CN106998072A - A kind of mixed energy storage system capacity configuration optimizing method for optimizing operation towards power distribution network - Google Patents

Abstract

The invention discloses a kind of mixed energy storage system capacity configuration optimizing method for optimizing operation towards power distribution network, first obtain light-preserved system participate in distribution network operation it is optimal exert oneself with photovoltaic is actual exert oneself between imbalance power, as the reference power of hybrid energy-storing, imbalance power is decomposed into by a series of intrinsic mode functions using set empirical mode decomposition, the instantaneous frequency time graph of each intrinsic mode function is obtained by recurrence Hilbert transform；According to power-type and the characteristic of energy type energy storage, crossover frequency is at least determined for principle with instantaneous frequency time graph aliasing, the high fdrequency component and low frequency component using power-type energy storage and energy type energy storage respectively to imbalance power are stabilized；Finally, it is considered to efficiency for charge-discharge, state-of-charge and overall life cycle cost, power-type and energy type energy storage rated power and rated capacity economic optimum allocation plan are proposed.The present invention can not only compensate new energy and exert oneself relative to conventional method, improve the digestion capability to new energy, and can be actively engaged in the optimization operation of power distribution network.

Description

A kind of mixed energy storage system capacity configuration optimizing method for optimizing operation towards power distribution network

Technical field

The invention belongs to participate in the technical field of energy storage of power distribution network optimization operation, and in particular to a kind of towards power distribution network optimization The mixed energy storage system capacity configuration optimizing method of operation.

Background technology

Clean energy resource generation technology by representative of photovoltaic, is just changing the electric power and energy knot in worldwide deeply Structure.However, photovoltaic generation there are problems that randomness, high permeability it is grid-connected can to distribution power flow distribution, electricity Energy quality etc., which is brought, to be had a strong impact on.Therefore, the random wave of storage symmetrical photovoltaic plant power output need to be configured in power distribution network It is dynamic.

Energy-storage system is engaged with photovoltaic, on the one hand can be effectively reduced intermittence and randomness that photovoltaic is exerted oneself, be made light The power output of volt tends to be steady, and reduces the impact to power network；On the other hand can to energy carry out across when dispatch, participate in distribution The optimization operation of net.Energy storage can be divided into energy type energy storage and the major class of power-type energy storage two by its power producing characteristics.With lead-acid accumulator, Its energy density is big for the energy type energy storage of representative for lithium battery etc., but power density is small, the response time is long, it is adaptable to stabilize high energy The low-frequency power fluctuations of amount.With ultracapacitor, super conductive magnetic storage energy etc. for the power-type energy storage of representative its power density is big, response Time is short, but energy density is small, it is adaptable to stabilize the high frequency power fluctuation of low energy.Using having energy type energy storage and power-type concurrently The mixed energy storage system compensation new energy power output of energy storage advantage turns into current research and the focus of application.

Conventional method stabilizes new energy using mixed energy storage system (Hybrid energy storage system, HESS) The fluctuation of source power output, does not consider that HESS participates in power distribution network optimization operation.The application of energy-storage system can not only compensate new energy Source is exerted oneself, and improves the digestion capability to new energy, and can be actively engaged in the optimization operation of power distribution network.

The content of the invention

In view of the above-mentioned problems, the present invention proposes that a kind of mixed energy storage system capacity optimization for optimizing operation towards power distribution network is matched somebody with somebody Method is put, the rated power and rated capacity of energy-storage system can be significantly reduced, is that HESS capacity is distributed rationally there is provided a kind of new Method, with certain reference value；Meanwhile, HESS application can not only stabilize the fluctuation that photovoltaic is exerted oneself, and promote photovoltaic energy Source is dissolved, and can participate in the optimization operation of power distribution network, can effectively improve the economy and reliability of distribution network operation.

Above-mentioned technical purpose is realized, above-mentioned technique effect is reached, the present invention is achieved through the following technical solutions：

A kind of mixed energy storage system capacity configuration optimizing method for optimizing operation towards power distribution network, comprises the following steps：

Set up light-preserved system and participate in power distribution network optimal operation model, solution obtains light-preserved system and participates in distribution network operation most It is excellent to exert oneself, and using light-preserved system participate in distribution network operation it is optimal exert oneself with photovoltaic is actual exert oneself between imbalance power as The reference power of mixed energy storage system；

By improving, Hilbert-Huang transform decomposition light-preserved system participation distribution network operation is optimal to exert oneself and photovoltaic plant reality Border exert oneself between imbalance power, imbalance power is divided into high and low frequency two parts wave component, and be utilized respectively Power-type energy storage and energy type energy storage compensation high-frequency fluctuation component and low-frequency fluctuation component；

Obtain the mixed energy storage system capacity for optimizing operation towards power distribution network and distribute strategy rationally.

Further, the light-preserved system, which participates in power distribution network optimal operation model, includes object function and constraints,

The object function is：

P_PESS.loss(t)=(1- η_ch)P_ch(t)+(1-η_dis)P_dis(t)+ξ(t)·ΔP_ESS.loss

In formula：T is the cycle of operation；Δ t is operation step-length；N is system node number, P_i(t) it is the injection of t period node is Active power；ξ (t) is that energy-storage system changes dimension from the t-1 periods to the charging and discharging state of t periods, and value is 0 or 1；η_ch、η_dis For the efficiency for charge-discharge of energy-storage system；P_ch(t)、P_dis(t) it is the charge-discharge electric power of t period energy-storage systems；ΔP_ESS.lossDuring for t The loss that section energy-storage system is produced when charging and discharging state switches.

Further, the constraints is：

P_PESS(t)=P_PV(t)+P_ESS(t)

P_PV(t)=P_PV.MPPT(t)

-P_ch.max≤P_ESS(t)≤P_dis.max

S (t+1)=(1- σ) S (t)+P_ESS(t)Δt

S_min≤S(t)≤S_max

S (0)=S (T)

In formula：I, j are system node number；U_i(t)、U_j(t) it is t periods node i and node j point voltage magnitudes；G_ij、B_ijPoint Transconductance and mutual susceptance that Wei be between node i and node j；δ_ij(t) it is the phase difference between t periods node i and node j；P_Ki (t)、Q_Ki(t) it is respectively that the outlet of t periods feeder line is active and idle；P_PESSi(t)、Q_PESSi(t) it is light-preserved system at t period node is Active and reactive power；P_Di(t)、Q_Di(t) it is the active and reactive power of load at t period node is；P_ij(t) it is t Line power between node i and node j；The subscript "-" and subscript " _ " of variable represent the upper and lower bound of variable respectively； P_PESSAnd Q (t)_PESS(t) be respectively t periods light-preserved system output active and reactive power；P_PV(t) it is t period photovoltaic generations Unit active power of output；P_MPPTIt is active according to maximum output during MPPT control strategies for photovoltaic DC/DC current transformers；P_ESS(t) For the active output of t period energy-storage systems, it is divided into electric discharge, free time, three kinds of states of charging；P_chAnd P (t)_dis(t) it is respectively the t periods The charge-discharge electric power of energy-storage system；η_chAnd η_disRespectively energy-storage system efficiency for charge-discharge；P_ch.maxAnd P_dis.maxRespectively charge and discharge Electrical power bound；S (t) is the dump energy of t period energy-storage systems；σ is the self-discharge rate of energy-storage system；S_minAnd S_maxRespectively For the bound of energy-storage system dump energy.

Further, it is described solution obtain light-preserved system participate in the optimal of distribution network operation exert oneself, be specially：

Power distribution network optimal operation model is participated in using two-dimension dynamic programming Algorithm for Solving light-preserved system, show that light-preserved system is joined Optimal with distribution network operation is exerted oneself；Specifically include following steps：

Step 2.1：Light-preserved system participation power distribution network optimal operation model is converted into the operable model of two-dimension dynamic programming, Including stage t, state vector S (s_Pt,s_Qt), decision vector U (u_Pt,u_Qt), strategy, state transition equation and target function；

Step 2.2：Initialization, light-preserved system and distribution network system primary data needed for input；

Step 2.3：Determine the permission state set of t-1 period light-preserved systems；

Step 2.4：It is determined that meeting the permission state set of t period light-preserved systems under constraints；

Step 2.5：Calculate correspondence decision variable and target function value；

Step 2.6：Whether judge index function is optimal, if so, then skipping to step 2.7；Otherwise, step 2.8 is skipped to；

Step 2.7：Preserve current state variable, decision variable and object function；

Step 2.8：Judge whether the permission state set of traversal t-1 period light-preserved systems, if so, skipping to step 2.9；It is no Then, step 2.3 is skipped to；

Step 2.9：Judge whether the permission state set of traversal t period light-preserved systems, if so, skipping to step 2.10；If It is no, skip to step 2.4；

Step 2.10：Judge whether t=T, finished if so, then calculating, output result；If it is not, then t=t+1, and skip to step Rapid 2.3；

It is described by light-preserved system participate in distribution network operation it is optimal exert oneself with photovoltaic is actual exert oneself between imbalance power Calculation formula be：

P_net(t)=P_Opt(t)-P_PV(t)

In formula：P_net(t) it is imbalance power；P_Opt(t) participate in that distribution network operation is optimal exerts oneself for light-preserved system；P_PV(t) For photovoltaic generation unit active power of output.

Further, the stage t is：One full schedule cycle T is divided into several periods, the single time is remembered Section is the stage, and stage sequence number is labeled as t, t ∈ { 1,2,3 ... T }, and the adjacent phases time difference is Δ t；

State S (the s_Pt,s_Qt) be：The dump energy S for choosing energy-storage system in light-preserved system is used as state S_Pt, and will Electricity difference between its discretization, adjacent states is Δ S；Choose the remaining reactive capability of light-preserved system inverter and be used as state S_Qt, and By its discretization, the reactive capability difference between adjacent states is Δ Q；

Decision-making U (the u_Pt,u_Qt) be：By the P in light-preserved system each period_PESSAnd Q (t)_PESS(t) become as decision-making Measure u_PtAnd u_Qt, it must is fulfilled for light-preserved system operation constraint；

The strategy is：The sequence of the decision variable composition in each stage；

The state transition equation includes：S_PtState transition equation and S_QtState transition equation；The S_PtState transfer Equation is S (t+1)=(1- σ) S (t)+P_ESS(t)Δt；The S_QtAbsolute transfer relationship is not present between adjacent two benches, it is permitted Perhaps state set is by current state S_PtDetermined with inverter capacity, its state transition equation is：

The target function is：It regard t phase targets function as stage target function V_t(S(s_Pt,s_Qt),U(u_Pt, u_Qt)), then the target function in t stages is：

In formula：D(u_Pt,u_Qt) represent t stage conditions permission decision-making set, V_tExpression stage target function.

Further, it is described by improve Hilbert-Huang transform decompose light-preserved system participate in distribution network operation it is optimal go out Power and photovoltaic plant is actual exert oneself between imbalance power, by imbalance power be divided into high and low frequency two parts fluctuation point Measure, detailed process is：

Using set empirical mode decomposition EEMD to signal P_net(t) decomposed, after set empirical mode decomposition P_net(t) it is represented by：

In formula：P_net(t) it is imbalance power signal；c_i,0(t) it is i-th of intrinsic mode function IMF；R (t) is remaining point Amount；For all intrinsic mode function averages；For all residual components averages；

To c_i,0(t) recurrence Hilbert transform, signal c are carried out_i,0(t) it is represented by：

In formula:c_i,0(t) it is i-th of intrinsic mode function IMF；a_i,j(t) it is intrinsic mode function c_i,j(t) amplitude letter Number；cosφ_i,m(t) it is recursive calculation to magnitude function a_i,j(t) when being intended to 1, intrinsic mode function c_i,m(t) instantaneous phase Cosine value；M is decomposition number of times；

The instantaneous frequency for obtaining intrinsic mode function IMF is：

In formula：ω_i(t) it is signal c_i,0(t) instantaneous frequency；φ_i,m(t) it is recursive calculation to magnitude function a_i,j(t) become To in 1 when, intrinsic mode function c_i,m(t) instantaneous phase, m is decomposition number of times；

Determine crossover frequency f_gIf, crossover frequency f_gBetween g and g+1 IMF, then c₁(t),c₂(t),…,c_g(t) regard For high-frequency fluctuation component, this portion of energy is stabilized by power-type energy storage；c_g+1(t),c_g+2(t),…,c_n(t) it is considered as low-frequency fluctuation Component, this portion of energy is stabilized by energy type energy storage, then the reference power of energy type energy storage and power-type energy storage is：

In formula：P_g,ref(t)、P_n,ref(t) be respectively power-type and energy type energy storage reference power；c_i(t) it is natural mode State function, its value substituted into is all intrinsic mode function averagesR (t) is residual components.

Further, it is described to be utilized respectively power-type energy storage and energy type energy storage compensation high-frequency fluctuation component and low-frequency fluctuation The detailed process of component is：

Set up energy type energy storage rated power configuration calculation formula be：

In formula：P_n,rateFor the rated power of energy type energy storage；t_n,0For energy type energy storage discharge and recharge initial time, T_nTo grind Study carefully duration；η_DC-DCAnd η_DC-ACFor the energy conversion efficiency of transverter；η_ncAnd η_ndFor the efficiency for charge-discharge of energy type energy storage；

Set up energy type energy storage rated capacity configuration calculation formula be：

In formula：E_n,rateFor energy type energy storage rated capacity；P_n(t) it is consideration transverter energy conversion efficiency and discharge and recharge The charge-discharge electric power of energy type energy storage after efficiency；ΔT_nFor energy type energy storage discharge and recharge time interval；SOC_n,0For energy type energy storage Initial state-of-charge；SOC_n,max、SOC_n,minFor the upper lower limit value of energy type energy storage charge state；

Work as SOC_n,0Meet equation, and E_n,rateWhen taking equal sign, the rated capacity of energy type energy storage is minimum；

Set up power-type energy storage rated power configuration calculation formula be：

In formula：P_g,rateFor the rated power of power-type energy storage；t_g,0For power-type energy storage discharge and recharge initial time；η_DC-DCWith η_DC-ACFor the energy conversion efficiency of transverter；η_gcAnd η_gdFor the efficiency for charge-discharge of power-type energy storage；

Set up power-type energy storage rated capacity configuration calculation formula be：

In formula：E_g,rateFor power-type energy storage rated capacity；P_g(t) it is consideration transverter energy conversion efficiency and discharge and recharge The charge-discharge electric power of power-type energy storage after efficiency；ΔT_gFor power-type energy storage discharge and recharge time interval；SOC_g,0For power-type energy storage Initial state-of-charge；SOC_g,max、SOC_g,minFor the upper lower limit value of power-type energy storage charge state；

Work as SOC_g,0Meet equation, and E_g,rateWhen taking equal sign, the rated capacity of power-type energy storage is minimum.

Further, the power-type energy storage includes：Ultracapacitor, flywheel energy storage, super conductive magnetic storage energy；The energy type Energy storage includes lithium ion battery.

Further, described a kind of mixed energy storage system capacity configuration optimizing method for optimizing operation towards power distribution network, The mixed energy storage system capacity configuration scheme for proposing to optimize operation towards power distribution network, including：Set up hybrid energy-storing complete Life cycle cost model, for evaluating the economy under different allocation plans, is held with obtaining most economical mixed energy storage system Configuration scheme is measured, detailed process is：

Min LCC=C_inv+C_rep+C_om+C_scr-C_res

C_inv=C_pinvP_rate+C_einvE_rate

C_scr=(C_pscrP_rate+C_escrE_rate)(n+1)(P/F,i,T)

C_res=σ_res(C_inv+C_rep)(P/F,i,T)

In formula：C_inv、C_rep、C_om、C_scr、C_resRespectively HESS initial outlay cost, renewal displacement cost, operation maintenance Cost, scrap processing cost and reclaim residual value；C_pinv、C_einv、C_prep、C_einv、C_pom、C_eom、C_pscr、C_escrRespectively unit power Cost coefficient, unit capacity cost coefficient, unit power update cost coefficient, unit capacity and update cost coefficient, unit power O＆M cost coefficient, unit capacity O＆M cost coefficient, unit power scrap processing cost coefficient and unit capacity scraps processing Cost coefficient；P_rate、E_rateRespectively HESS rated power and HESS rated capacity；(P/F, i, t)=(1+i)^-t；W_ess (t) it is year discharge and recharge；σ_resTo reclaim salvage value rate.

Further, the σ_resTake 3%~5%.

Beneficial effects of the present invention：

The present invention proposes a kind of mixed energy storage system capacity configuration optimizing method for optimizing operation towards power distribution network, can be notable The rated power and rated capacity of energy-storage system are reduced, is that HESS capacity is distributed rationally there is provided a kind of new method, with certain Reference value；Meanwhile, HESS application can not only stabilize the fluctuation that photovoltaic is exerted oneself, and promote dissolving for photovoltaic energy, and The optimization operation of power distribution network can be participated in, the economy and reliability of distribution network operation can be effectively improved.

Brief description of the drawings

Fig. 1 is the structure chart of light-preserved system in an embodiment of the present invention；

Fig. 2 is the photovoltaic plant schematic diagram containing HESS；

Fig. 3 is typical day operation curve map；

Fig. 4 is imbalance power EEMD decomposition result figures；

Fig. 5 is IMF instantaneous frequencys-time plot.

Embodiment

In order to make the purpose , technical scheme and advantage of the present invention be clearer, with reference to embodiments, to the present invention It is further elaborated.It should be appreciated that the specific embodiments described herein are merely illustrative of the present invention, it is not used to Limit the present invention.

The application principle of the present invention is explained in detail below in conjunction with the accompanying drawings.

A kind of mixed energy storage system capacity configuration optimizing method for optimizing operation towards power distribution network of the present invention, is obtained first Light-preserved system participate in distribution network operation it is optimal exert oneself with photovoltaic is actual exert oneself between imbalance power, as mixing store up The reference power of energy, is decomposed into a series of intrinsic mode functions, by passing using set empirical mode decomposition by imbalance power Hilbert transform is returned to obtain instantaneous frequency-time graph of each intrinsic mode function；According to power-type and energy type energy storage Characteristic, at least determines crossover frequency, using power-type energy storage and energy type energy storage with instantaneous frequency-time graph aliasing for principle The high fdrequency component and low frequency component to imbalance power are stabilized respectively；Finally, it is considered to efficiency for charge-discharge, state-of-charge and complete Life cycle cost, proposes power-type and energy type energy storage rated power and rated capacity economic optimum allocation plan.The present invention Relative to conventional method, it can not only compensate new energy and exert oneself, improve the digestion capability to new energy, and can be actively engaged in The optimization operation of power distribution network.

Embodiment one

Light-preserved system in the embodiment of the present invention is as shown in figure 1, C₁For unidirectional DC/DC converters, main realization boosting work( Can be with MPPT maximum power point tracking (MPPT)；C₂For two-way DC/DC converters, the controllable discharge and recharge of power is realized, makes system grid connection The active optimal active power to participate in power distribution network optimization operation；C₃For DC/AC inverters (i.e. photovoltaic DC-to-AC converter), for stable DC bus-bar voltage, is realized active and reactive optimal grid-connected；Photovoltaic generation unit (photovoltaic battery array in such as Fig. 1) and energy storage Unit is collected to common DC bus by respective DC/DC current transformers, then is connected to power distribution system through public DC/AC inverters System；It is assumed that light-preserved system is using to power network power output as positive direction.

A kind of mixed energy storage system capacity configuration optimizing method for optimizing operation towards power distribution network in the embodiment of the present invention, Comprise the following steps：

1st, set up light-preserved system and participate in power distribution network optimal operation model, solve and obtain light-preserved system participation distribution network operation It is optimal to exert oneself, and by light-preserved system participate in distribution network operation it is optimal exert oneself with photovoltaic is actual exert oneself between imbalance power make For the reference power of mixed energy storage system；

Foregoing light-preserved system, which participates in power distribution network optimal operation model, includes object function and constraints；

The embodiment of the present invention chooses the whole network loss minimization as object function, is stored up for the light proposed in the embodiment of the present invention System, the whole network network loss is except the loss P of transmission line of electricity_L.loss(t) (be represented by each node of system active injection power it With), in addition it is also necessary to consider the running wastage P of light-preserved system_PESS.loss(t) (the predominantly efficiency for charge-discharge of energy-storage system and its charge and discharge Energy loss when electricity condition is changed), therefore：The whole network loss minimization object function is：

P_PESS.loss(t)=(1- η_ch)P_ch(t)+(1-η_dis)P_dis(t)+ξ(t)·ΔP_ESS.loss

In formula：T is the cycle of operation；Δ t is operation step-length；N is system node number, P_i(t) it is the injection of t period node is Active power；ξ (t) is that energy-storage system changes dimension from the t-1 periods to the charging and discharging state of t periods, and value is 0 or 1；η_ch、η_dis For the efficiency for charge-discharge of energy-storage system；P_ch(t)、P_dis(t) it is the charge-discharge electric power of t period energy-storage systems；ΔP_ESS.lossDuring for t The loss that section energy-storage system is produced when charging and discharging state switches, preferably takes the 0.5% of energy-storage system rated capacity.

The constraints bag in the constraints of power distribution network optimization operation, the embodiment of the present invention is participated in view of light-preserved system Include：System load flow constraint, the constraint of feeder line outlet power, node voltage constraint, line power constraint and light-preserved system operation are about Beam；

The grid trend constraint is：

The substation feeder outlet power is constrained to：

The node voltage is constrained to：

The line power is constrained to：

The light-preserved system operation is constrained to：

P_PESS(t)=P_PV(t)+P_ESS(t)

P_PV(t)=P_PV.MPPT(t)

-P_ch.max≤P_ESS(t)≤P_dis.max

Because energy-storage system state of charge has absolute continuity in sequential, it is in strict accordance with time sequencing according to filling Electric discharge

Watt level carries out cumulative calculation, therefore light-preserved system operation constraint also includes：

S (t+1)=(1- σ) S (t)+P_ESS(t)Δt

S_min≤S(t)≤S_max

S (0)=S (T)

In above-mentioned formula：I, j are system node number；U_i(t)、U_j(t) it is t periods node i and node j point voltage magnitudes； G_ij、B_ijTransconductance and mutual susceptance respectively between node i and node j；δ_ij(t) it is the phase between t periods node i and node j Potential difference；P_Ki(t)、Q_Ki(t) it is respectively that t periods feeder line exports active and reactive power；P_PESSi(t)、Q_PESSi(t) it is t period nodes The active reactive of light-preserved system (i.e. inverter C3) is exerted oneself at i；P_Di(t)、Q_Di(t) for load active at t period node is and Reactive power；P_ij(t) it is the line power between t node i and node j；Subscript "-" and subscript " _ " difference table of variable Show the upper and lower bound of variable；P_PESSAnd Q (t)_PESS(t) be respectively t periods inverter output active and reactive power；P_PV (t) it is t period photovoltaic generation unit active power of output；P_MPPTIt is DC/DC current transformers according to maximum during MPPT control strategies Output is active；P_ESS(t) it is the active output of t period energy-storage systems, is divided into electric discharge, free time, three kinds of states of charging；P_ch(t) and P_dis(t) be respectively t period energy-storage systems charge-discharge electric power；η_chAnd η_disRespectively energy-storage system efficiency for charge-discharge；P_ch.maxWith P_dis.maxRespectively charge-discharge electric power bound；S (t) is the dump energy of t period energy-storage systems；σ is put certainly for energy-storage system Electric rate；S_minAnd S_maxThe respectively bound of energy-storage system dump energy.

The solution obtains light-preserved system participation the optimal of distribution network operation and exerted oneself, and is specially：Using two-dimension dynamic programming Algorithm for Solving light-preserved system participates in power distribution network optimal operation model, show that light-preserved system participates in the optimal of distribution network operation and exerted oneself； More specifically comprise the following steps：

Step 2.1：Light-preserved system participation power distribution network optimal operation model is converted into the operable model of two-dimension dynamic programming, Including stage t, state vector S (s_Pt,s_Qt), decision vector U (u_Pt,u_Qt), strategy, state transition equation and target function；

Step 2.2：Initialization, light-preserved system and distribution network system primary data needed for input；

Step 2.3：Determine the permission state set of t-1 period light-preserved systems；

Step 2.4：It is determined that meeting the permission state set of t period light-preserved systems under constraints；

Step 2.5：Calculate correspondence decision variable and target function value；

Step 2.6：Whether judge index function is optimal, if so, then skipping to step 2.7；Otherwise, step 2.8 is skipped to；

Step 2.7：Preserve current state variable, decision variable and object function；

Step 2.8：Judge whether the permission state set of traversal t-1 period light-preserved systems, if so, skipping to step 2.9；It is no Then, step 2.3 is skipped to；

Step 2.9：Judge whether the permission state set of traversal t period light-preserved systems, if so, skipping to step 2.10；If It is no, skip to step 2.4；

Step 2.10：Judge whether t=T, finished if so, then calculating, output result；If it is not, then t=t+1, and skip to step Rapid 2.3.

Further, the stage t is：One full schedule cycle T is divided into several periods, remembers that the period is Stage, stage sequence number is labeled as t, t ∈ { 1,2,3 ... T }, and the adjacent phases time difference is Δ t；

State S (the s_Pt,s_Qt) be：The dump energy S for choosing energy-storage system in light-preserved system is used as state S_Pt, and will Electricity difference between its discretization, adjacent states is Δ S；Choose the remaining reactive capability of light-preserved system inverter and be used as state S_Qt, and By its discretization, the reactive capability difference between adjacent states is Δ Q；

Decision-making U (the u_Pt,u_Qt) be：By the P in light-preserved system each period_PESSAnd Q (t)_PESS(t) become as decision-making Measure u_PtAnd u_Qt, it must is fulfilled for light-preserved system operation constraint；

The strategy is：The sequence of the decision variable composition in each stage；

The state transition equation includes：S_PtState transition equation and S_QtState transition equation；The S_PtState Equation of transfer is S (t+1)=(1- σ) S (t)+P_ESS(t) Δ t, (i.e. energy-storage system electricity and charge-discharge electric power relation equation)；Institute State S_QtIntersegmental in the absence of absolute transfer relationship in adjacent time, it allows state set by current state S_PtHold with inverter Amount determines that its state transition equation is

The target function is：It regard t phase targets function as stage target function V_t(S(s_Pt,s_Qt),U(u_Pt, u_Qt)), then the in-process metrics function in t stages is：

In formula：D(u_Pt,u_Qt) represent t stage conditions permission decision-making set, V_tExpression stage target function.

It is described by light-preserved system participate in distribution network operation it is optimal exert oneself with photovoltaic is actual exert oneself between imbalance power Calculation formula be：

P_net(t)=P_Opt(t)-P_PV(t)

In formula：P_net(t) it is imbalance power；P_Opt(t) participate in that distribution network operation is optimal exerts oneself for light-preserved system；P_PV(t) For photovoltaic generation unit active power of output.

2nd, decomposing light-preserved system by improving Hilbert-Huang transform participates in that distribution network operation is optimal exerts oneself and photovoltaic plant Imbalance power between actually exerting oneself, high and low frequency two parts wave component, and profit respectively are divided into by imbalance power With power-type energy storage and energy type energy storage compensation high-frequency fluctuation component and low-frequency fluctuation component；

In the embodiment of the present invention, participate in that distribution network operation is optimal exerts oneself and light using mixed energy storage system light-preserved system of dissolving Overhead utility is actual exert oneself between imbalance power, its structure is as shown in Figure 2.Photovoltaic battery array by DC/DC converters with Dc bus is connected, the mixed energy storage system being made up of energy type energy storage and power-type energy storage by two-way DC/DC converters with Dc bus is connected, absorbed by hybrid energy-storing, delivered power to stabilize photovoltaic plant real output, meet light-preserved system ginseng Optimal with distribution network operation is exerted oneself.

Imbalance power is obtained a series of intrinsic mode functions by the embodiment of the present invention by gathering empirical mode decomposition (IMF) all IMF instantaneous frequency-time graph, and using recurrence Hilbert transform is obtained.According to two adjacent IMF it Between frequency alias minimum principle determine the crossover frequency of imbalance power, imbalance power is resolved into high and low frequency two parts Wave component.In figure：P_n,refFor energy type energy storage power output reference value, i.e., all low frequency component sums；P_g,refFor power-type Energy storage power output reference value, i.e., all high fdrequency component sums；P_n,rate、P_g,rateRespectively energy type energy storage and power-type energy storage Specified charge-discharge electric power；SOC_n,max、SOC_n,minThe respectively maximum and minimum value of energy type energy storage charge state；SOC_n (t) it is the state-of-charge of energy type energy storage t；SOC_g,max、SOC_g,minThe respectively maximum of power-type energy storage charge state And minimum value；SOC_g(t) it is the state-of-charge of power-type energy storage t.

Improving Hilbert-Huang transform includes set empirical mode decomposition (EEMD) and recurrence Hilbert transform two parts. EEMD essence is into a series of data sequences with different characteristic time scale, i.e. intrinsic mode function by signal decomposition (IMF), each IMF is imbalance power P in fact_net(t) a kind of oscillation mode, is obtained by recurrence Hilbert transform Each IMF instantaneous frequency-time graph, so that the degree of accuracy of its instantaneous frequency of IMF is improved, and EEMD decomposable processes determine c_i (t) instantaneous frequency is generally higher than c_i+1(t) instantaneous frequency, therefore need to determine optimal in many instantaneous frequency-time graphs Crossover frequency so that frequency alias is minimum between two adjacent IMF, and imbalance power is divided into high and low frequency two parts ripple Dynamic component, and it is utilized respectively power-type energy storage and energy type energy storage compensation high-frequency fluctuation component and low-frequency fluctuation component.Below will The detailed process for being improved Hilbert-Huang transform decomposition is illustrated.

1) EEMD decomposes imbalance power

Light-preserved system participate in distribution network operation it is optimal exert oneself and photovoltaic plant is actual exert oneself between imbalance power be：

P_net(t)=P_Opt(t)-P_PV(t)

In formula：P_Opt(t) participate in that distribution network operation is optimal exerts oneself for light-preserved system；P_PV(t) it is the reality output of photovoltaic plant Power.

Because imbalance power is nonlinear change, the embodiment of the present invention is decomposed using EEMD to it, EEMD points Solution adds white noise acoustic disturbance signal to be decomposed, and carries out multiple empirical mode decomposition (EMD), obtained by repeatedly decomposing IMF population means be considered as final IMF.This method can be prevented effectively from modal overlap, improve EMD discomposing effects, specific steps It is as follows：

If primary signal is x (t), addition white Gaussian noise disturbance w (x) is obtained on x (t)：

s_j(t)=x (t)+w_j(x), j=1,2 ..., m

In formula：s_j(t) the imbalance power signal added for jth time after white Gaussian noise disturbance；w_j(x) to be separate White noise acoustic disturbance；M is that EMD carries out number of times.

s_j(t) signal resolves into n IMF, i.e. c by EMD_1j(t),c_2j(t),…,c_nj, and a remainder r (t)_j (t) it is

s_j(t)=c_1j(t)+c_2j(t)+…+c_nj(t)+r_j(t)

In formula：s_j(t) the imbalance power signal added for jth time after white Gaussian noise disturbance；c_nj(t) it is to pass through EMD N-th of the intrinsic mode function decomposed；r_j(t) remainder after being decomposed for EMD.

Decomposed by m EMD, obtain m group decomposition results：

[c₁₁(t),c₂₁(t),…c_i1(t),…c_n1(t),r₁(t)],

[c_1j(t),c_2j(t),…c_ij(t),…c_nj(t),r_j(t)],

[c_1m(t),c_2m(t),…c_im(t),…c_nm(t),r_m(t)]。

In formula：c_ij(t) i-th obtained of IMF is decomposed for jth time；r_j(t) obtained remainder is decomposed for jth time.

Obtained all IMF are decomposed by m EMD and residual components its averages is：

In formula：To decompose the average of obtained all IMF and residual components by m EMD；M is EMD points Solve number of times.

Therefore, the P after EEMD is decomposed_net(t) signal is represented by：

In formula：P_net(t) it is imbalance power signal；For by m EMD decompose obtained all IMF with The average of residual components.

2) IMF recurrence Hilbert transform

Primary signalIf c_i,0(t) it is H [c after Hilbert transform_i,0(t)], then：

Analytic signal z_i,0(t) it is represented by：

z_i,0(t)=c_i,0(t)+jH[c_i,0(t)]=a_i,0(t)exp[jφ_i,0(t)]

In formula：a_i,0(t) it is c_i,0(t) instantaneous amplitude,φ_i,0(t) it is c_i,0(t) Instantaneous phase,

By c_i,0(t) it is expressed as instantaneous amplitude a_i,0(t) with pure FM signal cos φ_i,0(t) product, then：

c_i,0(t)=a_i,0(t)cosφ_i,0(t)

In formula：c_i,0(t) it is instantaneous amplitude a_i,0(t) with pure FM signal cos φ_i,0(t) product.

With pure FM Function c_i,1(t)=cos φ_i,0(t) as new signal, and Hilbert transform is carried out to it, obtained To new instantaneous amplitude and pure FM signal.Recursive calculation is constantly carried out, its recurrence formula is：

Often corresponding magnitude function and phase function are just can obtain by a recursive calculation：

Recursive calculation is constantly carried out until magnitude function a_i,j(t) it is intended to 1, then c_i,m+1(t)=cos φ_i,m(t) it is, comprehensive Above-mentioned recursive computing steps, signal c_i,0(t) it is represented by：

Instantaneous frequency is represented by：

In formula：ω_i(t) it is instantaneous frequency；φ_i,m(t) it is instantaneous phase.

3) crossover frequency is determined

IMF obtains c by recurrence Hilbert transform_i(t) instantaneous frequency-time graph, due to i increase, c_i (t) waveform is more smooth, therefore f_i(t) instantaneous frequency is also lower.Adjacent f two-by-two_iAnd f (t)_i+1(t) always there is frequency to mix Folded phenomenon, therefore need to determine that optimal crossover frequency causes c_iAnd c (t)_i+1(t) energy aliasing is minimum.

If crossover frequency f_gBetween g and g+1 IMF, then c₁(t),c₂(t),…,c_g(t) it is considered as high-frequency fluctuation point Amount, this portion of energy is stabilized by power-type energy storage；c_g+1(t),c_g+2(t),…,c_n(t) it is considered as low-frequency fluctuation component, this part energy Amount is stabilized by energy type energy storage, then the reference power of energy type energy storage and power-type energy storage is：

In formula：P_g,ref(t)、P_n,ref(t) it is power-type and the reference power of energy type energy storage；c_i(t) it is natural mode of vibration letter Number, its value substituted into is all intrinsic mode function averagesTo be every in the m group decomposition result matrixes that are previously obtained The average of one row)；R (t) is residual components.

The charge-discharge characteristic of energy type energy storage and power-type energy storage is taken into full account, the use longevity of energy storage device is considerably increased Life, can effectively reduce system operation cost.

3rd, obtain the mixed energy storage system capacity for optimizing operation towards power distribution network and distribute strategy rationally.

It is described propose be towards the detailed process of mixed energy storage system capacity configuration scheme that power distribution network optimizes operation：

The rated power of energy type energy storage is set as P_n,rate, it is considered to transverter energy conversion efficiency and energy type energy storage Efficiency for charge-discharge, the rated power of energy type energy storage configures calculation formula and is：

In formula：P_n,rateFor the rated power of energy type energy storage；t_n,0For energy type energy storage discharge and recharge initial time；η_DC-DCWith η_DC-ACFor the energy conversion efficiency of transverter；η_ncAnd η_ndFor the efficiency for charge-discharge of energy type energy storage；P_n,ref(t) stored up for energy type The reference power of energy.

The rated capacity configuration of energy type energy storage is calculated：

If the initial state-of-charge of energy type energy storage is SOC_n,0, the state-of-charge at k moment is：

In formula：ΔT_nFor energy type energy storage discharge and recharge time interval；SOC_n,kFor the state-of-charge of k moment energy type energy storage； SOC_n,min、SOC_n,maxFor the bound of energy type energy storage charge state；P_n(t) it is consideration transverter energy conversion efficiency and charge and discharge The charge-discharge electric power of energy type energy storage after electrical efficiency, its calculation formula is：

In formula：P_n,ref(t) it is the reference power of energy type energy storage；η_DC-DCAnd η_DC-ACFor the energy conversion efficiency of transverter； η_ncAnd η_ndFor the efficiency for charge-discharge of energy type energy storage.

A length of T when research_nWhen, energy type energy storage rated capacity E_n,rateCalculation formula be

In formula：E_n,rateFor energy type energy storage rated capacity；P_n(t) it is consideration transverter energy conversion efficiency and discharge and recharge The charge-discharge electric power of energy type energy storage after efficiency；ΔT_nFor energy type energy storage discharge and recharge time interval；SOC_n,0For energy type energy storage Initial state-of-charge；SOC_n,min、SOC_n,maxFor the bound of energy type energy storage charge state

Work as SOC₀Meet equation, and E_n,rateWhen taking equal sign, the rated capacity of energy type energy storage is minimum.

In formula：P_n(t) it is the charge and discharge electric work of energy type energy storage after consideration transverter energy conversion efficiency and efficiency for charge-discharge Rate；ΔT_nFor energy type energy storage discharge and recharge time interval；SOC_n,0For the initial state-of-charge of energy type energy storage；SOC_n,max、 SOC_n,minFor the upper lower limit value of energy type energy storage charge state.

The rated power configuration of power-type energy storage is calculated：

The rated power of power-type energy storage is set as P_g,rate, it is considered to transverter energy conversion efficiency and power-type energy storage Efficiency for charge-discharge, the rated power of power-type energy storage configures calculation formula and is：

In formula：P_g,rateFor the rated power of power-type energy storage；t_g,0For power-type energy storage discharge and recharge initial time；η_DC-DCWith η_DC-ACFor the energy conversion efficiency of transverter；η_gcAnd η_gdFor the efficiency for charge-discharge of power-type energy storage；P_g,ref(t) stored up for power-type The reference power of energy.

The rated capacity configuration of power-type energy storage is calculated：

If the initial state-of-charge of power-type energy storage is SOC_g,0, the state-of-charge at k moment is：

In formula：ΔT_gFor power-type energy storage discharge and recharge time interval；SOC_g,kFor the state-of-charge of k moment power-type energy storage； SOC_g,min、SOC_g,maxFor the bound of power-type energy storage charge state；P_g(t) it is consideration transverter energy conversion efficiency and charge and discharge The charge-discharge electric power of power-type energy storage after electrical efficiency, its calculation formula is：

In formula：P_g,ref(t) it is the reference power of power-type energy storage；η_DC-DCAnd η_DC-ACFor the energy conversion efficiency of transverter； η_gcAnd η_gdFor the efficiency for charge-discharge of power-type energy storage.

A length of T when research_gWhen, power-type energy storage rated capacity E_g,rateCalculation formula be

In formula：E_g,rateFor power-type energy storage rated capacity；P_g(t) it is consideration transverter energy conversion efficiency and discharge and recharge The charge-discharge electric power of power-type energy storage after efficiency；ΔT_gFor power-type energy storage discharge and recharge time interval；SOC_g,0For power-type energy storage Initial state-of-charge；SOC_g,max、SOC_g,minFor the upper lower limit value of power-type energy storage charge state；

Work as SOC_g,0Meet equation, and E_g,rateWhen taking equal sign, the rated capacity of power-type energy storage is minimum.

In formula：E_g,rateFor power-type energy storage rated capacity；P_g(t) it is consideration transverter energy conversion efficiency and discharge and recharge The charge-discharge electric power of power-type energy storage after efficiency；ΔT_gFor power-type energy storage discharge and recharge time interval；SOC_g,0For power-type energy storage Initial state-of-charge；SOC_g,max、SOC_g,minFor the upper lower limit value of power-type energy storage charge state.

4th, hybrid energy-storing economic evaluation

The embodiment of the present invention considers the input cost of hybrid energy-storing equipment, sets up the complete of mixed energy storage system (HESS) Life cycle cost model,.Overall life cycle cost LCC refers to all produced in whole life cycle directly, indirectly Etc. all expenses, it is assumed that the HESS life-spans are T, discount rate is i, and HESS elements update displacement number of times for n, then life cycle management into This model is：

Min LCC=C_inv+C_rep+C_om+C_scr-C_res

C_inv=C_pinvP_rate+C_einvE_rate

C_scr=(C_pscrP_rate+C_escrE_rate)(n+1)(P/F,i,T)

C_res=σ_res(C_inv+C_rep)(P/F,i,T)

In formula：C_inv、C_rep、C_om、C_scr、C_resRespectively HESS initial outlay cost, renewal displacement cost, operation maintenance Cost, scrap processing cost and reclaim residual value；C_pinv、C_einv、C_prep、C_einv、C_pom、C_eom、C_pscr、C_escrRespectively unit power Cost coefficient, unit capacity cost coefficient, unit power update cost coefficient, unit capacity and update cost coefficient, unit power O＆M cost coefficient, unit capacity O＆M cost coefficient, unit power scrap processing cost coefficient and unit capacity scraps processing Cost coefficient；P_rate、E_rateRespectively HESS rated power and HESS rated capacity (calculate energy type energy storage when, P_rateJust Represent P_n,rate；When calculating power-type energy storage, P_rateJust represent P_g,rate, similarly push away to obtain E_rate)；(P/F, i, t)=(1+i)^-t； W_ess(t) it is year discharge and recharge；σ_resTo reclaim salvage value rate, 3%~5% is taken.

Embodiment two

(adopted using the real output data of Jiangsu Province's photovoltaic plant certain typical day in 2013 in the embodiment of the present invention The sample time is 1min), simulation analysis are carried out in Matlab.One day excellent is carried out to light-preserved system using 33 Node power distribution systems Change, be utilized respectively photovoltaic power Forecasting Methodology and load forecasting method obtains photovoltaic plant sunrise force curve and load day operation is bent Line, as shown in figure 3, the light-preserved system participation active distribution network optimal operation model that the present invention is carried is built, using two-dimentional dynamic Planning algorithm solving model, obtains light-preserved system participation the optimal of distribution network operation and exerts oneself.The typical daylight storage system is participated in Distribution network operation is optimal to exert oneself with the actual difference exerted oneself of photovoltaic as imbalance power data, and Fig. 4 gives imbalance power EEMD analysis results.As seen from the figure, intrinsic mode function c_i(t) amplitude and frequency reduces with i increase, and zero Value is oscillated around.Instantaneous frequency-the time graph obtained with reference to recurrence Hilbert transform, as shown in Figure 5.

Knowable to analysis, c₄And c (t)₅(t) aliasing is minimum, therefore takes i=4 as frequency partition separation, by uneven work( Rate is divided into high-low frequency weight, then power-type energy storage and energy type energy storage power output reference value are P_g,ref(t)=c₁(t)+c₂ (t)+c₃(t)+c₄(t)；P_n,ref(t)=c₅(t)+c₆(t)+r(t)。

With reference to the characteristic of power-type energy storage and energy type energy storage, the embodiment of the present invention is using the different configuration side of following four Case.

Scheme 1 is lithium ion battery；Scheme 2 is lithium ion battery and ultracapacitor；Scheme 3 is lithium ion battery and flown Take turns energy storage；Scheme 4 is lithium ion battery and super conductive magnetic storage energy.Each energy storage parameter as shown in table 1, is carried out different according to the above method The configuration of rated power and rated capacity under scheme, and estimate mixed energy storage system (HESS) life-span, it is assumed that HESS life-cycle Cycle is 10 years, then can calculate the cost of investment of HESS under each allocation plan, and this calculating process is prior art, is not gone to live in the household of one's in-laws on getting married State, specific result of calculation is as shown in table 2.

Table 1

Table 2

By analyzing table 2, the rated power and rated capacity of traditional Hilbert-Huang transform configuration are substantially high In the configuration result of the inventive method, such as scheme 2, the rated power and specified appearance of traditional Hilbert-Huang transform configuration Amount is respectively 2.18MW/0.37MWh, 3.60MW/0.19MWh, the rated power and rated capacity of the inventive method configuration Respectively 1.39MW/0.29MWh, 2.64MW/0.17MWh.It is far excellent using HESS schemes economy from the point of view of cost of investment In single energy storage scheme, because HESS schemes can greatly increase the service life of lithium ion battery, so as to reduce the investment of energy storage Cost, the wherein economy of scheme 2 are optimal, therefore can optimize the lower hybrid energy-storing capacity configuration of operation application as towards power distribution network Suggested design.

The general principle and principal character and advantages of the present invention of the present invention has been shown and described above.The technology of the industry Personnel are it should be appreciated that the present invention is not limited to the above embodiments, and the simply explanation described in above-described embodiment and specification is originally The principle of invention, without departing from the spirit and scope of the present invention, various changes and modifications of the present invention are possible, these changes Change and improvement all fall within the protetion scope of the claimed invention.The claimed scope of the invention by appended claims and its Equivalent thereof.

Claims (10)

1. it is a kind of towards power distribution network optimize operation mixed energy storage system capacity configuration optimizing method, it is characterised in that including with Lower step：

Set up light-preserved system participate in power distribution network optimal operation model, solve obtain light-preserved system participate in distribution network operation it is optimal go out Power, and using light-preserved system participate in distribution network operation it is optimal exert oneself with photovoltaic is actual exert oneself between imbalance power as mixing The reference power of energy-storage system；

By improve Hilbert-Huang transform decompose light-preserved system participate in distribution network operation it is optimal exert oneself with photovoltaic plant it is actual go out Imbalance power between power, is divided into high and low frequency two parts wave component, and be utilized respectively power by imbalance power Type energy storage and energy type energy storage compensation high-frequency fluctuation component and low-frequency fluctuation component；

Obtain the mixed energy storage system capacity for optimizing operation towards power distribution network and distribute strategy rationally.

2. a kind of mixed energy storage system capacity side of distributing rationally for optimizing operation towards power distribution network according to claim 1 Method, it is characterised in that：The light-preserved system, which participates in power distribution network optimal operation model, includes object function and constraints,

The object function is：

$F=min\underset{t=1}{\overset{T}{\Σ}}\left(P_{L.loss}\right(t)+P_{PESS.loss}(t\left)\right)\mathrm{\Δ}t$

$P_{L.loss}\left(t\right)=\underset{i=1}{\overset{N}{\Σ}}{P}_i\left(t\right)\mathrm{\Δ}t$

P_PESS.loss(t)=(1- η_ch)P_ch(t)+(1-η_dis)P_dis(t)+ξ(t)·ΔP_ESS.loss

In formula：T is the cycle of operation；Δ t is operation step-length；N is distribution system nodes, P_i(t) being injected with for t period node is Work(power；ξ (t) is that energy-storage system changes dimension from the t-1 periods to the charging and discharging state of t periods, and value is 0 or 1；η_ch、η_disFor The efficiency for charge-discharge of energy-storage system；P_ch(t)、P_dis(t) it is the charge-discharge electric power of t period energy-storage systems；ΔP_ESS.lossFor the t periods The loss that energy-storage system is produced when charging and discharging state switches.

3. a kind of mixed energy storage system capacity side of distributing rationally for optimizing operation towards power distribution network according to claim 2 Method, it is characterised in that：The constraints is：

$P_{Ki}\left(t\right)+P_{PESSi}\left(t\right)-P_{Di}\left(t\right)=U_i\left(t\right)\underset{j=1}{\overset{n-1}{\mathrm{\Σ}}}{U}_j\left(t\right)\left(G_{ij}{\mathrm{cos\δ}}_{ij}\left(t\right)+B_{ij}{\mathrm{sin\δ}}_{ij}\left(t\right)\right)$

$Q_{Ki}\left(t\right)+Q_{PESSi}\left(t\right)-Q_{Di}\left(t\right)=U_i\left(t\right)\underset{j=1}{\overset{n-1}{\mathrm{\Σ}}}{U}_j\left(t\right)\left(G_{ij}{\mathrm{sin\δ}}_{ij}\left(t\right)-B_{ij}{\mathrm{cos\δ}}_{ij}\left(t\right)\right)$

${\underset{\‾}{P}}_{Ki}\≤P_{Ki}\left(t\right)\≤{\stackrel{\‾}{P}}_{Ki}$

${\underset{\‾}{Q}}_{Ki}\≤Q_{Ki}\left(t\right)\≤{\stackrel{\‾}{Q}}_{Ki}$

${\underset{\‾}{U}}_i\≤U_i\left(t\right)\≤{\stackrel{\‾}{U}}_i$

$-{\stackrel{\‾}{P}}_{ij}\≤P_{ij}\left(t\right)\≤{\stackrel{\‾}{P}}_{ij}$

$\sqrt{{\left(P_{PESS}\right(t\left)\right)}^2+{\left(Q_{PESS}\right(t\left)\right)}^2}\≤S_{PESS.max}$

P_PESS(t)=P_PV(t)+P_ESS(t)

P_PV(t)=P_PV.MPPT(t)

-P_ch.max≤P_ESS(t)≤P_dis.max

S (t+1)=(1- σ) S (t)+P_ESS(t)Δt

S_min≤S(t)≤S_max

S (0)=S (T)

In formula：I, j are distribution network systems node number；U_i(t)、U_j(t) it is t periods node i and node j point voltage magnitudes；G_ij、B_ijPoint Transconductance and mutual susceptance that Wei be between node i and node j；δ_ij(t) it is the phase difference between t periods node i and node j；P_Ki (t)、Q_Ki(t) it is respectively that the outlet of t periods feeder line is active and idle；P_PESSi(t)、Q_PESSi(t) it is light-preserved system at t period node is Active and reactive power；P_Di(t)、Q_Di(t) it is the active and reactive power of load at t period node is；P_ij(t) it is t Line power between node i and node j；The subscript "-" and subscript " _ " of variable represent the upper and lower bound of variable respectively； P_PESSAnd Q (t)_PESS(t) be respectively t periods light-preserved system output active and reactive power；S_PESS.maxFor the specified appearance of inverter Amount；P_PV(t) it is t period photovoltaic generation unit active power of output；P_MPPTIt is photovoltaic DC/DC current transformers according to MPPT control strategies When maximum output it is active；P_ESS(t) it is the active output of t period energy-storage systems, is divided into electric discharge, free time, three kinds of states of charging； P_chAnd P (t)_dis(t) be respectively t period energy-storage systems charge-discharge electric power；η_chAnd η_disRespectively energy-storage system discharge and recharge is imitated Rate；P_ch.maxAnd P_dis.maxRespectively charge-discharge electric power bound；S (t) is the dump energy of t period energy-storage systems；σ is energy storage The self-discharge rate of system；S_minAnd S_maxThe respectively bound of energy-storage system dump energy.

4. a kind of mixed energy storage system capacity side of distributing rationally for optimizing operation towards power distribution network according to claim 1 Method, it is characterised in that：The solution obtains light-preserved system participation the optimal of distribution network operation and exerted oneself, and is specially：

Power distribution network optimal operation model is participated in using two-dimension dynamic programming Algorithm for Solving light-preserved system, show that light-preserved system participates in matching somebody with somebody The optimal of operation of power networks is exerted oneself；Specifically include following steps：

Step 2.1：Light-preserved system participation power distribution network optimal operation model is converted into the operable model of two-dimension dynamic programming, including Stage t, state vector S (s_Pt,s_Qt), decision vector U (u_Pt,u_Qt), strategy, state transition equation and target function；

Step 2.2：Initialization, light-preserved system and distribution network system primary data needed for input；

Step 2.3：Determine the permission state set of t-1 period light-preserved systems；

Step 2.4：It is determined that meeting the permission state set of t period light-preserved systems under constraints；

Step 2.5：Calculate correspondence decision variable and target function value；

Step 2.6：Whether judge index function is optimal, if so, then skipping to step 2.7；Otherwise, step 2.8 is skipped to；

Step 2.7：Preserve current state variable, decision variable and object function；

Step 2.8：Judge whether the permission state set of traversal t-1 period light-preserved systems, if so, skipping to step 2.9；Otherwise, Skip to step 2.3；

Step 2.9：Judge whether the permission state set of traversal t period light-preserved systems, if so, skipping to step 2.10；If it is not, jumping To step 2.4；

Step 2.10：Judge whether t=T, finished if so, then calculating, output result；If it is not, then t=t+1, and skip to step 2.3；

It is described by light-preserved system participate in distribution network operation it is optimal exert oneself with photovoltaic is actual exert oneself between imbalance power meter Calculating formula is：

P_net(t)=P_Opt(t)-P_PV(t)

In formula：P_net(t) it is imbalance power；P_Opt(t) participate in that distribution network operation is optimal exerts oneself for light-preserved system；P_PV(t) it is light Lie prostrate generator unit active power of output.

5. a kind of mixed energy storage system capacity side of distributing rationally for optimizing operation towards power distribution network according to claim 4 Method, it is characterised in that：

The stage t is：One full schedule cycle T is divided into several periods, remembers that the single period is stage, rank Section sequence number is labeled as t, t ∈ { 1,2,3 ... T }, and the adjacent phases time difference is Δ t；

State S (the s_Pt,s_Qt) be：The dump energy S for choosing energy-storage system in light-preserved system is used as state S_Pt, and by its from Electricity difference between dispersion, adjacent states is Δ S；Choose the remaining reactive capability of light-preserved system inverter and be used as state S_Qt, and by its Reactive capability difference between discretization, adjacent states is Δ Q；

Decision-making U (the u_Pt,u_Qt) be：By the P in light-preserved system each period_PESSAnd Q (t)_PESS(t) as decision variable u_Pt And u_Qt, it must is fulfilled for light-preserved system operation constraint；

The strategy is：The sequence of the decision variable composition in each stage；

The state transition equation includes：S_PtState transition equation and S_QtState transition equation；The S_PtState transfer Equation is S (t+1)=(1- σ) S (t)+P_ESS(t)Δt；The S_QtAbsolute transfer relationship is not present between adjacent two benches, its Allow state set by current state S_PtDetermined with inverter capacity, its state transition equation is：

The target function is：It regard t phase targets function as stage target function V_t(S(s_Pt,s_Qt),U(u_Pt,u_Qt)), then The target function in t stages is：

$F\left(S\right(s_{Pt},s_{Qt}\left)\right)=\underset{U(u_{Pt},u_{Qt})\∈D(u_{Pt},u_{Qt})}{min}\{V_t\left(S\right(s_{Pt},s_{Qt}),U(u_{Pt},u_{Qt}\left)\right)+F\left(S\right(s_{P(t-1)},s_{Q(t-1)}\left)\right)\}---\left(19\right)$

In formula：Represent the permission decision-making set of t stage conditions, V_tExpression stage target function.

6. a kind of mixed energy storage system capacity side of distributing rationally for optimizing operation towards power distribution network according to claim 1 Method, it is characterised in that：It is described by improve Hilbert-Huang transform decompose light-preserved system participate in distribution network operation it is optimal exert oneself and Photovoltaic plant is actual exert oneself between imbalance power, imbalance power is divided into high and low frequency two parts wave component, its Detailed process is：

Using set empirical mode decomposition EEMD to signal P_net(t) decomposed, the P after set empirical mode decomposition_net (t) it is represented by：

$P_{net}\left(t\right)=\underset{i=1}{\overset{n}{\Σ}}{c}_{i,0}\left(t\right)+r\left(t\right)=\underset{i=1}{\overset{n}{\Σ}}{\stackrel{\‾}{c}}_i\left(t\right)+\stackrel{\‾}{r}\left(t\right)$

In formula：P_net(t) it is imbalance power signal；c_i,0(t) it is i-th of intrinsic mode function IMF；R (t) is residual components；

For all intrinsic mode function averages；For all residual components averages；

To c_i,0(t) recurrence Hilbert transform, signal c are carried out_i,0(t) it is represented by：

$c_{i,0}\left(t\right)=\underset{j=1}{\overset{m}{\mathrm{\Π}}}{a}_{i,j}\left(t\right){\mathrm{cos\φ}}_{i,m}\left(t\right)$

In formula:c_i,0(t) it is i-th of intrinsic mode function IMF；a_i,j(t) it is intrinsic mode function c_i,j(t) magnitude function；

cosφ_i,m(t) it is recursive calculation to magnitude function a_i,j(t) when being intended to 1, intrinsic mode function c_i,m(t) instantaneous phase Position cosine value；M is decomposition number of times；

The instantaneous frequency for obtaining intrinsic mode function IMF is：

${\mathrm{\ω}}_i\left(t\right)=\frac{{\mathrm{d\φ}}_{i,m}\left(t\right)}{dt}$

In formula：ω_i(t) it is signal c_i,0(t) instantaneous frequency；φ_i,m(t) it is recursive calculation to magnitude function a_i,j(t) it is intended to 1 When,

Intrinsic mode function c_i,m(t) instantaneous phase, m is decomposition number of times；

Determine crossover frequency f_gIf, crossover frequency f_gBetween g and g+1 IMF, then c₁(t),c₂(t),…,c_g(t) it is considered as height Frequency wave component, this portion of energy is stabilized by power-type energy storage；c_g+1(t),c_g+2(t),…,c_n(t) it is considered as low-frequency fluctuation component, This portion of energy is stabilized by energy type energy storage, then the reference power of energy type energy storage and power-type energy storage is：

$\left\{\begin{array}{c}{P}_{g,ref}\left(t\right)=\underset{i=1}{\overset{g}{\mathrm{\Σ}}}{c}_i\left(t\right)\\ P_{n,ref}\left(t\right)=\underset{i=g+1}{\overset{n}{\mathrm{\Σ}}}{c}_i\left(t\right)+r\left(t\right)\end{array}\right.$

In formula：P_g,ref(t)、P_n,ref(t) be respectively power-type and energy type energy storage reference power；c_i(t) it is natural mode of vibration letter Number, its value substituted into is all intrinsic mode function averagesR (t) is residual components.

7. a kind of mixed energy storage system capacity side of distributing rationally for optimizing operation towards power distribution network according to claim 6 Method, it is characterised in that：It is described to be utilized respectively power-type energy storage and energy type energy storage compensation high-frequency fluctuation component and low-frequency fluctuation point The detailed process of amount is：

Set up energy type energy storage rated power configuration calculation formula be：

$P_{n,rate}=\max \left\{\left|\underset{t\∈\left(t_{n,0},t_{n,0}+T_n\right)}{\max }\left(P_{n,ref}\left(t\right)\right)\right|{\mathrm{\η}}_{DC-DC}{\mathrm{\η}}_{DC-AC}{\mathrm{\η}}_{nc},\frac{\left|\underset{t\∈\left(t_{n,0},t_{n,0}+T_n\right)}{\min }\left(P_{n,ref}\left(t\right)\right)\right|}{{\mathrm{\η}}_{DC-DC}{\mathrm{\η}}_{DC-AC}{\mathrm{\η}}_{nd}}\right\}$

In formula：P_n,rateFor the rated power of energy type energy storage；t_n,0For energy type energy storage discharge and recharge initial time, T_nDuring for research It is long；η_DC-DCAnd η_DC-ACFor the energy conversion efficiency of transverter；η_ncAnd η_ndFor the efficiency for charge-discharge of energy type energy storage；

Set up energy type energy storage rated capacity configuration calculation formula be：

$E_{n,rate}\≥\max \left\{\frac{\underset{t\∈T_n}{\max }{\∫}_0^{{\mathrm{k\ΔT}}_n}{P}_n\left(t\right)dt}{{\mathrm{SOC}}_{n,\max }-{\mathrm{SOC}}_{n,0}},\frac{\underset{t\∈T_n}{-\min }{\∫}_0^{{\mathrm{k\ΔT}}_n}{P}_n\left(t\right)dt}{{\mathrm{SOC}}_{n,0}-{\mathrm{SOC}}_{n,\min }}\right\}$

${\mathrm{SOC}}_{n,0}=\frac{\underset{t\∈T_n}{max}\[{\∫}_0^{{\mathrm{k\ΔT}}_n}{P}_n\left(t\right)dt\]{\mathrm{SOC}}_{n,min}-\underset{t\∈T_n}{min}\[{\∫}_0^{{\mathrm{k\ΔT}}_n}{P}_n\left(t\right)dt\]{\mathrm{SOC}}_{n,max}}{\underset{t\∈T_n}{max}\[{\∫}_0^{{\mathrm{k\ΔT}}_n}{P}_n\left(t\right)dt\]-\underset{t\∈T_n}{min}\[{\∫}_0^{{\mathrm{k\ΔT}}_n}{P}_n\left(t\right)dt\]}$

In formula：E_n,rateFor energy type energy storage rated capacity；P_n(t) after for consideration transverter energy conversion efficiency and efficiency for charge-discharge The charge-discharge electric power of energy type energy storage；ΔT_nFor energy type energy storage discharge and recharge time interval；SOC_n,0For the initial of energy type energy storage State-of-charge；SOC_n,max、SOC_n,minFor the upper lower limit value of energy type energy storage charge state；

Work as SOC_n,0Meet equation, and E_n,rateWhen taking equal sign, the rated capacity of energy type energy storage is minimum；

Set up power-type energy storage rated power configuration calculation formula be：

$P_{g,rate}=max\left\{\right|\underset{t\∈(t_{g,0},t_{g,0}+T_g)}{max}\left(P_{g,ref}\right(t\left)\right)|{\mathrm{\η}}_{DC-DC}{\mathrm{\η}}_{DC-AC}{\mathrm{\η}}_{gc},\frac{\left|\underset{t\∈(t_{g,0},t_{g,0}+T)}{\min }\left(P_{g,ref}\right(t\left)\right)\right|}{{\mathrm{\η}}_{DC-DC}{\mathrm{\η}}_{DC-AC}{\mathrm{\η}}_{gd}}\}$

In formula：P_g,rateFor the rated power of power-type energy storage；t_g,0For power-type energy storage discharge and recharge initial time；T_gDuring for research It is long；η_DC-DCAnd η_DC-ACFor the energy conversion efficiency of transverter；η_gcAnd η_gdFor the efficiency for charge-discharge of power-type energy storage；

Set up power-type energy storage rated capacity configuration calculation formula be：

$E_{g,rate}\≥\max \left\{\frac{\underset{t\∈T_g}{\max }{\∫}_0^{{\mathrm{k\ΔT}}_g}{P}_g\left(t\right)dt}{{\mathrm{SOC}}_{g,\max }-{\mathrm{SOC}}_{g,0}},\frac{\underset{t\∈T_g}{-\min }{\∫}_0^{{\mathrm{k\ΔT}}_g}{P}_g\left(t\right)dt}{{\mathrm{SOC}}_{g,0}-{\mathrm{SOC}}_{g,\min }}\right\}$

${\mathrm{SOC}}_{g,0}=\frac{\underset{t\∈T_g}{max}\[{\∫}_0^{{\mathrm{k\ΔT}}_g}{P}_g\left(t\right)dt\]{\mathrm{SOC}}_{g,min}-\underset{t\∈T_g}{min}\[{\∫}_0^{{\mathrm{k\ΔT}}_g}{P}_g\left(t\right)dt\]{\mathrm{SOC}}_{g,max}}{\underset{t\∈T_g}{\max }\[{\∫}_0^{{\mathrm{k\ΔT}}_g}{P}_g\left(t\right)dt\]-\underset{t\∈T_g}{\min }\[{\∫}_0^{{\mathrm{k\ΔT}}_g}{P}_g\left(t\right)dt\]}$

In formula：E_g,rateFor power-type energy storage rated capacity；P_g(t) after for consideration transverter energy conversion efficiency and efficiency for charge-discharge The charge-discharge electric power of power-type energy storage；ΔT_gFor power-type energy storage discharge and recharge time interval；SOC_g,0For the initial of power-type energy storage State-of-charge；SOC_g,max、SOC_g,minFor the upper lower limit value of power-type energy storage charge state；

Work as SOC_g,0Meet equation, and E_g,rateWhen taking equal sign, the rated capacity of power-type energy storage is minimum.

8. a kind of mixed energy storage system capacity side of distributing rationally for optimizing operation towards power distribution network according to claim 7 Method, it is characterised in that：The power-type energy storage includes：Ultracapacitor, flywheel energy storage, super conductive magnetic storage energy；The energy type storage Lithium ion battery can be included.

9. a kind of mixed energy storage system capacity side of distributing rationally for optimizing operation towards power distribution network according to claim 7 Method, it is characterised in that the proposition optimizes the mixed energy storage system capacity configuration scheme of operation, specific bag towards power distribution network Include：Hybrid energy-storing overall life cycle cost model is set up, it is most economical to obtain for evaluating the economy under different allocation plans Mixed energy storage system capacity configuration scheme, detailed process is：

Min LCC=C_inv+C_rep+C_om+C_scr-C_res

C_inv=C_pinvP_rate+C_einvE_rate

$C_{rep}=\underset{k=1}{\overset{n}{\Σ}}(C_{prep}{P}_{rate}+C_{einv}{E}_{rate})\[P/F,i,\left(\frac{kT}{n+1}\right)\]$

$C_{om}=C_{pom}{P}_{rate}(P/A,i,T)+\underset{t=1}{\overset{T}{\Σ}}{C}_{eom}{W}_{ess}\left(t\right)(P/F,i,T)$

C_scr=(C_pscrP_rate+C_escrE_rate)(n+1)(P/F,i,T)

C_res=σ_res(C_inv+C_rep)(P/F,i,T)

In formula：C_inv、C_rep、C_om、C_scr、C_resRespectively HESS initial outlay cost, update displacement cost, operation maintenance into Originally, scrap processing cost and reclaim residual value；C_pinv、C_einv、C_prep、C_einv、C_pom、C_eom、C_pscr、C_escrRespectively unit power into This coefficient, unit capacity cost coefficient, unit power update cost coefficient, unit capacity and update cost coefficient, unit power fortune Tie up cost coefficient, unit capacity O＆M cost coefficient, unit power scraps processing cost coefficient and unit capacity is scrapped and is processed into This coefficient；P_rate、E_rateRespectively HESS rated power and HESS rated capacity；(P/F, i, t)=(1+i)^-t；W_ess(t) For year discharge and recharge；σ_resTo reclaim salvage value rate.

10. a kind of mixed energy storage system capacity side of distributing rationally for optimizing operation towards power distribution network according to claim 9 Method, it is characterised in that：The σ_resTake 3%~5%.