CN104951614A - EV-charging-controllability considered unit combination model and modeling method - Google Patents

Abstract

The invention discloses an EV-charging-controllability considered unit combination model and a modeling method. According to the EV-charging-controllability considered unit combination model and the modeling method, the output of a unit and total loads of an EV integrated controller serve as control objects, the EV integrated controller serves as an interaction platform of a power grid dispatching center and EVs, the dispatching center issues a charging command to the EV integrated controller, and the EV integrated controller controls the EVs in an administration region of the EV integrated controller after receiving the charging command to meet the command; constraints such as user charging requirement constraints, dispatching capacity range constraints and unit-side constraints are comprehensively considered, the minimum sum of the running cost of the unit and the dispatching cost of the EVs serves as an objective function, and the EV-charging-controllability considered unit combination model is built; the dispatching cost of the EVs is considered in the objective function of the model, the user charging requirement constraints are considered in the constraint conditions, the requirement of users for the electric quantity of the EVs serves as one condition which must be met when the dispatching center controls the EVs to be charged, and the vehicle attributes of the EVs are sufficiently considered.

Description

A kind of Unit Combination model and modeling method taking into account charging electric vehicle controllability

Technical field

The invention belongs to power scheduling technical field, more specifically, relate to a kind of Unit Combination model and the modeling method of taking into account charging electric vehicle controllability.

Background technology

Electric automobile (Electric Vehicle, EV) replaces oil as its major impetus energy using new forms of energy electric power, has low, the eco-friendly feature of carbon emission, becomes the strategic industry direction of each Main Auto manufacturing power in the world gradually.

Unit Combination (unit commitment, UC) problem is the content of dispatching of power netwoks, and conventional rack combination is by dispatching center according to load prediction curve, formulates the start plan of unit so that unit operation cost is minimum for target; When relating to wind-power electricity generation, dispatching center then formulates the start plan of fired power generating unit according to load prediction curve and wind-powered electricity generation prediction curve, minimum for target with thermal power unit operation cost, and the spare capacity needs of system is then relevant with the precision of prediction of wind power output.

When considering the Optimization of Unit Commitment By Improved of EV access, prior art has following scheme, and one is by EV as pure load, predicts, join in traditional load prediction and carry out Unit Combination arrangement its charging load; Two is regard each EV as controllable unit, is directly controlled its charging by grid dispatching center, reaches the object reducing thermal power unit operation cost; Three is for control object is dispatched with EV Centralized Controller.

When extensive EV accesses electrical network, if using EV as pure load, superpose with traditional load and carry out Unit Combination arrangement, because EV charging load and traditional load have similar Behavior law, namely charging and other the daily routines of EV can be carried out after user comes home from work, so there will be the phenomenon at " Shang Jia peak, peak ", thus unit cost is significantly increased; And for dispatching center directly with the method that separate unit EV is control object, be feasible when EV negligible amounts, but when EV recoverable amount increases to tens thousand of, grid dispatching center directly and each EV to carry out the efficiency of information communication lower; For dispatching center with EV Centralized Controller for control object carries out the method for dispatching, existing technology does not consider user's request, at any time may interruptible load, makes troubles to user; The scheduling cost of electric automobile is not considered yet.

Summary of the invention

For above defect or the Improvement requirement of prior art, the invention provides a kind of Unit Combination model and the modeling method of taking into account charging electric vehicle controllability, its object is to dispatching of power netwoks scheme when electric automobile large-scale grid connection is provided.

For achieving the above object, according to one aspect of the present invention, provide a kind of Unit Combination model taking into account charging electric vehicle controllability, comprise constraint condition and objective function; Constraint condition comprises the constraint of unit side and the constraint of user side;

Objective function is:

$\min F(P_{Gi}^t,u_i^t,P_{Aggj}^t)=\underset{t=1}{\overset{T}{\Σ}}\underset{i=1}{\overset{N_g}{\Σ}}\[C_i^t\left(P_{Gi}^t\right)u_i^t+S_i^t(x_i^{offt},u_i^t,u_i^{t-1})\]+M\·\underset{t=1}{\overset{T}{\Σ}}\underset{j=1}{\overset{N_{Agg}}{\Σ}}\left|P_{Aggj}^t-P_{Aggj-pre}^t\right|$

Wherein, refer to unit operation cost, refer to genset fuel cost, refer to unit starting cost;

$S_i^t(x_i^{offt},u_i^t,u_i^{t-1})=\left\{\begin{array}{c}{S}_{ih}^t,T_i^{off}<x_i^{offt}\&le;(T_i^{off}+T_i^c)\\ \begin{array}{cc}and& u_i^t-u_i^{t-1}=1\end{array}\\ S_{ic}^t,T_i^{offt}>(T_i^{off}+T_i^c)\\ \begin{array}{cc}and& u_i^t-u_i^{t-1}=1\end{array}\\ 0,otherwise\end{array}\right.$

Wherein, refer to charging electric vehicle load scheduling cost;

$M=\left\{\begin{array}{cc}{M}^{\&prime;}& P_{Aggj}^t<P_{Aggj-pre}^t\\ 0& P_{Aggj}^t>P_{Aggj-pre}^t\end{array}\right.$

Wherein, i is the numbering of genset, and j is the numbering of EV Centralized Controller, and t is the numbering of period, be exerting oneself of i-th genset t period; be the state of i-th genset t period, when it is in running status be 1, otherwise be 0; it is i-th genset t period continuous stop time; be i-th genset t period cost; it is the start-up cost of i-th genset t period; N _gfor genset sum; N _aggfor EV Centralized Controller number; T for always to optimize the period, T=24; for a jth EV Centralized Controller is at the charging load of t period; for a jth EV Centralized Controller is at the charging predicted load of t period; it is the control object of Unit Combination model;

Wherein, M is EV unit compensation cost, refers to when EV should charge but obey dispatch command interruption charging, the cost of compensation that electrical network should be paid; And when EV originally do not charge but obey dispatch command charge time, electrical network does not need to pay cost of compensation; M ' is interruption 1MW electric automobile load and within 1 hour, should compensates to the cost of user;

Wherein, a _i, b _i, c _ibe i-th genset fuel cost coefficient; it is the warm start cost of i-th genset; it is the cold start-up cost of i-th genset; it is the minimum permission stop time of i-th genset; it is the cold start-up time of i-th genset.

This Unit Combination model taking into account charging electric vehicle controllability provided by the invention, with unit output and EV Centralized Controller load for control object, dispatches cost sum with unit operation cost and EV minimum for objective function; Charging electric vehicle load scheduling cost has been counted on the one hand in objective function; On the other hand, count the constraint of user side, to meet the charge requirement of electric automobile user.

Preferably, the constraint of unit side comprises account load balancing constraints, the constraint of unit output bound, spinning reserve constraint, Unit Commitment time-constrain and unit ramp loss, specific as follows:

Account load balancing constraints is: refer to that each generator output sum should be suitable with load, to meet workload demand;

Wherein, be the conventional load level of t period electric system, it is the charging load instruction value of a t period jth EV Centralized Controller;

Unit output bound is constrained to: refer to the bound scope that fired power generating unit is exerted oneself;

Wherein, the minimum load of unit i, for the maximum output of unit i;

Spinning reserve is constrained to: refer to the capacity for tackling load fluctuation that unit is reserved;

Wherein, R ^tfor t period system spinning reserve capacity demand;

Unit Commitment time-constrain relates to motor from number when shutting down to the minimum shutdown starting shooting, and generator shut down from starting shooting to minimum start time number, concrete:

Generator Status is from start to shutdown:

Generator Status is from shutdown to start:

Refer to when after generator outage, when need be fully parked with minimum shutdown, number could be started shooting again; And after generator start, then when needing to be full of minimum start, number could be shut down again;

Wherein, it is i-th time that genset ran continuously in the t period; it is the minimum permission on time of i-th genset;

Unit ramp loss is: refer to the variation range of exerting oneself of the fired power generating unit unit interval that the inertia due to fired power generating unit causes;

Wherein, for the power ascending amount of unit i limits, for the power drop amount of unit i limits.

Preferably, the constraint of user side comprises the constraint of user's energy requirement, the bound constraint of EV Centralized Controller charging load and the constraint of EV Centralized Controller charging load distribution ratio, specific as follows:

User's energy requirement is constrained to $\underset{t=1}{\overset{T}{\&Sigma;}}{P}_{Aggj}^t=\underset{t=1}{\overset{T}{\&Sigma;}}{P}_{Aggj-pre}^t;$

The constraint of user's energy requirement makes the charging load total amount in the whole scheduling slot of EV Centralized Controller equal with charging predicted load total amount, makes the residue delivery of EV Centralized Controller after being scheduled can meet its local EV and charges;

EV Centralized Controller charging load bound is constrained to:

$\left\{\begin{array}{c}{P}_{Aggj}^t\&le;P_{Aggj}^{tmax}=P_{Aggj-pre}^t+\mathrm{\&alpha;}\&CenterDot;P_{Aggj-up}^t\\ P_{Aggj}^t\&GreaterEqual;P_{Aggj}^{t\min }=P_{Aggj-pre}^t+\mathrm{\&alpha;}\&CenterDot;P_{Aggj-down}^t\end{array}\right.$

EV Centralized Controller charging load bound refers to the schedulable scope of Centralized Controller charging load; Wherein, the schedulable range limit of a jth EV Centralized Controller in t, the schedulable lower range limit of a jth EV Centralized Controller in t;

Wherein, can the go up capacitance-adjustable of a jth EV Centralized Controller in t, the descended capacitance-adjustable of a jth EV Centralized Controller in t;

The constraint of EV Centralized Controller charging load distribution ratio comprises following two kinds of situations:

A, when namely t needs to raise the charging predicted load of EV Centralized Controller, then each EV Centralized Controller carries out proportional distribution according to rise amount of capacity, should meet:

$\frac{P_{Aggj}^t-P_{Aggj-pre}^t}{\underset{j=1}{\overset{N_{Agg}}{\&Sigma;}}{P}_{Aggj}^t-\underset{j=1}{\overset{N_{Agg}}{\&Sigma;}}{P}_{Aggj-pre}^t}=\frac{P_{Aggj}^{t\max }-P_{Aggj-pre}^t}{\underset{j=1}{\overset{N_{Agg}}{\&Sigma;}}{P}_{Aggj}^{t\max }-\underset{j=1}{\overset{N_{Agg}}{\&Sigma;}}{P}_{Aggj-pre}^t}$

B, when namely t needs to lower the charging predicted load of EV Centralized Controller, then each EV Centralized Controller carries out proportional distribution according to downward amount of capacity, should meet:

$\frac{P_{Aggj}^t-P_{Aggj-pre}^t}{\underset{j=1}{\overset{N_{Agg}}{\&Sigma;}}{P}_{Aggj}^t-\underset{j=1}{\overset{N_{Agg}}{\&Sigma;}}{P}_{Aggj-pre}^t}=\frac{P_{Aggj}^{t\min }-P_{Aggj-pre}^t}{\underset{j=1}{\overset{N_{Agg}}{\&Sigma;}}{P}_{Aggj}^{t\min }-\underset{j=1}{\overset{N_{Agg}}{\&Sigma;}}{P}_{Aggj-pre}^t}.$

According to another aspect of the present invention, provide the above-mentioned modeling method taking into account the Unit Combination model of charging electric vehicle controllability, comprise the following steps:

(1) utilize EV Centralized Controller to carry out cluster to electric automobile, and obtain the EV charging plan of EV Centralized Controller and the predicted value of charging load;

Wherein, local all electric automobiles put together by referring to of cluster, are managed by an EV Centralized Controller; Dispatching center's management EV Centralized Controller, EV centralized control management separate unit EV; When EV will dispatch in dispatching center, charging instruction is issued to EV Centralized Controller, by its local each EV of control of EV centralized control implement body; EV Centralized Controller is equivalent to the interaction platform of each EV and grid dispatching center;

(2) the schedulable capacity of EV Centralized Controller is obtained according to the EV charging plan of EV Centralized Controller;

(3) the schedulable scope of EV Centralized Controller all moment chargings load is determined according to the schedulable capacity of EV Centralized Controller;

(4) according to the constraint of machine unit characteristic determination unit side, determine that user side retrains according to EV Centralized Controller charging predicted load with the schedulable scope of charging load;

And dispatch the minimum value of cost sum for objective function with unit operation cost and electric automobile, charge load for control object with unit output and EV Centralized Controller, set up the Unit Combination model taking into account charging electric vehicle controllability.

Preferably, above-mentioned steps 1 is specific as follows:

(1.1) EV Centralized Controller is utilized to carry out cluster to EV: to collect local EV Centralized Controller parameter and EV parameter by EV Centralized Controller, comprising: EV Centralized Controller number N _agg, the EV sum N in each EV Centralized Controller, each EV grid-connected moment t ₀with from net moment t _d, i-th EV battery rated capacity C ⁱ, charge power Pev _i, the grid-connected moment battery electric quantity with the demand electricity of the user from the net moment

(1.2) according to the EV charging plan that EV charging modes arranges EV Centralized Controller local, and the charging predicted load of this EV Centralized Controller is obtained

Wherein, EV charging modes comprises: unordered charging, tou power price charging and intelligent charge;

Wherein, unordered charging refers to and does not control electric automobile, inserts electrical network and just starts charging, stop charging when reaching user's request or leave electrical network once EV; Tou power price charging refers to the Price Mechanisms utilized in load valley phase reduction electricity price, guides EV user to charge at low rate period, reach the effect of peak load shifting with this; Intelligent charge refers to and to control the charging of EV according to optimization aim (level and smooth load curve, reduce abandon wind), reaches corresponding object.

Preferably, the process obtaining EV parameter in step (1.1) is specific as follows:

(1.1.1) the historical statistical data formation probability transition matrix of rule of going on a journey based on user, utilizes Markov chain and Monte Carlo simulated sampling to obtain each EV the running status situation of change of certain day, to obtain the grid-connected moment t of each EV ₀with from net moment t _d; Wherein, the grid-connected moment refers to that electric automobile accesses the moment of electrical network, refers to that electric automobile leaves the moment of electrical network from the net moment;

(1.1.2) based on the statistics acquisition probability densimetric curve of EV battery rated capacity, charge power, the battery electric quantity in grid-connected moment and the user's request electricity from the net moment; According to the battery rated capacity C of this curve sampling acquisition i-th EV ⁱ, charge power Pev _i, the grid-connected moment battery electric quantity with the demand electricity of the user from the net moment

Preferably, the schedulable capacity of the EV Centralized Controller in step 2, comprises and can go up capacitance-adjustable with can descend capacitance-adjustable according to following Procedure Acquisition:

(2.1) according to the driving information of EV Centralized Controller charging plan acquisition t i-th EV, the time departure t of EV is comprised _d, current electric quantity expected energy charge power Pev _iwith battery capacity C ⁱ;

(2.2) whether to be connected to the grid according to EV and the charge requirement of EV, to judge the controllability of EV;

(2.3) judge that whether t i-th EV be in charging; If not, then using the charging load of this EV as going up capacitance-adjustable; If so, then using the charging load of this EV as descending capacitance-adjustable;

(2.4) repeat step (2.2) and step (2.3), obtain the controllability of all EV in EV Centralized Controller, the load and cumulative all EV be connected to the grid charge, as the capacitance-adjustable gone up of EV Centralized Controller cumulative all EV charging loads excising out electrical network, as the descended capacitance-adjustable of EV Centralized Controller

Preferably, step (2.2) is specific as follows:

(2.2.1) judge whether t meets t ₀≤ t≤t _d, if not, show that EV does not access electrical network, EV is uncontrollable; If so, show that EV accesses electrical network, enter step (2.2.2);

(2.2.2) remove with the time interval of Δ T the electrical demand judging EV whether meet the condition of following formula:

$\left\{\begin{array}{c}{\mathrm{SOC}}_{need}^i-{\mathrm{SOC}}_t^i<\frac{\left(\right(t_d-t)\&times;\mathrm{\&Delta;}T\&times;{\mathrm{Pev}}_i)\&times;\mathrm{\&eta;}}{C^i}\&times;100\\ {\mathrm{SOC}}_t^i\&le;100\end{array}\right.$

If so, then EV has controllability; If not, then show that EV need from current time trickle charge always until user leaves just can meet consumers' demand, this EV must charge in this period, did not have controllability;

Wherein, t ₀for the EV grid-connected moment, t _dfor from net the moment; for the current electric quantity of battery, η is charge efficiency.

In general, the above technical scheme conceived by the present invention compared with prior art, can obtain following beneficial effect:

(1) the present invention utilizes EV Centralized Controller to carry out cluster to EV, and the schedulable scope of the load that charges according to schedulable procurement of reserve capacity EV Centralized Controller, with conventional power unit exert oneself and EV centralized manager load for control object, consider user's charge requirement, schedulable range of capacity and the constraint of unit side, set up the Unit Combination model taking into account charging electric vehicle controllability, achieve the object that auxiliary dispatching of power netwoks reduces unit operation cost; For realizing the scheduling of scale electric automobile auxiliary electrical network operation, there is important practical significance;

(2) Unit Combination model provided by the invention, owing to possessing complete unit side and the constraint condition of user side, has reacted the alternative mechanism of dispatching center and extensive electric automobile in practical application; Further, in objective function except the fuel cost of unit and start-up cost, also take into account the scheduling cost of electric automobile;

And in prior art, when occurring in electric system that system is gained merit not enough, system call person often forces the power supply of uncompensated interruption certain user to ensure the safe operation of electric system; But in Power Market, the service of termination needs to give customer interrupted certain compensation; Compared to prior art, the cost of compensation that the plan of change charging electric vehicle causes is considered within model by the present invention, makes model more perfect;

(3) Unit Combination model provided by the invention, consider the mechanics of in fact EV, grid dispatching center adopts the method indirectly controlled to electric automobile, electrical network is controlled separate unit electric automobile by Centralized Controller, always charge load for control object with EV Centralized Controller, EV Centralized Controller and grid dispatching center carry out information transmission, charging instruction is handed down to EV Centralized Controller by grid dispatching center, EV Centralized Controller controls its local EV to meet this instruction after receiving charging instruction, the scheduling cost of electric automobile is considered in the objective function of model, the charge requirement constraint of user is considered in constraint condition, using user to one of condition that the demand of EV electricity must meet as dispatching center when controlling charging electric vehicle, take into full account the vehicles attribute of electric automobile, must be based upon can meet under prerequisite that user travels demand to its scheduling.

Accompanying drawing explanation

Fig. 1 is the Unit Combination model modeling process flow diagram taking into account extensive charging electric vehicle controllability provided by the invention;

Fig. 2 is command value and the predicted value of (M ' difference=0,3,6,20,100,1000) EV Centralized Controller charging load under certain cost of compensation in the embodiment of the present invention;

Fig. 3 is command value and the predicted value of (M ' difference=0,3,6,20,100,1000) all EV Centralized Controller chargings load under certain cost of compensation in the embodiment of the present invention;

Fig. 4 is (M ' difference=0,3,6,20,100,1000) system total load curve under certain cost of compensation in the embodiment of the present invention;

Fig. 5 is command value and the predicted value of (M ' difference=0,3,6,20,100,1000) all EV Centralized Controller chargings load under different cost of compensation in the embodiment of the present invention.

Embodiment

In order to make object of the present invention, technical scheme and advantage clearly understand, below in conjunction with drawings and Examples, the present invention is further elaborated.Should be appreciated that specific embodiment described herein only in order to explain the present invention, be not intended to limit the present invention.In addition, if below in described each embodiment of the present invention involved technical characteristic do not form conflict each other and just can mutually combine.

The present invention, on the basis of conventional rack built-up pattern and interruptible load compensation method, establishes and considers that user's request retrain and takes into account the Unit Combination model containing extensive electric automobile of electric automobile controllability and scheduling cost thereof;

For the control of extensive electric automobile access electrical network, EV Centralized Controller is utilized to carry out cluster management to electric automobile, consider the constraint of unit side and user side constraint condition, specifically comprise: the constraint of account load balancing constraints, unit output bound, spinning reserve constraint, the constraint of Unit Commitment time-constrain, unit ramp loss, user's energy requirement, the constraint of EV Centralized Controller charging load bound, the constraint of EV Centralized Controller charging load distribution ratio;

Owing to possessing more complete constraint condition, Unit Combination model provided by the invention has reacted the alternative mechanism of dispatching center and extensive electric automobile in reality well, and, in objective function except the fuel cost of unit and start-up cost, also take into account the scheduling cost of electric automobile, thus the cost of compensation that the plan of change charging electric vehicle causes is considered within model, make model more perfect.

Because model provided by the invention is a Nonlinear Mixed Integer Programming Problem, in an embodiment, adopt YALMIP modeling software to carry out building of model in MATLAB, call CPLEX solver and calculate; For ease of solving, Partial Linear process is carried out to the objective function of model and constraint condition; The method has important practical significance for realizing the scheduling of scale electric automobile auxiliary electrical network operation in the future.

As shown in Figure 1, the invention provides the extensive electric automobile Optimization of Unit Commitment By Improved modeling method that contains taking into account electric automobile scheduling cost and comprise following step:

Step 1: the charging load of prediction EV Centralized Controller eV Centralized Controller collects local electric automobile parameter, selects charging electric vehicle mode, thus arranges the local EV charging plan of EV Centralized Controller and obtain EV Centralized Controller charging predicted load

The electric automobile parameter related in embodiment comprises: EV Centralized Controller number N _agg=3, the electric automobile sum N=20000 in each EV Centralized Controller; Based on historical statistical data formation probability transition matrix, utilize Markov chain and Monte Carlo simulated sampling to obtain the running status situation of change of each electric automobile in the middle of certain day, thus obtain the grid-connected moment t of corresponding each electric automobile ₀with from net moment t _d;

The probability density curve of Corpus--based Method data acquisition, sampling obtains the battery rated capacity C of i-th electric automobile ⁱ=60kWh, charge power Pev _ithe battery electric quantity obey being uniformly distributed between 3 ~ 4kW, starting when charging obedience blocks Gaussian distribution, and its average is 40, and variance is 20, minimum be 20, maximum be 50; The charge capacity that user expects is charge efficiency gets η=95%.EV in the EV Centralized Controller control strategy that charges is taked with level and smooth total load curve for target carries out the pattern of intelligent charge, concrete:

Intelligent charge objective function is minimum for target with daily load curve standard variance:

$\min \sqrt{\frac{1}{T}\underset{t=1}{\overset{T}{\&Sigma;}}{\&lsqb;\underset{i-1}{\overset{N}{\&Sigma;}}(x_t^i\&times;{\mathrm{Pev}}_i)+P_{LD}^t-P_{avg}\&rsqb;}^2}$

Wherein: N is the electric automobile quantity in this electric automobile Centralized Controller; represent whether electric automobile charges, represent and do not charge, represent charging; Pev _ifor the charge power of electric automobile, unit is kW; represent conventional load, unit is kW; P _avgfor taking into account the daily load mean value of electric automobile:

$P_{avg}=\frac{1}{T}\&CenterDot;\underset{t=1}{\overset{T}{\&Sigma;}}\&lsqb;\underset{i=1}{\overset{N}{\&Sigma;}}(x_t^i\&times;{\mathrm{Pev}}_i)+P_{LD}^t\&rsqb;$

In embodiment, constraint condition comprises the constraint of user's electrical demand, duration of charging constraint, concrete:

The constraint of user's electrical demand refers to that user is in the electrical demand value needing to reach user's setting from electricity during net, that is: ${\mathrm{SOC}}_{need}^i\&le;{\mathrm{SOC}}_{td}^i\&le;100\%;$

represent i-th electric automobile from net time battery electric quantity; represent that i-th electric automobile user is from desired battery electric quantity during net, this value can be set by the user.

Battery electric quantity obtains according to following formula:

${\mathrm{SOC}}_{t+1}^i={\mathrm{SOC}}_t^i+\frac{(x_t^i\&times;\mathrm{\&Delta;}T\&times;P_{Ci})\&times;\mathrm{\&eta;}}{C^i}\&times;100\%$

In formula: represent the electricity in t+1 moment; Δ T is material calculation; η is charge efficiency; C ⁱfor battery capacity.

Duration of charging constraint refers to that the charging moment of electric automobile should in the grid-connected moment with from netting between the moment, that is: t ₀≤ t≤t _d-1; t ₀represent the grid-connected moment; t _drepresent from the net moment.

Whether be charged as control variable with each electric automobile in each calculating moment, namely form T × N number of independent variable be expressed as follows with matrix X:

Step 2: the schedulable capacity of assessment EV Centralized Controller: when arranging EV charging plan properly, calculates a jth EV Centralized Controller according to the appraisal procedure of schedulable capacity and can go up capacitance-adjustable in t with can descend capacitance-adjustable

Step 3: the schedulable scope determining EV Centralized Controller load; Calculate the charging load schedulable scope of all EV Centralized Controllers in all moment according to the following formula; α=90% in embodiment.

$\left\{\begin{array}{c}{P}_{Aggj}^{t\max }=P_{Aggj-pre}^t+\mathrm{\&alpha;}\&CenterDot;P_{Aggj-up}^t\\ P_{Aggj}^{t\min }=P_{Aggj-pre}^t+\mathrm{\&alpha;}\&CenterDot;P_{Aggj-down}^t\end{array}\right.$

Step 4: set up the electric automobile Unit Combination model taking into account electric automobile scheduling cost; Consider unit side and the constraint of user side, cost sum is dispatched minimum for objective function with unit operation cost and electric automobile, with unit output and EV Centralized Controller charging load for control object, set up take into account electric automobile scheduling cost containing extensive electric automobile Unit Combination model;

YALMIP modeling software is adopted to build model in MATLAB, call CPLEX solver, the start-up mode of 10 units and the predicted value of Centralized Controller charging load and command value when getting unit operation cost and EV Centralized Controller cost of compensation under different cost of compensation, cost M '=3;

In embodiment, the unit parameter of 10 machine systems comprises the higher limit of exerting oneself of unit lower limit coal consumption cost quadratic term coefficient a _i, Monomial coefficient b _i, constant term coefficient c _i, the minimum permission on time minimum permission stop time initial start and stop state warm start cost cold start-up cost the cold start-up time

Unit output upper lower limit value, the coal consumption cost coefficient of 10 machine systems in embodiment is listed with following table 1;

Unit output upper lower limit value, the coal consumption cost coefficient parameter of 10 machine systems in table 1 embodiment

Table 2 lists that the unit of 10 machine systems in embodiment is minimum allows out/stop time, start-up cost coefficient;

In table 2 embodiment, the unit of 10 machine systems is minimum allows out/stop time, start-up cost figure parameters

Table 3 lists traditional load parameter, and T is 24 moment in one day, for traditional load in each moment, concrete as:

The traditional load parameter list of table 3

Contrast table 1 and table 3,10 unit total volumies are 1662MW, 1500MW when load peak is 12, and the ratio that load accounts for total volume is 90.25%, and ratio is higher, and a lot of unit of load boom period may be in full hair-like state.

Table 4 is unit operation cost and EV Centralized Controller cost of compensations under different cost of compensation in embodiment; Specifically comprise unit total operating cost, the start of unit fuel cost, unit cost, EV Centralized Controller cost of compensation;

Table 4 cost of compensation and unit operation cost and EV Centralized Controller cost of compensation relation list

As can be seen from Table 4, control is optimized to EV Centralized Controller charging load and can reduces unit operation total cost, but along with the increase of unit compensation expense M ', unit operation cost reduces amplitude and diminishes, and EV Centralized Controller cost of compensation first increases and then decreases, this is because when unit compensation expense M ' is less than certain value, although the charging load of scheduling EV Centralized Controller will pay certain cost, but significantly can reduce the fuel cost of unit, therefore a large amount of scheduling can be carried out to EV Centralized Controller charging load, but after unit compensation expense M ' is greater than certain value, the cost that scheduling EV Centralized Controller charging load is paid has exceeded the reduction of unit fuel cost, therefore can not scheduled EV Centralized Controller charging load.

When cost M '=3, the start-up mode of 10 units, at low ebb phase mainly the first five unit output, is all exerted oneself, shown in table 5 specific as follows at load boom period 10 units:

The start-up mode of 10 units during table 5 cost M '=3

Found out by table 5; because load accounts for the large percentage of total volume; No. 1 unit and No. 2 units minimum allow out/and stop time is long again; therefore these two units are in full hair-like state substantially; 3,4, No. 5 units are then supplemented when load is higher, and rear five unit capacities are less but minimumly allow out/stop time is shorter again, can the fluctuation of load-responsive quickly; therefore can be open at the load valley phase at load boom period to close again, switching on and shutting down are more frequent.

Under the cost of compensation of M '=3, a certain EV Centralized Controller charging predicted load and command value are as shown in Figure 2, all EV Centralized Controller charging predicted load and command value are as shown in Figure 3, system total load curve is illustrated in fig. 4 shown below, command value has carried out adjustment to a certain degree to predicted value, but is within schedulable scope; Contrast total load curve, can see, after being optimized, reduces load peak to Centralized Controller charging load.

Under different cost of compensation, the command value of Centralized Controller charging load and predicted value are illustrated in fig. 5 shown below, and along with the continuous increase of cost of compensation, the scheduling of grid dispatching center to electric automobile is more and more less.

Those skilled in the art will readily understand; the foregoing is only preferred embodiment of the present invention; not in order to limit the present invention, all any amendments done within the spirit and principles in the present invention, equivalent replacement and improvement etc., all should be included within protection scope of the present invention.

Claims (8)

1. take into account a Unit Combination model for charging electric vehicle controllability, it is characterized in that, the constraint condition of described model comprises the constraint of unit side and the constraint of user side;

The objective function of described model is:

, described in refer to unit operation cost, refer to genset fuel cost, refer to unit starting cost;

Described refer to charging electric vehicle load scheduling cost;

Described i is the numbering of genset, and j is the numbering of EV Centralized Controller, and t is the numbering of period, be exerting oneself of i-th genset t period; be the state of i-th genset t period, when it is in running status be 1, otherwise be 0; it is i-th genset t period continuous stop time; be i-th genset t period cost; it is the start-up cost of i-th genset t period; N _gfor genset sum; N _aggfor EV Centralized Controller number; T for always to optimize the period, T=24; for a jth EV Centralized Controller is at the charging load of t period; for a jth EV Centralized Controller is at the charging predicted load of t period;

Described M is EV unit compensation cost; M ' should compensate to the cost of user for 1 hour for interrupting 1MW electric automobile load;

Described a _i, b _i, c _ibe i-th genset fuel cost coefficient; it is the warm start cost of i-th genset; it is the cold start-up cost of i-th genset; it is the minimum permission stop time of i-th genset; it is the cold start-up time of i-th genset.

2. Unit Combination model as claimed in claim 1, is characterized in that, the constraint of described unit side comprises account load balancing constraints, the constraint of unit output bound, spinning reserve constraint, Unit Commitment time-constrain and unit ramp loss;

Described account load balancing constraints is: refer to that each generator output sum should be suitable with load, to meet workload demand;

Wherein, be the conventional load level of t period electric system, it is the charging load instruction value of a t period jth EV Centralized Controller;

Described unit output bound is constrained to: refer to the bound scope that fired power generating unit is exerted oneself;

Wherein, the minimum load of unit i, for the maximum output of unit i;

Described spinning reserve is constrained to: refer to the capacity for tackling load fluctuation that unit is reserved; Wherein, R ^tfor t period system spinning reserve capacity demand;

Described Unit Commitment time-constrain relates to motor from number when shutting down to the minimum shutdown starting shooting, and generator shut down from starting shooting to minimum start time number, concrete:

Generator Status is from start to shutdown:

Generator Status is from shutdown to start:

Wherein, it is i-th time that genset ran continuously in the t period; it is the minimum permission on time of i-th genset;

Described unit ramp loss is: refer to the variation range of exerting oneself of the fired power generating unit unit interval that the inertia due to fired power generating unit causes;

Wherein, for the power ascending amount of unit i limits, for the power drop amount of unit i limits.

3. Unit Combination model as claimed in claim 1 or 2, is characterized in that, the constraint of described user side comprises the constraint of user's energy requirement, the bound constraint of EV Centralized Controller charging load and the constraint of EV Centralized Controller charging load distribution ratio;

Described user's energy requirement is constrained to

Described EV Centralized Controller charging load bound is constrained to:

Wherein, the schedulable range limit of a jth EV Centralized Controller in t, the schedulable lower range limit of a jth EV Centralized Controller in t;

Wherein, can the go up capacitance-adjustable of a jth EV Centralized Controller in t, the descended capacitance-adjustable of a jth EV Centralized Controller in t;

The constraint of described EV Centralized Controller charging load distribution ratio comprises following two kinds of situations:

A, when namely t needs to raise the charging predicted load of EV Centralized Controller, then each EV Centralized Controller carries out proportional distribution according to rise amount of capacity, should meet:

B, when namely t needs to lower the charging predicted load of EV Centralized Controller, then each EV Centralized Controller carries out proportional distribution according to downward amount of capacity, should meet:

4. take into account a modeling method for the Unit Combination model of charging electric vehicle controllability, it is characterized in that, comprise the following steps:

(1) utilize EV Centralized Controller to carry out cluster to electric automobile, and obtain the EV charging plan of EV Centralized Controller and the predicted value of charging load;

(2) the schedulable capacity of EV Centralized Controller is obtained according to the EV charging plan of EV Centralized Controller;

(3) the schedulable scope of EV Centralized Controller all moment chargings load is determined according to the schedulable capacity of EV Centralized Controller;

(4) according to the constraint of machine unit characteristic determination unit side, determine that user side retrains according to EV Centralized Controller charging predicted load with the schedulable scope of charging load;

And dispatch the minimum value of cost sum for objective function with unit operation cost and electric automobile, charge load for control object with unit output and EV Centralized Controller, set up the Unit Combination model taking into account charging electric vehicle controllability.

5. modeling method as claimed in claim 4, it is characterized in that, described step (1) is specific as follows:

(1.1) collect local EV Centralized Controller parameter and EV parameter by EV Centralized Controller, comprising: EV Centralized Controller number N _agg, the EV sum N in each EV Centralized Controller, each EV grid-connected moment t ₀with from net moment t _d, i-th EV battery rated capacity C ⁱ, charge power Pev _i, the grid-connected moment battery electric quantity with the demand electricity of the user from the net moment

(1.2) according to the EV charging plan that EV charging modes arranges EV Centralized Controller local, and the charging predicted load of this EV Centralized Controller is obtained

6. modeling method as claimed in claim 5, is characterized in that, the process obtaining EV parameter in described step (1.1) is specific as follows:

(1.1.1) the historical statistical data formation probability transition matrix of rule of going on a journey based on user, utilizes Markov chain and Monte Carlo simulated sampling to obtain each EV the running status situation of change of certain day, to obtain the grid-connected moment t of each EV ₀with from net moment t _d; Wherein, the grid-connected moment refers to that electric automobile accesses the moment of electrical network, refers to that electric automobile leaves the moment of electrical network from the net moment;

(1.1.2) based on the statistics acquisition probability densimetric curve of EV battery rated capacity, charge power, the battery electric quantity in grid-connected moment and the user's request electricity from the net moment; According to the battery rated capacity C of this curve sampling acquisition i-th EV ⁱ, charge power Pev _i, the grid-connected moment battery electric quantity with the demand electricity of the user from the net moment .

7. modeling method as claimed in claim 4, it is characterized in that, the schedulable capacity of the EV Centralized Controller in described step (2), comprises and can go up capacitance-adjustable with can descend capacitance-adjustable obtain according to following steps:

(2.1) according to the driving information of EV Centralized Controller charging plan acquisition t i-th EV, the time departure t of EV is comprised _d, current electric quantity expected energy charge power Pev _iwith battery capacity C ⁱ;

(2.2) whether to be connected to the grid according to EV and the charge requirement of EV, to judge the controllability of EV;

(2.3) judge that whether t i-th EV be in charging; If not, then using the charging load of this EV as going up capacitance-adjustable; If so, then using the charging load of this EV as descending capacitance-adjustable;

(2.4) repeat step (2.2) and step (2.3), obtain the controllability of all EV in EV Centralized Controller, the load and cumulative all EV be connected to the grid charge, as the capacitance-adjustable gone up of EV Centralized Controller cumulative all EV charging loads excising out electrical network, as the descended capacitance-adjustable of EV Centralized Controller

8. modeling method as claimed in claim 7, it is characterized in that, described step (2.2) judges that the step of EV controllability is specific as follows:

(2.2.1) judge whether t meets t ₀≤ t≤t _d, if not, show that EV does not access electrical network, EV is uncontrollable; If so, show that EV accesses electrical network, enter step (2.2.2);

(2.2.2) remove with the time interval of Δ T the electrical demand judging EV whether meet the condition of following formula:

If so, then EV has controllability; If not, then show that EV need from current time trickle charge always until user leaves just can meet consumers' demand, this EV must charge in this period, did not have controllability;

Wherein, t ₀for the EV grid-connected moment, t _dfor from net the moment; for the current electric quantity of battery, η is charge efficiency.