CN106127337A - Unit Combination method based on the modeling of convertible frequency air-conditioner virtual robot arm - Google Patents

Abstract

The invention discloses a kind of Unit Combination method based on the modeling of convertible frequency air-conditioner virtual robot arm, comprise the steps: that (1) sets up thermodynamical model and the electrical model of single air conditioner；(2) by Load aggregation business, air-conditioning group is carried out centralized dispatching and control, set up the polymerization model of air-conditioning group；(3) carry out virtual robot arm modeling according to conventional rack characteristic, and set up user and respond the cost function of virtual robot arm under uncertain condition；(4) in ahead market, according to secondary daily load and temperature prediction, set up the Unit erriger built-up pattern of conventional rack and virtual robot arm, arrange time daily dispatch scheduling；(5) during actual schedule, consider the economic interests of Load aggregation business and participate in user fairness and the comfort level of virtual robot arm modeling, it is achieved load adjustment amount optimized distribution between users.The centralized dispatching that present invention achieves air-conditioning group controls, and decreases the amount of calculation of relevant traffic department and controls difficulty.

Description

Unit Combination method based on the modeling of convertible frequency air-conditioner virtual robot arm

Technical field

The present invention relates to a kind of Unit Combination method based on the modeling of convertible frequency air-conditioner virtual robot arm, particularly relate to convertible frequency air-conditioner Polymerization modeling and consideration user's probabilistic virtual robot arm modeling technique, belong to demand response and participate in conventional rack combination Application.

Background technology

Demand response technology is one of core technology of intelligent grid, alleviates supply and demand tension, increasing by demand response Strong system reply current rip kinetic force, raising running efficiency of system, minimizing relevant enterprise have lost and have maximized economic interests Become industry common cognition.In all flexible loads, thermal control load has hot storage capacity can be within a certain period of time because of it Transfer load can be that system provides multiple assistant service to receive extensive concern.Air conditioner load is that a kind of typical thermal control is born Lotus, its compressor experienced by from determining the frequency development to frequency conversion, and current convertible frequency air-conditioner is because of its higher efficiency share in the market It is gradually increased.Modeling and the control technique of research convertible frequency air-conditioner have preferable application prospect.

Air conditioner load has a very wide distribution, the scale of construction big, therefore needs certain technological means that it is carried out polymerization modeling, facilitates phase The United Dispatching of pass department and control, Load aggregation business is as a kind of business model being specifically designed to integration load side resource, no It is only capable of and represents middle-size and small-size burdened resource and provide various services for system, and can be by means of the senior measurement body of intelligent grid Load is measured and controls by system in real time, it is achieved efficiently utilizing and the maximization of economic benefit of resource.Meanwhile, air-conditioning Load is affected relatively big by the factor such as user behavior, weather condition, has bigger uncertainty, it is therefore necessary to entering unit Take into full account during row operation plan that user responds uncertainty, reduce the economic loss of relevant departments.

Summary of the invention

Goal of the invention: in order to overcome the deficiencies in the prior art, the present invention provides a kind of virtual based on convertible frequency air-conditioner The Unit Combination method of unit modeling, by fully excavating Demand-side resource, it is achieved the centralized and unified scheduling control of extensive load System, to reduce the frequent start-stop control of conventional rack.

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

A kind of Unit Combination method based on the modeling of convertible frequency air-conditioner virtual robot arm, comprises the steps:

(1) thermodynamical model and the electrical model of single air conditioner is set up according to conservation of energy principle and operation of air conditioner characteristic, I.e. set up the relation between the power P of air-conditioning and refrigerating capacity Q of air-conditioning；

(2) by Load aggregation business, air-conditioning group is carried out centralized dispatching and control, set up the polymerization model of air-conditioning group；

(3) load adjustment amount is carried out Potential Evaluation on the basis of polymerization model, enter according to conventional rack characteristic simultaneously Row virtual robot arm models, and sets up user and respond the cost function of virtual robot arm under uncertain condition；

(4) in ahead market, according to secondary daily load and temperature prediction, the combination machine of conventional rack and virtual robot arm is set up Group built-up pattern, arranges time daily dispatch scheduling；

(5) during actual schedule, economic interests and the participation virtual robot arm modeling of Load aggregation business is considered User fairness and comfort level, with the economic interests maximizing Load aggregation business and minimize user's non-comfort for target letter Number, it is achieved load adjustment amount optimized distribution between users.

Concrete, described step (1) comprises the steps:

(11) thermodynamical model of single air conditioner is set up:

$C_a\frac{{\mathrm{dT}}_{in}}{dt}=\frac{1}{R_1}(T_{out}-T_{in})+Q^{\′}-Q---\left(1\right)$

Wherein: T_outFor outdoor temperature, T_inFor indoor temperature, C_aFor the equivalent thermal capacitance of air-conditioning, R₁Equivalence resistance for air-conditioning Anti-, Q is the refrigerating capacity of air-conditioning, and Q' is the heat dissipation capacity of indoor object, and t is the time；

(12) electrical model of single air conditioner is set up:

Frequency f of air-conditioning subsequent time_t+1By frequency f of current time_tWith design temperature T_sWith current indoor temperature T_in,t Difference Δ T_tDetermining, the recurrence relation of frequency f of air-conditioning is as follows:

ΔT_t=T_s-T_in,t (2)

By the relational representation of the power P of air-conditioning and frequency f of air-conditioning it is:

P=k₁f+l₁ (4)

By the relational representation of refrigerating capacity Q of air-conditioning and frequency f of air-conditioning it is:

Q=k₂f+l₂ (5)

The relation between the power P of air-conditioning and refrigerating capacity Q of air-conditioning of setting up is:

$Q=\frac{k_2}{k_1}P+\frac{k_1{l}_2-l_1{k}_2}{k_1}---\left(6\right)$

Wherein: Δ T_minFor the minimum detection temperature difference set, Δ T_maxFor the maximum detection temperature difference set, k, k₁、l₁、 k₂And l₂It is constant coefficient.

Concrete, described step (2) comprises the steps:

Air-conditioning is to be stored in affiliated building with the form of heat energy by electric energy, and the highest energy storage capacity of indoor temperature is the least, room The lowest energy storage capacity of interior temperature is the biggest, and the comfort level scope of note user is [T_min,T_max]；If indoor temperature is T_maxTime energy storage capacity be 0, then indoor temperature is T_inTime energy storage capacity O_inFor:

O_in=C_a(T_max-T_in) (7)

The stored energy capacitance O of building is:

O=C_a(T_max-T_min) (8)

The state-of-charge SOC of definition air-conditioning is energy storage capacity O_inRatio with stored energy capacitance O:

$SOC=\frac{O_{in}}{O}=\frac{T_{max}-T_{in}}{T_{\max }-T_{\min }}---\left(9\right)$

When indoor temperature is maintained at T_inTime, refrigerating capacity Q that can obtain air-conditioning according to formula (1) is:

$Q=\frac{1}{R_1}(T_{out}-T_{in})+Q^{\′}---\left(10\right)$

Bringing formula (10) into formula (6), the power P obtaining air-conditioning is:

$P=\frac{k_1}{k_2{R}_1}(T_{out}-T_{in})+\frac{k_1{Q}^{\′}+k_2{l}_1-k_1{l}_2}{k_2}---\left(11\right)$

Bringing formula (9) into formula (11), the relation obtaining the power P of air-conditioning and the state-of-charge SOC of air-conditioning is:

P=α SOC+ β T_o+γ (12)

$\mathrm{\α}=\frac{k_1(T_{max}-T_{\min })}{k_2{R}_1}---\left(13\right)$

$\mathrm{\β}=\frac{k_1}{k_2{R}_1}---\left(14\right)$

$\mathrm{\γ}=\frac{-k_1{T}_{\max }+k_1{R}_1{Q}^{\′}+k_2{R}_1{l}_1-k_1{R}_1{l}_2}{k_2{R}_1}---\left(15\right)$

The excursion [0,1] of the state-of-charge SOC of air-conditioning is divided into N number of minizone, charged according to every air-conditioning All air-conditionings are divided in each minizone by state SOC, and the air-conditioning quantity added up in each minizone is respectively m₁,m₂,…, m_i,…,m_N, by unified for the state-of-charge of the air-conditioning in i-th minizone for SOC_i:

${\mathrm{SOC}}_i=\frac{1}{N}i-\frac{1}{2N}---\left(16\right)$

Set up air-conditioned polymerization model in i-th minizone:

P_i=α_iSOC_i+β_iT_o+γ_i (17)

${\mathrm{\α}}_i={\mathrm{\α}}_{i\_1}+{\mathrm{\α}}_{i\_2}+\mathrm{...}+{\mathrm{\α}}_{i\_k}+\mathrm{...}+{\mathrm{\α}}_{i\_m_i}---\left(18\right)$

${\mathrm{\β}}_i={\mathrm{\β}}_{i\_1}+{\mathrm{\β}}_{i\_2}+...+{\mathrm{\β}}_{i\_k}+...+{\mathrm{\β}}_{i\_m_i}---\left(19\right)$

${\mathrm{\γ}}_i={\mathrm{\γ}}_{i\_1}+{\mathrm{\γ}}_{i\_2}+...+{\mathrm{\γ}}_{i\_k}+...+{\mathrm{\γ}}_{i\_m_i}---\left(20\right)$

The total polymerization power P of whole air-conditioning group_totalFor:

$P_{otal}=\underset{i=1}{\overset{N}{\Σ}}{P}_i---\left(21\right)$

When outdoor temperature keeps constant, i-th the minizone in air-conditioned state integrated regulation in jth community Aggregate power changes delta P_i-jFor:

ΔP_i-j=α_i(SOC_j-SOC_i) (22)

Wherein: P_iFor aggregate power air-conditioned in i-th minizone, α_{i_k}、β_{i_k}And γ_{i_k}It is respectively i-th community α, β and γ of interior kth platform air-conditioning.

Concrete, described step (3) comprises the steps:

The schedulable power determining virtual robot arm is:

$P^{up}=\underset{i=1}{\overset{N}{\Σ}}{\mathrm{\α}}_i({\mathrm{SOC}}_N-{\mathrm{SOC}}_i)---\left(23\right)$

$P^{down}=\underset{i=1}{\overset{N}{\Σ}}{\mathrm{\α}}_i({\mathrm{SOC}}_i-{\mathrm{SOC}}_1)---\left(24\right)$

Wherein: P^upFor adjusting power on virtual robot arm, P^downFor adjusting power under virtual robot arm；

The cost setting up virtual robot arm is:

C=λ | Δ P | (25)

Wherein: C is the cost of schedule virtual unit, λ is that virtual robot arm unit dispatches cost, and Δ P is the meter of virtual robot arm Draw schedule power；

Consider that user responds uncertainty, if the actual schedule power Δ P' of virtual robot arm be planned dispatching power Δ P with Stochastic variable ω sum:

| Δ P'|=| Δ P |+ω (26)

When the actual schedule power of virtual robot arm is less than planned dispatching power, except traffic department need to be according to virtual robot arm Actual schedule power is paid outside Load aggregation business's expense, and Load aggregation business also needs to carry out certain economic benefit to traffic department Repay；When the actual schedule power of virtual robot arm is more than planned dispatching power, then traffic department adjusts according to the plan of virtual robot arm Degree power pays Load aggregation business's expense；Therefore consider that user responds the cost function C' of virtual robot arm under uncertain condition and is:

$C^{\&prime;}=\left\{\begin{array}{cc}\mathrm{\&lambda;}\left|\mathrm{\&Delta;}P\right|+\mathrm{\&tau;}\mathrm{\&omega;}& \mathrm{\&omega;}<0\\ \mathrm{\&lambda;}\left|\mathrm{\&Delta;}P\right|& \mathrm{\&omega;}\&GreaterEqual;0\end{array}\right.---\left(27\right)$

Wherein: τ is the compensation unit price that Load aggregation business pays traffic department；

Set up the expected cost C of virtual robot arm " (Δ P) be:

$C^{\&prime;\&prime;}\left(\mathrm{\&Delta;}P\right)=\mathrm{\&lambda;}\left|\mathrm{\&Delta;}P\right|+\underset{-\mathrm{\&infin;}}{\overset{0}{\&Integral;}}\mathrm{\&tau;}\mathrm{\&omega;}df\left(\mathrm{\&omega;}\right)---\left(28\right)$

Wherein: f (ω) is the probability density function of ω, obtain according to historical data statistics.

Concrete, described step (4) comprises the steps:

In ahead market, according to secondary daily load and temperature prediction, set up the Unit erriger of conventional rack and virtual robot arm Built-up pattern, arranges time daily dispatch scheduling, and object function is:

$\min F=\underset{t=1}{\overset{T}{\&Sigma;}}\underset{i=1}{\overset{N_G}{\&Sigma;}}\&lsqb;C_{Gi}\left(P_{Gi}^t\right)U_{Gi}^t+U_{Gi}^t(1-U_{Gi}^{t-1})S_{Gi}\&rsqb;+C^{\&prime;\&prime;}\left({\mathrm{\&Delta;P}}^t\right)---\left(29\right)$

Wherein: T is the total activation period, Δ P^tIt is the planned dispatching power of the t scheduling slot virtual robot arm, C " (Δ P^t) It is the expected cost of the t scheduling slot virtual robot arm, N_GFor conventional rack number of units,It is the t scheduling slot conventional rack The start and stop state of i,Represent start,Represent and shut down, S_GiFor the payment for initiation use of conventional rack i,For tradition machine Group i exerts oneself,Meritorious producing cost function for conventional rack i:

$C_{Gi}\left(P_{Gi}^t\right)=a_{Gi}{\left(P_{Gi}^t\right)}^2+b_{Gi}{P}_{Gi}^t+c_{Gi}---\left(30\right)$

Wherein: a_Gi、b_GiAnd c_GiCost coefficient for conventional rack i；

The system restriction that object function is corresponding is as follows:

A () system loading balances

${\mathrm{\&Delta;P}}^t+\underset{i=1}{\overset{N_G}{\&Sigma;}}{P}_{Gi}^t{U}_{Gi}^t=P_L^t---\left(31\right)$

Wherein:It it is the workload demand of the t scheduling slot prediction；

B () system reserve retrains

${\mathrm{\&Delta;P}}_{max}^t+\underset{i=1}{\overset{N_G}{\&Sigma;}}{P}_{Gi\_max}^t{U}_{Gi}^t\&GreaterEqual;P_L^t+P_R^t---\left(32\right)$

Wherein:It is the EIAJ of the t scheduling slot virtual robot arm,It is the t scheduling slot tradition The EIAJ of unit i,It it is the stand-by requirement of t scheduling slot；

(c) conventional rack exert oneself bound constraint

$P_{Gi\_\min }^t\&le;P_{Gi}^t\&le;P_{Gi\_\max }^t---\left(33\right)$

Wherein:It it is the minimum load of the t scheduling slot conventional rack i；

(d) conventional rack Climing constant

$P_{Gi}^t{U}_{Gi}^t-P_{Gi}^{t-1}{U}_{Gi}^{t-1}\&le;R_{up,i}---\left(34\right)$

$P_{Gi}^{t-1}{U}_{Gi}^{t-1}-P_{Gi}^t{U}_{Gi}^t\&le;R_{down,i}---\left(35\right)$

Wherein: R_up,iFor the maximum upward slope speed of conventional rack i, R_down,iMaximum downslope speed for conventional rack i；

(e) conventional rack startup-shutdown time-constrain

$\left(X_{Gi}^{on}\right(t-1)-T_{Gi}^{on})(U_{Gi}^{t-1}-U_{Gi}^t)\&GreaterEqual;0---\left(36\right)$

$\left(X_{Gi}^{off}\right(t-1)-T_{Gi}^{off})(U_{Gi}^t-U_{Gi}^{t-1})\&GreaterEqual;0---\left(37\right)$

Wherein:For conventional rack i in the accumulative available machine time of t-1 scheduling slot,For passing System unit i in the accumulative downtime of t-1 scheduling slot,For the minimum start duration of conventional rack i,For The minimum of conventional rack i shuts down duration；

E () virtual robot arm retrains

$-P^{down}\left(T_{out}^t\right)\&le;{\mathrm{\&Delta;P}}^t\&le;P^{up}\left(T_{out}^t\right)---\left(38\right)$

Wherein:It is that the t scheduling slot virtual robot arm is in outdoor temperatureTime lower adjusting power,It is that the t scheduling slot virtual robot arm is in outdoor temperatureTime upper adjusting power.

Concrete, described step (5) comprises the steps:

During actual schedule, the object function of the economic interests maximizing Load aggregation business is:

Wherein: σ is the compensation unit price that Load aggregation business pays user；

Consider the economic interests of Load aggregation business and participate in user fairness and the comfort level of virtual robot arm modeling, making total work The object function of rate adjustment amount sum minimum is:

$\min \left|\underset{i=1}{\overset{N}{\&Sigma;}}{\mathrm{\&nu;}}_i{\mathrm{\&Delta;SOC}}_i\right|---\left(40\right)$

${\mathrm{\&nu;}}_i=\frac{x_i}{x}---\left(41\right)$

Wherein: x_iFor the controlled number of times of air-conditioning of i-th minizone, x is the controlled number of times of all air-conditionings.

Beneficial effect: the present invention is directed to the feature that the convertible frequency air-conditioner scale of construction has a very wide distribution greatly, it is provided that a kind of empty based on frequency conversion Adjust the Unit Combination method of virtual robot arm modeling, propose the polymerization modeling method of a kind of air-conditioning group, it is achieved that the concentration of air-conditioning group Scheduling controlling, decreases the amount of calculation of relevant traffic department and controls difficulty；On the basis of the polymerization of air-conditioning group models, according to biography The operation characteristic of system unit, establishes virtual robot arm model, and considers that user responds uncertainty, set up virtual robot arm uncertain Scheduling cost under implementations, decreases the risk of traffic department；When air-conditioning group is controlled, maximizing Load aggregation Users'comfort and fairness has been taken into account, it is achieved that the optimization between users of load adjustment amount divides on the basis of business's economic interests Join.

Accompanying drawing explanation

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

Fig. 2 is air-conditioning polymerization model；

Fig. 3 Unit Combination Scheduling Framework.

Detailed description of the invention

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

It is illustrated in figure 1 a kind of Unit Combination method based on the modeling of convertible frequency air-conditioner virtual robot arm, below with regard to each step Illustrated.

Step one: set up the thermodynamical model of single air conditioner and electric mould according to conservation of energy principle and operation of air conditioner characteristic Type, i.e. sets up the relation between the power P of air-conditioning and refrigerating capacity Q of air-conditioning.

(11) thermodynamical model of single air conditioner is set up:

$C_a\frac{{\mathrm{dT}}_{in}}{dt}=\frac{1}{R_1}(T_{out}-T_{in})+Q^{\&prime;}-Q---\left(1\right)$

Wherein: T_outFor outdoor temperature, T_inFor indoor temperature, C_aFor the equivalent thermal capacitance of air-conditioning, R₁Equivalence resistance for air-conditioning Anti-, Q is the refrigerating capacity of air-conditioning, and Q' is the heat dissipation capacity of indoor object, and t is the time；

(12) electrical model of single air conditioner is set up:

Frequency f of air-conditioning subsequent time_t+1By frequency f of current time_tWith design temperature T_sWith current indoor temperature T_in,t Difference Δ T_tDetermining, the recurrence relation of frequency f of air-conditioning is as follows:

ΔT_t=T_s-T_in,t (2)

By the relational representation of the power P of air-conditioning and frequency f of air-conditioning it is:

P=k₁f+l₁ (4)

By the relational representation of refrigerating capacity Q of air-conditioning and frequency f of air-conditioning it is:

Q=k₂f+l₂ (5)

The relation between the power P of air-conditioning and refrigerating capacity Q of air-conditioning of setting up is:

$Q=\frac{k_2}{k_1}P+\frac{k_1{l}_2-l_1{k}_2}{k_1}---\left(6\right)$

Wherein: Δ T_minFor the minimum detection temperature difference set, Δ T_maxFor the maximum detection temperature difference set, k, k₁、l₁、 k₂And l₂It is constant coefficient.

Step 2: air-conditioning group is carried out centralized dispatching and control by Load aggregation business, sets up the polymerization model of air-conditioning group.

Air-conditioning is to be stored in affiliated building with the form of heat energy by electric energy, and the highest energy storage capacity of indoor temperature is the least, room The lowest energy storage capacity of interior temperature is the biggest, and the comfort level scope of note user is [T_min,T_max]；If indoor temperature is T_maxTime energy storage capacity be 0, then indoor temperature is T_inTime energy storage capacity O_inFor:

O_in=C_a(T_max-T_in) (7)

The stored energy capacitance O of building is:

O=C_a(T_max-T_min) (8)

The state-of-charge SOC of definition air-conditioning is energy storage capacity O_inRatio with stored energy capacitance O:

$SOC=\frac{O_{in}}{O}=\frac{T_{max}-T_{in}}{T_{\max }-T_{\min }}---\left(9\right)$

When indoor temperature is maintained at T_inTime, refrigerating capacity Q that can obtain air-conditioning according to formula (1) is:

$Q=\frac{1}{R_1}(T_{out}-T_{in})+Q^{\&prime;}---\left(10\right)$

Bringing formula (10) into formula (6), the power P obtaining air-conditioning is:

$P=\frac{k_1}{k_2{R}_1}(T_{out}-T_{in})+\frac{k_1{Q}^{\&prime;}+k_2{l}_1-k_1{l}_2}{k_2}---\left(11\right)$

Bringing formula (9) into formula (11), the relation obtaining the power P of air-conditioning and the state-of-charge SOC of air-conditioning is:

P=α SOC+ β T_o+γ (12)

$\mathrm{\&alpha;}=\frac{k_1(T_{max}-T_{\min })}{k_2{R}_1}---\left(13\right)$

$\mathrm{\&beta;}=\frac{k_1}{k_2{R}_1}---\left(14\right)$

$\mathrm{\&gamma;}=\frac{-k_1{T}_{\max }+k_1{R}_1{Q}^{\&prime;}+k_2{R}_1{l}_1-k_1{R}_1{l}_2}{k_2{R}_1}---\left(15\right)$

The excursion [0,1] of the state-of-charge SOC of air-conditioning is divided into N number of minizone, charged according to every air-conditioning All air-conditionings are divided in each minizone by state SOC, and the air-conditioning quantity added up in each minizone is respectively m₁,m₂,…, m_i,…,m_N, by unified for the state-of-charge of the air-conditioning in i-th minizone for SOC_i:

${\mathrm{SOC}}_i=\frac{1}{N}i-\frac{1}{2N}---\left(16\right)$

Set up air-conditioned polymerization model in i-th minizone:

P_i=α_iSOC_i+β_iT_o+γ_i (17)

${\mathrm{\&alpha;}}_i={\mathrm{\&alpha;}}_{i\_1}+{\mathrm{\&alpha;}}_{i\_2}+\mathrm{...}+{\mathrm{\&alpha;}}_{i\_k}+\mathrm{...}+{\mathrm{\&alpha;}}_{i\_m_i}---\left(18\right)$

${\mathrm{\&beta;}}_i={\mathrm{\&beta;}}_{i\_1}+{\mathrm{\&beta;}}_{i\_2}+...+{\mathrm{\&beta;}}_{i\_k}+...+{\mathrm{\&beta;}}_{i\_m_i}---\left(19\right)$

${\mathrm{\&gamma;}}_i={\mathrm{\&gamma;}}_{i\_1}+{\mathrm{\&gamma;}}_{i\_2}+...+{\mathrm{\&gamma;}}_{i\_k}+...+{\mathrm{\&gamma;}}_{i\_m_i}---\left(20\right)$

The total polymerization power P of whole air-conditioning group_totalFor:

$P_{otal}=\underset{i=1}{\overset{N}{\&Sigma;}}{P}_i---\left(21\right)$

When outdoor temperature keeps constant, i-th the minizone in air-conditioned state integrated regulation in jth community Aggregate power changes delta P_i-jFor:

ΔP_i-j=α_i(SOC_j-SOC_i) (22)

Wherein: P_iFor aggregate power air-conditioned in i-th minizone, α_{i_k}、β_{i_k}And γ_{i_k}It is respectively i-th community α, β and γ of interior kth platform air-conditioning.

Step 3: on the basis of polymerization model, load adjustment amount is carried out Potential Evaluation, simultaneously special according to conventional rack Property carries out virtual robot arm modeling, and sets up user and respond the cost function of virtual robot arm under uncertain condition.

The schedulable power determining virtual robot arm is:

$P^{up}=\underset{i=1}{\overset{N}{\&Sigma;}}{\mathrm{\&alpha;}}_i({\mathrm{SOC}}_N-{\mathrm{SOC}}_i)---\left(23\right)$

$P^{down}=\underset{i=1}{\overset{N}{\&Sigma;}}{\mathrm{\&alpha;}}_i({\mathrm{SOC}}_i-{\mathrm{SOC}}_1)---\left(24\right)$

Wherein: P^upFor adjusting power on virtual robot arm, P^downFor adjusting power under virtual robot arm；

The cost setting up virtual robot arm is:

C=λ | Δ P | (25)

Wherein: C is the cost of schedule virtual unit, λ is that virtual robot arm unit dispatches cost, and Δ P is the meter of virtual robot arm Draw schedule power；

Consider that user responds uncertainty, if the actual schedule power Δ P' of virtual robot arm be planned dispatching power Δ P with Stochastic variable ω sum:

| Δ P'|=| Δ P |+ω (26)

When the actual schedule power of virtual robot arm is less than planned dispatching power, except traffic department need to be according to virtual robot arm Actual schedule power is paid outside Load aggregation business's expense, and Load aggregation business also needs to carry out certain economic benefit to traffic department Repay；When the actual schedule power of virtual robot arm is more than planned dispatching power, then traffic department adjusts according to the plan of virtual robot arm Degree power pays Load aggregation business's expense；Therefore consider that user responds the cost function C' of virtual robot arm under uncertain condition and is:

$C^{\&prime;}=\left\{\begin{array}{cc}\mathrm{\&lambda;}\left|\mathrm{\&Delta;}P\right|+\mathrm{\&tau;}\mathrm{\&omega;}& \mathrm{\&omega;}<0\\ \mathrm{\&lambda;}\left|\mathrm{\&Delta;}P\right|& \mathrm{\&omega;}\&GreaterEqual;0\end{array}\right.---\left(27\right)$

Wherein: τ is the compensation unit price that Load aggregation business pays traffic department；

Set up the expected cost C of virtual robot arm " (Δ P) be:

$C^{\&prime;\&prime;}\left(\mathrm{\&Delta;}P\right)=\mathrm{\&lambda;}\left|\mathrm{\&Delta;}P\right|+\underset{-\mathrm{\&infin;}}{\overset{0}{\&Integral;}}\mathrm{\&tau;}\mathrm{\&omega;}df\left(\mathrm{\&omega;}\right)---\left(28\right)$

Wherein: f (ω) is the probability density function of ω, obtain according to historical data statistics.

Step 4: in ahead market, according to secondary daily load and temperature prediction, sets up the connection of conventional rack and virtual robot arm Close Unit Combination model, arrange time daily dispatch scheduling.

In ahead market, according to secondary daily load and temperature prediction, set up the Unit erriger of conventional rack and virtual robot arm Built-up pattern, arranges time daily dispatch scheduling, and object function is:

$\min F=\underset{t=1}{\overset{T}{\&Sigma;}}\underset{i=1}{\overset{N_G}{\&Sigma;}}\&lsqb;C_{Gi}\left(P_{Gi}^t\right)U_{Gi}^t+U_{Gi}^t(1-U_{Gi}^{t-1})S_{Gi}\&rsqb;+C^{\&prime;\&prime;}\left({\mathrm{\&Delta;P}}^t\right)---\left(29\right)$

Wherein: T is the total activation period, Δ P^tIt is the planned dispatching power of the t scheduling slot virtual robot arm, C " (Δ P^t) It is the expected cost of the t scheduling slot virtual robot arm, N_GFor conventional rack number of units,It is the t scheduling slot conventional rack The start and stop state of i,Represent start,Represent and shut down, S_GiFor the payment for initiation use of conventional rack i,For tradition machine Group i exerts oneself,Meritorious producing cost function for conventional rack i:

$C_{Gi}\left(P_{Gi}^t\right)=a_{Gi}{\left(P_{Gi}^t\right)}^2+b_{Gi}{P}_{Gi}^t+c_{Gi}---\left(30\right)$

Wherein: a_Gi、b_GiAnd c_GiCost coefficient for conventional rack i；

The system restriction that object function is corresponding is as follows:

A () system loading balances

${\mathrm{\&Delta;P}}^t+\underset{i=1}{\overset{N_G}{\&Sigma;}}{P}_{Gi}^t{U}_{Gi}^t=P_L^t---\left(31\right)$

Wherein:It it is the workload demand of the t scheduling slot prediction；

B () system reserve retrains

${\mathrm{\&Delta;P}}_{max}^t+\underset{i=1}{\overset{N_G}{\&Sigma;}}{P}_{Gi\_max}^t{U}_{Gi}^t\&GreaterEqual;P_L^t+P_R^t---\left(32\right)$

Wherein:It is the EIAJ of the t scheduling slot virtual robot arm,It is the t scheduling slot tradition The EIAJ of unit i,It it is the stand-by requirement of t scheduling slot；

(c) conventional rack exert oneself bound constraint

$P_{Gi\_\min }^t\&le;P_{Gi}^t\&le;P_{Gi\_\max }^t---\left(33\right)$

Wherein:It it is the minimum load of the t scheduling slot conventional rack i；

(d) conventional rack Climing constant

$P_{Gi}^t{U}_{Gi}^t-P_{Gi}^{t-1}{U}_{Gi}^{t-1}\&le;R_{up,i}---\left(34\right)$

$P_{Gi}^{t-1}{U}_{Gi}^{t-1}-P_{Gi}^t{U}_{Gi}^t\&le;R_{down,i}---\left(35\right)$

Wherein: R_up,iFor the maximum upward slope speed of conventional rack i, R_down,iMaximum downslope speed for conventional rack i；

(e) conventional rack startup-shutdown time-constrain

$\left(X_{Gi}^{on}\right(t-1)-T_{Gi}^{on})(U_{Gi}^{t-1}-U_{Gi}^t)\&GreaterEqual;0---\left(36\right)$

$\left(X_{Gi}^{off}\right(t-1)-T_{Gi}^{off})(U_{Gi}^t-U_{Gi}^{t-1})\&GreaterEqual;0---\left(37\right)$

Wherein:For conventional rack i in the accumulative available machine time of t-1 scheduling slot,For passing System unit i in the accumulative downtime of t-1 scheduling slot,For the minimum start duration of conventional rack i,For The minimum of conventional rack i shuts down duration；

E () virtual robot arm retrains

$-P^{down}\left(T_{out}^t\right)\&le;{\mathrm{\&Delta;P}}^t\&le;P^{up}\left(T_{out}^t\right)---\left(38\right)$

Wherein:It is that the t scheduling slot virtual robot arm is in outdoor temperatureTime lower adjusting power,It is that the t scheduling slot virtual robot arm is in outdoor temperatureTime upper adjusting power.

Step 5: during actual schedule, the economic interests considering Load aggregation business are set up with participating in virtual machine The user fairness of mould and comfort level, with the economic interests maximizing Load aggregation business and minimize user's non-comfort as mesh Scalar functions, it is achieved load adjustment amount optimized distribution between users.

During actual schedule, the object function of the economic interests maximizing Load aggregation business is:

Wherein: σ is the compensation unit price that Load aggregation business pays user；

Consider the economic interests of Load aggregation business and participate in user fairness and the comfort level of virtual robot arm modeling, making total work The object function of rate adjustment amount sum minimum is:

$min\left|\underset{i=1}{\overset{N}{\&Sigma;}}{\mathrm{\&nu;}}_i{\mathrm{\&Delta;SOC}}_i\right|---\left(40\right)$

${\mathrm{\&nu;}}_i=\frac{x_i}{x}---\left(41\right)$

Wherein: x_iFor the controlled number of times of air-conditioning of i-th minizone, x is the controlled number of times of all air-conditionings.

The above is only the preferred embodiment of the present invention, it should be pointed out that: for the ordinary skill people of the art For Yuan, under the premise without departing from the principles of the invention, it is also possible to make some improvements and modifications, these improvements and modifications also should It is considered as protection scope of the present invention.

Claims (6)

1. a Unit Combination method based on the modeling of convertible frequency air-conditioner virtual robot arm, it is characterised in that: comprise the steps:

(1) set up thermodynamical model and the electrical model of single air conditioner according to conservation of energy principle and operation of air conditioner characteristic, i.e. build Relation between power P and refrigerating capacity Q of air-conditioning of vertical air-conditioning；

(2) by Load aggregation business, air-conditioning group is carried out centralized dispatching and control, set up the polymerization model of air-conditioning group；

(3) load adjustment amount is carried out Potential Evaluation on the basis of polymerization model, carry out void according to conventional rack characteristic simultaneously Intend unit modeling, and set up user and respond the cost function of virtual robot arm under uncertain condition；

(4) in ahead market, according to secondary daily load and temperature prediction, the Unit erriger group of conventional rack and virtual robot arm is set up Matched moulds type, arranges time daily dispatch scheduling；

(5) during actual schedule, the economic interests of Load aggregation business and the user participating in virtual robot arm modeling are considered Fairness and comfort level, with the economic interests maximizing Load aggregation business and minimize user's non-comfort as object function, Realize load adjustment amount optimized distribution between users.

Unit Combination method based on the modeling of convertible frequency air-conditioner virtual robot arm the most according to claim 1, it is characterised in that: institute State step (1) to comprise the steps:

(11) thermodynamical model of single air conditioner is set up:

$C_a\frac{{\mathrm{dT}}_{in}}{dt}=\frac{1}{R_1}(T_{out}-T_{in})+Q^{\&prime;}-Q---\left(1\right)$

Wherein: T_outFor outdoor temperature, T_inFor indoor temperature, C_aFor the equivalent thermal capacitance of air-conditioning, R₁For the equiva lent impedance of air-conditioning, Q is The refrigerating capacity of air-conditioning, Q' is the heat dissipation capacity of indoor object, and t is the time；

(12) electrical model of single air conditioner is set up:

Frequency f of air-conditioning subsequent time_t+1By frequency f of current time_tWith design temperature T_sWith current indoor temperature T_in,tDifference ΔT_tDetermining, the recurrence relation of frequency f of air-conditioning is as follows:

ΔT_t=T_s-T_in,t (2)

By the relational representation of the power P of air-conditioning and frequency f of air-conditioning it is:

P=k₁f+l₁ (4)

By the relational representation of refrigerating capacity Q of air-conditioning and frequency f of air-conditioning it is:

Q=k₂f+l₂ (5)

The relation between the power P of air-conditioning and refrigerating capacity Q of air-conditioning of setting up is:

$Q=\frac{k_2}{k_1}P+\frac{k_1{l}_2-l_1{k}_2}{k_1}---\left(6\right)$

Wherein: Δ T_minFor the minimum detection temperature difference set, Δ T_maxFor the maximum detection temperature difference set, k, k₁、l₁、k₂And l₂ It is constant coefficient.

Unit Combination method based on the modeling of convertible frequency air-conditioner virtual robot arm the most according to claim 1, it is characterised in that: institute State step (2) to comprise the steps:

Air-conditioning is to be stored in affiliated building with the form of heat energy by electric energy, and the highest energy storage capacity of indoor temperature is the least, Indoor Temperature Spending the lowest energy storage capacity the biggest, the comfort level scope of note user is [T_min,T_max]；If indoor temperature is T_maxTime energy storage capacity be 0, then Indoor temperature is T_inTime energy storage capacity O_inFor:

O_in=C_a(T_max-T_in) (7)

The stored energy capacitance O of building is:

O=C_a(T_max-T_min) (8)

The state-of-charge SOC of definition air-conditioning is energy storage capacity O_inRatio with stored energy capacitance O:

$SOC=\frac{O_{in}}{O}=\frac{T_{max}-T_{in}}{T_{\max }-T_{\min }}---\left(9\right)$

When indoor temperature is maintained at T_inTime, refrigerating capacity Q that can obtain air-conditioning according to formula (1) is:

$Q=\frac{1}{R_1}(T_{out}-T_{in})+Q^{\&prime;}---\left(10\right)$

Bringing formula (10) into formula (6), the power P obtaining air-conditioning is:

$P=\frac{k_1}{k_2{R}_1}(T_{out}-T_{in})+\frac{k_1{Q}^{\&prime;}+k_2{l}_1-k_1{l}_2}{k_2}---\left(11\right)$

Bringing formula (9) into formula (11), the relation obtaining the power P of air-conditioning and the state-of-charge SOC of air-conditioning is:

P=α SOC+ β T_o+γ (12)

$\mathrm{\&alpha;}=\frac{k_1(T_{max}-T_{\min })}{k_2{R}_1}---\left(13\right)$

$\mathrm{\&beta;}=\frac{k_1}{k_2{R}_1}---\left(14\right)$

$\mathrm{\&gamma;}=\frac{-k_1{T}_{max}+k_1{R}_1{Q}^{\&prime;}+k_2{R}_1{l}_1-k_1{R}_1{l}_2}{k_2{R}_1}---\left(15\right)$

The excursion [0,1] of the state-of-charge SOC of air-conditioning is divided into N number of minizone, according to the state-of-charge of every air-conditioning All air-conditionings are divided in each minizone by SOC, and the air-conditioning quantity added up in each minizone is respectively m₁,m₂,…, m_i,…,m_N, by unified for the state-of-charge of the air-conditioning in i-th minizone for SOC_i:

${\mathrm{SOC}}_i=\frac{1}{N}i-\frac{1}{2N}---\left(16\right)$

Set up air-conditioned polymerization model in i-th minizone:

P_i=α_iSOC_i+β_iT_o+γ_i (17)

${\mathrm{\&alpha;}}_i={\mathrm{\&alpha;}}_{i\_1}+{\mathrm{\&alpha;}}_{i\_2}+...+{\mathrm{\&alpha;}}_{i\_k}+...+{\mathrm{\&alpha;}}_{i\_m_i}---\left(18\right)$

${\mathrm{\&beta;}}_i={\mathrm{\&beta;}}_{i\_1}+{\mathrm{\&beta;}}_{i\_2}+...+{\mathrm{\&beta;}}_{i\_k}+...+{\mathrm{\&beta;}}_{i\_m_i}---\left(19\right)$

${\mathrm{\&gamma;}}_i={\mathrm{\&gamma;}}_{i\_1}+{\mathrm{\&gamma;}}_{i\_2}+...+{\mathrm{\&gamma;}}_{i\_k}+...+{\mathrm{\&gamma;}}_{i\_m_i}---\left(20\right)$

The total polymerization power P of whole air-conditioning group_totalFor:

$P_{total}=\underset{i=1}{\overset{N}{\&Sigma;}}{P}_i---\left(21\right)$

When outdoor temperature keeps constant, i-th the minizone in air-conditioned state integrated regulation to gathering in jth community Close changed power Δ P_i-jFor:

ΔP_i-j=α_i(SOC_j-SOC_i) (22)

Wherein: P_iFor aggregate power air-conditioned in i-th minizone, α_{i_k}、β_{i_k}And γ_{i_k}It is respectively in i-th minizone α, β and γ of kth platform air-conditioning.

Unit Combination method based on the modeling of convertible frequency air-conditioner virtual robot arm the most according to claim 3, it is characterised in that: institute State step (3) to comprise the steps:

The schedulable power determining virtual robot arm is:

$P^{up}=\underset{i=1}{\overset{N}{\&Sigma;}}{\mathrm{\&alpha;}}_i({\mathrm{SOC}}_N-{\mathrm{SOC}}_i)---\left(23\right)$

$P^{down}=\underset{i=1}{\overset{N}{\&Sigma;}}{\mathrm{\&alpha;}}_i({\mathrm{SOC}}_i-{\mathrm{SOC}}_1)---\left(24\right)$

Wherein: P^upFor adjusting power on virtual robot arm, P^downFor adjusting power under virtual robot arm；

The cost setting up virtual robot arm is:

C=λ | Δ P | (25)

Wherein: C is the cost of schedule virtual unit, λ is that virtual robot arm unit dispatches cost, and Δ P is that the plan of virtual robot arm is adjusted Degree power；

Consider that user responds uncertainty, if the actual schedule power Δ P' of virtual robot arm is that planned dispatching power Δ P is with random Variable ω sum:

| Δ P'|=| Δ P |+ω (26)

When the actual schedule power of virtual robot arm is less than planned dispatching power, except traffic department need to be according to the reality of virtual robot arm Schedule power is paid outside Load aggregation business's expense, and Load aggregation business also needs to carry out certain economic compensation to traffic department； When the actual schedule power of virtual robot arm is more than planned dispatching power, then traffic department is according to the planned dispatching merit of virtual robot arm Rate pays Load aggregation business's expense；Therefore consider that user responds the cost function C' of virtual robot arm under uncertain condition and is:

$C^{\&prime;}=\left\{\begin{array}{cc}\mathrm{\&lambda;}\left|\mathrm{\&Delta;}P\right|+\mathrm{\&tau;}\mathrm{\&omega;}& \mathrm{\&omega;}<0\\ \mathrm{\&lambda;}\left|\mathrm{\&Delta;}P\right|& \mathrm{\&omega;}\&GreaterEqual;0\end{array}\right.---\left(27\right)$

Wherein: τ is the compensation unit price that Load aggregation business pays traffic department；

Set up the expected cost C of virtual robot arm " (Δ P) be:

$C^{\&prime;\&prime;}\left(\mathrm{\&Delta;}P\right)=\mathrm{\&lambda;}\left|\mathrm{\&Delta;}P\right|+\underset{-\mathrm{\&infin;}}{\overset{0}{\&Integral;}}\mathrm{\&tau;}\mathrm{\&omega;}df\left(\mathrm{\&omega;}\right)---\left(28\right)$

Wherein: f (ω) is the probability density function of ω, obtain according to historical data statistics.

Unit Combination method based on the modeling of convertible frequency air-conditioner virtual robot arm the most according to claim 4, it is characterised in that: institute State step (4) to comprise the steps:

In ahead market, according to secondary daily load and temperature prediction, set up the Unit erriger combination of conventional rack and virtual robot arm Model, arranges time daily dispatch scheduling, and object function is:

$\min F=\underset{t=1}{\overset{T}{\&Sigma;}}\underset{i=1}{\overset{N_G}{\&Sigma;}}\&lsqb;C_{Gi}\left(P_{Gi}^t\right)U_{Gi}^t+U_{Gi}^t(1-U_{Gi}^{t-1})S_{Gi}\&rsqb;+C^{\&prime;\&prime;}\left({\mathrm{\&Delta;P}}^t\right)---\left(29\right)$

Wherein: T is the total activation period, Δ P^tIt is the planned dispatching power of the t scheduling slot virtual robot arm, C " (Δ P^t) it is t The expected cost of individual scheduling slot virtual robot arm, N_GFor conventional rack number of units,It is opening of the t scheduling slot conventional rack i Stop state,Represent start,Represent and shut down, S_GiFor the payment for initiation use of conventional rack i,For conventional rack i Exert oneself,Meritorious producing cost function for conventional rack i:

$C_{Gi}\left(P_{Gi}^t\right)=a_{Gi}{\left(P_{Gi}^t\right)}^2+b_{Gi}{P}_{Gi}^t+c_{Gi}---\left(30\right)$

Wherein: a_Gi、b_GiAnd c_GiCost coefficient for conventional rack i；

The system restriction that object function is corresponding is as follows:

A () system loading balances

${\mathrm{\&Delta;P}}^t+\underset{i=1}{\overset{N_G}{\&Sigma;}}{P}_{Gi}^t{U}_{Gi}^t=P_L^t---\left(31\right)$

Wherein:It it is the workload demand of the t scheduling slot prediction；

B () system reserve retrains

${\mathrm{\&Delta;P}}_{max}^t+\underset{i=1}{\overset{N_G}{\mathrm{\&Sigma;}}}{P}_{Gi\_\max }^t{U}_{Gi}^t\&GreaterEqual;P_L^t+P_R^t---\left(32\right)$

Wherein:It is the EIAJ of the t scheduling slot virtual robot arm,It is the t scheduling slot conventional rack i EIAJ,It it is the stand-by requirement of t scheduling slot；

(c) conventional rack exert oneself bound constraint

$P_{Gi\_\min }^t\&le;P_{Gi}^t\&le;P_{Gi\_max}^t---\left(33\right)$

Wherein:It it is the minimum load of the t scheduling slot conventional rack i；

(d) conventional rack Climing constant

$P_{Gi}^t{U}_{Gi}^t-P_{Gi}^{t-1}{U}_{Gi}^{t-1}\&le;R_{up,i}---\left(34\right)$

$P_{Gi}^{t-1}{U}_{Gi}^{t-1}-P_{Gi}^t{U}_{Gi}^t\&le;R_{down,i}---\left(35\right)$

Wherein: R_up,iFor the maximum upward slope speed of conventional rack i, R_down,iMaximum downslope speed for conventional rack i；

(e) conventional rack startup-shutdown time-constrain

$\left(X_{Gi}^{on}\right(t-1)-T_{Gi}^{on})(U_{Gi}^{t-1}-U_{Gi}^t)\&GreaterEqual;0---\left(36\right)$

$\left(X_{Gi}^{off}\right(t-1)-T_{Gi}^{off})(U_{Gi}^t-U_{Gi}^{t-1})\&GreaterEqual;0---\left(37\right)$

Wherein:For conventional rack i in the accumulative available machine time of t-1 scheduling slot,For tradition machine I is in the accumulative downtime of t-1 scheduling slot for group,For the minimum start duration of conventional rack i,For tradition The minimum of unit i shuts down duration；

E () virtual robot arm retrains

$-P^{down}\left(T_{out}^t\right)\&le;{\mathrm{\&Delta;P}}^t\&le;P^{up}\left(T_{out}^t\right)---\left(38\right)$

Wherein:It is that the t scheduling slot virtual robot arm is in outdoor temperatureTime lower adjusting power,For The t scheduling slot virtual robot arm in outdoor temperature isTime upper adjusting power.

Unit Combination method based on the modeling of convertible frequency air-conditioner virtual robot arm the most according to claim 5, it is characterised in that: institute State step (5) to comprise the steps:

During actual schedule, the object function of the economic interests maximizing Load aggregation business is:

Wherein: σ is the compensation unit price that Load aggregation business pays user；

Consider the economic interests of Load aggregation business and participate in user fairness and the comfort level of virtual robot arm modeling, making general power adjust The object function of whole amount sum minimum is:

$min\left|\underset{i=1}{\overset{N}{\mathrm{\&Sigma;}}}{\mathrm{\&nu;}}_i{\mathrm{\&Delta;SOC}}_i\right|---\left(40\right)$

${\mathrm{\&nu;}}_i=\frac{x_i}{x}---\left(41\right)$

Wherein: x_iFor the controlled number of times of air-conditioning of i-th minizone, x is the controlled number of times of all air-conditionings.