限、充放功率在限值范围的前提下独立平抑波动,当充放电功率越限时,LiB动作辅助协作,不越限时则不动作;
    LiB优先启动,VRB在满足限定条件下动作,对于非VRB优先动作状况,由LiB启动并承担平抑任务,对于VRB的启动,取决于LiB充放电功率的变化速率及其SOC,辅助LiB平抑目标或调整自身SOC以处于较优运行状态。
  4. 根据权利要求3所述的基于未来信息的光伏功率波动的动态平抑方法,其中,所述步骤S3中的充放电策略的约束条件包括充放电功率约束、SOC约束,其中,
    充放电功率约束为:
    Pmax-discha-LiB<P(t)<Pmax-cha-LiB
    Pmax-discha-VRB<P(t)<Pmax-cha-VRB
    P(t)=PVRB(t)+PLiB(t);
    SOC约束为:
    SOCmin-LiB<SOCLiB(t)<SOCmax-LiB
    SOCmin-VRB<SOCVRB(t)<SOCmax-VRB
  5. 根据权利要求4所述的基于未来信息的光伏功率波动的动态平抑方法,其中,所述步骤S3中的充放电策略,“VRB主要动作,LiB在限定条件下启动”时的充放电模型对应具体运行模式为:
    Figure PCTCN2016091982-appb-100004
    式中,Ei(i=1,2,3,4)为递进控制步长对应的两个充放区间的能量,且
    Figure PCTCN2016091982-appb-100005
    [ts-i,te-i]为各充放区间的始末时刻;Ei>0则代表HESS吸收能量,否则为释放能量;[Emin-discha,Emin-cha]分别为VRB优先启动对应的充放能量区间;SOCVRB(t)为VRB的SOC瞬时值,SOCmax-VRB、SOCmin-VRB分别为VRB的SOC运行上下限值;当满足Pmax-discha-VRB<P(t)<Pmax-cha-VRB时,其中Pmax-cha-VRB、Pmax-discha-VRB分别为VRB的最大充、放电功率,本控制步长内VRB独立完成功率平抑;否则,LiB辅助启动协调平抑。
  6. 根据权利要求4所述的基于未来信息的光伏功率波动的动态平抑方法,其中,所述步骤S3中的充放电策略,“LiB优先启动,VRB在满足限定条件下动作”中VRB辅助启动条件为:
    Figure PCTCN2016091982-appb-100006
    式中,SOCmax-LiB、SOCmin-LiB分别为LiB的SOC运行上下限值;Pmax-cha-LiB、Pmax-discha-LiB分别为LiB的最大充、放电功率;ΔP(t)为充放电功率的变化率,且ΔP(t)=P(t)-P(t-1);ΔPmax-cha-LiB、ΔPmax-discha-LiB分别为LiB的最大充放电功率变化率;SOCLiB(t)为LiB的SOC瞬时值。
  7. 根据权利要求6所述的基于未来信息的光伏功率波动的动态平抑方法,其中,所述VRB辅助启动条件中,LiB的SOC运行上下限值SOCmax-LiB、SOCmin-LiB分别为0.9和0.2。
  8. 根据权利要求2所述的基于未来信息的光伏功率波动的动态平抑方法,其中,所述步骤S1目标函数中,LiB和VRB的最佳运行SOCOLiB和SOCOVRB分别为0.6和0.5。
PCT/CN2016/091982 2016-03-23 2016-07-28 基于未来信息的光伏功率波动的动态平抑方法 WO2017161787A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201610169978.X 2016-03-23
CN201610169978.XA CN107230988A (zh) 2016-03-23 2016-03-23 基于未来信息的光伏功率波动的动态平抑方法

Publications (1)

Publication Number Publication Date
WO2017161787A1 true WO2017161787A1 (zh) 2017-09-28

Family

ID=59899192

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2016/091982 WO2017161787A1 (zh) 2016-03-23 2016-07-28 基于未来信息的光伏功率波动的动态平抑方法

Country Status (2)

Country Link
CN (1) CN107230988A (zh)
WO (1) WO2017161787A1 (zh)

Cited By (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN108321447A (zh) * 2018-01-31 2018-07-24 上海交通大学 基于荷电状态均衡逼近算法的多电池调度方法及系统
CN110198052A (zh) * 2019-07-11 2019-09-03 国网甘肃省电力公司经济技术研究院 一种光热-风电联合并网发电协调控制方法
CN110311386A (zh) * 2019-07-26 2019-10-08 国网江苏省电力有限公司淮安供电分公司 一种基于pso新能源电站自发无功和svg无功补偿的容量优化配置方法
CN110797872A (zh) * 2019-11-18 2020-02-14 华润智慧能源有限公司 用户侧储能容量配置方法、装置、设备及存储介质
CN111162551A (zh) * 2020-01-15 2020-05-15 国网内蒙古东部电力有限公司 一种基于风电功率超短期预测的蓄电池充放电控制方法
CN111307158A (zh) * 2020-03-19 2020-06-19 哈尔滨工程大学 一种auv三维航路规划方法
CN111628497A (zh) * 2020-05-22 2020-09-04 青海大学 一种面向电网稳定性的动态负载管理方法及计算机设备
CN112036603A (zh) * 2020-07-28 2020-12-04 南京航空航天大学 一种基于双堆燃料电池的混合储能系统能量管理策略

Families Citing this family (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110086181B (zh) * 2019-05-31 2020-08-11 浙江大学 一种离网运行时电-热联供微电网中功率自适应协同控制方法
CN114825381A (zh) * 2022-05-22 2022-07-29 国网甘肃省电力公司电力科学研究院 一种风光新能源基地光热电站容量配置方法

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN103366314A (zh) * 2013-07-18 2013-10-23 山东大学 考虑目标分解及其互补平抑的风电场复合储能容量规划方法
CN103928938A (zh) * 2014-02-24 2014-07-16 国家电网公司 考虑功率预测的储能电站优化控制方法
US20150022007A1 (en) * 2013-07-16 2015-01-22 Hongtao Ma Wind turbine generator controller responsive to grid frequency change
CN105226682A (zh) * 2015-11-04 2016-01-06 国家电网公司 一种混合储能平抑光伏功率波动方法

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20150022007A1 (en) * 2013-07-16 2015-01-22 Hongtao Ma Wind turbine generator controller responsive to grid frequency change
CN103366314A (zh) * 2013-07-18 2013-10-23 山东大学 考虑目标分解及其互补平抑的风电场复合储能容量规划方法
CN103928938A (zh) * 2014-02-24 2014-07-16 国家电网公司 考虑功率预测的储能电站优化控制方法
CN105226682A (zh) * 2015-11-04 2016-01-06 国家电网公司 一种混合储能平抑光伏功率波动方法

Cited By (12)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN108321447A (zh) * 2018-01-31 2018-07-24 上海交通大学 基于荷电状态均衡逼近算法的多电池调度方法及系统
CN108321447B (zh) * 2018-01-31 2020-09-29 上海交通大学 基于荷电状态均衡逼近算法的多电池调度方法及系统
CN110198052A (zh) * 2019-07-11 2019-09-03 国网甘肃省电力公司经济技术研究院 一种光热-风电联合并网发电协调控制方法
CN110198052B (zh) * 2019-07-11 2022-05-03 国网甘肃省电力公司经济技术研究院 一种光热-风电联合并网发电协调控制方法
CN110311386A (zh) * 2019-07-26 2019-10-08 国网江苏省电力有限公司淮安供电分公司 一种基于pso新能源电站自发无功和svg无功补偿的容量优化配置方法
CN110797872A (zh) * 2019-11-18 2020-02-14 华润智慧能源有限公司 用户侧储能容量配置方法、装置、设备及存储介质
CN110797872B (zh) * 2019-11-18 2024-03-19 华润智慧能源有限公司 用户侧储能容量配置方法、装置、设备及存储介质
CN111162551A (zh) * 2020-01-15 2020-05-15 国网内蒙古东部电力有限公司 一种基于风电功率超短期预测的蓄电池充放电控制方法
CN111307158A (zh) * 2020-03-19 2020-06-19 哈尔滨工程大学 一种auv三维航路规划方法
CN111628497A (zh) * 2020-05-22 2020-09-04 青海大学 一种面向电网稳定性的动态负载管理方法及计算机设备
CN111628497B (zh) * 2020-05-22 2022-04-29 青海大学 一种面向电网稳定性的动态负载管理方法及计算机设备
CN112036603A (zh) * 2020-07-28 2020-12-04 南京航空航天大学 一种基于双堆燃料电池的混合储能系统能量管理策略

Also Published As

Publication number Publication date
CN107230988A (zh) 2017-10-03

Similar Documents

Publication Publication Date Title
WO2017161787A1 (zh) 基于未来信息的光伏功率波动的动态平抑方法
WO2021114849A1 (zh) 增加新能源发电波动的海岛电网储能系统分层控制方法
Xu et al. The control of lithium‐ion batteries and supercapacitors in hybrid energy storage systems for electric vehicles: a review
WO2017161785A1 (zh) 基于储能运行状态的光伏功率稳定输出控制方法
CN103337001B (zh) 考虑最佳期望输出与荷电状态的风电场储能容量优化方法
CN108110795B (zh) 风电场群外送输电容量及储能配置联合优化方法
CN104319793B (zh) 一种平抑长短期风电功率波动的风储协调控制方法
CN105140942B (zh) 考虑荷电状态偏差的混合储能优化功率分配方法
WO2015014011A1 (zh) 考虑充放电倍率的多类型电池储能电站能量管理方法
CN105226689A (zh) 考虑运行和维护的多类型储能系统能量管理方法和系统
CN103928938A (zh) 考虑功率预测的储能电站优化控制方法
CN111064214A (zh) 基于电动汽车两阶段滚动式策略的配电网优化调度方法
CN106505604A (zh) 接入区域配电网的光伏储能联合运行单元优化配置方法
CN110450677B (zh) 一种基于电池老化状态估计的复合储能电动汽车的能量管理方法
CN106299511B (zh) 电动汽车充电站储能容量优化方法
CN105680478A (zh) 基于超短期预测的储能系统平滑新能源发电控制方法
CN105226694A (zh) 基于模糊经验模态分解的储能平滑新能源发电控制方法
CN115733180A (zh) 基于自适应模态数vmd算法的混合储能系统调频控制方法及系统
CN114819373A (zh) 一种基于合作博弈的共享混合储能电站的储能规划方法
CN103915851B (zh) 一种递进步长和期望输出均可变的储能系统优化控制方法
CN104283224B (zh) 一种限制风电波动率的储能系统平滑风电功率控制方法
Li et al. Optimization strategy of secondary frequency modulation based on dynamic loss model of the energy storage unit
CN103401260B (zh) 一种复合储能的超前优化递进控制方法
CN114201916A (zh) 一种基于参数优化的微网混合储能系统剩余功率分解方法
CN109004642B (zh) 用于平抑分布式电源功率波动的配网分布式储能评价方法

Legal Events

Date Code Title Description
NENP Non-entry into the national phase

Ref country code: DE

121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 16895137

Country of ref document: EP

Kind code of ref document: A1

32PN Ep: public notification in the ep bulletin as address of the adressee cannot be established

Free format text: NOTING OF LOSS OF RIGHTS PURSUANT TO RULE 112(1) EPC (EPO FORM 1205A DATED 22/02/2019)

122 Ep: pct application non-entry in european phase

Ref document number: 16895137

Country of ref document: EP

Kind code of ref document: A1