CN106156921A - Based on the electric automobile photovoltaic charge station energy storage selection of configuration method that Copula is theoretical - Google Patents

Abstract

A kind of electric automobile photovoltaic charge station energy storage selection of configuration method theoretical based on Copula is disclosed herein, comprise the following steps: choose photovoltaic cells and charging electric vehicle load to go out power rate be stochastic variable, by measured data normalization, build the marginal distribution of each variable；Theoretical based on Copula, choose Gumbel-Copula and Clayton-Copula and build mixing Copula function to describe the dependency of asymmetric rear tail characteristic between variable；The year net load amount of sampled analog photovoltaic charge station on the basis of both combine probability density function of exerting oneself；Under the constraint of the condition such as stability bandwidth, confidence level, set up the energy storage Optimal Allocation Model with the photovoltaic electric minimum object function of vehicle charging station annual operating and maintenance cost；Matlab programming is optimized calculating to optimum stored energy capacitance.Consider runnability and economy, the integrated system energy storage allocation models correctness and feasibility considering dependency is analyzed, selects optimal allocation scheme, thus improve the efficiency of light energy utilization and systematic economy benefit.

Description

Based on the electric automobile photovoltaic charge station energy storage selection of configuration method that Copula is theoretical

Technical field

The present invention relates to energy storage selection of configuration field, a kind of electricity theoretical based on Copula Electrical automobile photovoltaic charge station energy storage selection of configuration method.

Background technology

Energy crisis and environmental problem are day by day serious, regenerative resource and electric automobile (Electric Vehicles, EVs) utilization shows great potential in terms of energy-saving and emission-reduction.Photovoltaic generation undulatory property, intermittent feature Make it dissolve on a large scale to acquire a certain degree of difficulty.A large amount of electric automobile chargings at random can increase the weight of the burden of electrical network, and The carbon emission amount indirectly produced is not preponderated, and improves the energy and environmental problem inconspicuous.In city Under environment, build this typical integrated mode of electric automobile photovoltaic charge station and both can realize regenerative resource On-site elimination reduces again the adverse effect of charging electric vehicle Load on Electric Power Grid, has certain development prospect With exploration meaning.

Energy storage technology possesses the time-shift ability to power and energy, but high energy storage cost always stores up Technology can develop very important problem.The energy-storage system of configuration optimum can improve the raising light of photovoltaic charge station Can utilization rate, reduce the impact of bulk power grid, increase economic efficiency.For the capacity configuration of energy-storage system, Some achievements in research have been there are in terms of improving operation of power networks economy stability.

The stroke characteristic of electric automobile has randomness, but the charging load of a large amount of electric automobile produces cluster Effect i.e. has statistical property, provides auxiliary clothes even with V2G (vehicle to grid) technology to electrical network Business, photovoltaic is exerted oneself and charging electric vehicle load has the typical feature that certain dependency is integrated system.Mesh Front traditional energy storage configuration method does not all consider the dependency impact on energy-storage system optimization allocation.

Summary of the invention

It is an object of the invention to, for this typical integrated system of photovoltaic electric vehicle charging station, propose one Plant the energy storage configuration method that consider dependency theoretical based on Copula, to improve the efficiency of light energy utilization and system Business efficiency.

For achieving the above object, the technical solution used in the present invention is:

1) choose photovoltaic cells and charging electric vehicle load to go out power rate be stochastic variable, measured data is entered Row normalized, builds the marginal distribution of each variable；

2) theoretical based on Copula, choose Gumbel-Copula and Clayton-Copula and build mixing Copula function describes photovoltaic and exerts oneself and the dependency of the asymmetric rear tail characteristic of charging electric vehicle load；

3) sampled analog photovoltaic charge station year net load amount on the basis of both combine probability density function of exerting oneself；

4) under the constraint of the conditions such as power-balance, state-of-charge, stability bandwidth, confidence level, set up with photovoltaic The energy storage Optimal Allocation Model of the minimum object function of electric automobile charging station annual operating and maintenance cost；

5) program in matlab environment, stored energy capacitance is optimized calculating.

Selective allocation plan specifically includes: considers the single batteries to store energy model of dependency, consider phase Close property hybrid energy-storing model, do not consider the single batteries to store energy model of dependency and do not consider dependency Hybrid energy-storing model.

Technical scheme has the advantages that

Technical scheme, theoretical on the basis of dependency based on Copula, by considering operation Performance and Financial cost, the feasibility running photovoltaic electric vehicle charging station integrated system is analyzed, and gives Go out the allocation plan of practicality, thus reached to improve the efficiency of light energy utilization and the purpose of business efficiency.

Below by drawings and Examples, technical scheme is described in further detail.

Accompanying drawing explanation

Fig. 1 is the photovoltaic electric vehicle charging station energy storage configuration choosing considering dependency described in the embodiment of the present invention Selection method flow chart

Fig. 2 is that photovoltaic is exerted oneself frequency histogram and probability density curve comparison diagram

Fig. 3 is that electric automobile load is exerted oneself frequency histogram and probability density curve comparison diagram

Fig. 4 is that photovoltaic is exerted oneself and the bivariate frequency rectangular histogram of the load that charges

Fig. 5 is that photovoltaic is exerted oneself and the joint probability density of the load that charges

Fig. 6 is electric automobile photovoltaic charge station year net load analog data

Detailed description of the invention

Below in conjunction with the drawings and specific embodiments, the present invention is further elaborated.

This example electric automobile charging station is provided with 120 charging piles, and the rated power of separate unit charging pile is 10kW； The rated capacity of photovoltaic generation unit is 800kW；Energy-storage units uses lead-acid accumulator and ultracapacitor to constitute, Component parameters is shown in Table 1.

Table 1 energy storage device parameter

1) choose photovoltaic cells and charging electric vehicle load to go out power rate be stochastic variable, number will be surveyed According to being normalized.By frequency histogram, it is nonnormal distribution that Fig. 2, Fig. 3 understand sample data, Then estimate marginal distribution with non parametric tests, use Density Estimator method herein.If x1, x2 ..., Xn is the sample of stochastic variable X, and probability density function f (x) computing formula of stochastic variable X is as follows:

$f_h\left(x\right)=\frac{1}{\mathrm{nh}}\underset{j=1}{\overset{n}{\mathrm{\Σ}}}K\left(\frac{x-x_j}{h}\right)=\frac{1}{n}\underset{j=1}{\overset{n}{\mathrm{\Σ}}}{K}_h(x-x_j)$

2) photovoltaic is exerted oneself and the dependency that charges between load has asymmetrical rear tail characteristic such as Fig. 4 institute Show.Characteristic according to each function choose the relevant Gumbel-Copula of Asymmetric Tail and Clayton-Copula linear combination builds mixing Copula function and is fitted.Form is as follows:

C (u, v, θ)=ω₁C_G(u, v；θ₁)+ω₂C_C(u, v；θ₂)

$C_G(u,v;{\mathrm{\θ}}_1)=\exp \{-{\left[{(-\log {\left(u\right)}^{{\mathrm{\θ}}_1}+(-\log \left(u\right))}^{{\mathrm{\θ}}_1}\right]}^{\frac{1}{{\mathrm{\θ}}_1}}\}$

$C_C(u,v;{\mathrm{\θ}}_2)={(u^{-{\mathrm{\θ}}_2}+v^{-{\mathrm{\θ}}_2}-1)}^{-\frac{1}{{\mathrm{\θ}}_2}}$

U=F in formula_PV(P_PV), v=F_EV(P_EV)；ω₁, ω₂For the weight coefficient of single Copula function, and ω₁+ω₂=1；θ₁, θ₂Relevance parameter for Gumbel-Copula and Clayton-Copula.Use maximum Expect that (EM) algorithm estimates parameter ω₁, ω₂。

3) joint probability density function formula is obtained as follows:

H (x, y)=[ω₁c_G(u, v；θ₁)+ω₂c_C(u, v；θ₂)]f_pv(x)f_ev(y)

$c_G(u,v;{\mathrm{\θ}}_1)=\frac{C_G(u,v;{\mathrm{\θ}}_1){(\ln u\·\ln v)}^{\frac{1}{{\mathrm{\θ}}_1}-1}}{\mathrm{uv}{[{(-\ln u)}^{\frac{1}{{\mathrm{\θ}}_1}}+{(-\ln v)}^{\frac{1}{{\mathrm{\θ}}_1}}]}^{2-{\mathrm{\θ}}_1}}\{{[{(-\ln u)}^{\frac{1}{{\mathrm{\θ}}_1}}+{(-\ln v)}^{\frac{1}{{\mathrm{\θ}}_1}}]}^{{\mathrm{\θ}}_1}+\frac{1}{{\mathrm{\θ}}_1}-1\}$

$c_C(u,v;{\mathrm{\θ}}_2)=(1+{\mathrm{\θ}}_2){\left(\mathrm{uv}\right)}^{-{\mathrm{\θ}}_2-1}{(u^{-{\mathrm{\θ}}_2}+v^{-{\mathrm{\θ}}_2}-1)}^{-2-\frac{1}{{\mathrm{\θ}}_2}}$

ω₁=0.3038, ω₂=0.6962, θ₁=4.2788, θ₂=8.6953.Fig. 5 be both combine probability of exerting oneself Density function, the year net load amount of sampled analog photovoltaic charge station on this functional foundations, as shown in Figure 6.

4) energy storage Optimal Allocation Model is set up.

A, set up object function in terms of following two, calculate cost.

The average annual operating cost of A1, energy-storage system

C₁=C_in+C_op+C_ma+C_di=(1+k_opba+k_maba+k_diba)·k_debamf_ba+(1+k_opsc+k_disc)k_desnf_sc

Wherein C_in、C_op、C_ma、C_diRespectively purchase, run, safeguard and processing cost, simply consider, Converted herein as different proportionality coefficients.k_opba、k_maba、k_dibaRepresent that lead-acid accumulator runs, ties up respectively Protect, processing cost coefficient.k_opsc、k_diseRepresent ultracapacitor operation, processing cost coefficient respectively.

A2 system year power purchase expense.

$C_2=\underset{t=1}{\overset{T}{\mathrm{\Σ}}}\mathrm{pr}\left(t\right)(P_g\left(t\right)\·\mathrm{\Δt})$

Pr (t) and P in formula_gT () represents t period bulk power grid and the mutual electricity price of system and mean power respectively； Δ t is the duration of t period.

The whole integrated charge station minimum object function of annual operating and maintenance cost, it may be assumed that

Min C=min (C₁+C₂)

B constraints

The B1 system power equilibrium of supply and demand.

P_g(t)+P_pv(t)+P_ev(t)+P_ba(t)+P_uc(t)=0

Exchange power P between integrated charge station and bulk power grid_gT (), photovoltaic are exerted oneself P_pvT (), EVs charging load is exerted oneself P_evT (), energy storage are exerted oneself P_ba(t) and P_ucT () should keep balance the moment.

B2 power constraint.Battery be output as zero or rated value and the output of ultracapacitor should be less than In its maximum.

P_ba(t)=0 or P_ba(t)=P_ban

P_sc(t)≤P_scmax

B3 super-charge super-discharge can reduce the SOC of the service life of energy-storage travelling wave tube, battery and ultracapacitor all

Should be in the range of reasonably limiting.

E_bamin≤E_ba(t)≤E_bamax

E_scmin≤E_sc(t)≤E_scmax

The constraint of the grid-connected power swing of B4

α≤α_max

The constraint of B5 confidence level.Grid-connected power-performance mutually restricts with cost, need to be in certain confidence level

The lower balance realizing performance and cost.

η≥η_min

Alternative allocation plan specifically includes: consider the single batteries to store energy model of dependency, consideration The hybrid energy-storing model of dependency, do not consider the single batteries to store energy model of dependency and do not consider dependency Hybrid energy-storing model.

5) improvement invasive weed algorithm based on differential evolution strategy is used to be programmed meter in matlab environment Calculate, choose the allocation plan of optimum.Set stability bandwidth limit value 4%, during confidence level 97%, consider dependency As shown in table 2 with not considering to optimize result of calculation in the case of dependency.

Table 2 stored energy capacitance optimum results

Compared with not considering dependency, it is considered to dependency decreases energy storage on the premise of tallying with the actual situation The configuration capacity of device, improves the economy of system.Hybrid energy-storing mode can slow down battery depreciation, subtracts The alternative costs of little element, thus obtain higher whole economic efficiency.

The all data of the technical program are all that conversion is to considering as a whole in each year

The foregoing is only the preferred embodiments of the present invention, be not limited to the present invention, although with reference to aforementioned The present invention has been described in detail by embodiment, and for a person skilled in the art, it is the most permissible Technical scheme described in foregoing embodiments is modified, or wherein portion of techniques feature is carried out With replacing.All within the spirit and principles in the present invention, any modification, equivalent substitution and improvement etc. made, Should be included within the scope of the present invention.

Claims (6)

1. based on the electric automobile photovoltaic charge station energy storage selection of configuration method that Copula is theoretical, comprise with Lower step:

1) choose photovoltaic cells and charging electric vehicle load to go out power rate be stochastic variable, measured data is entered Row normalized, builds the marginal distribution of each variable；

2) theoretical based on Copula, choose Gumbel-Copula and Clayton-Copula and build mixing Copula Function describes photovoltaic and exerts oneself and the dependency of the asymmetric rear tail characteristic of charging electric vehicle load；

3) the year net load amount of sampled analog photovoltaic charge station on the basis of both combine probability density function of exerting oneself；

4) under the constraint of the conditions such as power-balance, state-of-charge, stability bandwidth, confidence level, set up with photovoltaic The energy storage Optimal Allocation Model of the minimum object function of electric automobile charging station annual operating and maintenance cost；

5) matlab programming is optimized calculating to optimum stored energy capacitance.

The electric automobile photovoltaic charge station energy storage configuration theoretical based on Copula the most according to claim 1 System of selection, it is characterised in that step 1) in x₁、x₂、...、x_nFor the sample of stochastic variable X, at random Probability density function f (x) computing formula of variable X is:

$f_h\left(x\right)=\frac{1}{\mathrm{nh}}\underset{j=1}{\overset{n}{\mathrm{\Σ}}}K\left(\frac{x-x_j}{h}\right)=\frac{1}{n}{K}_h(x-x_j)$

The electric automobile photovoltaic charge station energy storage configuration theoretical based on Copula the most according to claim 1 System of selection, it is characterised in that step 2) in choose Asymmetric Tail according to the characteristic of each function relevant Gumbel-Copula and Clayton-Copula linear combination builds mixing Copula function and is fitted.Shape Formula is as follows:

C (u, v, θ)=ω₁C_G(u, v；θ₁)+ω₂C_C(u, v；θ₂)

$C_G(u,v;{\mathrm{\θ}}_1)=\exp \{-[(-\log {\left(u\right)}^{{\mathrm{\θ}}_1}+{(-\log \left(u\right))}^{{\mathrm{\θ}}_1}{]}^{\frac{1}{{\mathrm{\θ}}_1}}\}$

$C_c(u,v;{\mathrm{\θ}}_2)={(u^{-{\mathrm{\θ}}_2}+v^{-{\mathrm{\θ}}_2}-1)}^{-\frac{1}{{\mathrm{\θ}}_2}}$

U=F in formula_PV(P_PV), v=F_EV(P_EV)；ω₁, ω₂For the weight coefficient of single Copula function, and ω₁+ω₂=1；θ₁, θ₂Relevance parameter for Gumbel-Copula and Clayton-Copula.Use maximum Expect that (EM) algorithm estimates parameter ω₁, ω₂。

The electric automobile photovoltaic charge station energy storage configuration theoretical based on Copula the most according to claim 1 System of selection, it is characterised in that step 3) to obtain joint probability density function formula as follows:

H (x, y)=[ω₁c_G(u, v；θ₁)+ω₂c_C(u, v；θ₂)]f_pv(x)f_ev(y)

$c_G(u,v;{\mathrm{\θ}}_1)=\frac{C_G(u,v;{\mathrm{\θ}}_1){(\ln u\·\ln v)}^{\frac{1}{{\mathrm{\θ}}_1}-1}}{\mathrm{uv}{[{(-\ln u)}^{\frac{1}{{\mathrm{\θ}}_1}}+{(-\ln v)}^{\frac{1}{{\mathrm{\θ}}_1}}]}^{2-{\mathrm{\θ}}_1}}\{{[{(-\ln u)}^{\frac{1}{{\mathrm{\θ}}_1}}+{(-\ln v)}^{\frac{1}{{\mathrm{\θ}}_1}}]}^{{\mathrm{\θ}}_1}+\frac{1}{{\mathrm{\θ}}_1}-1\}$

$c_C(u,v;{\mathrm{\θ}}_2)=(1+{\mathrm{\θ}}_2){\left(\mathrm{uv}\right)}^{-{\mathrm{\θ}}_2-1}{(u^{-{\mathrm{\θ}}_2}+v^{-{\mathrm{\θ}}_2}-1)}^{-2-\frac{1}{{\mathrm{\θ}}_2}}$

The electric automobile photovoltaic charge station energy storage configuration theoretical based on Copula the most according to claim 1 System of selection, it is characterised in that step 4) set up object function in terms of following two, calculate cost.

1. the average annual operating cost of energy-storage system

C₁=C_in+C_op+C_ma+C_di=(1+k_opba+k_maba+k_diba)·k_debamf_ba+(1+k_opsc+k_disc)k_descnf_sc

Wherein C_in、C_op、C_ma、C_dtRespectively purchase, run, safeguard and processing cost, simply consider, Converted herein as different proportionality coefficients.k_opba、k_maba、k_dtbaRepresent that lead-acid accumulator runs, ties up respectively Protect, processing cost coefficient.k_opsc、k_discRepresent ultracapacitor operation, processing cost coefficient respectively.

2. system year power purchase expense.

$C_2=\underset{t=1}{\overset{T}{\mathrm{\Σ}}}\mathrm{pr}\left(t\right)(P_g\left(t\right)\·\mathrm{\Δt})$

Pr (t) and P in formula_gT () represents t period bulk power grid and the mutual electricity price of system and mean power respectively； Δ t is the duration of t period.

The whole integrated charge station minimum object function of annual operating and maintenance cost, it may be assumed that

Min C=min (C₁+C₂)

The electric automobile photovoltaic charge station energy storage configuration theoretical based on Copula the most according to claim 1 System of selection, it is characterised in that step 4) constraints includes following five aspects:

1. the system power equilibrium of supply and demand.

P_g(t)+P_pv(t)+P_ev(t)+P_ba(t)+P_uc(t)=0

Exchange power P between integrated charge station and bulk power grid_gT (), photovoltaic are exerted oneself P_pvT (), EVs charging load is exerted oneself P_evT (), energy storage are exerted oneself P_ba(t) and P_ucT () should keep balance the moment.

2. power constraint.Battery be output as zero or rated value and the output of ultracapacitor should be less than In its maximum.

P_ba(t)=0 or P_ba(t)=P_ban

P_sc(t)≤P_{sc max}

3. super-charge super-discharge can reduce the SOC of the service life of energy-storage travelling wave tube, battery and ultracapacitor all Should be in the range of reasonably limiting.

E_{ba min}≤E_ba(t)≤E_{ba max}

E_{sc min}≤E_sc(t)≤E_{sc max}

The constraint of the most grid-connected power swing

α≤α_max

5. the constraint of confidence level.Grid-connected power-performance mutually restricts with cost, need to be in certain confidence level The lower balance realizing performance and cost.

η≥η_min