CN106961124A - A kind of economic load dispatching model responded based on price demand - Google Patents

Abstract

It is an object of the invention to provide a kind of economic load dispatching model responded based on price demand, the present invention is primarily based on the relation between price elastic coefficient analysis price type demand response scheduling cost and interactive response amount.Secondly, the interactive response amount state algorithm of price type demand response is described, and as one of expected distribution principle.On this basis, Network Security Constraints and the constraint of price type demand response interactive response satisfaction are considered simultaneously, target is minimised as with the price type demand response scheduling cost of grid side, Real time optimal dispatch model is set up, it is ensured that the safe operation of power network in Real-Time Scheduling.

Description

A kind of economic load dispatching model responded based on price demand

Technical field

The present invention relates to a kind of economic load dispatching model responded based on price demand.

Background technology

Wind-powered electricity generation is the most generation mode of economic development prospect in regenerative resource, however, due to the fluctuation of wind-powered electricity generation itself The unfriendly feature such as property, intermittence, anti-peak regulation, low schedulability, the rapid growth of its installed capacity fortune traditional to power network Row control model produces significant impact.To tackle this challenge, interaction should be coordinated between power supply, power network and load three, this It is the important development direction of intelligent grid.But in general, the main mould for using generating follow load of operation of power networks control Formula, the workload demand of prediction is met by dispatching the start-stop of generating set with exerting oneself.But when large-scale wind power accesses power train During system, generating becomes uncontrollable in itself.Therefore, many current research begin to focus on Steam Generator in Load Follow power generation mode, pass through load Side resource active response participates in power network interaction, to realize the lifting of wind-powered electricity generation comprehensive utilization ratio.

Demand response (demand response, DR) refers to that power consumer is directed to market price signal or incentive mechanism master The dynamic behavior for changing original consumption power mode.Demand response resource, which participates in power network interaction, mainly following a few classes：1) price type DR, guides user's reasonable adjusting by price signal (such as tou power price, Spot Price) and improves power structure and power mode； 2) contract can be interrupted by signing；3) direct load control mode.When wind-powered electricity generation is actual exerts oneself more than predicted value, demand response can be passed through Resource increase power consumption fills up system power vacancy；Conversely, when wind-powered electricity generation is actual exert oneself less than predicted value when, demand response can be passed through Resource cuts down power consumption and carrys out stabilizing system unbalanced power amount.In recent years, having carried out both at home and abroad is used to demand response balance wind The correlative study of electro-mechanical wave, focuses more on the Potentials of demand response balance wind-powered electricity generation fluctuation, participates in the need of wind electricity digestion Seek response scheduling strategy and to the influence of spinning reserve capacity in terms of.

Wind electricity digestion problem is tackled, price type demand response participates in power network Real-Time Scheduling by price signal, to keep electricity The power equilibrium of supply and demand.However, the change of wind power output and load, makes electric network swim and node voltage face out-of-limit risk, so that net Network security constraint turns into the key factor of restriction wind electricity digestion.

The content of the invention

For above technical problem, it is an object of the invention to provide a kind of economic load dispatching mould responded based on price demand Type, is primarily based on the relation between price elastic coefficient analysis price type demand response scheduling cost and interactive response amount.Secondly, The interactive response amount state algorithm of price type demand response is described, and as one of expected distribution principle.On this basis, Network Security Constraints and the constraint of price type demand response interactive response satisfaction are considered simultaneously, are rung with the price type demand of grid side Cost should be dispatched and be minimised as target, Real time optimal dispatch model is set up, it is ensured that the safe operation of power network in Real-Time Scheduling.

The present invention is achieved by the following technical solutions：

Step 1, introducing price demand factor, build the price demand response model based on demand elasticity, pass through electricity in real time Valency guiding user changes the change of electricity consumption behavior response wind power output；

Step 2, the object function by setting up Robust Interval economic load dispatching model and power-balance, conventional power unit exert oneself, Spinning reserve, conventional power unit climbing rate, network security, wind power output constraints, Shandong is dissolved into by price demand response model Rod interval economic load dispatching model, is built the large-scale wind power responded based on price demand and dissolved Robust Interval in a few days economic load dispatching mould Type.

Preferably, in step 1, the price demand response model based on demand elasticity is as follows：

Response formula of the user under different electricity prices be：

Response is constrained to：

In t period nodes j power selling income f (L_{F, jt}+L_d,jt) be：

In formula：ε_jtFor t period nodes j workload demand self-elasticity coefficient, the electricity consumption demand pair of user is reflected The sensitivity that electricity price changes, generally known parameters；L_f,jtFor t period nodes j predicted load；L_d,jtFor the t periods Response load variations power of the node j in different electricity prices；ρ_jtFor t period nodes j electricity price；ρ_ref,jtTo refer to electricity price；For node j user's peak response capacity.

Preferably, to method approximate processing of the formula (3) using successive linearization, the expression formula that it is linearized is as follows：

In formula：NM is the division number after linearisation；Decision variable d_mjtFor t period nodes j m sections of load work( Rate, and meet 0≤d_mjt≤D_mjt, D_mjtFor m sections of maximum load power；ρ_mjtElectricity price is segmented for corresponding m.

Beneficial effects of the present invention：

1st, robust optimization, can be in given uncertain parameter bounded as a kind of decision-making technique for solving uncertain problem A ROBUST OPTIMAL SOLUTIONS is tried to achieve in set so that constraints is met under all values of uncertain parameter.With base The probability Distribution Model for knowing uncertain parameter is needed to compare in the Uncertainty Analysis Method of probability theory, robust Optimal methods are examined The uncertain information of worry is the excursion of uncertain parameter, and this category information is easier to obtain in practice, and robust optimizes The solution efficiency of method is higher, is more suitable for the occasion of extensive application on site.

2nd, using price demand respond as conventional power unit regulating power effective supplement, flexible modulation node load, quickly Wind power output change is responded, the influence of the not true property of wind-powered electricity generation is reduced, reduction abandons wind power and increases social welfare.

Brief description of the drawings

Fig. 1 is the machine test system line charts of IEEE-RTS 26 of the invention.

Specific embodiment

Below in conjunction with the embodiment of the present invention and accompanying drawing, the technical scheme in the embodiment of the present invention is carried out clear, complete Ground is described, it is clear that described embodiment is only a part of embodiment of the invention, rather than whole embodiments.Based on this Embodiment in invention, the every other reality that those of ordinary skill in the art are obtained under the premise of creative work is not made Example is applied, the scope of protection of the invention is belonged to.

The in a few days economic load dispatching model responded based on price demand, is comprised the following steps：

Step 1, introducing price demand factor, build the price demand response model based on demand elasticity, pass through electricity in real time Valency guiding user changes the change of electricity consumption behavior response wind power output, improves the wind electricity digestion capability of power network；

Step 2, the object function by setting up Robust Interval economic load dispatching model and power-balance, conventional power unit exert oneself, Spinning reserve, conventional power unit climbing rate, network security, wind power output constraints, Shandong is dissolved into by price demand response model Rod interval economic load dispatching model, is built the large-scale wind power responded based on price demand and dissolved Robust Interval in a few days economic load dispatching mould Type.

Traditional Robust Interval economic load dispatching model can refer to《Automation of Electric Systems》Volume 38 in October 25 day in 2014 20th phase, the object function and power-balance, conventional power unit of the invention by changing traditional Robust Interval economic load dispatching model Exert oneself, (other parts of model, which are not done, to be changed for spinning reserve, conventional power unit climbing rate, network security, wind power output constraints Become), price demand response model is dissolved into traditional Robust Interval economic load dispatching model, builds what is responded based on price demand Large-scale wind power is dissolved Robust Interval in a few days economic load dispatching model.

It is specifically introduced below：It is preferred that, in step 1, price demand response be by price signal (such as tou power price, Spot Price) guide user's reasonable adjusting and improve power structure and power mode.The Spot Price update cycle is compared with tou power price It is shorter, it can be 1h or shorter, can effectively pass on electricity price signal, guiding user changes electricity consumption behavior, response system operation shape State changes.The respondent behavior of user is described to be the basis for formulating the scheduling strategy for considering demand response, and the present invention, which is used, to be based on needing The customer response model of elasticity is sought, response formula and response constraint of the user under different electricity prices are as follows：

Response formula of the user under different electricity prices be：

Response is constrained to：

In：ε_jtFor t period nodes j workload demand self-elasticity coefficient, the electricity consumption demand of user is reflected to electricity The sensitivity that valency changes, generally known parameters；L_f,jtFor t period nodes j predicted load；L_d,jtFor the t periods Response load variations power of the point j in different electricity prices；ρ_jtFor t period nodes j electricity price；ρ_ref,jtTo refer to electricity price； For node j user's peak response capacity.

Power selling income f (L of the grid company in t period nodes j_{F, jt}+L_d,jt) be：

F (Lf, jt+Ld, jt) is the secondary convex letter of demand response load (Lf, jt+Ld, jt) it can be seen from formula (3) Number, can typically use the method approximate processing of successive linearization, and the expression formula that it is linearized is as follows：

In formula：NM is the division number after linearisation；Decision variable d_mjtFor t period nodes j m sections of load work( Rate, and meet 0≤d_mjt≤D_mjt, D_mjtFor m sections of maximum load power；ρ_mjtElectricity price is segmented for corresponding m.

Wherein,, will by changing the object function and constraints of traditional Robust Interval economic load dispatching model in step 2 The price demand response model that step 1 is set up incorporates traditional Robust Interval economic load dispatching model, builds and considers that price demand is rung The large-scale wind power answered is dissolved Robust Interval economic load dispatching model, and proposes the method for solving based on linear programming.Modeling process In only consider the uncertainty of wind-powered electricity generation, have ignored load and Demand Side Response resource uncertainty and conventional power unit it is random Stop transport, it is specific as follows：

1) object function：

Traditional meter and abandon wind factor Robust Interval dynamic economic dispatch model it is general to minimize going out for conventional power unit Power cost is object function with wind punishment cost is abandoned, and is introduced after price demand response, from social welfare angle modification object function.

The cost of exerting oneself that social welfare can subtract conventional power unit by power selling income is obtained with wind punishment cost is abandoned, specific table It is shown as：Object function is set up with social welfare angle：

In formula：NT is total fixed number of economic load dispatching；NJ is system loading interstitial content；NI and NK are respectively conventional machine Group number and conventional power unit secondary convex function exert oneself cost it is linearized after division number；NW wind power plant numbers；c_kiWith c_min, i be respectively conventional power unit i cost of exerting oneself it is linearized after kth section cost slope and minimum load cost, and meet c1i≤c_2i≤…≤cNKi；p_kitFor t period conventional power unit i kth section active power output；WithPoint Not Wei wind power plant w predictions exert oneself and bound and allow bound of exerting oneself；Bound deviation punishment cost systems of the Vw to wind power plant w Number；

2) constraints：

Introduce after price demand response constraint, add the decision variable dmjt of demand response load, therefore also need to pair Power-balance constraint, spinning reserve constraint and the Network Security Constraints of traditional Robust Interval economic load dispatching model are adjusted, and are had Body is analyzed as follows：

2-a) power-balance constraint：

In formula：Due to using DC flow model herein, therefore ignore the influence of via net loss；p_itAnd p_min, i is respectively The active power output and minimum load of conventional power unit i t periods；For the t period wind power plants w economic optimum plan of exerting oneself；

2-b) the bound constraint of exerting oneself of conventional power unit

In formula：p_min,i、p_g,1i、p_g,2i、…、p_g,(NK-1)iAnd p_max,iTo be [p by i-th conventional power unit scope of exerting oneself_min,i, p_max,i] linearize the power branch after NK sections；

2-c) spinning reserve is constrained：

In formula：WithThe upper and lower spinning reserve capacity that respectively t period units i is provided；r_i ^uAnd r_i ^dRespectively machine Upper and lower creep speeds of the group i in unit time period；Δ T is the duration of each scheduling slot；

Wherein, the spinning reserve constraint violation of conventional power unit can cause to abandon wind, from the angle of system dynamic response capability, wind Electricity, which exerts oneself to be mutated, can cause the variable capacity of conventional power unit to reduce, and when system adjustable capacity is smaller, its level of security is lower, this Scape is also more severe, is consequently formed criterion formula (11)-(14) of most severe scene：

In formula：W is wind power plant set；WithThe upper and lower rotation that system is provided respectively under t periods most severe scene Turn spare capacity；WithWind-powered electricity generation generating optimization variable in the severe scene of respectively upper and lower spinning reserve constraint； WithThe upper and lower spinning reserve capacity requirement of system of respectively t periods；

2-d) the climbing rate constraint of conventional power unit：

From the angle of system dynamic response capability, the climbing rate of conventional power unit is constrained under most severe scene：

In formula：WithRespectively conventional power unit i going out in the t periods under formula (11), (13) two kinds of most severe scenes The upper and lower adjustment amount of power；WithConventional power unit i is in t-1 respectively under formula (11), (13) two kinds of most severe scenes The upper and lower adjustment amount of exerting oneself of section；p_i,t-1For conventional power unit i the t-1 periods active power output；

2-e) the Network Security Constraints based on DC power flow algorithm：

Network Security Constraints, which are violated to also result in, abandons wind, from the angle of network security, wind power output meeting in border value Line load rate is caused to reach maximum, line load rate is higher, and system security level is then lower, and the scene is also more severe, by This forms most severe scene criterion formula (18) and formula (19)：

In formula：NB is system node number；NL is number of branches；SF_lbRepresent transfer distribution factors of the node b to branch road l； I ∞ b, w ∞ b and j ∞ b represent the unit i, wind power plant w and load j being connected with node b respectively；WithDuring respectively t Duan Zhilu l maximum forward and reverse effective power flow power；WithThe respectively forward and reverse effective power flow power constraint of branch road Severe scene in wind-powered electricity generation generating optimization variable；U_lFor branch road l effective power flow limit value；

2-f) wind-powered electricity generation allows interval and wind power output constraint of exerting oneself：

By in step 1 and the 2 double-deck Robust Interval Optimal Operation Models built, the coupled relation between layer model causes up and down The model can not direct solution.But the requirement due to upper layer model to underlying model is only the pact of minimax generating capacity Beam, thus according to the principle of duality of linear programming, the dual problem that the constraint can be equivalent to lower floor's Optimized model meets corresponding It is required that.The bi-level optimal model of former problem can be converted into individual layer linear programming model to solve accordingly.I.e. according to linear programming The principle of duality will be dissolved the double-deck excellent of Robust Interval in a few days economic load dispatching model based on the large-scale wind power that price demand is responded Change model conversation and optimize solution for individual layer linear programming model.

In former Optimized model, the formula (11), (13), (18) and (19) of lower floor's Optimized model is used into corresponding antithesis respectively The constraint of problem is replaced, to ensure that the dual objective function of lower floor's Optimized model is the upper bound or the lower bound of former Optimized model.Wherein, Dual variable after the variable of lower floor Optimized model is and changed before antithesis conversion is respectively α wt, β wt, δ wt and after conversion Optimized model form it is as follows：

It is right in formula (15) and (16) that left end item in formula (22) and (23) is substituted into respectivelyWithIt is replaced, obtains With the individual layer linear programming for Robust Interval in a few days economic load dispatching model equivalency of being dissolved based on the large-scale wind power that price demand is responded Model：

In formula, s.t. is subject to abbreviation, is affined implication, and the first row formula is meant that：Work as target When function (social welfare=power selling income-cost of exerting oneself-abandon wind punishment cost) obtains maximum, correspondence bound variable p_kit, d_mjt,α_wt, β_wtδ_wt Acquired value.

The robust optimization of the present invention, can be in given uncertain ginseng as a kind of decision-making technique for solving uncertain problem A ROBUST OPTIMAL SOLUTIONS is tried to achieve in number bounded set so that constraints is expired under all values of uncertain parameter Foot.Compared with the Uncertainty Analysis Method based on probability theory needs to know the probability Distribution Model of uncertain parameter, robust is excellent The uncertain information that change method considers is the excursion of uncertain parameter, and this category information is easier to obtain in practice, and The solution efficiency of robust Optimal methods is higher, is more suitable for the occasion of extensive application on site.The present invention responds price demand As effective supplement of conventional power unit regulating power, flexible modulation node load, the change of quick response wind power output reduces wind-powered electricity generation The influence of not true property, reduction abandons wind power and increases social welfare.

Finally, above example and accompanying drawing are merely illustrative of the technical solution of the present invention and unrestricted, although by above-mentioned The present invention is described in detail for embodiment, it is to be understood by those skilled in the art that can in form and carefully Various changes are made to it on section, without departing from claims of the present invention limited range.

Claims (3)

1. a kind of economic load dispatching model responded based on price demand, it is characterised in that：

Step 1, introducing price demand factor, build the price demand response model based on demand elasticity, are drawn by Spot Price Lead user and change the change of electricity consumption behavior response wind power output；

Step 2, the object function by setting up Robust Interval economic load dispatching model and power-balance, conventional power unit are exerted oneself, rotated Standby, conventional power unit climbing rate, network security, wind power output constraints, robust area is dissolved into by price demand response model Between economic load dispatching model, build the large-scale wind power responded based on price demand and dissolve Robust Interval in a few days economic load dispatching model.

2. a kind of economic load dispatching model responded based on price demand according to claim 1, it is characterised in that：In step In 1, the price demand response model based on demand elasticity is as follows：

Response formula of the user under different electricity prices be：

${\mathrm{\ϵ}}_{jt}=\frac{\left(\right(L_{f,jt}+L_{d,jt})-L_{f,jt})/L_{f,jt}}{({\mathrm{\ρ}}_{jt}-{\mathrm{\ρ}}_{ref,jt})/{\mathrm{\ρ}}_{ref,jt}}---\left(1\right)$

Response is constrained to：

$-L_{d,jt}^{\max }\≤L_{d,jt}\≤L_{d,jt}^{\max }---\left(2\right)$

In t period nodes j power selling income f (L_{F, jt}+L_d,jt) be：

$f(L_{f,jt}+L_{d,jt})=\frac{{\mathrm{\ρ}}_{ref,jt}}{2{\mathrm{\ϵ}}_{jt}{L}_{f,jt}}{(L_{f,jt}+L_{d,jt})}^2+(1-\frac{1}{{\mathrm{\ϵ}}_{jt}}){\mathrm{\ρ}}_{ref,jt}(L_{f,jt}+L_{d,jt})---\left(3\right)$

In formula：ε_jtFor t period nodes j workload demand self-elasticity coefficient, the electricity consumption demand of user is reflected to electricity price The sensitivity of variation, generally known parameters；L_f,jtFor t period nodes j predicted load；L_d,jtFor t period nodes Response load variations power of the j in different electricity prices；ρ_jtFor t period nodes j electricity price；ρ_ref,jtTo refer to electricity price；For Node j user's peak response capacity.

3. a kind of economic load dispatching model responded based on price demand according to claim 1, it is characterised in that：To formula (3) using the method approximate processing of successive linearization, the expression formula that it is linearized is as follows：

$f(L_{f,jt}+L_{d,jt})=\underset{m=1}{\overset{NM}{\Σ}}{\mathrm{\ρ}}_{mjt}{d}_{mjt}---\left(4\right)$

$L_{f,jt}+L_{d,jt}=\underset{m=1}{\overset{NM}{\Σ}}{d}_{mjt}---\left(5\right)$

In formula：NM is the division number after linearisation；Decision variable d_mjtFor t period nodes j m sections of load power, and Meet 0≤d_mjt≤D_mjt, D_mjtFor m sections of maximum load power；ρ_mjtElectricity price is segmented for corresponding m.