CN106056264A - Time-of-use electricity price optimization method with load development being considered - Google Patents

Abstract

The invention relates to a time-of-use electricity price optimization method with load development being considered, which comprises the steps of (1) acquiring a typical daily load curve of an electric power system of the current year, and carrying out time segment division; (2) predicting a typical daily load curve of the following two years based on the typical daily load curve of the current year; (3) building a time-of-use electricity price optimization model with the load development being considered, solving under certain constraint conditions and releasing the electricity price to a user side. The time-of-use electricity price optimization method with the load development being considered can formulate more reasonable time-of-use electricity price in allusion to load development conditions in a certain area. The advantages are that the time effectiveness of the time-of-use electricity price is prolonged, and the ability of stimulating users to carry out expected responses is ensured within a longer period of time.

Description

A kind of tou power price optimization method considering that load develops

Technical field

The invention belongs to demand response technical field, be specifically related to a kind of tou power price optimization side considering that load develops Method.

Background technology

Tou power price is that power industry implements dsm, and for encouraging user to change power mode, peak load cutting, to reach The rate of load condensate to peak load shifting, improving power system and economic means that is stable and that take.As the effective demand of one Side management means, tou power price is widely used abroad, at home, also has a lot of place carrying out timesharing electricity Valency.Concretely, tou power price is exactly to be divided into peak, flat, three kinds of paddy one day 24 hours according to the peak and low valley of load curve Period, and correspond to peak, flat, three kinds of electricity prices of paddy therewith.There is higher peak electricity price the peak period, and there is relatively low paddy electricity the paddy period Valency, thus guide user to avoid peak high price electricity consumption as far as possible, multiplex low price paddy electricity, to reach peak load shifting, to improve load curve Purpose.Current Electricity Demand rapid development, frequently, this results in the tou power price made in load composition situation change Over time, become the arousal effect of user is died down.

It would therefore be highly desirable to solution the problems referred to above.

Summary of the invention

Goal of the invention: in order to overcome the deficiencies in the prior art, the present invention provides one to contain typical day load curve Prediction, and considered a kind of tou power price optimization considering that load develops of current loads situation and future load change Method.

Technical scheme: the present invention provides a kind of tou power price optimization method considering that load develops, and comprises the following steps:

(1) obtain the typical day load curve of power system then and carry out Time segments division: choosing this area third season In three months, the daily load curve meansigma methods of temperature second highest day is typical day load curve then, when marking off peak for this curve Section, peace period paddy period；

(2) based on the typical day load curve of 2 years after Typical Day Load Curve Forecasting then: for typical case then Daily load curve utilizes Typical Day Load Curve Forecasting method to carry out Second Year and the Typical Day Load Curve Forecasting of the 3rd year；

(3) set up the tou power price Optimized model of consideration load development and solve under certain constraints: when solving Typical day load curve first against Second Year and the 3rd year calculates tou power price, then that result of calculation is upper and lower as electricity price Limit constraint, again solves for typical day load curve then, i.e. obtains final tou power price；Optimization aim is by three partial weightings And obtain:

Min(F₁+F₂+F₃) (1)

F₁Peak load for typical day load curve:

$F_1=L_{max}=\underset{1\≤i\≤24}{max}(L_i\±{\mathrm{\ΔL}}_{i1}\±{\mathrm{\ΔL}}_{i2})---\left(2\right)$

Wherein Δ L_i1With Δ L_i2Implication be: Δ L_i1=Δ L_pf,ΔL_i2=Δ L_gf, it is the peak period when the i period, L_iRepresent The loading of i period, Δ L after enforcement tou power price_pfRepresent the load being transferred to the peak period after implementing tou power price by section at ordinary times Amount, Δ L_gfRepresent and be transferred to the loading of peak period by the paddy period after implementing tou power price；

F₂For typical day load curve peak-valley difference:

$F_2=L_{\max }-L_{\min }=\underset{1\≤i\≤24}{\max }(L_i\±{\mathrm{\ΔL}}_{i1}\±{\mathrm{\ΔL}}_{i2})-\underset{1\≤i\≤24}{\min }(L_i\±{\mathrm{\ΔL}}_{i1}\±{\mathrm{\ΔL}}_{i2})---\left(3\right)$

Wherein Δ L_i1With Δ L_i2Implication be: Δ L_i1=Δ L_fp,ΔL_i2=Δ L_gp, it is the peak period when the i period, Δ L_fpTable Show and be transferred to the loading of section at ordinary times, Δ L by the peak period after implementing tou power price_gpRepresent and turned by the paddy period after implementing tou power price Move to the loading of section at ordinary times；

F₃For the load variations amount of adjacent moment in daily load curve:

$F_3=\underset{i=1}{\overset{24}{\Σ}}|(L_i\±{\mathrm{\ΔL}}_{i1}\±{\mathrm{\ΔL}}_{i2})-(L_{i-1}\±{\mathrm{\ΔL}}_{(i-1)1}\±{\mathrm{\ΔL}}_{(i-1)2})|---\left(4\right)$

Wherein Δ L_i1With Δ L_i2Implication be: Δ L_i1=Δ L_fg,ΔL_i2=Δ L_pg, it is the peak period when the i period, Δ L_fgTable Show and be transferred to the loading of paddy period, Δ L by the peak period after implementing tou power price_pgRepresent and turned by section at ordinary times after implementing tou power price Move to the loading of paddy period；

(4) electricity price is issued to user side；It is issued to user side, electricity price information bag by communication port after making electricity price Include Time segments division information and electricity price value information.

Wherein, typical day load curve Time segments division described in step (1) method particularly includes:

(1.1) determine minimum load point and the peak load point of typical day load curve, be designated as a point and b point respectively；Wherein It is 100% that a point is in the probability of paddy period, and the probability being in the peak period is 0, and b point is in the probability of peak period and is 100%, the probability being in the paddy period is 0；

(1.2) calculate remaining each point according to type membership function less than normal and be in the probability of paddy period, according to type bigger than normal Membership function calculates remaining each point and is in the probability of peak period；Peak period probability according to every bit may with the paddy period Property, if this peak period probability is much larger than paddy period probability, this point belongs to the peak period, if this paddy period probability Belonging to the paddy period much larger than peak period probability then this point, remaining point belongs to section at ordinary times；

(1.3) combine the duration of every period of independent period more than or equal to two hours, and peak period, at ordinary times section, the paddy period total Duration is all higher than, equal to six hours, to obtain the Time segments division result of typical day load curve.

Preferably, the Typical Day Load Curve Forecasting method described in step (2) is:

(2.1) typical day load curve is decomposed, be decomposed into development component and the sign characterizing its law of development The wave component of its fluctuation pattern, development component takes peak load then, and wave component takes each point load and peak load Ratio；Wherein development component utilizes Lycoperdon polymorphum Vitt (Grey Model-GM) (1,1) model to obtain, and wave component utilizes BP (Back Propagation) neural network model obtains；

(2.2) synthesize with the predictive value of wave component according to the predictive value of development component, obtain typical case's daily load bent Line predictive value.

Further, the loading Δ L of peak period it is transferred to after implementing tou power price described in step (3) by section at ordinary times_pf, real It is transferred to the loading Δ L of peak period by the paddy period after executing tou power price_gf, implement to be transferred at ordinary times by the peak period after tou power price The loading Δ L of section_fp, implement to be transferred to the loading Δ L of section at ordinary times by the paddy period after tou power price_gp, implement after tou power price The loading Δ L of paddy period it is transferred to by the peak period_fg, implement to be transferred to after tou power price the loading Δ of paddy period by section at ordinary times L_pgComputational methods be:

It is transferred to the loading Δ L of peak period by section at ordinary times after implementing tou power price_pfComputational methods be:

ΔL_pf=μ_fp×L_f (5)

${\mathrm{\μ}}_{fp}=\left\{\begin{array}{cc}0& (0\≤{\mathrm{\Δ}}_{fp}\≤a_{fp})\\ K_{fp}*({\mathrm{\Δ}}_{fp}-a_{fp})& (a_{fp}\≤{\mathrm{\Δ}}_{fp}\≤b_{fp})\\ {\mathrm{\μ}}_{fp\max }& ({\mathrm{\Δ}}_{fp}\≥b_{fp})\end{array}\right.---\left(6\right)$

μ_fpFor peak ordinary telegram price differential cause unit time period load decrement that the peak period produces with when carrying out tou power price leading peak The ratio of section average load, L_fFor carrying out tou power price leading peak period average load；Electricity price difference Δ_fp=P_f-P_p, P_f、P_pIt is respectively The electricity price at peak, at ordinary times section, a_fp、b_fpFor electricity price difference segmentation parameter, K_fpIt is in calculating factor during second segment for electricity price difference；

It is transferred to the loading Δ L of peak period by the paddy period after implementing tou power price_gfComputational methods be:

ΔL_gf=μ_fg×L_f (7)

${\mathrm{\μ}}_{fg}=\left\{\begin{array}{cc}0& (0\≤{\mathrm{\Δ}}_{fg}\≤a_{fg})\\ K_{fg}*({\mathrm{\Δ}}_{fg}-a_{fg})& (a_{fg}\≤{\mathrm{\Δ}}_{fg}\≤b_{fg})\\ {\mathrm{\μ}}_{fg\max }& ({\mathrm{\Δ}}_{fg}\≥b_{fg})\end{array}\right.---\left(8\right)$

μ_fgThe specific load decrement that the peak period produces is caused to be put down with carrying out the tou power price leading peak period for electricity price between peak and valley The all ratio of load, electricity price difference Δs_fg=P_f-P_g, P_f、P_gIt is respectively peak, the electricity price of paddy period, a_fg、b_fgJoin for electricity price difference section Number, K_fgIt is in calculating factor during second segment for electricity price difference；

It is transferred to the loading Δ L of section at ordinary times by the peak period after implementing tou power price_fpComputational methods be:

${\mathrm{\ΔL}}_{fp}={\mathrm{\ΔL}}_{pf}\×\frac{N_f}{N_p}---\left(9\right)$

N in formula_f, N_pIt is respectively the time hop count of peak period and section at ordinary times；

It is transferred to the loading Δ L of section at ordinary times by the paddy period after implementing tou power price_gpComputational methods be:

ΔL_gp=μ_pg×L_p (10)

${\mathrm{\μ}}_{pg}=\left\{\begin{array}{cc}0& (0\≤{\mathrm{\Δ}}_{pg}\≤a_{pg})\\ K_{pg}*({\mathrm{\Δ}}_{pg}-a_{pg})& (a_{pg}\≤{\mathrm{\Δ}}_{pg}\≤b_{pg})\\ {\mathrm{\μ}}_{pg\max }& ({\mathrm{\Δ}}_{pg}\≥b_{pg})\end{array}\right.---\left(11\right)$

μ_pgFor Pinggu electricity price difference cause specific load decrement that section at ordinary times produces with carry out tou power price before Duan Ping at ordinary times The all ratio of load, L_pFor carrying out before tou power price section average load, electricity price difference Δ at ordinary times_pg=P_p-P_g, P_p、P_gThe most flat, The electricity price of paddy period, a_pg、b_pgFor electricity price difference segmentation parameter, K_pgIt is in calculating factor during second segment for electricity price difference；

It is transferred to the loading Δ L of paddy period by the peak period after implementing tou power price_fgComputational methods be:

${\mathrm{\ΔL}}_{fg}={\mathrm{\ΔL}}_{gf}\×\frac{N_f}{N_g}---\left(12\right)$

N in formula_f, N_gIt is respectively the time hop count of peak period and paddy period；

It is transferred to the loading Δ L of paddy period by section at ordinary times after implementing tou power price_pgComputational methods be:

${\mathrm{\ΔL}}_{pg}={\mathrm{\ΔL}}_{gp}\×\frac{N_p}{N_g}---\left(13\right)$

N in formula_p, N_gIt is respectively the time hop count of Duan Yugu period at ordinary times.

Wherein, the constraints in described step (3) includes that the income from sales of Utilities Electric Co. is constant, the sale of electricity of Utilities Electric Co. Amount keeps constant, peak period, paddy period tou power price constraint, electrical price pattern constraint and user's power purchase expense numerically about Bundle；

The income from sales of Utilities Electric Co. is constant:

The income from sales carrying out the supplier of electricity before tou power price is:

M_NON=Q × P_NON (14)

M_NONFor the Utilities Electric Co.'s income from sales before unexecuted tou power price, Q and P_NONIt is respectively the sale of electricity before tou power price Amount and electricity price；After carrying out tou power price, the income from sales of supplier of electricity is:

M_TOU=Q_fTOU×P_f+Q_pTOU×P_p+Q_gTOU×P_g (15)

Q_fTOU、Q_pTOU、Q_gTOUFor carrying out tou power price postpeak period, at ordinary times section and the power consumption of paddy period, supplier of electricity is made a profit Constraints be:

M_TOU≥M_NON-M_SAVE (16)

Wherein M_SAVEThe Electricity Investment saved for supplier of electricity after carrying out tou power price and power supply cost；

The electricity sales amount of Utilities Electric Co. keeps constant, i.e. power consumption before and after tou power price keeps constant:

Q_NON=Q_fTOU+Q_pTOU+Q_gTOU (17)

Wherein Q_fTOU、Q_pTOUAnd Q_gTOUFor carrying out tou power price postpeak, flat, the power consumption of paddy period, Q_NONFor tou power price Front power consumption；

Peak period, the constraint numerically of paddy period tou power price: according to the principle that marginal cost is theoretical, i.e. power system exists The electricity price of paddy period is more than or equal to operation marginal cost P in this period_MaCOST, and the electricity price of peak period is not higher than little generating Cost price P of unit generation_GeCOST, i.e.

P_g≥P_MaCOST (18)

P_f≤P_GeCOST (19)

This constraint is updated according to the typical day load curve optimum results of Second Year and the 3rd year；

Electrical price pattern retrains, i.e. electrical price pattern is:

1.2 ＜ P_f/P_g≤4 (20)

User's power purchase expense restriction: due to user paid power purchase expense be the power selling income of Utilities Electric Co., therefore This constraints can be expressed as

M_TOU≤M_NON (21)。

Beneficial effect: compared with the prior art, the present invention has a following remarkable advantage: the first timesharing of this consideration load development Electricity price optimization method based on typical day load curve then utilize Typical Day Load Curve Forecasting method Accurate Prediction Second Year with The typical day load curve of the 3rd year, its method is reasonable and error is little；Furthermore the tou power price optimization side of this consideration load development Method can be formulated more for the concrete development of regional load by setting up the tou power price Optimized model considering load development For rational tou power price, proper extension tou power price timeliness, ensure to stimulate user to make Expected Response within the longer time Ability, for tou power price formulate provide a kind of optimum resolving ideas.

Accompanying drawing explanation

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

Fig. 2 is Typical Day Load Curve Forecasting flow chart；

Fig. 3 is the tou power price Optimizing Flow figure considering load development；

Fig. 4 is Second Year typical day load curve peak period user's response results；

Fig. 5 is Second Year typical day load curve paddy period user's response results；

Fig. 6 is the 3rd year typical day load curve peak period user response results；

Fig. 7 is the 3rd year typical day load curve paddy period user's response results.

Detailed description of the invention

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

As it is shown in figure 1, a kind of tou power price optimization method considering that load develops of the present invention, comprise the following steps:

(1) obtain the typical day load curve of power system then and carry out Time segments division: choose this area big industry, General industry and commerce and resident three class user formulate tou power price, choose temperature second highest day in three months third seasons of this three class user Daily load curve meansigma methods be typical day load curve then.

First determine minimum load point and the peak load point of typical day load curve, be designated as a point and b point respectively；Wherein a It is 100% that point is in the probability of paddy period, and the probability being in the peak period is 0, and b point is in the probability of peak period and is 100%, the probability being in the paddy period is 0；Then calculate remaining each point according to type membership function less than normal and be in the paddy period Probability, calculate remaining each point according to π membership function bigger than normal and be in the probability of peak period；During peak according to every bit Section probability and paddy period probability, if this peak period probability is much larger than paddy period probability, when this point belongs to peak Section, if this paddy period probability is much larger than peak period probability, this point belongs to the paddy period, and remaining point belongs to section at ordinary times； Duration in combination with every period of independent period is more than or equal to two hours, and total duration of peak period, at ordinary times section, paddy period is all higher than Equal to six hours, the Time segments division result of typical day load curve can be obtained:

(2) based on the typical day load curve of 2 years after Typical Day Load Curve Forecasting then: for typical case then Daily load curve utilizes Typical Day Load Curve Forecasting method to carry out Second Year and the Typical Day Load Curve Forecasting of the 3rd year；

First typical day load curve is decomposed, be decomposed into and characterize the development component of its law of development and characterize it The wave component of fluctuation pattern, development component takes peak load then, and wave component takes the ratio of each point load and peak load Value；Wherein development component utilizes Lycoperdon polymorphum Vitt (Grey Model-GM) (1,1) model conventional in Mid-long term load forecasting to obtain, i.e. The time series formed with the development component several years ago observed constructs a kind of first-order difference Differential Equation Model, it was predicted that The development component of the coming years；Wave component uses BP (Back Propagation) nerve net conventional in short-term load forecasting Network model obtains, and i.e. utilizes historical volatility component to make as the input layer in neural network model, the wave component of the coming years For output layer, concrete pre-flow gauge is as shown in Figure 2.Predictive value according to development component closes with the predictive value of wave component Become, obtain Typical Day Load Curve Forecasting value.

(3) set up the tou power price Optimized model of consideration load development and solve under certain constraints, solving total stream Journey figure is as shown in Figure 3: when solving, the typical day load curve first against Second Year and the 3rd year calculates tou power price, then Result of calculation is retrained as electricity price bound, again solves for typical day load curve then, i.e. obtain final timesharing electricity Valency.Optimization aim is to minimize the peak load of typical day load curve, is minimised as typical day load curve peak-valley difference, minimizes Load variations amount three partial weighting of adjacent moment in daily load curve and obtain；

Optimization aim is obtained by three partial weightings:

Min(F₁+F₂+F₃) (1)

F₁Peak load for typical day load curve:

$F_1=L_{max}=\underset{1\≤i\≤24}{max}(L_i\±{\mathrm{\ΔL}}_{i1}\±{\mathrm{\ΔL}}_{i2})---\left(2\right)$

Wherein Δ L_i1With Δ L_i2Implication be: Δ L_i1=Δ L_pf,ΔL_i2=Δ L_gf, it is the peak period when the i period, L_iRepresent The loading of i period, Δ L after enforcement tou power price_pfRepresent the load being transferred to the peak period after implementing tou power price by section at ordinary times Amount, Δ L_gfRepresent and be transferred to the loading of peak period by the paddy period after implementing tou power price；

F₂For typical day load curve peak-valley difference:

$F_2=L_{\max }-L_{\min }=\underset{1\≤i\≤24}{\max }(L_i\±{\mathrm{\ΔL}}_{i1}\±{\mathrm{\ΔL}}_{i2})-\underset{1\≤i\≤24}{\min }(L_i\±{\mathrm{\ΔL}}_{i1}\±{\mathrm{\ΔL}}_{i2})---\left(3\right)$

Wherein Δ L_i1With Δ L_i2Implication be: Δ L_i1=Δ L_fp,ΔL_i2=Δ L_gp, it is the peak period when the i period, Δ L_fpTable Show and be transferred to the loading of section at ordinary times, Δ L by the peak period after implementing tou power price_gpRepresent and turned by the paddy period after implementing tou power price Move to the loading of section at ordinary times；

F₃For the load variations amount of adjacent moment in daily load curve:

$F_3=\underset{i=1}{\overset{24}{\Σ}}|(L_i\±{\mathrm{\ΔL}}_{i1}\±{\mathrm{\ΔL}}_{i2})-(L_{i-1}\±{\mathrm{\ΔL}}_{(i-1)1}\±{\mathrm{\ΔL}}_{(i-1)2})|---\left(4\right)$

Wherein Δ L_i1With Δ L_i2Implication be: Δ L_i1=Δ L_fg,ΔL_i2=Δ L_pg, it is the peak period when the i period, Δ L_fgTable Show and be transferred to the loading of paddy period, Δ L by the peak period after implementing tou power price_pgRepresent and turned by section at ordinary times after implementing tou power price Move to the loading of paddy period；

Wherein, the loading Δ L of peak period it is transferred to after implementing tou power price described in step (3) by section at ordinary times_pf, implement It is transferred to the loading Δ L of peak period by the paddy period after tou power price_gf, implement to be transferred to section at ordinary times by the peak period after tou power price Loading Δ L_fp, implement to be transferred to the loading Δ L of section at ordinary times by the paddy period after tou power price_gp, implement after tou power price by The peak period is transferred to the loading Δ L of paddy period_fg, implement to be transferred to after tou power price the loading Δ of paddy period by section at ordinary times L_pgComputational methods be:

It is transferred to the loading Δ L of peak period by section at ordinary times after implementing tou power price_pfComputational methods be:

ΔL_pf=μ_fp×L_f (5)

${\mathrm{\μ}}_{fp}=\left\{\begin{array}{cc}0& (0\≤{\mathrm{\Δ}}_{fp}\≤a_{fp})\\ K_{fp}*({\mathrm{\Δ}}_{fp}-a_{fp})& (a_{fp}\≤{\mathrm{\Δ}}_{fp}\≤b_{fp})\\ {\mathrm{\μ}}_{fp\max }& ({\mathrm{\Δ}}_{fp}\≥b_{fp})\end{array}\right.---\left(6\right)$

μ_fpFor peak ordinary telegram price differential cause unit time period load decrement that the peak period produces with when carrying out tou power price leading peak The ratio of section average load, L_fFor carrying out tou power price leading peak period average load；Electricity price difference Δ_fp=P_f-P_p, P_f、P_pIt is respectively The electricity price at peak, at ordinary times section, a_fp、b_fpFor electricity price difference segmentation parameter, K_fpIt is in calculating factor during second segment for electricity price difference；

It is transferred to the loading Δ L of peak period by the paddy period after implementing tou power price_gfComputational methods be:

ΔL_gf=μ_fg×L_f (7)

${\mathrm{\μ}}_{fg}=\left\{\begin{array}{cc}0& (0\≤{\mathrm{\Δ}}_{fg}\≤a_{fg})\\ K_{fg}*({\mathrm{\Δ}}_{fg}-a_{fg})& (a_{fg}\≤{\mathrm{\Δ}}_{fg}\≤b_{fg})\\ {\mathrm{\μ}}_{fg\max }& ({\mathrm{\Δ}}_{fg}\≥b_{fg})\end{array}\right.---\left(8\right)$

μ_fgThe specific load decrement that the peak period produces is caused to be put down with carrying out the tou power price leading peak period for electricity price between peak and valley The all ratio of load, electricity price difference Δs_fg=P_f-P_g, P_f、P_gIt is respectively peak, the electricity price of paddy period, a_fg、b_fgJoin for electricity price difference section Number, K_fgIt is in calculating factor during second segment for electricity price difference；

It is transferred to the loading Δ L of section at ordinary times by the peak period after implementing tou power price_fpComputational methods be:

${\mathrm{\ΔL}}_{fp}={\mathrm{\ΔL}}_{pf}\×\frac{N_f}{N_p}---\left(9\right)$

N in formula_f, N_pIt is respectively the time hop count of peak period and section at ordinary times；

It is transferred to the loading Δ L of section at ordinary times by the paddy period after implementing tou power price_gpComputational methods be:

ΔL_gp=μ_pg×L_p (10)

${\mathrm{\μ}}_{pg}=\left\{\begin{array}{cc}0& (0\≤{\mathrm{\Δ}}_{pg}\≤a_{pg})\\ K_{pg}*({\mathrm{\Δ}}_{pg}-a_{pg})& (a_{pg}\≤{\mathrm{\Δ}}_{pg}\≤b_{pg})\\ {\mathrm{\μ}}_{pg\max }& ({\mathrm{\Δ}}_{pg}\≥b_{pg})\end{array}\right.---\left(11\right)$

μ_pgFor Pinggu electricity price difference cause specific load decrement that section at ordinary times produces with carry out tou power price before Duan Ping at ordinary times The all ratio of load, L_pFor carrying out before tou power price section average load, electricity price difference Δ at ordinary times_pg=P_p-P_g, P_p、P_gThe most flat, The electricity price of paddy period, a_pg、b_pgFor electricity price difference segmentation parameter, K_pgIt is in calculating factor during second segment for electricity price difference；

It is transferred to the loading Δ L of paddy period by the peak period after implementing tou power price_fgComputational methods be:

${\mathrm{\ΔL}}_{fg}={\mathrm{\ΔL}}_{gf}\×\frac{N_f}{N_g}---\left(12\right)$

N in formula_f, N_gIt is respectively the time hop count of peak period and paddy period；

It is transferred to the loading Δ L of paddy period by section at ordinary times after implementing tou power price_pgComputational methods be:

${\mathrm{\ΔL}}_{pg}={\mathrm{\ΔL}}_{gp}\×\frac{N_p}{N_g}---\left(13\right)$

N in formula_p, N_gIt is respectively the time hop count of Duan Yugu period at ordinary times.

In step of the present invention (3), constraints includes that the income from sales electricity sales amount constant, Utilities Electric Co. of Utilities Electric Co. is protected Hold constant, peak period, paddy period tou power price constraint, electrical price pattern constraint and user's power purchase expense restriction numerically；

The income from sales of Utilities Electric Co. is constant:

The income from sales carrying out the supplier of electricity before tou power price is:

M_NON=Q × P_NON (14)

M_NONFor the Utilities Electric Co.'s income from sales before unexecuted tou power price, Q and P_NONIt is respectively the sale of electricity before tou power price Amount and electricity price；After carrying out tou power price, the income from sales of supplier of electricity is:

M_TOU=Q_fTOU×P_f+Q_pTOU×P_p+Q_gTOU×P_g (15)

Q_fTOU、Q_pTOU、Q_gTOUFor carrying out tou power price postpeak period, at ordinary times section and the power consumption of paddy period, supplier of electricity is made a profit Constraints be:

M_TOU≥M_NON-M_SAVE (16)

Wherein M_SAVEThe Electricity Investment saved for supplier of electricity after carrying out tou power price and power supply cost, typically use interest concessions system The method of number calculates, and is taken as 0 in this model for the time being；

The electricity sales amount of Utilities Electric Co. keeps constant, i.e. power consumption before and after tou power price keeps constant:

Q_NON=Q_fTOU+Q_pTOU+Q_gTOU (17)

Wherein Q_fTOU、Q_pTOUAnd Q_gTOUFor carrying out tou power price postpeak, flat, the power consumption of paddy period, Q_NONFor tou power price Front power consumption；

Peak period, the constraint numerically of paddy period tou power price: according to the principle that marginal cost is theoretical, i.e. power system exists The electricity price of paddy period is more than or equal to operation marginal cost P in this period_MaCOST, and the electricity price of peak period is not higher than little generating Cost price P of unit generation_GeCOST, i.e.

P_g≥P_MaCOST (18)

P_f≤P_GeCOST (19)

This constraint is updated according to the typical day load curve optimum results of Second Year and the 3rd year；

Electrical price pattern retrains, i.e. electrical price pattern is:

1.2 ＜ P_f/P_g≤4 (20)

User's power purchase expense restriction: due to user paid power purchase expense be the power selling income of Utilities Electric Co., therefore This constraints can be expressed as

M_TOU≤M_NON (21)

Assume before implementing tou power price, to this area with carrying out average electricity price per family, big industry 0.633yuan/kWh； General industry and commerce 0.829yuan/kWh；Resident 0.518yuan/kWh, it is assumed that system is in operation marginal cost P of paddy period_MaCOST For 0.27yuan/kWh, in addition, it is assumed that small generation set is at the cost of electricity-generating valency P of peak period_GeCOSTFor 1.48yuan/kWh.This Invention utilizes and only considers that the tou power price (referred to as " electricity price one ") that typical day load curve optimization then is formulated is sent out with consideration load Exhibition and optimize the tou power price (referred to as " electricity price two ") of formulation and contrast.Table 1 is to optimize the electricity price result formulated:

1 two kinds of tou power price numerical value contrasts of table

The peak of response, low-valley interval load curve be as shown in Figure 4 and Figure 5 under different electricity prices for Second Year load curve. The peak of response, low-valley interval load curve be as shown in Figure 6 and Figure 7 under different electricity prices for 3rd yearly load curve.3 years loads Response condition under different electricity prices, employs maximum/minimum load, peak-valley ratio and load fluctuation rate index and weighs, Concrete result of calculation is shown in Table 2.

The 3rd year load of the table 2 Second Year response condition to different electricity prices

By index listed in result of implementation Fig. 4～Fig. 7 and table 2 it can be seen that in the case of implementing electricity price two, maximum is negative Lotus is relatively implemented the situation of electricity price one and decreases, and minimum load rises, and load fluctuation rate reduces, and peak-valley ratio all increases. Therefore tou power price that model provided by the present invention calculates is preferable to following load adaptability, the after tou power price is carried out 2 years with within the 3rd year, still user can be played incentive action.

4) electricity price is issued to user side；It is issued to user side, electricity price information bag by communication port after making electricity price Include Time segments division information and electricity price value information.

Claims (5)

1. the tou power price optimization method considering that load develops, it is characterised in that: comprise the following steps:

(1) obtain the typical day load curve of power system then and carry out Time segments division: choosing this area third season three In month, the daily load curve meansigma methods of temperature second highest day is typical day load curve then, for this curve mark off the peak period, Peace period paddy period；

(2) based on the typical day load curve of 2 years after Typical Day Load Curve Forecasting then: bear for typical case's day then Lotus curve utilizes Typical Day Load Curve Forecasting method to carry out Second Year and the Typical Day Load Curve Forecasting of the 3rd year；

(3) set up the tou power price Optimized model of consideration load development and solve under certain constraints: when solving first Typical day load curve for Second Year and the 3rd year calculates tou power price, then using result of calculation as electricity price bound about Bundle, again solves for typical day load curve then, i.e. obtains final tou power price；Optimization aim is by three partial weightings :

Min(F₁+F₂+F₃) (1)

F₁Peak load for typical day load curve:

$F_1=L_{max}=\underset{1\≤i\≤24}{max}(L_i\±{\mathrm{\ΔL}}_{i1}\±{\mathrm{\ΔL}}_{i2})---\left(2\right)$

Wherein Δ L_i1With Δ L_i2Implication be: Δ L_i1=Δ L_pf,ΔL_i2=Δ L_gf, it is the peak period when the i period, L_iRepresent and implement The loading of i period after tou power price, Δ L_pfRepresent the loading being transferred to the peak period after implementing tou power price by section at ordinary times, Δ L_gfRepresent and be transferred to the loading of peak period by the paddy period after implementing tou power price；

F₂For typical day load curve peak-valley difference:

$F_2=L_{max}-L_{min}=\underset{1\≤i\≤24}{max}(L_i\±{\mathrm{\ΔL}}_{i1}\±{\mathrm{\ΔL}}_{i2})-\underset{1\≤i\≤24}{min}(L_i\±{\mathrm{\ΔL}}_{i1}\±{\mathrm{\ΔL}}_{i2})---\left(3\right)$

Wherein Δ L_i1With Δ L_i2Implication be: Δ L_i1=Δ L_fp,ΔL_i2=Δ L_gp, it is the peak period when the i period, Δ L_fpRepresent real It is transferred to the loading of section at ordinary times, Δ L by the peak period after executing tou power price_gpRepresent and be transferred to by the paddy period after implementing tou power price The loading of section at ordinary times；

F₃For the load variations amount of adjacent moment in daily load curve:

$F_3=\underset{i=1}{\overset{24}{\mathrm{\Σ}}}|(L_i\±{\mathrm{\ΔL}}_{i1}\±{\mathrm{\ΔL}}_{i2})-(L_{i-1}\±{\mathrm{\ΔL}}_{(i-1)1}\±{\mathrm{\ΔL}}_{(i-1)2})|---\left(4\right)$

Wherein Δ L_i1With Δ L_i2Implication be: Δ L_i1=Δ L_fg,ΔL_i2=Δ L_pg, it is the peak period when the i period, Δ L_fgRepresent real It is transferred to the loading of paddy period, Δ L by the peak period after executing tou power price_pgRepresent and be transferred to by section at ordinary times after implementing tou power price The loading of paddy period；

(4) electricity price is issued to user side；It is issued to user side by communication port, when electricity price information includes after making electricity price Section division information and electricity price value information.

A kind of tou power price optimization method considering that load develops the most according to claim 1, it is characterised in that step (1) the typical day load curve Time segments division described in method particularly includes:

(1.1) determine minimum load point and the peak load point of typical day load curve, be designated as a point and b point respectively；Wherein a point The probability being in the paddy period is 100%, and the probability being in the peak period is 0, and it is 100% that b point is in the probability of peak period, The probability being in the paddy period is 0；

(1.2) calculate remaining each point according to type membership function less than normal and be in the probability of paddy period, be subordinate to according to type bigger than normal Function calculates remaining each point and is in the probability of peak period；Peak period probability according to every bit and paddy period probability, If this peak period probability is much larger than paddy period probability, this point belongs to the peak period, if this paddy period probability is long-range Belonging to the paddy period in peak period probability then this point, remaining point belongs to section at ordinary times；

(1.3) combine the duration of every period of independent period to be more than or equal to two hours, and peak period, at ordinary times section, total duration of paddy period It is all higher than, equal to six hours, the Time segments division result of typical day load curve to be obtained.

A kind of tou power price optimization method considering that load develops the most according to claim 1, it is characterised in that step (2) the Typical Day Load Curve Forecasting method described in is:

(2.1) typical day load curve is decomposed, be decomposed into and characterize the development component of its law of development and characterize its ripple The wave component of dynamic rule, development component takes peak load then, and wave component takes the ratio of each point load and peak load； Wherein development component utilizes Lycoperdon polymorphum Vitt (Grey Model-GM) (1,1) model to obtain, and wave component utilizes BP (Back Propagation) neural network model obtains；

(2.2) synthesize with the predictive value of wave component according to the predictive value of development component, obtain typical day load curve pre- Measured value.

A kind of tou power price optimization method considering that load develops the most according to claim 1, it is characterised in that step (3) the loading Δ L of peak period it is transferred to after implementing tou power price described in by section at ordinary times_pf, when implementing after tou power price by paddy Section is transferred to the loading Δ L of peak period_gf, implement to be transferred to the loading Δ L of section at ordinary times by the peak period after tou power price_fp, real It is transferred to the loading Δ L of section at ordinary times by the paddy period after executing tou power price_gp, when implementing to be transferred to paddy by the peak period after tou power price The loading Δ L of section_fg, implement to be transferred to after tou power price the loading Δ L of paddy period by section at ordinary times_pgComputational methods be:

It is transferred to the loading Δ L of peak period by section at ordinary times after implementing tou power price_pfComputational methods be:

ΔL_pf=μ_fp×L_f (5)

${\mathrm{\μ}}_{fp}=\left\{\begin{array}{cc}0& (0\≤{\mathrm{\Δ}}_{fp}\≤a_{fp})\\ K_{fp}*({\mathrm{\Δ}}_{fp}-a_{fp})& (a_{fp}\≤{\mathrm{\Δ}}_{fp}\≤b_{fp})\\ {\mathrm{\μ}}_{fp\max }& ({\mathrm{\Δ}}_{fp}\≥b_{fp})\end{array}\right.---\left(6\right)$

μ_fpThe unit time period load decrement that the peak period produces is caused to be put down with carrying out the tou power price leading peak period for peak ordinary telegram price differential The all ratio of load, L_fFor carrying out tou power price leading peak period average load；Electricity price difference Δ_fp=P_f-P_p, P_f、P_pBe respectively peak, The electricity price of section, a at ordinary times_fp、b_fpFor electricity price difference segmentation parameter, K_fpIt is in calculating factor during second segment for electricity price difference；

It is transferred to the loading Δ L of peak period by the paddy period after implementing tou power price_gfComputational methods be:

ΔL_gf=μ_fg×L_f (7)

${\mathrm{\μ}}_{fg}=\left\{\begin{array}{cc}0& (0\≤{\mathrm{\Δ}}_{fg}\≤a_{fg})\\ K_{fg}*({\mathrm{\Δ}}_{fg}-a_{fg})& (a_{fg}\≤{\mathrm{\Δ}}_{fg}\≤b_{fg})\\ {\mathrm{\μ}}_{fg\max }& ({\mathrm{\Δ}}_{fg}\≥b_{fg})\end{array}\right.---\left(8\right)$

μ_fgSpecific load decrement that the peak period produces is caused with to carry out the tou power price leading peak period the most negative for electricity price between peak and valley The ratio of lotus, electricity price difference Δ_fg=P_f-P_g, P_f、P_gIt is respectively peak, the electricity price of paddy period, a_fg、b_fgFor electricity price difference segmentation parameter, K_fg It is in calculating factor during second segment for electricity price difference；

It is transferred to the loading Δ L of section at ordinary times by the peak period after implementing tou power price_fpComputational methods be:

${\mathrm{\ΔL}}_{fp}={\mathrm{\ΔL}}_{pf}\×\frac{N_f}{N_p}---\left(9\right)$

N in formula_f, N_pIt is respectively the time hop count of peak period and section at ordinary times；

It is transferred to the loading Δ L of section at ordinary times by the paddy period after implementing tou power price_gpComputational methods be:

ΔL_gp=μ_pg×L_p (10)

${\mathrm{\μ}}_{pg}=\left\{\begin{array}{cc}0& (0\≤{\mathrm{\Δ}}_{pg}\≤a_{pg})\\ K_{pg}*({\mathrm{\Δ}}_{pg}-a_{pg})& (a_{pg}\≤{\mathrm{\Δ}}_{pg}\≤b_{pg})\\ {\mathrm{\μ}}_{pgmax}& ({\mathrm{\Δ}}_{pg}\≥b_{pg})\end{array}\right.---\left(11\right)$

μ_pgCause before specific load decrement that section at ordinary times produces and implementation tou power price section at ordinary times the most negative for Pinggu electricity price difference The ratio of lotus, L_pFor carrying out before tou power price section average load, electricity price difference Δ at ordinary times_pg=P_p-P_g, P_p、P_gIt is respectively flat, Gu Shi The electricity price of section, a_pg、b_pgFor electricity price difference segmentation parameter, K_pgIt is in calculating factor during second segment for electricity price difference；

It is transferred to the loading Δ L of paddy period by the peak period after implementing tou power price_fgComputational methods be:

${\mathrm{\ΔL}}_{fg}={\mathrm{\ΔL}}_{gf}\×\frac{N_f}{N_g}---\left(12\right)$

N in formula_f, N_gIt is respectively the time hop count of peak period and paddy period；

It is transferred to the loading Δ L of paddy period by section at ordinary times after implementing tou power price_pgComputational methods be:

${\mathrm{\ΔL}}_{pg}={\mathrm{\ΔL}}_{gp}\×\frac{N_p}{N_g}---\left(13\right)$

N in formula_p, N_gIt is respectively the time hop count of Duan Yugu period at ordinary times.

A kind of tou power price optimization method considering that load develops the most according to claim 1, it is characterised in that described step Suddenly the constraints in (3) include the income from sales of Utilities Electric Co. electricity sales amount constant, Utilities Electric Co. keep constant, peak period, Paddy period tou power price constraint, electrical price pattern constraint and user's power purchase expense restriction numerically；

The income from sales of Utilities Electric Co. is constant:

The income from sales carrying out the supplier of electricity before tou power price is:

M_NON=Q × P_NON (14)

M_NONFor the Utilities Electric Co.'s income from sales before unexecuted tou power price, Q and P_NONBe respectively the electricity sales amount before tou power price and Electricity price；After carrying out tou power price, the income from sales of supplier of electricity is:

M_TOU=Q_fTOU×P_f+Q_pTOU×P_p+Q_gTOU×P_g (15)

Q_fTOU、Q_pTOU、Q_gTOUFor carrying out tou power price postpeak period, at ordinary times section and the power consumption of paddy period, the pact that supplier of electricity is made a profit Bundle condition is:

M_TOU≥M_NON-M_SAVE (16)

Wherein M_SAVEThe Electricity Investment saved for supplier of electricity after carrying out tou power price and power supply cost；

The electricity sales amount of Utilities Electric Co. keeps constant, i.e. power consumption before and after tou power price keeps constant:

Q_NON=Q_fTOU+Q_pTOU+Q_gTOU (17)

Wherein Q_fTOU、Q_pTOUAnd Q_gTOUFor carrying out tou power price postpeak, flat, the power consumption of paddy period, Q_NONBefore tou power price Power consumption；

Peak period, the constraint numerically of paddy period tou power price: according to the principle that marginal cost is theoretical, i.e. power system is when paddy The electricity price of section is more than or equal to operation marginal cost P in this period_MaCOST, and the electricity price of peak period is not higher than small generation set Cost price P of generating_GeCOST, i.e.

P_g≥P_MaCOST (18)

P_f≤P_GeCOST (19)

This constraint is updated according to the typical day load curve optimum results of Second Year and the 3rd year；

Electrical price pattern retrains, i.e. electrical price pattern is:

1.2 ＜ P_f/P_g≤4 (20)

User's power purchase expense restriction: due to user paid power purchase expense be the power selling income of Utilities Electric Co., so about Bundle condition can be expressed as

M_TOU≤M_NON (21)。