CN104578160B - A kind of microgrid energy control method - Google Patents

Abstract

The invention discloses a kind of microgrid energy control method, described microgrid energy control method comprises the following steps: according to the probability distribution of wind speed, intensity of illumination, pricing and optimized distributionl, determine microgrid contract period operating cost Z that in electricity market, microgrid calculates in kth time period_Tk；Stochastic variable is carried out Lowest Confidence Interval process, increases imbalance power penalty term, determine microgrid contract period operating cost Z_TkExpression formula under confidence level α；Optimal confidence level is determined according to expected cost；Determine that optimal change confidence level is interval according to becoming confidence level optimisation strategy.The present invention, according to the probabilistic model of the factors such as different distributions formula power supply, thermoelectricity load, pricing, is made that accurate description to the microgrid energy Controlling model in market environment；The present invention proposes the viewpoint using expected cost to weigh confidence level selection scheme, and microgrid economical operation in distribution side Power Market is made guidance.

Description

A kind of microgrid energy control method

Technical field

The present invention relates to distributed power source and microgrid field, particularly relate to a kind of microgrid energy control method.

Background technology

Microgrid is a small electrical system collected by distributed power source, energy-storage units, load, changer etc., It is ordinarily connected in low pressure or medium voltage distribution network, there is the method for operation and schedulable performance flexibly, can provide a user with simultaneously Electric energy and heat energy, be an autonomous system being capable of self-contr ol.Microgrid needs to be similar to the energy control of tradition bulk power grid Distributed power source and energy storage device are carried out running control by system processed, it is achieved electric energy, heat energy, the efficient utilization of regenerative resource, And ensure micro net power quality and reduce microgrid operating cost.Countries in the world in the definition to microgrid all with the comprehensive utilization of energy By the basic feature for microgrid, using microgrid as a kind of new distributed energy organizational form and structure, convenient regenerative resource The access of system, realize maximally utilizing of Demand-side control and the existing energy.

Meanwhile, deepening constantly of electrical network market reform provides new opportunity for micro-grid connection operation.Along with electric power Opening of market, microgrid by participating in the scheduling of electric energy with identity-independent, by the two-way interaction with power distribution company, favorably In realizing the economic allocation of electric load, reduce the loss of transmission and distribution networks.Meanwhile, to enhance electric power networks reliable for microgrid The ability of power supply, when system malfunctions as stand-by power supply, help system restores electricity ability.Additionally, need as user Seeking a kind of effective form that side responds, microgrid is elastic according to the time of self composition and load, carries out interactive energy-saving distribution, logical Spend adjustment power trade time, the beneficially peak load shifting of power system；According to the difference of peak interval of time pricing, adjust certainly The electricity consumption behavior of body, it is possible to help user to obtain more economic benefit.

At present, Chinese scholars manages for the microgrid energy in market environment and has carried out substantial amounts of research, the work done Be concentrated mainly on scheduling model and optimization method two aspect, the Financial cost of comprehensive microgrid, technical costs, Environmental costs etc. because of Element, is converted into multi-objective problem non-linear single-objective problem and solves, or set up Model for Multi-Objective Optimization, formulates distributed electrical The economic control strategy in source.The present invention determines microgrid operation reserve for target, for electricity market with contract period economic interests optimum Microgrid economical operation under the conditions of high development provides to be instructed.

Summary of the invention

The invention provides a kind of microgrid energy control method, the present invention is according to different distributions formula power supply, thermoelectricity load, friendship The probabilistic model of the factors such as easy electricity price, is made that accurate description to the microgrid energy Controlling model in market environment, as detailed below Describe:

A kind of microgrid energy control method, described microgrid energy control method comprises the following steps:

According to the probability distribution of wind speed, intensity of illumination, pricing and optimized distributionl, determine microgrid in electricity market In microgrid contract period operating cost Z that kth time period calculates_Tk；

Stochastic variable is carried out Lowest Confidence Interval process, increases imbalance power penalty term, determine microgrid contract period Operating cost Z_TkExpression formula under confidence level α；

Optimal confidence level is determined according to expected cost；Determine according to change confidence level optimisation strategy and most preferably become confidence level Interval.

The described probability distribution according to wind speed, intensity of illumination, pricing and optimized distributionl, determines in electricity market Microgrid contract period operating cost Z that microgrid calculates in kth time period_TkStep particularly as follows:

Determine microgrid and distribution network electric energy mechanism of exchange in electricity market, set a kind of contract period electric energy clearing form；

Calculate miniature gas turbine cost of electricity-generating；Calculate wind-driven generator and photovoltaic cell capable of generating power cost；Calculating accumulator Operating cost；Calculate power trade cost；Calculate the microgrid operating cost of the i-th period；The microgrid calculated in contract period T runs into This；

Thermal load demands P in i-th period_HLξi, electrical load requirement P_ELξiProbability distribution intended by normal distribution Close；

Wind-driven generator peak power output P in i-th period_WTξiProbability distribution be Two-parameter Weibull Distribution；I-th Photovoltaic cell peak power output P in period_PVξiProbability distribution be Beta distribution；

The pricing in the i-th period probability distribution in the underload period is normal distribution, middle high load capacity period general Rate is distributed as logarithm normal distribution；

Determined in the microgrid contract period operating cost expression formula that kth time period calculates by above-mentioned steps.

The technical scheme that the present invention provides provides the benefit that: the present invention is directed in Power Market microgrid user and joins The Bilateral contracts transaction that electricity company exists, according to the probability mould of the factors such as different distributions formula power supply, thermoelectricity load, pricing Type, is made that accurate description to the microgrid energy Controlling model in market environment, proposes a kind of economic tune based on confidence interval Degree Optimized model, defines imbalance power to characterize the interval of peer-to-peer constraint and processes the imbalance problem caused, use Genetic algorithm solves, and analyzes the confidence level impact on microgrid economical operation in the case of different prediction deviations, proposes Expected cost is used to weigh the viewpoint of confidence level selection scheme, to microgrid economical operation in distribution side Power Market Make guidance.

Accompanying drawing explanation

Fig. 1 is the pricing figure of microgrid and power distribution network day part；

Fig. 2 is that distributed power source is exerted oneself and electric load figure；

Fig. 3 is the microgrid Financial cost figure under different confidence level；

Fig. 4 is the microgrid Financial cost figure in the case of different prediction deviation；

Fig. 5 is the flow chart of a kind of microgrid energy control method.

Detailed description of the invention

For making the object, technical solutions and advantages of the present invention clearer, below embodiment of the present invention is made further Ground describes in detail.

The present invention proposes a kind of economic load dispatching Optimized model based on confidence interval, defines imbalance power to characterize The interval of peer-to-peer constraint processes the imbalance problem caused, and uses genetic algorithm to solve, analyzes confidence level and exist Impact on microgrid economical operation in the case of different prediction deviations, proposes to use expected cost to weigh confidence level selection scheme Viewpoint, makes guidance to microgrid economical operation in distribution side Power Market, described below:

101: according to the probability distribution of wind speed, intensity of illumination, pricing and optimized distributionl, determine in electricity market Microgrid contract period operating cost Z that microgrid calculates in kth time period_Tk；

1011: determine microgrid and distribution network electric energy mechanism of exchange in electricity market, set a kind of contract period electric energy clearing side Formula；

(1) microgrid and power distribution network both sides set up bilateral contract, and both sides are about scheduled on purchase of electricity W in a length of T of contract period_BT, Electricity sales amount W_ST, here contract period is divided into the period of N number of a length of Δ t, and thinks that the electricity price in arbitrary Δ t does not changes Become；

(2) for without departing from W_BTPurchase of electricity enjoy discounted cost, beyond part according to actual power purchase price C_BiTie Calculate；

(3) for without departing from W_STElectricity sales amount enjoy extra profit, beyond part according to actual sale of electricity price C_SiTie Calculate；

(4) if the accumulative transaction electricity in the contract end of term fails to reach W_BT、W_ST, microgrid is regarded as differential section and buys Using, microgrid should be according to punishment electricity price C_FPay required amount of money Z_F。

1012: calculate miniature gas turbine cost of electricity-generating；

$Z_{\mathrm{MTi}}=f\left(\frac{P_{\mathrm{MTi}}}{P_{\mathrm{MT}.N}}\right)P_{\mathrm{MTi}}\mathrm{\Δt}-f\left(\frac{P_{\mathrm{HLi}}}{P_{\mathrm{MT}.N}}\right)P_{\mathrm{HLi}}\mathrm{\Δt}$

Wherein, P_MT.N、P_MTiFor the nominal output of miniature gas turbine with actual exert oneself, P_HLiFor meeting the miniature of thermic load Gas turbine minimum load.F (m)=am²+ bm+c is the combustion gas expense of miniature gas turbine unit power, $m=\frac{P_{\mathrm{MTi}}}{P_{\mathrm{MT}.N}}$ Or

1013: calculate wind-driven generator and photovoltaic cell capable of generating power cost；

Z_WTi=C_WTV_WTiP_WTiΔt

Z_PVi=C_PVV_PViP_PViΔt

Wherein, P_WTi、C_WTIt is exerting oneself and O&M cost coefficient of blower fan, P_PVi、C_PVIt is exerting oneself and operation and maintenance expenses of photovoltaic cell With coefficient, V_WTi、V_PViFor " 0 or 1 " variable, represent whether blower fan, photovoltaic cell put into operation.

1014: calculating accumulator operating cost；

Z_BATi=C_BAT|P_BATiΔt|

Wherein, P_BATiFor the charge-discharge electric power of accumulator, C_BATAmortization charge coefficient for accumulator.

1015: calculate power trade cost；

Z_Gi=K_BiC_BiW_Bi-K_SiC_SiW_Si

Wherein, W_Bi、W_Si、C_Bi、C_Si、K_Bi、K_SiIt is that the i-th period purchase of electricity, electricity sales amount, power purchase valency, sale of electricity valency and power purchase are excellent Favour coefficient and the preferential coefficient of sale of electricity.Transaction electricity is without departing from K during Contract Energy_Bi＜ 1, K_Si＞ 1, otherwise K_Bi=1, K_Si=1.

1016: calculate the microgrid operating cost of the i-th period；

$\begin{array}{c}{Z}_{\mathrm{Ti}}=C_{\mathrm{Bi}}{K}_{\mathrm{Bi}}{W}_{\mathrm{Bi}}-C_{\mathrm{Si}}{K}_{\mathrm{Si}}{W}_{\mathrm{Si}}+C_{\mathrm{WT}}{V}_{\mathrm{WTi}}{P}_{\mathrm{WTi}}\mathrm{\Δt}+C_{\mathrm{PV}}{V}_{\mathrm{PVi}}{P}_{\mathrm{PVi}}\mathrm{\Δt}\\ +f\left(\frac{P_{\mathrm{WTi}}}{P_{\mathrm{MT}.N}}\right)P_{\mathrm{MTi}}\mathrm{\Δt}-f\left(\frac{P_{\mathrm{HLi}}}{P_{\mathrm{MT}.N}}\right)P_{\mathrm{HLi}}\mathrm{\Δt}+C_{\mathrm{BAT}}\left|P_{\mathrm{BATi}}\mathrm{\Δt}\right|\end{array}$

Will be with Z_TiOptimized model for object function is defined as segmentation local optimum.

1017: the microgrid operating cost calculated in contract period T is；

$\begin{array}{c}{Z}_T=\underset{i=1}{\overset{N}{\mathrm{\Σ}}}[C_{\mathrm{Bi}}{K}_{\mathrm{Bi}}{W}_{\mathrm{Bi}}-C_{\mathrm{Si}}{K}_{\mathrm{Si}}{W}_{\mathrm{Si}}+C_{\mathrm{WT}}{V}_{\mathrm{WTi}}{P}_{\mathrm{WTi}}\mathrm{\Δt}+C_{\mathrm{PV}}{V}_{\mathrm{PVi}}{P}_{\mathrm{PVi}}\mathrm{\Δt}+f\left(\frac{P_{\mathrm{MTi}}}{P_{\mathrm{MT}.N}}\right)P_{\mathrm{MTi}}\mathrm{\Δt}-\\ f\left(\frac{P_{\mathrm{HLi}}}{P_{\mathrm{MT}.N}}\right)P_{\mathrm{HLi}}\mathrm{\Δt}+C_{\mathrm{BAT}}\left|P_{\mathrm{BATi}}\mathrm{\Δt}\right|]+C_{\mathrm{BF}}g(W_{\mathrm{BT}}-\underset{i=1}{\overset{N}{\mathrm{\Σ}}}{W}_{\mathrm{Bi}})+C_{\mathrm{SF}}g(W_{\mathrm{ST}}-\underset{i=1}{\overset{N}{\mathrm{\Σ}}}{W}_{\mathrm{Si}})\end{array}$

Wherein, C_BFFor power purchase promise breaking penalty coefficient, C_SFFor sale of electricity promise breaking penalty coefficient, g (x) is threshold function table.If x More than zero, x takesOrThen g (x)=x, otherwise g (x)=0.Will be with Z_TOptimization for object function Model is global optimization afterwards.

Thermal load demands P in 1018: the i-th periods_HLξi, electrical load requirement P_ELξiProbability distribution by normal distribution come Being fitted, wherein identifier ξ represents that this variable is predictive value；

Wind-driven generator peak power output P in 1019: the i-th periods_WTξiProbability distribution be that two-parameter Weibull divides Cloth；

$f\left(P_{\mathrm{WT\ξi}}\right)=\frac{B}{A}{\left(\frac{P_{\mathrm{WT\ξi}}}{A}\right)}^{B-1}{e}^{-{\left(\frac{P_{\mathrm{WT\ξi}}}{A}\right)}^B}$

Wherein, A is scale parameter, and B is form parameter.

Photovoltaic cell peak power output P in 1020: the i-th periods_PVξiProbability distribution be Beta distribution；

$f\left(P_{\mathrm{PV\ξi}}\right)=\frac{\mathrm{\Γ}(a+b)}{\mathrm{\Γ}\left(a\right)\mathrm{\Γ}\left(b\right)}{\left(\frac{P_{\mathrm{PV\ξi}}}{P_{\mathrm{PVi}}}\right)}^{a-1}{(1-\frac{P_{\mathrm{PV\ξi}}}{P_{\mathrm{PVi}}})}^{b-1}$

Wherein a, b are the form parameter of Beta distribution.Г is Gamma function.

The pricing in 1021: the i-th periods probability distribution in the underload period is normal distribution, the middle high load capacity period Probability distribution be logarithm normal distribution.Its middle-low load period is taken as 0 o'clock to 7 o'clock, and the middle high load capacity period is 7 o'clock to 24 o'clock.

1022: determine in the microgrid contract period operating cost expression formula that kth time period calculates.

The microgrid contract period operating cost that kth time period calculates is:

$\begin{array}{c}{Z}_{\mathrm{Tk}}=\underset{i=k}{\overset{N}{\mathrm{\Σ}}}[C_{\mathrm{B\ξi}}{K}_{\mathrm{Bi}}{W}_{\mathrm{Bi}}-C_{\mathrm{S\ξi}}{K}_{\mathrm{Si}}{W}_{\mathrm{Si}}+C_{\mathrm{WT}}{V}_{\mathrm{WTi}}{P}_{\mathrm{WT\ξi}}\mathrm{\Δt}+C_{\mathrm{PV}}{V}_{\mathrm{PVi}}{P}_{\mathrm{PV\ξi}}\mathrm{\Δt}+f\left(\frac{P_{\mathrm{MTi}}}{P_{\mathrm{MT}.N}}\right)P_{\mathrm{MTi}}\mathrm{\Δt}-\\ f\left(\frac{P_{\mathrm{HL\ξi}}}{P_{\mathrm{MT}.N}}\right)P_{\mathrm{HL\ξi}}\mathrm{\Δt}+C_{\mathrm{BAT}}\left|P_{\mathrm{BATi}}\mathrm{\Δt}\right|]+C_{\mathrm{BF}}g(W_{\mathrm{BT}}-\underset{i=1}{\overset{N}{\mathrm{\Σ}}}{W}_{\mathrm{Bi}})+C_{\mathrm{SF}}g(W_{\mathrm{ST}}-\underset{i=1}{\overset{N}{\mathrm{\Σ}}}{W}_{\mathrm{Si}})\end{array}$

102: to stochastic variable (C in 101_Bξi,C_Sξi,P_WTξi,P_PVξi,P_HLξi,P_ELξi) carry out at Lowest Confidence Interval Reason, increases imbalance power penalty term in object function, determines microgrid contract period operating cost Z_TkTable under confidence level α Reach formula；

1021: with Z_TkOptimum for constraints to miniature gas turbine in the Optimized model of object function foundation:

P_MTi≥P_HLξi

P_MT.Min≤P_MTi≤P_MT.N

Wherein, P_HLξiMiniature gas turbine corresponding to i period microgrid thermal load demands goes out force value.

1022: with Z_TkOptimum for constraints to accumulator in the Optimized model of object function foundation:

|P_BATi|≤P_BAT.Max

30%≤SOC_i≤ 100%

Wherein, P_BAT.MaxThe maximum charge-discharge electric power allowed by accumulator, SOC_iIt it is the charged shape of accumulator of the i-th period State.

1023: with Z_TkOptimum for constraints to bargain transaction in the Optimized model of object function foundation:

W_Bi·W_Si=0

1024: with Z_TkOptimum for constraints to power-balance in the Optimized model of object function foundation:

$\frac{W_{\mathrm{Bi}}-W_{\mathrm{Si}}}{\mathrm{\Δt}}=P_{\mathrm{EL\ξi}}-V_{\mathrm{WTi}}{P}_{\mathrm{WT\ξi}}-V_{\mathrm{PVi}}{P}_{\mathrm{PV\ξi}}-P_{\mathrm{MTi}}-P_{\mathrm{BATi}}$

Wherein, P_ELξiMicrogrid electrical load requirement for the i period.

1025: stochastic variable Lowest Confidence Interval under α confidence level determines that principle is:

$\left\{\begin{array}{c}F\left(x_{\mathrm{\α}}^{+}\right)-F\left(x_{\mathrm{\α}}^{-}\right)=\mathrm{\α}\\ f\left(x_{\mathrm{\α}}^{+}\right)=f\left(x_{\mathrm{\α}}^{-}\right)\\ x_{\mathrm{\α}}^{-}\≤x_{\max }\≤x_{\mathrm{\α}}^{+}\end{array}\right.$

Wherein, F (x), f (x) are respectively the probability-distribution function of stochastic variable x, probability density function,Respectively Represent the up-and-down boundary of variable x confidence interval under confidence level α, x_maxFor the variable corresponding to probability density value maximum Value.

HereinafterPoint Do not represent the electric load P of the i-th period_ELξi, thermic load P_HLξi, blower fan peak power output P_WTξi, photovoltaic peak power output P_PVξi, purchase electricity price C_Bξi, sale of electricity electricity price C_SξiConfidence interval under confidence level α.

1026: theoretical according to confidence interval, definition imbalance power is:

$\mathrm{\Δ}{P}_i=\frac{W_{\mathrm{Bi}}-W_{\mathrm{Si}}}{\mathrm{\Δt}}-[P_{\mathrm{EL\αi}}^{-}-V_{\mathrm{WTi}}{P}_{\mathrm{WT\αi}}^{+}-V_{\mathrm{PVi}}{P}_{\mathrm{PV\αi}}^{+}-P_{\mathrm{MTi}}-P_{\mathrm{BATi}}]$

Wherein, imbalance power penalty term is the regulation expense considering may cause because of load and unbalanced power, if Power supply power supply capacity is underestimated, then can increase power purchase regulation expenseIf power supply power supply capacity overestimate, mistake In wanting to buy the probability increase that electricity makes sale of electricity break a contract, therefore penalty term W that amendment is broken a contract_SiFor g (W_Si-ΔP_iΔt_i)。

1027: set up confidence interval expression formula.

$\begin{array}{c}\min Z_{\mathrm{Tk}}=\underset{i=k}{\overset{N}{\mathrm{\Σ}}}[C_{\mathrm{B\αi}}^{+}{K}_{\mathrm{Bi}}{W}_{\mathrm{Bi}}-C_{\mathrm{S\αi}}^{-}{K}_{\mathrm{Si}}{W}_{\mathrm{Si}}+C_{\mathrm{WT}}{V}_{\mathrm{WTi}}{P}_{\mathrm{WT\αi}}^{+}\mathrm{\Δt}+C_{\mathrm{PV}}{V}_{\mathrm{PVi}}{P}_{\mathrm{PV\αi}}^{+}\mathrm{\Δt}+f\left(\frac{P_{\mathrm{MTi}}}{P_{\mathrm{MT}.N}}\right)P_{\mathrm{MTi}}\mathrm{\Δt}-f\left(\frac{P_{\mathrm{HL\αi}}^{-}}{P_{\mathrm{MT}.N}}\right)P_{\mathrm{HL\αi}}^{-}\mathrm{\Δt}+\\ C_{\mathrm{BAT}}\left|P_{\mathrm{BATi}}\mathrm{\Δt}\right|]+\underset{i=k}{\overset{N}{\mathrm{\Σ}}}{C}_{\mathrm{B\αi}}^{+}\mathrm{\Δ}{P}_i\mathrm{\Δ}{t}_i+C_{\mathrm{BF}}g(W_{\mathrm{BT}}-\underset{i=1}{\overset{N}{\mathrm{\Σ}}}{W}_{\mathrm{Bi}})+C_{\mathrm{SF}}g(W_{\mathrm{ST}}-\underset{i=1}{\overset{k-1}{\mathrm{\Σ}}}{W}_{\mathrm{Si}}-\underset{i=k}{\overset{N}{\mathrm{\Σ}}}g(W_{\mathrm{Si}}-\mathrm{\Δ}{P}_i\mathrm{\Δ}{t}_i))\end{array}$

103: determine optimal confidence level according to expected cost；

1031: use genetic algorithm to solve, to variable V_WTi、V_PVi, use simple binary coding；To variable W_Bi、 W_Si、P_MTi、P_BATi, use real coding, and they are carried out linear transformation be mapped on [0,1] interval.

1032: use genetic algorithm to solve, use wheel disc gaming act to select father individual.According to Z_TkOrder from big to small Arrangement individuality, forms queue M.The select probability P of individual i_iBy its position W in queue M_iDetermined:

$P_i=\frac{2W_i}{Q(Q+1)}$

Wherein, Q is Population Size.

1033: use genetic algorithm to solve, it is contemplated that decision variable comprises continuous variable and discrete variable simultaneously, In the middle of using, the mode of restructuring realizes crossover operator, uses simple perturbation scheme to realize mutation operator.Regulation crossover probability is 0.6 ～change in the range of 0.9, mutation probability changes in the range of 0.01～0.1.

1034: definition δ is the relative error of the maximum of probability density prediction value of before and after's prediction period, and reference value is actual Value.Set different error | δ |, different confidence levels α, repeat said process, obtain in different error levels, different confidence Microgrid contract period operating cost Z under level_T(|δ|,α)；

1035: definition expected cost；

Expected cost definition method is

$Z_{\mathrm{EX}}\left(\mathrm{\α}\right)=\frac{\∫w\left(\right|\mathrm{\δ}\left|\right)Z_T\left(\right|\mathrm{\δ}|,\mathrm{\α})\mathrm{d\δ}}{\∫w\left(\right|\mathrm{\δ}\left|\right)\mathrm{d\δ}}$

Wherein, (| δ | is Z to w_TThe weight coefficient of (| δ |, α), is actual value average probability in the probability distribution supposed Density.

104: determine that optimal change confidence level is interval according to becoming confidence level optimisation strategy.

Judge that the microgrid with contract period economic interests as target runs optimal confidence level according to expected cost.For farther out The bigger confidence level of slot setup, can improve prediction fault-tolerant ability, and for the less confidence of nearer slot setup Level, can obtain the management scheme of relatively small economy cost, and based on this, the present invention proposes to become confidence level optimisation strategy.

The optimal confidence level interval that becomes is [0.3,0.7], and specific implementation method is

${\mathrm{\α}}_i=0.3+\frac{0.7-0.3}{N}(i-k)$

Wherein, α_iIt is the confidence level that should use the i-th period, hop count when N is total in contract period；K, i represent tune respectively Spend period and future time period.

The feasibility of this method is verified below with concrete test, described below:

The present invention is with the miniature gas turbine of a 65kW, and its minimum technology is exerted oneself 15kW, the wind-power electricity generation of a 30kW Machine, the microgrid that the accumulator of the photovoltaic cell of one group of 20kW and one group of 200kW h is constituted is that example illustrates.Set and close Contract period is in terms of sky, and agreement W_BT=260kW h, W_ST=175kW h, as shown in Figure 1, contract is electric for pricing situation of change K in amount_Bi=0.8, K_Si=1.2.The electric load P of microgrid typical case's day_ELi, the Gas Turbine Output P of thermal load demands_HLi, blower fan EIAJ P_WTi, photovoltaic cell EIAJ P_PViAs shown in Figure 2.Relevant cost parameter value is as shown in table 1, and probability distribution is joined Number is as shown in table 2.

Table 1

Parameter	a	b	c	CWT	CPV	CBAT	CF
								Value	-0.4039	-0.0061	0.71	0.3	0.3	0.5	1.0

Table 2

Note: P_ELi、P_HLi、P_WTi、P_PVi、C_Bi、C_SiFor the actual value in Fig. 2, Fig. 1.

Use genetic algorithm and based on confidence interval Optimized model, microgrid Financial cost in contract period is optimized, its Result is as it is shown on figure 3, and be drawn on Fig. 3 by segmentation local optimum and global optimization result afterwards.

As can be seen from Figure 3: 1. microgrid Financial cost based on confidence interval significantly lower than segmentation local optimum and is higher than Global optimization afterwards；2. in the case of | the δ | that exists is given, Z_T=f (α) is a V-type curve, the confidence level bottom curve along with The increase of | δ | and increase；3. in the case of α is given, the biggest Z of | δ |_TThe biggest；4. confidence level is the lowest, Z_TAffected by | δ | The biggest.Here the corresponding contract end of term maximum predicted deviation of | δ | of adjacent time interval is [0,100%].

Z is calculated according to defined expected cost_EX(α), Fig. 3 is seen.It can be seen that at occasion the biggest or the least for α, Z_EX (α) the biggest, α microgrid Financial cost in the range of 0.3～0.7 is the most relatively low.

For this example, become confidence level shift gears into

$\mathrm{\α}={\mathrm{\α}}_{\min }+\frac{{\mathrm{\α}}_{\max }-{\mathrm{\α}}_{\min }}{N}(i-k)$

Wherein, α_max、α_minIt it is confidence level maximum, minimum；Hop count when N is total in contract period；K, i represent scheduling respectively Period and future time period.

It is optimized according to change confidence level method and solves, see Fig. 4.Become confidence level optimum results as shown in table 3.

Table 3

α	0.7	[0.1,0.9]	[0.2,0.8]	[0.3,0.7]	[0.4,0.6]
						ZT(first)	189.52	146.87	134.69	119.02	122.96

By the Cost comparisons of contract period, become confidence level optimization method in the range of [0.3,0.7] and there is reality The property used.

In sum, the present invention is directed to the Bilateral contracts that in Power Market, microgrid user and power distribution company exist hand over Easily, according to the probabilistic model of the factors such as different distributions formula power supply, thermoelectricity load, pricing, to the microgrid energy in market environment Amount administrative model is made that accurate description, proposes a kind of economic load dispatching Optimized model based on confidence interval, defines imbalance Power processes, to characterize the interval of peer-to-peer constraint, the imbalance problem caused, and uses genetic algorithm to solve, analyzes Confidence level impact on microgrid economical operation in the case of different prediction deviations, proposes to use expected cost to weigh confidence level The viewpoint of selection scheme, makes guidance to microgrid economical operation in distribution side Power Market.

It will be appreciated by those skilled in the art that accompanying drawing is the schematic diagram of a preferred embodiment, the invention described above embodiment Sequence number, just to describing, does not represent the quality of embodiment.

The foregoing is only presently preferred embodiments of the present invention, not in order to limit the present invention, all spirit in the present invention and Within principle, any modification, equivalent substitution and improvement etc. made, should be included within the scope of the present invention.

Claims (1)

1. a microgrid energy control method, it is characterised in that described microgrid energy control method comprises the following steps:

According to the probability distribution of wind speed, intensity of illumination, pricing and optimized distributionl, determine that in electricity market, microgrid is in kth Microgrid contract period operating cost Z that period calculates_Tk；

Stochastic variable is carried out Lowest Confidence Interval process, increases imbalance power penalty term, determine that microgrid contract period runs Cost Z_TkExpression formula under confidence level α；

Optimal confidence level is determined according to expected cost；Optimal change confidence level district is determined according to becoming confidence level optimisation strategy Between；

Wherein, the described probability distribution according to wind speed, intensity of illumination, pricing and optimized distributionl, determine in electricity market Microgrid contract period operating cost Z that microgrid calculates in kth time period_TkStep particularly as follows:

Determine microgrid and distribution network electric energy mechanism of exchange in electricity market, set a kind of contract period electric energy clearing form；

Calculate miniature gas turbine cost of electricity-generating；Calculate wind-driven generator and photovoltaic cell capable of generating power cost；Calculating accumulator runs Cost；Calculate power trade cost；Calculate the microgrid operating cost of the i-th period；Calculate the microgrid operating cost in contract period T；

Thermal load demands P in i-th period_HLξi, electrical load requirement P_ELξiProbability distribution be fitted by normal distribution；

Wind-driven generator peak power output P in i-th period_WTξiProbability distribution be Two-parameter Weibull Distribution；I-th period Interior photovoltaic cell peak power output P_PVξiProbability distribution be Beta distribution；

The pricing in the i-th period probability distribution in the underload period is normal distribution, and the probability of middle high load capacity period divides Cloth is logarithm normal distribution；

Determined in the microgrid contract period operating cost expression formula that kth time period calculates by above-mentioned steps；

Wherein, k and i is respectively scheduling slot and future time period；

With Z_TkSet up miniature gas turbine in the optimum Optimized model for object function respectively, accumulator, bargain transaction, power are put down The constraints of weighing apparatus；

Stochastic variable is carried out Lowest Confidence Interval process,

Determine that principle is:

$\left\{\begin{array}{c}F\left(x_{\mathrm{\α}}^{+}\right)-F\left(x_{\mathrm{\α}}^{-}\right)=\mathrm{\α}\\ f\left(x_{\mathrm{\α}}^{+}\right)=f\left(x_{\mathrm{\α}}^{-}\right)\\ x_{\mathrm{\α}}^{-}\≤x_{\max }\≤x_{\mathrm{\α}}^{+}\end{array}\right.$

F (x), f (x) are respectively the probability-distribution function of stochastic variable x, probability density function,Represent that variable x exists respectively The up-and-down boundary of the confidence interval under confidence level α, x_maxFor the variate-value corresponding to probability density value maximum；

Increase imbalance power penalty term, determine microgrid contract period operating cost Z_TkExpression formula under confidence level α；

Using genetic algorithm to solve, definition δ is the relative error of the maximum of probability density prediction value of before and after's prediction period, if Fixed different error | δ |, different confidence levels α, obtain the microgrid contract period under different error levels, different confidence level Operating cost Z_T(|δ|,α)；

Definition expected cost

Wherein, w (| δ |) is Z_TThe weight coefficient of (| δ |, α), is actual value average probability density in the probability distribution supposed；

Judge that the microgrid with contract period economic interests as target runs optimal confidence level according to expected cost；

For the confidence level that slot setup farther out is bigger, for the confidence level that nearer slot setup is less, according to change Confidence level optimisation strategy determines that optimal change confidence level is interval.