CN105117802A - Central air-conditioner energy storage characteristic-based power market optimal dispatching strategy - Google Patents

Abstract

The invention discloses a central air-conditioner energy storage characteristic-based power market optimal dispatching strategy. According to the power market optimal dispatching strategy, a storage battery model is established according to the thermodynamic model of a building to which a central air conditioner belongs; in the day-ahead market, a load aggregator performs electricity purchase arrangement according to clearing electricity price or load prediction condition of each time section of the next day, so that minimum day-ahead market electricity purchase cost of the load aggregator can be realized; since errors exist in load prediction of the load aggregator, difference between actual electricity purchase quantity and planned electricity purchase quantity is required to be balanced by means of real-time market, and the load aggregator predicts electricity price, load and outdoor temperature at time sections in the further according to current real-time electricity price, load and outdoor temperature; and an optimal dispatching model is established with minimum electricity purchase cost adopted as a target function and charging and discharging power of a storage battery adopted as decision-making variables, so that charging and discharging can be re-arranged for the storage battery, and the behaviors of the load aggregator in the real-time market can be optimized.

Description

A kind of electricity market Optimized Operation strategy based on central air conditioner energy storage characteristic

Technical field

The invention belongs to the application technology of central air-conditioning load in electricity market, be specifically related to the Optimized Operation strategy of Load aggregation business and the energy storage characteristic of air conditioner load.

Background technology

Air conditioner load is comparatively large at the proportion of power equipment terminal, and scheduling mode is flexible, and participating in system cloud gray model and have a high potential, is a kind of important demand response resource.Air conditioner load has hot storage capacity, by rational control device, can the scheduling of responding system rapidly, and for system provides excellent energy reserve service.Load aggregation business is a kind of business model being specifically designed to integration load side resource, middle-size and small-size burdened resource can not only be represented and participate in electricity market, and real-time Measurement & Control can be carried out by means of the advanced measuring system of intelligent grid to load, realize the efficiency utilization of resource and the maximization of economic benefit.

Along with the appearance of " the some suggestions about deepening power system reform further ", Power Market In China reform is deepened day by day, ahead market and Real-time markets operating mechanism will be more and more ripe, for the raising of the efficiency utilization and relevant enterprise economic benefit that realize load side resource provides a favourable opportunity.Meanwhile, the new forms of energy such as wind energy, sun power access the proposition of electrical network and energy internet concept in a large number, more and more higher requirement is proposed to energy-storage travelling wave tube, but the cost of traditional energy-storage travelling wave tube (as accumulator) is often higher, economy is poor, and air conditioner load is that the energy-storage travelling wave tube in electricity market provides another kind of possibility.

Summary of the invention

Goal of the invention: in order to reduce the power purchase expense that Load aggregation business increases because load prediction there is the problems such as error, the invention provides a kind of electricity market Optimized Operation strategy based on buildings energy storage characteristic belonging to central air conditioner.The energy storage characteristic of Load aggregation business buildings belonging to central air conditioner sets up its battery model; In ahead market, Load aggregation business is making the power purchase arrangement of each period of next day according to the clear electricity price of secondary sunrise and load prediction situation; In Real-time markets, Load aggregation business by Spot Price, load, outdoor temperature to the electricity price of future time period, load and outdoor temperature, and with the charge-discharge electric power of accumulator for decision variable, minimum for objective function with power purchase expense, set up Optimal Operation Model, realize the optimum Transaction algorithm of Real-time markets of Load aggregation business, make the maximization of economic benefit of Load aggregation business.

Technical scheme: for achieving the above object, the technical solution used in the present invention is:

Based on an electricity market Optimized Operation strategy for central air conditioner energy storage characteristic, comprise the steps:

(1) thermodynamical model of buildings belonging to central air conditioner, set up battery model:

B＝{O,SOE,P _d,C _max,D _max}(1)

In formula: B represents the parameter sets of battery model, and O represents stored energy capacitance, SOE represents lotus energy state, P _drepresent floating charging power, C _maxrepresent maximum charge power, D _maxrepresent maximum discharge power;

(2) Load aggregation business predicts the cleaing price of each period of next day and load, and by arranging the purchase of electricity of each period of next day to the charge and discharge control of battery model, make Load aggregation business power purchase next day expense minimum, form operation plan a few days ago:

${\mathrm{minF}}_1=\underset{i=1}{\overset{N}{\Σ}}\[E_s\left(i\right)p_{ad}^{\′}\left(i\right)+\mathrm{\λ}|C\left(i\right)-D\left(i\right)|\]---\left(2\right)$

E _s(i)＝L _ad'(i)+C(i)-D(i)(3)

In formula: F ₁represent that Load aggregation business to dispatch the total power purchase expense of prediction in market, E a few days ago _si () represents the prediction purchase of electricity of a few days ago dispatching period i in market, p' _adi () represents the prediction cleaing price a few days ago dispatching period i in market, L _ad' (i) represent the prediction load a few days ago dispatching period i in market; λ represents the scheduling expense of battery model, the charge power of battery model when C (i) represents period i, the discharge power of battery model when D (i) represents period i, and N represents period sum;

(3) when Load aggregation business to submit to a few days ago after operation plan to system operator, the actual cleaing price of each period of next day and known, then:

$F_2=\underset{i=1}{\overset{N}{\Σ}}\[E_s\left(i\right)p_{ad}\left(i\right)+\mathrm{\λ}|C\left(i\right)-D\left(i\right)|\]---\left(4\right)$

In formula: F ₂represent that Load aggregation business to dispatch the total power purchase expense of reality in market, p a few days ago _adi () represents the actual cleaing price a few days ago dispatching period i in market;

(4) because the load prediction of Load aggregation business to next day exists error, need to rely on Real-time markets to balance the deviation between real-time purchase of electricity and prediction purchase of electricity; In Real-time markets, the real-time purchase of electricity of Load aggregation business Real-time Obtaining period i, in real time cleaing price, Real-time Load and real-time outdoor temperature are optimized scheduling, and detailed process is:

(41) when Load aggregation business is optimized scheduling, with the real-time purchase of electricity of period i, in real time cleaing price, Real-time Load and in real time outdoor temperature to the purchase of electricity of period i+1 ~ period i+n in Real-time markets, cleaing price, load and and outdoor temperature predict;

(42) Load aggregation business with the charge-discharge electric power of battery model for decision variable, minimum for target with power purchase expense, set up Optimal Operation Model:

$\begin{array}{c}\min F_3\\ =\[\left(E_a\right(i)-E_s\left(i\right))p_{rt}\left(i\right)+\mathrm{\λ}|C\left(i\right)-D\left(i\right)|\]\\ =\underset{k=i+1}{\overset{i+n}{\mathrm{\Σ}}}\[({E_a}^{\′}\left(k\right)-E_s\left(k\right)){p_{rt}}^{\′}\left(k\right)+\mathrm{\λ}|C\left(k\right)-D\left(k\right)|\]\end{array}---\left(5\right)$

E _a(i)＝L _a(i)+C(i)-D(i)(6)

E _a'(k)＝L _a'(k)+C(k)-D(k)(7)

In formula: F ₃represent the total power purchase expense of the prediction of Load aggregation business in Real-time markets; E _ai () represents the real-time purchase of electricity of period i, p _rti () represents the real-time cleaing price of period i, L _ai () represents the Real-time Load of period i; E _a' (k) represent the prediction purchase of electricity of period k in Real-time markets, p _rt' (k) represent the prediction cleaing price of period k in Real-time markets, L _a' (k) represent the prediction load of period i in Real-time markets; N represents that period sum is optimized in retrogressing;

(43) adopt linear programming method to solve the formula (5) as objective function, according to solving result, charge-discharge electric power control is carried out to battery model during period i;

(44) i=i+1, returns step (41).

In described step (2) and step (42), all require 0≤C (i)≤C _maxand 0≤D (i)≤D _max.

Beneficial effect: the present invention is directed to Load aggregation business and purchase sale of electricity behavior in ahead market and Real-time markets, a kind of electricity market Optimized Operation strategy based on central air conditioner energy storage characteristic is proposed, its advantage is on buildings thermodynamical model basis belonging to air-conditioning, its energy storage characteristic of abundant excavation, set up battery model, ahead market and Real-time markets is participated under the scheduling and control of Load aggregation business, not only reduce the dependence to the higher traditional energy-storage travelling wave tube (as accumulator) of cost, delay the investment of energy storage device in electric system, improve the economy of electric system operation, meanwhile, too increase the economic benefit of Load aggregation business.

Accompanying drawing explanation

Fig. 1 is the general flow chart of the inventive method;

Fig. 2 is Load aggregation business operation system;

The battery model of Fig. 3 buildings belonging to air-conditioning.

Embodiment

Below in conjunction with accompanying drawing, the present invention is further described.

Based on an electricity market Optimized Operation strategy for central air conditioner energy storage characteristic, comprise the steps:

(1) thermodynamical model of buildings belonging to central air conditioner, set up battery model:

B＝{O,SOE,P _d,C _max,D _max}(1)

In formula: B represents the parameter sets of battery model, and O represents stored energy capacitance, SOE represents lotus energy state, P _drepresent floating charging power, C _maxrepresent maximum charge power, D _maxrepresent maximum discharge power;

(2) Load aggregation business predicts the cleaing price of each period of next day and load, and by arranging the purchase of electricity of each period of next day to the charge and discharge control of battery model, make Load aggregation business power purchase next day expense minimum, form operation plan a few days ago:

${\mathrm{minF}}_1=\underset{i=1}{\overset{N}{\Σ}}\[E_s\left(i\right)p_{ad}^{\′}\left(i\right)+\mathrm{\λ}|C\left(i\right)-D\left(i\right)|\]---\left(2\right)$

E _s(i)＝L _ad'(i)+C(i)-D(i)(3)

In formula: F ₁represent that Load aggregation business to dispatch the total power purchase expense of prediction in market, E a few days ago _si () represents the prediction purchase of electricity of a few days ago dispatching period i in market, p' _adi () represents the prediction cleaing price a few days ago dispatching period i in market, L _ad' (i) represent the prediction load a few days ago dispatching period i in market; λ represents the scheduling expense of battery model, the charge power of battery model when C (i) represents period i, the discharge power of battery model when D (i) represents period i, and N represents period sum; 0≤C (i)≤C _maxand 0≤D (i)≤D _max;

(3) when Load aggregation business to submit to a few days ago after operation plan to system operator, the actual cleaing price of each period of next day and known, then:

$F_2=\underset{i=1}{\overset{N}{\Σ}}\[E_s\left(i\right)p_{ad}\left(i\right)+\mathrm{\λ}|C\left(i\right)-D\left(i\right)|\]---\left(4\right)$

In formula: F ₂represent that Load aggregation business to dispatch the total power purchase expense of reality in market, p a few days ago _adi () represents the actual cleaing price a few days ago dispatching period i in market;

(4) because the load prediction of Load aggregation business to next day exists error, need to rely on Real-time markets to balance the deviation between real-time purchase of electricity and prediction purchase of electricity; In Real-time markets, the real-time purchase of electricity of Load aggregation business Real-time Obtaining period i, in real time cleaing price, Real-time Load and real-time outdoor temperature are optimized scheduling, and detailed process is:

(41) when Load aggregation business is optimized scheduling, with the real-time purchase of electricity of period i, in real time cleaing price, Real-time Load and in real time outdoor temperature to the purchase of electricity of period i+1 ~ period i+n in Real-time markets, cleaing price, load and and outdoor temperature predict;

(42) Load aggregation business with the charge-discharge electric power of battery model for decision variable, minimum for target with power purchase expense, set up Optimal Operation Model:

$\begin{array}{c}\min F_3\\ =\[\left(E_a\right(i)-E_s\left(i\right))p_{rt}\left(i\right)+\mathrm{\λ}|C\left(i\right)-D\left(i\right)|\]\\ =\underset{k=i+1}{\overset{i+n}{\mathrm{\Σ}}}\[({E_a}^{\′}\left(k\right)-E_s\left(k\right)){p_{rt}}^{\′}\left(k\right)+\mathrm{\λ}|C\left(k\right)-D\left(k\right)|\]\end{array}---\left(5\right)$

E _a(i)＝L _a(i)+C(i)-D(i)(6)

E _a'(k)＝L _a'(k)+C(k)-D(k)(7)

In formula: F ₃represent the total power purchase expense of the prediction of Load aggregation business in Real-time markets; E _ai () represents the real-time purchase of electricity of period i, p _rti () represents the real-time cleaing price of period i, L _ai () represents the Real-time Load of period i; E _a' (k) represent the prediction purchase of electricity of period k in Real-time markets, p _rt' (k) represent the prediction cleaing price of period k in Real-time markets, L _a' (k) represent the prediction load of period i in Real-time markets; N represents that period sum is optimized in retrogressing; 0≤C (i)≤C _maxand 0≤D (i)≤D _max;

(43) adopt linear programming method to solve the formula (5) as objective function, according to solving result, charge-discharge electric power control is carried out to battery model during period i;

(44) i=i+1, returns step (41).

In described step (1), the thermodynamical model of affiliated buildings is:

$\frac{{\mathrm{dT}}_{in}}{dt}=\mathrm{\α}(T_o-T_{in})+\mathrm{\γ}-\mathrm{\μ}Q---(1-1)$

$T_{in}=\left(T_{in}\right(0)-\frac{{\mathrm{\αT}}_o+\mathrm{\γ}-\mathrm{\μ}Q}{\mathrm{\α}})e^{-\mathrm{\α}t}+\frac{{\mathrm{\αT}}_o+\mathrm{\γ}-\mathrm{\μ}Q}{\mathrm{\α}}---(1-2)$

In formula: T _inrepresent indoor temperature, T _orepresent outdoor temperature, Q represents central air conditioner refrigerating capacity, and α, γ and μ are coefficient, T _in(0) indoor temperature of initial time is represented; T represents the time.

In described step (1), the process of establishing of battery model is as follows:

(1.1) stored energy capacitance O

If the indoor temperature interval meeting human comfort is [T _min, T _max], the cold of buildings belonging to central air conditioner is stored in room air and indoor solid and (is similar to accumulator by power storage in electric capacity), and indoor temperature is in T _mintime storage cold maximum, indoor temperature is in T _maxtime storage cold minimum; Indoor temperature is in T _maxtime storage cold be designated as 0, then when indoor temperature is in T _mintime storage cold be:

$O_{in}=\frac{T_{max}-T_{in}}{\mathrm{\μ}}---(1-3)$

In formula: represent that buildings often raises the energy required for 1 DEG C; It can thus be appreciated that stored energy capacitance O is:

$O=\frac{T_{max}-T_{min}}{\mathrm{\μ}}---(1-4)$

(1.2) lotus energy state SOE

The ratio of the current storage cold of lotus energy state representation and stored energy capacitance, reflects the energy storage state of battery model:

$SOE=\frac{O_{in}}{O}=\frac{T_{max}-T_{in}}{T_{\max }-T_{\min }}---(1-5)$

The heat energy of electric energy conversion for ease of storing that air-conditioning can will be not easy to store, it is also in storing electricity in itself that buildings belonging to air-conditioning stores cold, therefore SOE also reflects the storing electricity situation of battery model simultaneously, SOE value is larger, and battery model storing electricity is larger.

Formula (1-5) is brought into (1-2) and the time dependent rule of lotus energy state can be obtained:

$SOE\left(i\right)=SOE\left(0\right)\·e^{-\mathrm{\α}t}+\frac{\mathrm{\α}{(T_{max}-T)}_o-\mathrm{\γ}+\mathrm{\μ}Q}{\mathrm{\α}(T_{max}-T_{min})}(1-e^{-\mathrm{\α}t})---(1-6)$

In formula: reach lotus energy state during steady state (SS) during SOE (i) represents period i, SOE (0) represents the lotus energy state of initial time.

(1.3) floating charging power P _d

Buildings belonging to central air conditioner, due to reasons such as indoor/outdoor temperature-difference radiation, personnel's heat radiation, heat radiation of electrical apparatus, can produce heat, indoor temperature is raised, and lotus energy state value reduces, and namely battery model exists internal resistance and self-discharge processes; Constant in order to maintain lotus energy state value, central air conditioner needs from electrical network, to absorb electricity to produce equal cold, and take away the heat that buildings increases, the electric power setting up central air conditioner is the floating charging power of battery model:

P _d＝P(1-7)

In formula: P represents the electric power of central air conditioner, P is the function of SOE.

(1.4) maximum charge power C _max

If the lotus energy state of current battery model is SOE (i), when being increased to lotus energy state SOE (i+1) by lotus energy state SOE (i), known according to formula (1-1), the cold required for maintenance lotus energy state SOE (i+1) is:

$Q=\frac{\mathrm{\α}\[T_o-T_{max}+SOE(i+1)\·(T_{max}-T_{min})\]+\mathrm{\γ}}{\mathrm{\μ}}---(1-8)$

By converter technique, handpiece Water Chilling Units, chilled water pump, blower fan are controlled, make refrigerating capacity be adjusted to Q, according to formula (1-2) calculate increase to lotus energy state SOE (i+1) by lotus energy state SOE (i) time the required time be:

$t=-\frac{ln\frac{\mathrm{\α}(T_{max}-T_o)-\mathrm{\γ}+\mathrm{\μ}Q-\mathrm{\α}\·SOE(i+1)\·(T_{max}-T_{min})}{\mathrm{\α}(T_{max}-T_o)-\mathrm{\γ}+\mathrm{\μ}Q-\mathrm{\α}\·SOE\left(i\right)\·(T_{max}-T_{\min })}}{\mathrm{\α}}=\mathrm{\∞}---(1-9)$

Formula (1-9) explanation, lotus energy state is along with time infinite approach SOE (i+1) but can not equal SOE (i+1), this is determined by battery model self-characteristic, therefore establish an Adjustment precision △ SOE, when lotus energy state is increased to SOE (i+1)-△ SOE, can think that lotus energy state has been adjusted to SOE (i+1); In order to realize effective control of battery model, lotus energy state need be made in a scheduling slot to reach stable, need make:

t≤△t(1-10)

Solve and can obtain:

$SOE(i+1)\≤SOE\left(i\right)+\frac{T_{max}-\mathrm{\Δ}SOE\·(T_{max}-T_{\min })}{(T_{max}-T_{\min })e^{-\mathrm{\α}\mathrm{\Δ}t}}---(1-11)$

Therefore establish that lotus energy state is maximum to be increased to in like manner can obtain that lotus energy state is maximum to be reduced to therefore the variation range that in a scheduling slot, battery model lotus energy state allows is in the scope of these two values.

In order to increase the lotus energy state of battery model, needing to charge to battery model, now show as the floating charging power increasing battery model, therefore the maximum charge power of battery model being:

$C_{\max }=P_d\left(\min \right(1,SOE\left(i\right)+\frac{T_{max}-\mathrm{\Δ}SOE\·(T_{max}-T_{\min })}{(T_{max}-T_{\min })e^{-\mathrm{\α}\mathrm{\Δ}t}}\left)\right)-P_d\left(SOE\right(i\left)\right)---(1-12)$

In formula: △ SOE represents scheduling accuracy, △ t represents the duration of a period, P _dfloating charging power when (SOE (i)) expression lotus energy state is SOE (i).

(1.5) maximum discharge power D _max

In like manner in maximum charge power C _maxreckoning process, maximum discharge power D _max:

$D_{\max }=P_d\left(SOE\right(i)-P_d(\max \left(0,SOE\left(i\right)-\frac{T_{max}-\mathrm{\Δ}SOE\·(T_{max}-T_{\min })}{(T_{max}-T_{\min })e^{-\mathrm{\α}\mathrm{\Δ}t}}\right))---(1-13)$

Formula (1-3) ~ (1-7), (1-12) and (1-13) form the battery model of buildings belonging to central air conditioner jointly.

When charging to battery model, the SOE (i+1) after charging and current SOE (i) meet following relation:

P _d(SOE(i+1))＝P _d(SOE(i))+C(i)(1-14)

When discharging to battery model, the SOE (i+1) after electric discharge and current SOE (i) meet following relation:

P _d(SOE(i+1))＝P _d(SOE(i))-D(i)(1-15)

After charge-discharge electric power control is carried out to accumulator, can solve equation (1-14) and (1-15), upgrade the parameter sets of battery model.

One is contained to the group of M battery model, its maximum charge power is the maximum charge power C of all M battery model _maxsum, its maximum discharge power is the maximum discharge power D of all M battery model _maxsum.

In described step (43), the optimum results of Load aggregation business is C (i) or D (i), is described the strategy that battery model carries out charge-discharge electric power control for charge power control:

(43.1) the lotus energy state size according to value of M battery model is sorted, SOE ¹<SOE ²< ... <SOE ^z< ... <SOE ^m, the maximum charge power of all M battery model current scheduling period is respectively in order

(43.2) following program is performed:

$C=C\left(i\right)-(W-C_{max}^z)$

${\mathrm{SOE}}_{new}^z=solve{(}^{\&prime;}{P}_d\left({\mathrm{SOE}}_{new}^z\right)=P_d\left({\mathrm{SOE}}^z\right)+C^{\&prime;},{\mathrm{SOE}}_{new}^z)$

(43.3) when requiring that charge power is C, concrete charging schedules is: lotus energy state is SOE ¹, SOE ²sOE ^z-1battery model be adjusted to its maximum lotus energy state SOE ^zbe adjusted to the lotus energy state of other accumulators remains unchanged; Upgrade the parameter sets of the battery model after charge-discharge electric power control.

The control strategy of discharge power is similar to the control strategy of charge power, repeats no more herein.

Electricity market Optimized Operation strategy based on buildings energy storage characteristic belonging to central air-conditioning load provided by the invention, on the battery model basis of buildings belonging to central air conditioner, in ahead market, Load aggregation business makes power purchase arrangement according to the clear electricity price of secondary sunrise and load prediction situation; In Real-time markets, utilize the electricity price of the electricity price of present period, load, outdoor temperature prediction future time period, load and outdoor temperature, utilize the charge-discharge characteristic of accumulator with power purchase network minimal for objective function sets up the Optimized Operation strategy of Real-time markets, while reducing the energy storage device dependence higher to traditional cost, also improve the economic benefit of Load aggregation business.

The above is only the preferred embodiment of the present invention; be noted that for those skilled in the art; under the premise without departing from the principles of the invention, can also make some improvements and modifications, these improvements and modifications also should be considered as protection scope of the present invention.

Claims (2)

1., based on an electricity market Optimized Operation strategy for central air conditioner energy storage characteristic, it is characterized in that: comprise the steps:

(1) thermodynamical model of buildings belonging to central air conditioner, set up battery model:

B＝{O,SOE,P _d,C _max,D _max}(1)

In formula: B represents the parameter sets of battery model, and O represents stored energy capacitance, SOE represents lotus energy state, P _drepresent floating charging power, C _maxrepresent maximum charge power, D _maxrepresent maximum discharge power;

(2) Load aggregation business predicts the cleaing price of each period of next day and load, and by arranging the purchase of electricity of each period of next day to the charge and discharge control of battery model, make Load aggregation business power purchase next day expense minimum, form operation plan a few days ago:

${\mathrm{minF}}_1=\underset{i=1}{\overset{N}{\&Sigma;}}\&lsqb;E_s\left(i\right)p_{ad}^{\&prime;}\left(i\right)+\mathrm{\&lambda;}|C\left(i\right)-D\left(i\right)|\&rsqb;---\left(2\right)$

E _s(i)＝L _ad'(i)+C(i)-D(i)(3)

In formula: F ₁represent that Load aggregation business to dispatch the total power purchase expense of prediction in market, E a few days ago _si () represents the prediction purchase of electricity of a few days ago dispatching period i in market, p' _adi () represents the prediction cleaing price a few days ago dispatching period i in market, L _ad' (i) represent the prediction load a few days ago dispatching period i in market; λ represents the scheduling expense of battery model, the charge power of battery model when C (i) represents period i, the discharge power of battery model when D (i) represents period i, and N represents period sum;

(3) when Load aggregation business to submit to a few days ago after operation plan to system operator, the actual cleaing price of each period of next day and known, then:

$F_2=\underset{i=1}{\overset{N}{\&Sigma;}}\&lsqb;E_s\left(i\right)p_{ad}\left(i\right)+\mathrm{\&lambda;}|C\left(i\right)-D\left(i\right)|\&rsqb;---\left(4\right)$

In formula: F ₂represent that Load aggregation business to dispatch the total power purchase expense of reality in market, p a few days ago _adi () represents the actual cleaing price a few days ago dispatching period i in market;

(4) because the load prediction of Load aggregation business to next day exists error, need to rely on Real-time markets to balance the deviation between real-time purchase of electricity and prediction purchase of electricity; In Real-time markets, the real-time purchase of electricity of Load aggregation business Real-time Obtaining period i, in real time cleaing price, Real-time Load and real-time outdoor temperature are optimized scheduling, and detailed process is:

(41) when Load aggregation business is optimized scheduling, with the real-time purchase of electricity of period i, in real time cleaing price, Real-time Load and in real time outdoor temperature to the purchase of electricity of period i+1 ~ period i+n in Real-time markets, cleaing price, load and and outdoor temperature predict;

(42) Load aggregation business with the charge-discharge electric power of battery model for decision variable, minimum for target with power purchase expense, set up Optimal Operation Model:

$\begin{array}{c}\min F_3\\ =\&lsqb;\left(E_a\right(i)-E_s\left(i\right))p_{rt}\left(i\right)+\mathrm{\&lambda;}|C\left(i\right)-D\left(i\right)|\&rsqb;\\ =\underset{k=i+1}{\overset{i+n}{\mathrm{\&Sigma;}}}\&lsqb;({E_a}^{\&prime;}\left(k\right)-E_s\left(k\right)){p_{rt}}^{\&prime;}\left(k\right)+\mathrm{\&lambda;}|C\left(k\right)-D\left(k\right)|\&rsqb;\end{array}---\left(5\right)$

E _a(i)＝L _a(i)+C(i)-D(i)(6)

E _a'(k)＝L _a'(k)+C(k)-D(k)(7)

In formula: F ₃represent the total power purchase expense of the prediction of Load aggregation business in Real-time markets; E _ai () represents the real-time purchase of electricity of period i, p _rti () represents the real-time cleaing price of period i, L _ai () represents the Real-time Load of period i; E _a' (k) represent the prediction purchase of electricity of period k in Real-time markets, p _rt' (k) represent the prediction cleaing price of period k in Real-time markets, L _a' (k) represent the prediction load of period i in Real-time markets; N represents that period sum is optimized in retrogressing;

(43) adopt linear programming method to solve the formula (5) as objective function, according to solving result, charge-discharge electric power control is carried out to battery model during period i;

(44) i=i+1, returns step (41).

2. the electricity market Optimized Operation strategy based on central air conditioner energy storage characteristic according to claim 1, it is characterized in that: in described step (1), the thermodynamical model of affiliated buildings is:

$\frac{{\mathrm{dT}}_{in}}{dt}=\mathrm{\&alpha;}(T_o-T_{in})+\mathrm{\&gamma;}-\mathrm{\&mu;}Q---(1-1)$

$T_{in}=\left(T_{in}\right(0)-\frac{{\mathrm{\&alpha;T}}_o+\mathrm{\&gamma;}-\mathrm{\&mu;}Q}{\mathrm{\&alpha;}})e^{-\mathrm{\&alpha;}t}+\frac{{\mathrm{\&alpha;T}}_o+\mathrm{\&gamma;}-\mathrm{\&mu;}Q}{\mathrm{\&alpha;}}---(1-2)$

In formula: T _inrepresent indoor temperature, T _orepresent outdoor temperature, Q represents central air conditioner refrigerating capacity, and α, γ and μ are coefficient, T _in(0) indoor temperature of initial time is represented; T represents the time;

The process of establishing of battery model is as follows:

(1.1) stored energy capacitance O

If the indoor temperature interval meeting human comfort is [T _min, T _max], the cold of buildings belonging to central air conditioner is stored in room air and indoor solid, and indoor temperature is in T _mintime storage cold maximum, indoor temperature is in T _maxtime storage cold minimum; Indoor temperature is in T _maxtime storage cold be designated as 0, then when indoor temperature is in T _mintime storage cold be:

$O_{in}=\frac{T_{max}-T_{in}}{\mathrm{\&mu;}}---(1-3)$

In formula: represent that buildings often raises the energy required for 1 DEG C; It can thus be appreciated that stored energy capacitance O is:

$O=\frac{T_{max}-T_{min}}{\mathrm{\&mu;}}---(1-4)$

(1.2) lotus energy state SOE

The ratio of the current storage cold of lotus energy state representation and stored energy capacitance, reflects the energy storage state of battery model:

$SOE=\frac{O_{in}}{O}=\frac{T_{max}-T_{in}}{T_{\max }-T_{\min }}---(1-5)$

Formula (1-5) is brought into (1-2) and the time dependent rule of lotus energy state can be obtained:

$SOE\left(i\right)=SOE\left(0\right)\&CenterDot;e^{-\mathrm{\&alpha;}t}+\frac{\mathrm{\&alpha;}{(T_{max}-T)}_o-\mathrm{\&gamma;}+\mathrm{\&mu;}Q}{\mathrm{\&alpha;}(T_{\max }-T_{\min })}(1-e^{-\mathrm{\&alpha;}t})---(1-6)$

In formula: reach lotus energy state during steady state (SS) during SOE (i) represents period i, SOE (0) represents the lotus energy state of initial time;

(1.3) floating charging power P _d

The electric power setting up central air conditioner is the floating charging power of battery model:

P _d＝P(1-7)

In formula: P represents the electric power of central air conditioner, P is the function of SOE;

(1.4) maximum charge power C _max

$C_{\max }=P_d\left(\min \right(1,SOE\left(i\right)+\frac{T_{max}-\mathrm{\&Delta;}SOE\&CenterDot;(T_{max}-T_{\min })}{(T_{max}-T_{\min })e^{-\mathrm{\&alpha;}\mathrm{\&Delta;}t}}\left)\right)-P_d\left(SOE\right(i\left)\right)---(1-8)$

In formula: △ SOE represents scheduling accuracy, △ t represents the duration of a period, P _dfloating charging power when (SOE (i)) expression lotus energy state is SOE (i);

When charging to battery model, the SOE (i+1) after charging and current SOE (i) meet following relation:

P _d(SOE(i+1))＝P _d(SOE(i))+C(i)(1-9)

(1.5) maximum discharge power D _max

$D_{\max }=P_d\left(SOE\right(i)-P_d(\max \left(0,SOE\left(i\right)-\frac{T_{max}-\mathrm{\&Delta;}SOE\&CenterDot;(T_{max}-T_{\min })}{(T_{max}-T_{\min })e^{-\mathrm{\&alpha;}\mathrm{\&Delta;}t}}\right))---(1-10)$

When discharging to battery model, the SOE (i+1) after electric discharge and current SOE (i) meet following relation:

P _d(SOE(i+1))＝P _d(SOE(i))-D(i)(1-11)

After charge-discharge electric power control is carried out to accumulator, equation (1-10) and (1-11) are solved, upgrade the parameter sets of battery model.