CN106779487A - A kind of virtual plant dynamic economic dispatch model based on prim al- dual interior point m ethod - Google Patents

Abstract

A kind of virtual plant dynamic economic dispatch model based on prim al- dual interior point m ethod, comprises the following steps：（1）Virtual plant is divided into business type virtual plant and poly-talented virtual plant；And build virtual plant of the invention accordingly and run framework substantially；（2）Model objective function is set up, with virtual plant maximization of economic benefit as target；（3）Set up model constraints；Including：Node power equilibrium equation, the units limits of distributed power source, distributed power source exert oneself Climing constant, energy storage device charge/discharge power and capacity-constrained, can reduction plans constraint and line power constraint；（4）Above-mentioned model is solved using the full vector prim al- dual interior point m ethod based on former problem disturbed conditions.Virtual plant dynamic economic dispatch is non-linear, multiple constraint a higher-dimension mathematical optimization problem, and the present invention carries out Efficient Solution to it using the prim al- dual interior point m ethod of former problem disturbed conditions, obtains better effects.

Description

A kind of virtual plant dynamic economic dispatch model based on primal dual interior point method

Technical field

The present invention relates to a kind of virtual plant dynamic economic dispatch model based on primal dual interior point method, belong to power scheduling Technical field.

Background technology

To promote the utilization of clean energy resource, country has been directed to the renewable energy technologies such as photovoltaic generation, wind-power electricity generation, releases A series of practicable subsidy policys and transformation project is cleaned, connect with the extensive new energy cluster for realizing mains side Enter, the power supply characteristic of distribution side is also highlighted gradually.However, numerous new energy equipment are accessed, many problems are brought, these problems master Have：

(1) distributed new capacity is smaller, uncontrollable, effective coordination and control is lacked between equipment, it is difficult to substitute Traditional thermal power plant.

(2) generation of electricity by new energy flexibility and controllability are relatively low, and Systems Operator cannot carry out efficiently market operation to it And management, rely solely on new energy equipment and passively participate in the market operation in itself, the operation and throwing of new energy equipment will be caused Money cost increases, and relatively relies on the subsidy policy of government, seriously limits the paces of its development.

(3) generation of electricity by new energy (for example, wind-power electricity generation and photovoltaic generating system) is exerted oneself with obvious intermittent and fluctuation Property, larger uncertain factor is brought to electric power system dispatching operation, with the quick increase of its grid connection capacity and swashing for country Policy propulsion is encouraged, increasingly stern challenge will be brought to the stable operation of system.

In generation of electricity by new energy is efficiently participated in the operation of energy market and is controlled, and it is allowed to and traditional thermal power plant Realize preferably complementary and coordinate.Document《Micro-capacitance sensor technology summary》[Yang Xin's method, Su Jian, Lv Zhipeng wait [J] China motor Engineering journal, 2014,34 (1)：57-70] using microgrid flexibly, efficiently management and control new energy equipment, with various distributed power sources Application on the spot be control targe, improve power distribution network safety in operation and reliability, but it considers not enough to market efficiency.Text Offer《Active distribution network technology and its progress》[You Yi, Liu Dong, Yu Wenpeng wait [J] Automation of Electric Systems, 2012,36 (18)：10-16] propose to build active distribution network, trend is effectively managed using flexible network adjustment technology, to realize to new energy Source generating equipment is carried out from primal coordination and management and control, is the higher stage of following intelligent distribution network development, but its construction cost is high, week Phase is more long, virtual plant [Nerea Ruiz,Cobelo, Jos é Oyarzabal.A Direct Load Control Model for Virtual Power Plant Management [J] .IEEE Trans on Power Systems, 2009, 24(2)：959-966] introducing, for this problem provides new solution.

The content of the invention

It is an object of the present invention to provide a kind of virtual plant dynamic economic dispatch model based on primal dual interior point method, To solve during virtual plant economic load dispatching, the influence that various intermittent new energy are accessed, and wavy load is overcome to power train The not influence that system is caused.

To achieve the above object, the technical scheme is that providing a kind of virtual plant based on primal dual interior point method Dynamic economic dispatch model, realizes in accordance with the following steps：

(1) according to the difference of virtual plant function, business type virtual plant (CVPP) and poly-talented virtual electricity are classified as Factory (TVPP)；Business type virtual plant serves as market service and formulation in electricity market, performs generation schedule function；It is poly-talented Virtual plant realizes that grid connection security operation is checked and management function.And build virtual plant of the invention accordingly and run structure substantially Frame；

(2) model objective function is set up；The virtual plant dynamic economic dispatch model is with virtual plant economic benefit Maximum turns to target, in a distributed manner the conduct such as Chong discharge powers, interruptible load of power supply unit output, energy storage regulation and control resource, Cost and distributed power source cost of electricity-generating are cut down comprising purchases strategies, sale of electricity income, load；

(3) model constraints is set up；The virtual plant must is fulfilled for power network security operation constraint, new energy and sets Standby itself constraint and the premise of node power balance, the constraints of virtual plant dynamic economic dispatch model include：Node Power balance equation, the units limits of distributed power source, distributed power source are exerted oneself Climing constant, energy storage device charge/discharge power With capacity-constrained, can reduction plans constraint and line power constrain；

(4) above-mentioned model is solved using the full vector primal dual interior point method based on former problem disturbed conditions, can will be above-mentioned Model simplification is Non-Linear Programming form：

obj.min f(x)

S.t.h (x)=0

In formula, f (x) is object function；H (x) is equality constraints functions；G (x) is inequality constraints function,gWithRespectively It is its upper and lower limit.

Then the nonlinear programming problem is solved using original dual interior point.

The object function of the virtual plant dynamic economic dispatch model is expressed as follows：

In formula, hop count when T is；ρ_E,tIt is wholesale electricity price；ρ_d,tIt is zero potential energy；P_E,tIt is VPP to the purchase/sell of energy market Electricity；S_DIt is load bus set；S_CFor can reduction plans node set；S_esIt is energy storage node set；S_dgIt is distributed power source Node set；P_dg,i,tIt is the active power of node i distributed power source；P_D,i,t：The load power of node i；P_es,i,tIt is node i The charge/discharge active power of energy storage；P_c,i,tFor node i can reduction plans power；C_dg,iFor the distributed power source of node i generates electricity Cost function；C_es,iIt is the charge/discharge cost function of the energy storage device of node i；C_c,iFor the load of node i cuts down cost function.

The node power equilibrium equation is set up as follows：

For arbitrary node i, the active power sum for flowing into, flowing out node is zero；The active power for wherein flowing into includes The power purchase power of energy market, distributed power source are exerted oneself, controllable burden cuts down power and energy storage discharge power, the wattful power of outflow Rate includes load power and branch road corresponding node outflow power；Node power equilibrium equation is expressed as：

In formula, i, j ∈ S_B, S_BIt is system node set；S_EIt is energy market node set；Ω_iCoupled by with node i The set of branch road corresponding node；P_ij,tFor branch road (i, j) i ends are active.

The units limits of the distributed power source are as follows：

Exerting oneself for distributed power source must can meet distributed power source safe operating conditions；

In formula,P _dg,iIt is the lower limit of the distributed power source unit active power output of node i；It is the distributed power source of node i The upper limit of unit output, t=1:T.

Distributed power source Climing constant of exerting oneself is as follows：

Distributed power source is exerted oneself to change and is limited by increasing or reducing amplitude, be can be expressed as：

P_dg,i,t-P_dg,i,t-1≥R _i,

In formula,It is the distributed power source unit active power output maximum climbing speed of node i；R _iIt is the distributed electrical of node i Source unit active power output maximum fall off rate, t=1:T.

The energy storage device charge/discharge power and capacity-constrained are as follows：

In formula,It is the upper limit of node i energy storage device discharge power；It is upper for node i energy storage device charge power Limit；It is the maximum capacity of node i energy storage device；Cap_es,i,tIt is the stored energy capacitance of node i day part.

It is described can reduction plans constraint it is as follows：

In formula,For node i can the reduction plans upper limit of the power.

The line power constraint and loss constraint are as follows：

In formula, (i, j) ∈ S_L：S_LIt is line set.

The step of solution nonlinear programming problem using original dual interior point, is as follows：

(1) first, by introducing slack l, u is relaxed, and inequality constraints is changed into equality constraint.Equation will be contained about The optimization problem of beam is converted into unconstrained optimization problem, and Lagrangian form that will be after model conversion is as follows：

In formula, y, z, w is Lagrange multiplier；μ is obstacle constant (Discontinuous Factors)；L, u are slack variable；R is not for Equation number.

(2) by First Order Optimality Condition, i.e., the necessary condition that problem minimum is present is Lagrangian to all changes The partial derivative of amount and multiplier is 0, derives μ containing Discontinuous Factors>0 KKT equation groups：

L_y=h (x)=0

L_z=g (x)-l-g=0

Wherein, L, U, Z, W are the full diagonal matrix of r × r.

(3) above-mentioned Nonlinear System of Equations equation is processed using Newton method, can be exchanged into following autonomy ordinary differential system：

In formula, It is the Hessian matrix of f (x)； WithThe linear superposition of each function Hessian matrix in respectively h (x) and g (x).

(4) update equation is the matrix of higher-dimension in above formula, but it is seen that, it has the openness of height, when solving The meeting considerable drain time, therefore, turn to brief update equation and solve again：

In formula,

Compared to prior art, the invention has the advantages that：The present invention proposes a kind of based on point in original-antithesis The virtual plant dynamic economic dispatch model of method, according to virtual plant difference functionally, is classified as poly-talented virtual plant With business type virtual plant.Maximization of economic benefit with virtual plant sets up virtual plant dynamic economic dispatch mould as target Type, the in a distributed manner conduct such as Chong discharge powers, interruptible load of power supply unit power output, energy storage device regulates and controls resource, On the premise of meeting power network security operation constraint, new energy equipment itself constraint, realize that the target of virtual plant is optimal.It is empty It is non-linear, multiple constraint a higher-dimension mathematical optimization problem to intend power plant's dynamic economic dispatch, and the present invention is using former problem disturbance The primal dual interior point method of condition carries out Efficient Solution to it, obtains better effects.The present invention also can adapt to market-oriented ring for research The new energy management and control mode in border provides important guidance.

Brief description of the drawings

Fig. 1 is the basic operation frame of virtual plant in the embodiment of the present invention.

Basic inputs and output of the Fig. 2 for needed for the CVPP participations marketing activity in the embodiment of the present invention.

Basic inputs and output of the Fig. 3 for needed for the TVPP participations marketing activity in the embodiment of the present invention.

Fig. 4 is energy market and retail market price situation in the embodiment of the present invention.

Fig. 5 is virtual plant profit situation in the embodiment of the present invention.

Fig. 6 is the cost of electricity-generating situation of day part distributed power source in the embodiment of the present invention.

Specific embodiment

Below in conjunction with the accompanying drawings, technical scheme is specifically described.

A kind of Evaluation Method of Distribution Systems Reliability based on equipment life cycle management of the present embodiment, it is specific as follows：

First, the difference according to virtual plant (VPP) functionally, is participated in by market and system management function is by virtual plant Be subdivided into two classes, be respectively serve as market service and formulation in electricity market, the business type that performs generation schedule function it is virtual Power plant (CVPP) and the poly-talented virtual plant (TVPP) for realizing grid connection security operation check, monitoring and management function.

Secondly, the maximum economic benefit with virtual plant (VPP) is as target, comprising VPP from the purchases strategies of energy market Or sale of electricity income (depending on the energy flow direction of VPP and energy market) from VPP to energy market, to load sale of electricity income, Load cuts down cost, distributed power source cost of electricity-generating and energy storage device charge/discharge cost.Set up virtual plant dynamic economic dispatch The object function of model.

Then, in market operation is participated in, its operator must assure that system meets power network security fortune to virtual plant Row, new energy equipment self-technique is feasible and the premise of node power balance, and these constraintss must meet in real time.Set up The constraints of virtual plant dynamic economic dispatch model.

Finally, using a kind of primal dual interior point method based on former problem disturbed conditions, existing method is overcome to solve big rule The undesirable defect of the speed of modular system, full vectorization solves dynamic economic dispatch problem, and one is proposed in solution procedure Data structure --- the brief update equation of new update equation is planted, the matrix dimension for participating in calculating is greatly reduced, carried significantly Solving speed high, has good result for solving large scale system optimization problem.

In order to allow those skilled in the art to further appreciate that the technical scheme, illustrated with reference to specific embodiment.

1st, the classification of virtual plant

According to the difference of virtual plant function, be classified as business type virtual plant (Commercial VPP, CVPP) and Poly-talented virtual plant (Technical VPP, TVPP).CVPP serves as market service and formulation in electricity market, performs hair Electric planning function；TVPP realizes that grid connection security operation is checked and management function.Virtual plant operation frame such as Fig. 1 of the present embodiment It is shown.

CVPP refers to the virtual plant under business earnings angle, and influence of the virtual plant to power distribution network is not considered, will be distributed Formula generation assets are accessed in the form of investment combination, in it is participated in electricity market as traditional power plants.Fig. 2 show Basic input and output needed for the CVPP participation marketing activities.TVPP then considers all kinds of virtual plant participant polymerizations to local electricity The Real Time Effect of net, while representing the cost and operation characteristic of investment combination.Fig. 3 is shown needed for the TVPP participation marketing activities Basic input and output.

2nd, virtual plant dynamic economic dispatch model

A kind of virtual plant dynamic economic dispatch model based on primal dual interior point method is set up, is realized in accordance with the following steps：

Step S1：Maximum economic benefit with virtual plant as target, comprising VPP from the purchases strategies of energy market or VPP is to the sale of electricity income (depending on the energy flow direction of VPP and energy market) of energy market, to load sale of electricity income, negative Lotus cuts down cost, distributed power source cost of electricity-generating and energy storage device charge/discharge cost.Object function is expressed as follows：

Object function includes five parts, and five parts are respectively：VPP is from the purchases strategies or VPP of energy market to energy The sale of electricity income (depending on the energy flow direction of VPP and energy market) in source marketLoad sale of electricity incomeLoad cuts down costDistributed power source cost of electricity-generatingEnergy storage device charge/discharge cost

In formula, T：When hop count during system call；ρ_E,t：Wholesale electricity price；ρ_d,t：Zero potential energy；P_E,t：VPP is to energy city Purchase/the electricity sales amount of field；S_D：Common load access node set；S_C：Can reduction plans access node set；S_es：Energy storage device connects Ingress set；S_dg：Distributed power source access node set；P_dg,i,t：T periods node i accesses the wattful power of distributed power source Rate；P_D,i,t：The burden with power power of t period node is；P_es,i,t：The charge/discharge active power of the energy storage device of t period node is； P_c,i,t：T period node is can reduction plans power；C_dg,i：The distributed power source cost of electricity-generating function of node i；C_es,i：Node i Energy storage device charge/discharge cost function；C_c,i：Cost function is cut down in the burden with power of node i.

Step S2：In market operation is participated in, its operator must assure that system meets power network security to virtual plant Operation, new energy equipment self-technique is feasible and the premise of node power balance, and these constraintss must meet in real time.It is empty The constraints for intending power plant's centralization Optimal Operation Model is as follows：

1) node power equilibrium equation

For arbitrary node i, the active power sum for flowing into, flowing out node is zero, wherein the active power for flowing into includes VPP is cut by the power purchase power (existing only in both interaction nodes) of energy market, distributed power source generated output, controllable burden Subtract power and energy storage device discharge power, the active power of outflow includes load power and branch road corresponding node outflow power.Section Point power balance equation is expressed as：

In formula, i, j ∈ S_B：S_BIt is system node set；S_E：Energy market node set；Ω_i：Couple branch with node i The set of road corresponding node；P_ij,t：The active power of the i ends t periods of branch road (i, j).

2) the active power constraint of distributed power source

The active power of distributed power source is necessarily limited to can guarantee that institute under distributed power source long-time safe operating conditions Between maximum, the minimum load that can reach.

In formula,P _dg,i：The lower limit of the distributed power source unit active power output of node i；The distributed power source machine of node i The upper limit of group active power output.

3) distributed power source is exerted oneself Climing constant

Distributed power source is exerted oneself to change and is limited by increasing or reducing amplitude, and it is embodied as：

In formula,The distributed power source unit active power output maximum climbing speed of node i；R _i：The distributed electrical of node i Source unit active power output maximum fall off rate.

4) energy storage device charge/discharge power and capacity-constrained

In formula,The upper limit of node i energy storage device discharge power；The upper limit of node i energy storage device charge power；The maximum capacity of node i energy storage device；Cap_es,i,t：The stored energy capacitance of node i day part.

5) can reduction plans constraint

In formula,Node i can the reduction plans upper limit of the power.

6) line power constraint and loss constraint

In formula, (i, j) ∈ S_L：S_LIt is line set.

3rd, full vector primal dual interior point method

Above-mentioned model is solved using a kind of primal dual interior point method based on former problem disturbed conditions, the method is applied to place Reason Large-scale Optimization Problems, full vectorization solves dynamic economic dispatch problem, and a kind of new repairing is proposed in solution procedure Data structure --- the brief update equation of positive equation, greatly reduces the matrix dimension for participating in calculating, and substantially increases solution Speed, has good result for solving large scale system optimization problem.Primal dual interior point method introduced below is only required and sought Slack variable and Lagrange multiplier meet simple condition during excellent, you can instead of what must be solved in feasible zone originally It is required that, greatly simplify calculating process.

Above-mentioned model can be reduced to following Non-Linear Programming form：

obj.min f(x)

S.t.h (x)=0

In formula, f (x) is object function；H (x) is equality constraints functions；G (x) is inequality constraints function,gWithRespectively It is its upper and lower limit.

The step of solving the nonlinear programming problem using original dual interior point is as follows：

1) first, by introducing slack l, u is relaxed, and inequality constraints is changed into equality constraint.Equation will be contained about The optimization problem of beam is converted into unconstrained optimization problem, and Lagrangian form that will be after model conversion is as follows：

In formula, y, z, w is Lagrange multiplier；μ is obstacle constant (Discontinuous Factors)；L, u are slack variable；R is not for Equation number.

2) by First Order Optimality Condition, i.e., the necessary condition that problem minimum is present is Lagrangian to all variables And the partial derivative of multiplier is 0, derives μ containing Discontinuous Factors>0 KKT equation groups：

L_y=h (x)=0

L_z=g (x)-l-g=0

Wherein, L, U, Z, W are the full diagonal matrix of r × r.

3) Newton method processes above-mentioned Nonlinear System of Equations equation, can be exchanged into following autonomy ordinary differential system：

In formula, It is the Hessian matrix of f (x)；WithThe linear superposition of each function Hessian matrix in respectively h (x) and g (x).

4) update equation is the matrix of higher-dimension in above formula, but it is seen that, it has the openness of height, the meeting when solving The considerable drain time, therefore, turn to brief update equation and solve again：

In formula,

The l-G simulation test of the present embodiment is carried out on MATLAB 2013a, and simulation example is IEEE-33 Node power distribution systems, Energy market and retail market price situation are as shown in figure 4,24 period load proportion situations of system are as shown in table 1.This implementation Official holiday is set only when load proportion is higher than 70.0%, and ability schedulable can reduction plans.

The day part load proportion situation of table 1

When	Duty ratio	When	Duty ratio	When	Duty ratio
						1	43.75	9	81.25	17	62.50
2	46.88	10	87.50	18	81.25
						3	53.13	11	93.75	19	93.75
4	59.38	12	87.50	20	100.00
						5	62.50	13	81.25	21	81.25
6	68.75	14	77.50	22	68.75
						7	71.88	15	75.00	23	56.25
8	75.00	16	65.63	24	50.00

Fig. 5 show virtual plant yield curve, and the scheduling scheme of the present embodiment virtual plant does not just think of economic benefit, But consider the interests of its participant and the reliability of system electricity consumption.And, energy storage device is carried out in load peak bottom Put/charge, the waste of wind/light resource is avoided to a certain extent, play an important roll to stabilizing system power swing.

Fig. 6 show the cost of electricity-generating situation of day part distributed power source.7-15 period energy markets electricity price is relatively Height, the 18-21 periods are the load period higher, to meet the demand of virtual plant power-balance, now need to start cost of electricity-generating Distributed power source higher, now by cutting down, a certain amount of load saves system operation cost to virtual plant and maintenance is commonly used The electricity consumption reliability at family.

Above is presently preferred embodiments of the present invention, all changes made according to technical solution of the present invention, produced function work During with scope without departing from technical solution of the present invention, protection scope of the present invention is belonged to.

Claims (9)

1. a kind of virtual plant dynamic economic dispatch model based on primal dual interior point method, it is characterised in that the model according to Following steps are realized：

(1) according to the difference of virtual plant function, the virtual plant is divided into business type virtual plant and poly-talented virtual electricity Factory；Business type virtual plant serves as market service and formulation in electricity market, performs generation schedule function；Poly-talented virtual electricity Factory realizes that grid connection security operation is checked and management function；And build the virtual plant of the economic load dispatching model accordingly and run substantially Framework；

(2) model objective function is set up；The virtual plant dynamic economic dispatch model is maximum with virtual plant economic benefit Turn to target, in a distributed manner power supply unit output, the Chong discharge powers of energy storage and interruptible load as regulation and control resource, comprising Purchases strategies, sale of electricity income, load cut down cost and distributed power source cost of electricity-generating；

(3) model constraints is set up；The virtual plant must is fulfilled for power network security operation constraint, new energy equipment certainly The premise that body is constrained and node power is balanced, the constraints of virtual plant dynamic economic dispatch model includes：Node power Equilibrium equation, the units limits of distributed power source, distributed power source are exerted oneself Climing constant, energy storage device charge/discharge power and appearance Amount constraint, can reduction plans constraint and line power constraint；

(4) above-mentioned model is solved using the full vector primal dual interior point method based on former problem disturbed conditions, can be by above-mentioned model Non-Linear Programming form is reduced to, the nonlinear programming problem is then solved using original dual interior point.

2. a kind of virtual plant dynamic economic dispatch model based on primal dual interior point method according to claim 1, it is special Levy and be, the object function of the virtual plant dynamic economic dispatch model is expressed as follows：

$\underset{P_{dg},P_c,P_D,P_{es}}{\underset{{\mathrm{\ρ}}_E,{\mathrm{\ρ}}_d,P_E,}{max}}\left\{\begin{array}{c}-{\mathrm{\ρ}}_{E,t}\underset{t=1}{\overset{T}{\Σ}}{P}_{E,t}+\\ \underset{t=1}{\overset{T}{\Σ}}{\mathrm{\ρ}}_{d,t}(\underset{i\∈S_D}{\Σ}{P}_{D,i,t}-\underset{i\∈Sc}{\Σ}{P}_{c,i,t})-\\ \underset{t=1}{\overset{T}{\Σ}}\left(\underset{i\∈Sdg}{\Σ}{C}_{dg,i}\right(P_{dg,i,t})+\underset{i\∈Ses}{\Σ}{C}_{es,i}(P_{es,i,t}\left)\right)\\ -\underset{t=1}{\overset{T}{\Σ}}\left(\underset{i\∈Sc}{\Σ}{C}_{c,i}\right(P_{c,i,t}\left)\right)\end{array}\right\}$

In formula, hop count when T is；ρ_E,tIt is wholesale electricity price；ρ_d,tIt is zero potential energy；P_E,tIt is VPP to the purchase/electricity sales amount of energy market； S_DIt is load bus set；S_CFor can reduction plans node set；S_esIt is energy storage node set；S_dgIt is distributed electrical source node set Close；P_dg,i,tIt is the active power of node i distributed power source；P_D,i,t：The load power of node i；P_es,i,tIt is the energy storage of node i Charge/discharge active power；P_c,i,tFor node i can reduction plans power；C_dg,iIt is the distributed power source cost of electricity-generating letter of node i Number；C_es,iIt is the charge/discharge cost function of the energy storage device of node i；C_c,iFor the load of node i cuts down cost function.

3. a kind of virtual plant dynamic economic dispatch model based on primal dual interior point method according to claim 1, it is special Levy and be, the node power equilibrium equation is set up as follows：

For arbitrary node i, the active power sum for flowing into, flowing out node is zero；The active power for wherein flowing into includes the energy The power purchase power in market, distributed power source are exerted oneself, controllable burden cuts down power and energy storage discharge power, the active power bag of outflow Include load power and branch road corresponding node outflow power；Node power equilibrium equation is expressed as：

$0=\underset{i\∈S_E}{\Σ}{P}_{E,t}+\underset{i\∈S_{dg}}{\Σ}{P}_{dg,i,t}+\underset{i\∈S_c}{\Σ}{P}_{c,i,t}+\underset{i\∈S_{es}}{\Σ}{P}_{es,i,t}-\underset{i\∈S_D}{\Σ}{P}_{D,i,t}-\underset{j\∈{\mathrm{\Ω}}_i}{\Σ}{P}_{ij,t}$

In formula, i, j ∈ S_B, S_BIt is system node set；S_EIt is energy market node set；Ω_iCoupled branch road pair by with node i Answer the set of node；P_ij,tFor branch road (i, j) i ends are active.

4. a kind of virtual plant dynamic economic dispatch model based on primal dual interior point method according to claim 1, it is special Levy and be, the units limits of the distributed power source are as follows：

Exerting oneself for distributed power source must can meet distributed power source safe operating conditions；

${\underset{\‾}{P}}_{dg,i}\≤P_{dg,i,t}\≤{\stackrel{\‾}{P}}_{dg,i},\∀i\∈S_{dg}$

In formula,P _dg,iIt is the lower limit of the distributed power source unit active power output of node i；It is the distributed power source unit of node i The upper limit exerted oneself, t=1:T.

5. a kind of virtual plant dynamic economic dispatch model based on primal dual interior point method according to claim 1, it is special Levy and be, distributed power source Climing constant of exerting oneself is as follows：

Distributed power source is exerted oneself to change and is limited by increasing or reducing amplitude, be can be expressed as：

$P_{dg,i,t}-P_{dg,i,t-1}\≤{\stackrel{\‾}{R}}_i$

$P_{dg,i,t}-P_{dg,i,t-1}\≥{\underset{\‾}{R}}_i,\∀i\∈S_{dg}$

In formula,It is the distributed power source unit active power output maximum climbing speed of node i；R _iIt is the distributed power source machine of node i Group active power output maximum fall off rate, t=1:T.

6. a kind of virtual plant dynamic economic dispatch model based on primal dual interior point method according to claim 1, it is special Levy and be, the energy storage device charge/discharge power and capacity-constrained are as follows：

$P_{es,i,t}\≤{\stackrel{\‾}{P}}_{es,i}^{dch},P_{es,i,t}\≥0$

$-P_{es,i,t}\≤{\stackrel{\‾}{P}}_{es,i}^{ch},P_{es,i,t}\≤0$

$0\≤{\mathrm{Cap}}_{es,i,t}\≤{\mathrm{Cap}}_{es,i}^{Max},\∀i\∈S_{es}$

In formula,It is the upper limit of node i energy storage device discharge power；It is the upper limit of node i energy storage device charge power；It is the maximum capacity of node i energy storage device；Cap_es,i,tIt is the stored energy capacitance of node i day part.

7. a kind of virtual plant dynamic economic dispatch model based on primal dual interior point method according to claim 1, it is special Levy and be, it is described can reduction plans constraint it is as follows：

$0\≤P_{c,i,t}\≤{\stackrel{\‾}{P}}_{c,i}$

In formula,For node i can the reduction plans upper limit of the power.

8. a kind of virtual plant dynamic economic dispatch model based on primal dual interior point method according to claim 1, it is special Levy and be, the line power constraint and loss constraint are as follows：

${\underset{\‾}{P}}_{ij}\≤P_{ij,t}\≤{\stackrel{\‾}{P}}_{ij}$

In formula, (i, j) ∈ S_L：S_LIt is line set.

9. a kind of virtual plant dynamic economic dispatch model based on primal dual interior point method according to claim 1, it is special Levy and be, it is described using original dual interior point solve the nonlinear programming problem the step of it is as follows：

(1) by introducing slack l, u is relaxed, and inequality constraints is changed into equality constraint；By the optimization containing equality constraint Problem is converted into unconstrained optimization problem, and Lagrangian form that will be after model conversion is as follows：

$\begin{array}{c}L=f\left(x\right)-y^{T}h\left(x\right)-z^T\[g\left(x\right)-l-\underset{\‾}{g}\]\\ -w^T\[g\left(x\right)+u-\stackrel{\‾}{g}\]-\mathrm{\μ}\underset{j=1}{\overset{r}{\Σ}}In\left(l_r\right)-\mathrm{\μ}\underset{j=1}{\overset{r}{\Σ}}In\left(u_r\right)\end{array},$

In formula, y, z, w is Lagrange multiplier；μ is obstacle constant (Discontinuous Factors)；L, u are slack variable；R is inequality Number；

(2) by First Order Optimality Condition, i.e., the necessary condition that problem minimum is present be Lagrangian to all variables and The partial derivative of multiplier is 0, derives μ containing Discontinuous Factors>0 KKT equation groups：

L_y=h (x)=0

L_z=g (x)-l-g=0

$L_w=g\left(x\right)+u-\stackrel{\‾}{g}=0$

$L_l^{\mathrm{\μ}}=LZe-\mathrm{\μ}e=0$

$L_u^{\mathrm{\μ}}=UWe+\mathrm{\μ}e=0$

$\begin{array}{c}{L}_x={\▿}_{x}f\left(x\right)-{\▿}_x^{T}h\left(x\right)y-{\▿}_x^{T}g\left(x\right)(z+w)\\ -{\▿}_{x}g\left(x\right)\left(L^{-1}\right(L_l^{\mathrm{\μ}}+{\mathrm{ZL}}_z)+U^{-1}(L_u^{\mathrm{\μ}}+{\mathrm{WL}}_w\left)\right)=0\end{array}$

Wherein, L, U, Z, W are the full diagonal matrix of r × r；

(3) above-mentioned Nonlinear System of Equations equation is processed using Newton method, can be exchanged into following autonomy ordinary differential system：

$\left[\begin{array}{c}\stackrel{\·}{x}\\ \stackrel{\·}{y}\\ \stackrel{\·}{l}\\ \stackrel{\·}{u}\\ \stackrel{\·}{z}\\ \stackrel{\·}{w}\end{array}\right]={\left[\begin{array}{cccccc}{H}^{\′}(\·)& \▿h\left(x\right)& 0& 0& \▿g\left(x\right)& \▿g\left(x\right)\\ \▿h^T\left(x\right)& 0& 0& 0& 0& 0\\ \▿g^T\left(x\right)& 0& -I& 0& 0& 0\\ \▿g^T\left(x\right)& 0& 0& I& 0& 0\\ 0& 0& Z& 0& L& 0\\ 0& 0& 0& W& 0& U\end{array}\right]}^{-1}\left[\begin{array}{c}{L}_x\\ -L_y\\ -L_z\\ -L_w\\ -L_l^{\mathrm{\μ}}\\ -L_u^{\mathrm{\μ}}\end{array}\right]$

In formula, It is the Hessian matrix of f (x)；WithThe linear superposition of each function Hessian matrix in respectively h (x) and g (x)；

(4) update equation is the matrix of higher-dimension in above formula, but with the openness of height, the meeting considerable drain time when solving, Therefore, brief update equation is turned to solve again：

In formula,