CN105741027A - Optimization dispatching method for virtual power plant with electric vehicle - Google Patents

Abstract

The invention discloses an optimization dispatching method for a virtual power plant with an electric vehicle. The method comprises the following steps of 1) establishing an interaction model of the virtual power plant with the electric vehicle and a power grid; 2) establishing a random dispatching model for optimization dispatching of the virtual power plant with the electric vehicle; and 4) establishing a robust random dispatching model for optimization dispatching of the virtual power plant with the electric vehicle. Compared with the prior art, the method has the advantages of high robustness of a virtual power plant dispatching decision scheme, high calculation efficiency and the like.

Description

A kind of virtual plant Optimization Scheduling containing electric automobile

Technical field

The present invention relates to a kind of virtual plant Optimization Scheduling containing electric automobile, belong to virtual plant optimization and run field.

Background technology

Electric automobile (plug-inelectricvehicle, PEV) is as a kind of new traffic tool, and its large-scale application has significant advantage in alleviating the crises such as global energy shortage, environmental pollution and climate change.Meanwhile, along with deepening continuously of new forms of energy revolution, obtain quick development with the distributed power source (distributedgeneration, DG) that wind-powered electricity generation, photovoltaic are principal mode, estimate to be up to 200,000,000 kilowatts to the Wind turbines installed capacity of the year two thousand twenty China.But, extensive electric automobile will produce the load growth of a new round when accessing grid charging, aggravate network load peak-valley difference, and add the control difficulty of operation of power networks as a class moving load.On the other hand, all kinds of DG skewness on geographical position, it is subject to the natural cause impacts such as weather, its intermittence exerted oneself and uncertainty and the power-balance of power system, via net loss etc. are produced negative effect.

Proposing of virtual plant (virtualpowerplant, VPP) is solve the problems referred to above to provide new thinking.VPP does not change the mode that all kinds of DG is grid-connected, but by different types of elements such as advanced Coordinated Control, intelligent metering technology and ICT polymerization all kinds of DG, PEV, energy-storage systems, the coordination optimization being realized multiple distributed energy by the software algorithm on upper strata is run, thus promoting the configuration of resource reasonably optimizing and utilizing.It is polymerized electric automobile by VPP and wind energy turbine set participates in operation of power networks, cell apparatus in electric automobile is utilized to have the characteristic of distributed energy storage unit, electric automobile and wind energy turbine set are coordinated Optimized Operation, can effectively alleviate the unordered discharge and recharge of electric automobile and the uncertain negative effect produced to electrical network of wind energy turbine set wind power output on the one hand, improve the access capacity of electric automobile and wind-powered electricity generation；Enriching operation and the control device of power system on the other hand, VPP can participate in peak load shifting, provide the assistant service such as frequency stable and spare capacity.

Summary of the invention

The present invention provides a kind of virtual plant Optimization Scheduling containing electric automobile, and the virtual plant optimal scheduling decision scheme strong robustness, the computational efficiency that draw are high.

For solving above-mentioned technical problem, the technical solution adopted in the present invention is as follows:

A kind of virtual plant Optimization Scheduling containing electric automobile, comprises the following steps:

Step 1) set up the virtual plant containing electric automobile and electrical network interaction models；

Step 2) set up the random schedule model of virtual plant Optimized Operation containing electric automobile；

Step 3) set up the robust random schedule model of virtual plant Optimized Operation containing electric automobile；

Step 4) solve the virtual plant optimal scheduling scheme containing electric automobile.

The virtual plant optimal scheduling scheme containing electric automobile can be drawn by solving robust stochastic model.

Described virtual plant polymerizable electric automobile, wind-powered electricity generation distributed power supply participate in the operation of electrical network and electricity market.

Described virtual plant can include electric automobile, wind energy turbine set and gas turbine.

It is preferred that, described virtual plant is the forms of electricity generation that the traditional energy that electric automobile, wind energy turbine set and gas turbine form is combined with regenerative resource.

In order to improve computational efficiency further, the described virtual plant Optimization Scheduling containing electric automobile, it is characterised in that:

Step 1) set up the virtual plant containing electric automobile and electrical network interaction models；

Step 2) set up the random schedule model of virtual plant Optimized Operation containing electric automobile:

The object function of random schedule model:

$\begin{array}{cc}\min & C=\underset{s=1}{\overset{S}{\Σ}}p\left(s\right)\underset{t=1}{\overset{T}{\Σ}}\[\underset{m=1}{\overset{M}{\Σ}}{c}_{m,t}{P}_{m,t,s}+c_{g,t}{P}_{g,t}+\underset{j=1}{\overset{J}{\Σ}}{c}_{j,t}{P}_{j,t,s}\]\end{array}$

In formula: C is VPP day operation cost function, S is wind power output scene collection, the probability that p (s) occurs for scene s correspondence, and hop count when T is total in dispatching cycle, M is gas turbine unit number, and J is PEV Centralized Controller number, c_m,t、c_g,t、c_j,tRespectively gas turbine, power purchase and electric automobile discharge and recharge cost, P_m,t,s、P_g,t、P_j,t,sRespectively t period Gas Turbine Output, purchase of electricity, electric automobile charge-discharge electric power, for decision variable；

Constraints:

$x_{j,t,s}^{ch}+x_{j,t,s}^{dch}\≤1,\∀j,t,s$

$E_{j,t,s}=-E_{j,t-1,s}+{\mathrm{\η}}_{ch}-p_{j,t,s}^{ch}-1/{\mathrm{\η}}_{dch}{p}_{j,t,s}^{dch},\∀j,t,s$

$p_{j,t,s}^{ch}\≤P_j^{ch,\max }\·x_{j,t,s}^{ch},\∀j,t,s$

$p_{j,t,s}^{dch}\≤P_i^{dch,max}\·x_{j,t,s}^{dch},\∀j,t,s$

$P_{j,t,s}-p_{j,t,s}^{ch}{x}_{j,t,s}^{ch}+p_{j,t,s}^{dch}{x}_{j,t,s}^{dch},\∀j,t,s$

${\mathrm{SOC}}_j^{\min }\≤{\mathrm{SOC}}_{j,t,s}\≤{\mathrm{SOC}}_j^{\max },\∀j,t,s$

$p_{j,t,s}^{ch}\·{\mathrm{\η}}_{ch}\≤E_j^{\max }-E_{j,t,s},\∀j,t,s$

$1/{\mathrm{\η}}_{dch}\·p_{j,t,s}^{dch}\≤E_{j,t,s},\∀j,t,s$

In formula: forBoolean variableRepresent whether equivalent PEV fills (putting) electricity, be that (putting) electricity is filled in 1 expression, otherwise be 0；E_j,t,sRepresent dump energy in battery,Represent battery capacity；η_ch(η_dch) represent that battery fills (putting) electrical efficiency；Represent that battery fills (putting) electrical power,Represent that battery fills the higher limit of (putting) electrical power；SOC_j,t,sRepresent the state-of-charge of battery,Represent state-of-charge upper lower limit value；

$P_m^{\min }\≤P_{m,t,s}\≤P_m^{max},\∀m,t,s$

$P_{m,t,s}-P_{m,t+1,s}\≤{\mathrm{RD}}_m,\∀m,t,s$

$P_{m,t+1,s}-P_{m,t,s}\≤{\mathrm{RU}}_m,\∀m,t,s$

$P_{g,t}^{\min }\≤\left|P_{g,t}\right|\≤P_{g,t}^{\max },\∀t$

$0\≤P_{w,t,s}\≤P_{w,t,s}^{\max },\∀w,t,s$

$P_j^{\min }\≤P_{j,t,s}\≤P_j^{\max },\∀j,t,s$

In formula:Exert oneself bound for gas turbine m, RU_m(RD_m) the upwards climbing rate (downwards) for unit；For the bound of VPP power purchase,Exert oneself for wind energy turbine set maximum wind,Bound for equivalent PEV charge-discharge electric power；

$\underset{m=1}{\overset{M}{\Σ}}{P}_{m,t,s}+P_{g,t}+\underset{j=1}{\overset{J}{\Σ}}{P}_{j,t,s}+P_{w,t,s}=\underset{l=1}{\overset{L}{\Σ}}{P}_{l,t},\∀t,s$

In formula: P_w,t,sFor wind energy turbine set wind power output, L is virtual plant inside terminals number of users, P_l,tPower demand for terminal use；

Step 3) set up the robust random schedule model of virtual plant Optimized Operation containing electric automobile:

In formula: variable zd_j,t,s,Pd_j,t,s,zc_j,t,s,Pc_j,t,s, X, Y is the aid decision variable introduced in conversion process of equal value, without concrete physical significance, and zd_j,t,s,Pd_j,t,sCorresponding discharge condition, zc_j,t,s,Pc_j,t,sCorresponding charged state, parameter σ_d,σ_cDeviation factor for mutual power deviation prediction average.

Step 4) utilize CPLEX solver to solve the virtual plant optimal scheduling scheme containing electric automobile in GAMS software platform.

In order to improve computational efficiency further, in random schedule model, adopt the uncertainty of scenario simulation technical finesse wind power output, electric automobile charge-discharge electric power；In robust random schedule model, the method that robust optimizes is adopted to process the uncertainty of electric automobile charge-discharge electric power.

The NM technology of the present invention is all with reference to prior art.

Beneficial effect: the present invention adopts the uncertainty of scenario simulation technical finesse wind power output, adopts the method that robust optimizes to process the uncertainty of electric automobile charge-discharge electric power, has virtual plant scheduling decision scheme strong robustness, computational efficiency advantages of higher.

Accompanying drawing explanation

Fig. 1 is the flow chart of the present invention；

Fig. 2 is schedulable electric automobile quantity figure in dispatching cycle；

Fig. 3 is that confidence level is to object function and robust Relationship of Coefficients figure；

Fig. 4 is robust stochastic model optimal decision conceptual scheme；

Fig. 5 is the difference relationship figure of robust stochastic model and stochastic model optimal decision scheme；

Fig. 6 is virtual plant and electrical network interaction models.

Detailed description of the invention

In order to be more fully understood that the present invention, it is further elucidated with present disclosure below in conjunction with embodiment, but present disclosure is not limited solely to the following examples.

The flow chart of the present invention virtual plant Optimization Scheduling containing electric automobile is as it is shown in figure 1, include:

Step 1) set up the virtual plant containing electric automobile and electrical network interaction models.

Step 2) set up the random schedule model of virtual plant Optimized Operation containing electric automobile.

The object function of random schedule model:

$\begin{array}{cc}\min & C=\underset{s=1}{\overset{S}{\Σ}}p\left(s\right)\underset{t=1}{\overset{T}{\Σ}}\[\underset{m=1}{\overset{M}{\Σ}}{c}_{m,t}{P}_{m,t,s}+c_{g,t}{P}_{g,t}+\underset{j=1}{\overset{J}{\Σ}}{c}_{j,t}{P}_{j,t,s}\]\end{array}$

In formula: C is VPP day operation cost function, S is wind power output scene collection, the probability that p (s) occurs for scene s correspondence, and hop count when T is total in dispatching cycle, M is gas turbine unit number, and J is PEV Centralized Controller number, c_m,t、c_g,t、c_j,tRespectively gas turbine, power purchase and electric automobile discharge and recharge cost, P_m,t,s、P_g,t、P_j,t,sRespectively t period Gas Turbine Output, purchase of electricity, electric automobile charge-discharge electric power, for decision variable；

Constraints:

$x_{j,t,s}^{ch}+x_{j,t,s}^{dch}\≤1,\∀j,t,s$

$E_{j,t,s}=-E_{j,t-1,s}+{\mathrm{\η}}_{ch}-p_{j,t,s}^{ch}-1/{\mathrm{\η}}_{dch}-p_{j,t,s}^{dch},\∀j,t,s$

$p_{j,t,s}^{ch}\≤P_j^{ch,\max }\·x_{j,t,s}^{ch},\∀j,t,s$

$p_{j,t,s}^{dch}\≤P_j^{dch,max}\·x_{j,t,s}^{dch},\∀j,t,s$

$P_{j,t,s}=p_{j,t,s}^{ch}{x}_{j,t,s}^{ch}+p_{j,t,s}^{dch}{x}_{j,t,s}^{dch},\∀j,t,s$

${\mathrm{SOC}}_j^{\min }\≤{\mathrm{SOC}}_{j,t,s}\≤{\mathrm{SOC}}_j^{max},\∀j,t,s$

$p_{j,t,s}^{ch}\·{\mathrm{\η}}_{ch}\≤E_j^{\max }-E_{j,t,s},\∀j,t,s$

$1/{\mathrm{\η}}_{dch}\·p_{j,t,s}^{dch}\≤E_{j,t,s},\∀j,t,s$

In formula: forBoolean variableRepresent whether equivalent PEV fills (putting) electricity, be that (putting) electricity is filled in 1 expression, otherwise be 0；E_j,t,sRepresent dump energy in battery,Represent battery capacity；η_ch(η_dch) represent that battery fills (putting) electrical efficiency；Represent that battery fills (putting) electrical power,Represent that battery fills the higher limit of (putting) electrical power；SOC_j,t,sRepresent the state-of-charge of battery,Represent state-of-charge upper lower limit value；

$P_m^{\min }\≤P_{m,t,s}\≤P_m^{max},\∀m,t,s$

$P_{m,t,s}-P_{m,t+1,s}\≤{\mathrm{RD}}_m,\∀m,t,s$

$P_{m,t+1,s}-P_{m,t,s}\≤{\mathrm{RU}}_m,\∀m,t,s$

$P_{g,t}^{\min }\≤\left|P_{g,t}\right|\≤P_{g,t}^{\max },\∀t$

$0\≤P_{w,t,s}\≤P_{w,t,s}^{\max },\∀w,t,s$

$P_j^{\min }\≤P_{j,t,s}\≤P_j^{\max },\∀j,t,s$

In formula:Exert oneself bound for gas turbine m, RU_m(RD_m) the upwards climbing rate (downwards) for unit；For the bound of VPP power purchase,Exert oneself for wind energy turbine set maximum wind,Bound for equivalent PEV charge-discharge electric power.

$\underset{m=1}{\overset{M}{\Σ}}{P}_{m,t,s}+P_{g,t}+\underset{j=1}{\overset{J}{\Σ}}{P}_{j,t,s}+P_{w,t,s}=\underset{l=1}{\overset{L}{\Σ}}{P}_{l,t},\∀t,s$

In formula: P_w,t,sFor wind energy turbine set wind power output, L is virtual plant inside terminals number of users, P_l,tPower demand for terminal use.

Step 3) set up the robust random schedule model of virtual plant Optimized Operation containing electric automobile:

In formula: variable zd_j,t,s,Pd_j,t,s,zc_j,t,s,Pc_j,t,s, X, Y is the aid decision variable introduced in conversion process of equal value, without concrete physical significance, and zd_j,t,s,Pd_j,t,sCorresponding discharge condition, zc_j,t,s,Pc_j,t,sCorresponding charged state, parameter σ_d,σ_cDeviation factor for mutual power deviation prediction average.

Step 4) solve the virtual plant optimal scheduling scheme containing electric automobile.

If Fig. 6 is virtual plant and electrical network interaction models, in figure, PEV Centralized Controller is as the mutual interface of electric automobile and electrical network, has the double function servicing user and electrical network.PEV Centralized Controller carries out charge-discharge electric power prediction according to historical datas such as the user's request of schedulable electric automobile, running data and trip customs, VPP controls the charge-discharge electric power prediction of coordination center integration load prediction, wind power output prediction and PEV Centralized Controller and is optimized scheduling, formulating the generation schedule of generating set, PEV Centralized Controller discharge and recharge plan and exchange power planning with power distribution company, schedulable electric automobile is carried out orderly discharge and recharge by the dispatch command that PEV Centralized Controller issues according to control coordination center.

One embodiment of the present of invention is described below:

Considering to be formed VPP by electric automobile, wind energy turbine set and gas turbine, gas turbine parameter is as shown in table 1.

The parameter of table 1 gas turbine

If VPP comprises 2 place's PEV Centralized Controllers, PEV Centralized Controller contains 200,100 registration schedulable electric automobiles respectively, assume that the battery size in schedulable electric automobile is identical, battery parameter is taken from the fertile indigo plant of Chevrolet and is reached electric automobile, respectively charging upper limit value is 14kWh, and electric discharge lower limit is 4kWh, and charge-discharge electric power is 2.1kW, efficiency for charge-discharge is 90%, and fuel consumption when electric automobile normally travels is 30.2mile/gallon.Based on certain electric automobile demonstration project of Germany, in dispatching cycle, each moment obeys the electric automobile quantity of VPP scheduling as in figure 2 it is shown, can draw the mutual power of PEV Centralized Controller place equivalence PEV and electrical network according to historical datas such as the user's request of electric automobile, running data and trip customs to utilize statistical analysis to draw.

In robust Stochastic Optimization Model, choose the confidence interval that confidence level the is β uncertain collection of structure of equivalent PEV and electrical network mutual power prediction value；The upper limit assuming VPP and power distribution company Power Exchange is 1000kW.

Under different confidence levels, the relation between virtual plant day operation cost and robust coefficient is as shown in Figure 3.It can be seen that confidence level is more high, day operation cost is more steep with the relation curve of robust coefficient, and namely in robust stochastic model, when indeterminacy section range of convergence is bigger, object function will be produced large effect by the small change of robust coefficient.When confidence level is relatively low, the change of robust coefficient is on the impact of object function inconspicuous.

The difference relationship of the optimal decision scheme of robust stochastic model and robust stochastic model and stochastic model optimal decision scheme is as shown in Figure 4, Figure 5.As can be seen from the figure, the robust optimization mutual power on VPP optimal scheduling decision scheme medium value PEV and electrical network and the purchase of electricity from power distribution company affect bigger, oppose that the impact of exerting oneself of gas turbine is less mutually, this shows that the change of mutual power is little to Gas Turbine Output project impact, the optimum of gas turbine is exerted oneself and is had stronger robustness under uncertain environment, the generation schedule formulating the conventional power unit such as gas turbine is had Engineering Guidance meaning by this, the operating cost produced due to the operation of unit frequent start-stop can be avoided, reduce VPP day operation cost.

Claims (5)

1. the virtual plant Optimization Scheduling containing electric automobile, it is characterised in that: comprise the following steps:

Step 1) set up the virtual plant containing electric automobile and electrical network interaction models；

Step 2) set up the random schedule model of virtual plant Optimized Operation containing electric automobile；

Step 3) set up the robust random schedule model of virtual plant Optimized Operation containing electric automobile；

Step 4) solve the virtual plant optimal scheduling scheme containing electric automobile.

2. the virtual plant Optimization Scheduling containing electric automobile as claimed in claim 1, it is characterised in that: described virtual plant includes electric automobile, wind energy turbine set and gas turbine.

3. the virtual plant Optimization Scheduling containing electric automobile as claimed in claim 1 or 2, it is characterised in that: described virtual plant is the forms of electricity generation that the traditional energy that electric automobile, wind energy turbine set and gas turbine form is combined with regenerative resource.

4. the virtual plant Optimization Scheduling containing electric automobile as claimed in claim 1 or 2, it is characterised in that: comprise the following steps:

Step 1) set up the virtual plant containing electric automobile and electrical network interaction models；

Step 2) set up the random schedule model of virtual plant Optimized Operation containing electric automobile:

The object function of random schedule model:

$\min C=\underset{s=1}{\overset{S}{\Σ}}p\left(s\right)\underset{t=1}{\overset{T}{\Σ}}\[\underset{m=1}{\overset{M}{\Σ}}{c}_{m,t}{P}_{m,t,s}+c_{g,t}{P}_{g,t}+\underset{j=1}{\overset{J}{\Σ}}{c}_{j,t}{P}_{j,t,s}\]$

In formula: C is VPP day operation cost function, S is wind power output scene collection, the probability that p (s) occurs for scene s correspondence, and hop count when T is total in dispatching cycle, M is gas turbine unit number, and J is PEV Centralized Controller number, c_m,t、c_g,t、c_j,tRespectively gas turbine, power purchase and electric automobile discharge and recharge cost, P_m,t,s、P_g,t、P_j,t,sRespectively t period Gas Turbine Output, purchase of electricity, electric automobile charge-discharge electric power, for decision variable；

Constraints:

$x_{j,t,s}^{ch}+x_{j,t,s}^{dch}\≤1,\∀j,t,s$

$E_{j,t,s}=-E_{j,t-1,s}+{\mathrm{\η}}_{ch}-p_{j,t,s}^{ch}-1/{\mathrm{\η}}_{dch}{p}_{j,t,s}^{dch},\∀j,t,s$

$p_{j,t,s}^{ch}\≤P_j^{ch,\max }\·x_{j,t,s}^{ch},\∀j,t,s$

$p_{j,t,s}^{dch}\≤P_j^{dch,max}\·x_{j,t,s}^{dch},\∀j,t,s$

$P_{j,t,s}=p_{j,t,s}^{ch}{x}_{j,t,s}^{ch}+p_{j,t,s}^{dch}{x}_{j,t,s}^{dch},\∀j,t,s$

${\mathrm{SOC}}_j^{\min }\≤{\mathrm{SOC}}_{j,t,s}\≤{\mathrm{SOC}}_j^{max},\∀j,t,s$

$p_{j,t,s}^{ch}\·{\mathrm{\η}}_{ch}\≤E_j^{\max }-E_{j,t,s},\∀j,t,s$

$1/{\mathrm{\η}}_{dch}\·p_{j,t,s}^{dch}\≤E_{j,t,s},\∀j,t,s$

In formula: forBoolean variableRepresent whether equivalent PEV fills (putting) electricity, be that (putting) electricity is filled in 1 expression, otherwise be 0；E_j,t,sRepresent dump energy in battery,Represent battery capacity；h_ch(h_dch) represent that battery fills (putting) electrical efficiency；Represent that battery fills (putting) electrical power,Represent that battery fills the higher limit of (putting) electrical power；SOC_j,t,sRepresent the state-of-charge of battery,Represent state-of-charge upper lower limit value；

$P_m^{\min }\≤P_{m,t,s}\≤P_m^{max},\∀m,t,s$

$P_{m,t,s}-P_{m,t+1,s}\≤{\mathrm{RD}}_m,\∀m,t,s$

$P_{m,t+1,s}-P_{m,t,s}\≤{\mathrm{RU}}_m,\∀m,t,s$

$P_{g,t}^{\min }\≤\left|P_{g,t}\right|\≤P_{g,t}^{\max },\∀t$

$0\≤P_{w,t,s}\≤P_{w,t,s}^{\max },\∀w,t,s$

$P_j^{\min }\≤P_{j,t,s}\≤P_j^{max},\∀j,t,s$

In formula:Exert oneself bound for gas turbine m, RU_m(RD_m) the upwards climbing rate (downwards) for unit；For the bound of VPP power purchase,Exert oneself for wind energy turbine set maximum wind,Bound for equivalent PEV charge-discharge electric power；

$\underset{m=1}{\overset{M}{\Σ}}{P}_{m,t,s}+P_{g,t}+\underset{j=1}{\overset{J}{\Σ}}{P}_{j,t,s}+P_{w,t,s}=\underset{l=1}{\overset{L}{\Σ}}{P}_{l,t},\∀t,s$

In formula: P_w,t,sFor wind energy turbine set wind power output, L is virtual plant inside terminals number of users, P_l,tPower demand for terminal use；

Step 3) set up the robust random schedule model of virtual plant Optimized Operation containing electric automobile:

In formula: variable zd_j,t,s,Pd_j,t,s,zc_j,t,s,Pc_j,t,s, X, Y is the aid decision variable introduced in conversion process of equal value, without concrete physical significance, and zd_j,t,s,Pd_j,t,sCorresponding discharge condition, zc_j,t,s,Pc_j,t,sCorresponding charged state, parameter σ_d,σ_cDeviation factor for mutual power deviation prediction average；

Step 4) utilize CPLEX solver to solve the virtual plant optimal scheduling scheme containing electric automobile in GAMS software platform.

5. the virtual plant Optimization Scheduling containing electric automobile described in claim 4, it is characterised in that in random schedule model, adopts the uncertainty of scenario simulation technical finesse wind power output, electric automobile charge-discharge electric power；In robust random schedule model, the method that robust optimizes is adopted to process the uncertainty of electric automobile charge-discharge electric power.