CN102810860B - Standby volume analytical method under coordinating dispatch mode of batch-type energy and conventional energy - Google Patents

Abstract

The invention discloses a standby volume analytical method under a coordinating dispatch mode of batch-type energy and conventional energy. The method comprises the following five steps of calculating a standby past sample data region, obtaining similar cluster time intervals in a planed date, carrying out statics on a forecast output power deviation distribution of each time interval, obtaining a probability density function of the standby volume of wind energy, and calculating a rotation standby volume of a system under the coordinating dispatch mode of the batch-type energy and the conventional energy. With the adoption of the analytical method, according to a consideration of the coordinating dispatch of assembling units of the batch-type energy and the conventional energy, statics and analysis on data of past wind-power operation are carried out to form the current requirement on the rotation standby volume, safety guarantee is provided for the coordinating dispatch of the batch-type energy and the conventional energy, a reasonable standby volume is provided, and the realization of economical requirement on operation of a power system is facilitated. The method has the characteristics of low calculation intensity and strong adaptability, and is more suitable for popularization and application of connecting with dispatch mechanisms with larger power of wind power.

Description

Reserve capacity analytical method under wind energy and conventional energy resource coordinated dispatching mode

Technical field

The invention belongs to dispatching automation of electric power systems technical field, be specifically related to reserve capacity analytical method under a kind of wind energy and conventional energy resource coordinated dispatching mode.

Background technology

In recent years, energy-saving and emission-reduction also contain that climate warming has been face one, whole world challenge and important issue jointly, the Chinese government pays much attention to the energy-saving and emission-reduction work of power industry, propose to implement energy-saving power generation dispatching at power domain, improve power industry energy use efficiency, reduce environmental pollution, promote the energy and electric power structural adjustment, this is power industry implement scientific view of development, the major action of building a harmonious socialist society is the inevitable choice of building a resource-conserving and environment-friendly society.

With wind-powered electricity generation be the emerging energy of representative because of its non-polluting renewable characteristic, and without greenhouse gas emission, become the important directions of energy development gradually.Wind-powered electricity generation is as one of the most ripe utilization of new energy resources mode of technology, the quick growth of being doubled for continuous 4 years has been realized under the support energetically of country, 2,600 ten thousand kilowatts are reached by the total installed capacity of 2009 China's wind-powered electricity generation in the end of the year, rank the second in the world, expects the year two thousand twenty total installation of generating capacity and will reach 1.5 hundred million kilowatts.Be that the new forms of energy of representative just progressively become the important energy resources of China with wind-powered electricity generation, meeting energy demand, improve energy resource structure, reduce environmental pollution, preserve the ecological environment, promote socio-economic development etc. in play a significant role.

But wind-powered electricity generation has typical intermittent characteristic, power supply reliability is lower compared with conventional energy resource, because power production process carries out continuously, for ensureing that still can normally run in the situations such as overhaul of the equipments, unit fault, load fluctuation is appearring in electric power system, the system of being necessary for reserves certain reserve capacity, improves the reliability and stability of system cloud gray model; Along with the continuous increase of the wind energy quantity be incorporated into the power networks and scale, the randomness of wind-powered electricity generation, fluctuation and intermittence, make the uncertain factor in power system operation process also constantly increase, more highlight the importance of standby configuration.

In conventional energy resource generation schedule, alternative plan manages in proportion or regulates for subsequent use according to absolute-value sense initialization system spinning reserve and AGC (automatic generation amount controls to be a critical function in EMS EMS), calculate for generation schedule analysis for subsequent use a few days ago and supervision early warning, after extensive wind energy networks, except it is pollution-free, renewable and advantages of environment protection, new forms of energy have intermittence simultaneously, the shortcomings such as randomness, operation of power networks reliability is reduced, this adds certain risk to power system operation, for ensureing the fail safe of the rear electrical network of wind energy networking, stability, and the requirement of response intelligent grid economy, need to obtain system reserve capacity by setting up the reserved type module adapting to the access of extensive wind energy, eliminate the adverse effect of energy batch (-type) characteristic to electric power netting safe running and power supply quality, it is the problem of current primary solution.

Summary of the invention

The object of the invention is to overcome in prior art when after extensive wind energy networking, there is intermittence, the shortcoming of randomness etc., operation of power networks reliability is reduced, power system operation is made to add the problem of risk, reserve capacity analytical method under wind energy provided by the invention and conventional energy resource coordinated dispatching mode, the reserve capacity of system is obtained by setting up the reserved type module adapting to the access of extensive wind energy, eliminate the adverse effect of energy batch (-type) characteristic to electric power netting safe running and power supply quality, ensure wind energy network after the safety and stability of electrical network.

In order to solve the problems of the technologies described above, the technical solution adopted in the present invention is:

Reserve capacity analytical method under a kind of wind energy and conventional energy resource coordinated dispatching mode, is characterized in that: comprise the following steps,

It is interval that step (1) calculates historical sample data for subsequent use

Adopt the wind power prediction error of method to electric power system of directly statistics to carry out statistical analysis, select the history wind power output power sequence similar to date of enterprise, interval as calculating historical sample data for subsequent use;

Step (2) obtains the similar cluster period in date of enterprise

Period cluster is carried out in the historical sample data interval that step (1) obtains, obtains the similitude clustering period in date of enterprise;

The prediction that step (3) adds up each period is exerted oneself deviation profile

According to the similitude clustering period in the needs of scheduling, current wind energy power output situation and date of enterprise, add up the deviation profile of the prediction of each period;

Step (4) draws the probability density function of the reserve capacity of wind energy

With represent the probability density function of the spare capacity needs of the wind energy of corresponding period, with wind power prediction probability-distribution function f by mistake _t(wp ^error) identical, relation as formula (1),

$f_T^{\mathrm{reserve}}\left({\mathrm{wp}}^{\mathrm{error}}\right)=f_T\left({\mathrm{wp}}^{\mathrm{error}}\right)---\left(1\right)$

Namely the expression (2) of the reserve capacity of wind energy is,

${\∫}_{-\∞}^{R_{\mathrm{wind},h}}\frac{1}{\sqrt{2\mathrm{\π}}\·{\mathrm{\σ}}_{\mathrm{wind},h}}\·\exp (-\frac{{(\mathrm{\ξ}-{\mathrm{\μ}}_{\mathrm{wind},h})}^2}{2\·{\mathrm{\σ}}_{\mathrm{wind},h}^2})\·\mathrm{d\ξ}=\mathrm{\α}---\left(2\right)$

Wherein R _{wind, h}for the wind energy reserve capacity of h period, μ _{wind, h}for the wind power prediction error distribution of h period is expected, σ _{wind, h}for the wind power prediction error distribution variance of h period, α is the level of confidence that the reserve capacity of wind energy meets system cloud gray model, and ξ is integration variable;

Step (5) calculates the spinning reserve capacity of system under wind energy and conventional energy resource coordinated dispatching mode

After wind power integration, the spinning reserve capacity of system is divided into for subsequent use and wind energy two parts for subsequent use of conventional energy resource, the conventional energy resource employing experience for subsequent use estimation technique, percentage according to load is arranged, the reserve capacity of wind energy is that formula (2) represents, under wind energy and conventional energy resource coordinated dispatching mode, the spinning reserve capacity expression formula of system is as formula (3)

$\left\{\begin{array}{c}{R}_{\mathrm{total},h}=R_{\mathrm{cus},h}+R_{\mathrm{wind},h}\\ {\∫}_{-\∞}^{R_{\mathrm{wind},h}}\frac{1}{\sqrt{2\mathrm{\π}}\·{\mathrm{\σ}}_{\mathrm{wind},h}}\·\exp (-\frac{{(\mathrm{\ξ}-{\mathrm{\μ}}_{\mathrm{wind},h})}^2}{2\·{\mathrm{\σ}}_{\mathrm{wind},h}^2})\·\mathrm{d\ξ}=\mathrm{\α}\end{array}\right.---\left(3\right)$

Wherein R _{total, h}for the reserve capacity of the system of h period, R _{cus, h}for the reserve capacity of the conventional energy resource of h period, R _{wind, h}for the wind energy reserve capacity of h period.

Reserve capacity analytical method under aforesaid wind energy and conventional energy resource coordinated dispatching mode, is characterized in that: described step (1) selects the history wind power output power sequence similar to date of enterprise, and the method wherein selected, comprises the following steps,

(1) as the foundation judging similarity, formula (4) is used to express similar Euclidean distance d _i,j:

$d_{i,j}=\sqrt{{|\frac{x_{i,1}}{{\mathrm{Cap}}_i}-\frac{x_{j,1}}{{\mathrm{Cap}}_j}|}^2+{|\frac{x_{i,2}}{{\mathrm{Cap}}_i}-\frac{x_{j,2}}{{\mathrm{Cap}}_j}|}^2+...{|\frac{x_{i,96}}{{\mathrm{Cap}}_i}-\frac{x_{j,96}}{{\mathrm{Cap}}_j}|}^2}---\left(4\right)$

Wherein set i as plan day, j is history day, then Xi (xi, 1, xi, 2 ... .xi, 96), Xj (xj, 1, xj, 2 ... .xj, 96) i is expressed as, the j wind power output sequence of two days, Capi, Capj represent i respectively, the installed capacity of j sampling unit on the two, criterion Euclidean distance d _i,jrepresent the similitude of ratio in mean geometrical distance of i, the j wind power output of two days and the installed capacity of the sampling unit of history;

(2) what provide similarity determines threshold value Z, determines threshold value Z at [0, max{d _i,j] in interval, i is plan day, j ∈ { all history dates };

(3) that lists according to (2) determines threshold value Z, if di, j<Z, then the power output sequence of history day and the prediction of corresponding plan day sequence of exerting oneself is set of metadata of similar data, interval as calculating historical sample data for subsequent use.

Reserve capacity analytical method under aforesaid wind energy and conventional energy resource coordinated dispatching mode, is characterized in that: the method that period cluster is carried out in the historical sample data interval that step (1) obtains by described step (2) is specific as follows,

(1) carry out hierarchical clustering to the day part power output sequence in historical sample data interval, obtain K division, a calculating K center divided, respectively as the initial cluster centre of clustering algorithm;

(2) for the power output sequence of each period, the Euclidean distance of the K asking it a to correspond to cluster centre, is grouped into the cluster at the shortest place, center of distance, is formed a new K cluster;

(3) for forming K new cluster, averaging method is utilized to recalculate the cluster centre of each cluster calculation;

(4) iteration is carried out in repetition step (2), (3), until class members no longer changes, then iteration terminates, and exports the similar cluster period.

Reserve capacity analytical method under aforesaid wind energy and conventional energy resource coordinated dispatching mode, is characterized in that: the prediction of step (3) described statistics each period exerts oneself the method for deviation profile for for period T, and time range is [t _t0, t _tN], be provided with N number of time point, t _t0for the starting point of period T, t _tNfor the end point of period T, adjacent time point be spaced apart 15 minutes, the length of historical sample data is M days, then for period T, total M*N output deviation data, deviation data Normal Distribution, utilizes Maximum Likelihood Estimation Method, can obtain the expectation and variance of deviation profile according to formula (5), (6), the expectation estimation value of the deviation profile of predicated error is:

$\widetilde{\mathrm{\&mu;}}=\frac{1}{M*N}\&CenterDot;\underset{i=1}{\overset{M*N}{\mathrm{\&Sigma;}}}{\mathrm{wp}}_{t,i}^{\mathrm{error}}---\left(5\right)$

The estimate of variance of the deviation profile of predicated error is:

${\widetilde{\mathrm{\&sigma;}}}^2=\frac{1}{M*N}\&CenterDot;\underset{i=1}{\overset{M*N}{\mathrm{\&Sigma;}}}{({\mathrm{wp}}_{t,i}^{\mathrm{error}}-\widetilde{\mathrm{\&mu;}})}^2---\left(6\right)$

Wherein in formula (2), (3) for wind power prediction error amount.

The invention has the beneficial effects as follows:

1) through considering the coordinated scheduling of wind energy and conventional energy resource, consider uncertainty and the fluctuation of Wind turbines in advance, ensure that the enforceability of wind energy and conventional energy resource coordinated scheduling, ensure wind fire coordinated scheduling power system security, economic and stable operation;

2) according to historical forecast data and the history practical operation situation of wind-powered electricity generation, electric power system is load prediction data a few days ago, wind power prediction data etc., by carrying out predicated error analysis to wind power historical service data, form the plan of needs of spinning reserve capacity a few days ago, ensure the secure accessing of wind power, contribute to the safety and economic operation better instructing electric power system;

3) on the basis of fail safe ensureing the rear electric power system of wind energy access, rational reserve capacity is provided, for the coordinated scheduling of wind energy and conventional energy resource provides safety guarantee, contributes to the cost-effectiveness requirement realizing power system operation.

Accompanying drawing explanation

Fig. 1 is the flow chart of reserve capacity analytical method under wind energy of the present invention and conventional energy resource coordinated dispatching mode.

Fig. 2 is flow chart historical sample data interval being carried out to period cluster of the present invention.

Embodiment

Below in conjunction with Figure of description, the present invention is further illustrated.

As shown in Figure 1, reserve capacity analytical method under wind energy of the present invention and conventional energy resource coordinated dispatching mode, there is the intelligent recognition function of time-shared fashion for subsequent use, stand-by requirement is analyzed to be needed to carry out statistical analysis to a large amount of wind-powered electricity generation history datas, screen, select the historical data similar to plan day, as having the close date day with plan, or the wind-powered electricity generation historical data on close date for many years, identical month, the historical datas such as identical season are as data source, adopt Euclidean distance as the judgment basis of degree of similarity, determine that historical sample data is interval, as the calculated data source of reserve capacity, and adopt Kmeans clustering algorithm to mark off similar historical period, ensure accuracy and the reliability of statistical computation, also according to wind-powered electricity generation stand-by requirement analytical parameters, wind-powered electricity generation history data and wind power output prediction data, under analysis meets wind-powered electricity generation access situation entirely, system increases stand-by requirement newly, comprise spinning reserve and lower rotation stand-by requirement, and the fluctuation confidential interval of wind-powered electricity generation, rational reserve capacity is provided, contribute to the cost-effectiveness requirement realizing power system operation, as shown in Figure 1, comprise the following steps:

The first step, calculates historical sample data for subsequent use interval

The wind power prediction error of method to electric power system of directly statistics is adopted to carry out statistical analysis, select the history wind power output power sequence similar to date of enterprise, interval as calculating historical sample data for subsequent use, select the history wind power output power sequence method similar to date of enterprise, specifically comprise the following steps

(1) as the foundation judging similarity, formula (4) is used to express similar Euclidean distance d _i,j:

$d_{i,j}=\sqrt{{|\frac{x_{i,1}}{{\mathrm{Cap}}_i}-\frac{x_{j,1}}{{\mathrm{Cap}}_j}|}^2+{|\frac{x_{i,2}}{{\mathrm{Cap}}_i}-\frac{x_{j,2}}{{\mathrm{Cap}}_j}|}^2+...{|\frac{x_{i,96}}{{\mathrm{Cap}}_i}-\frac{x_{j,96}}{{\mathrm{Cap}}_j}|}^2}---\left(4\right)$

Wherein set i as plan day, j is history day, then Xi (xi, 1, xi, 2 ... .xi, 96), Xj (xj, 1, xj, 2 ... .xj, 96) i is expressed as, the j wind power output sequence of two days, Capi, Capj represent i respectively, the installed capacity of j sampling unit on the two, criterion Euclidean distance d _i,jrepresent the similitude of ratio in mean geometrical distance of i, the j wind power output of two days and the installed capacity of the sampling unit of history, d here _i,jless, similarity is larger;

(2) what provide similarity determines threshold value Z, determines threshold value Z at [0, max{d _i,j] in interval, i is plan day, j ∈ { all history dates };

(3) according to 2) list determine threshold value Z, if di, j<Z, then the power output sequence of history day and the prediction of corresponding plan day sequence of exerting oneself is set of metadata of similar data, interval as calculating historical sample data for subsequent use;

Second step, obtains the similar cluster period in date of enterprise

Period cluster is carried out in the historical sample data interval first step obtained, and obtains the similitude clustering period in date of enterprise, and as shown in Figure 2, the method for historical sample data interval being carried out to period cluster is specific as follows,

1) carry out hierarchical clustering to the day part power output sequence in historical sample data interval, obtain K division, a calculating K center divided, respectively as the initial cluster centre of Kmeans clustering algorithm;

2) for the power output sequence of each period, the Euclidean distance of the K asking it a to correspond to cluster centre, is grouped into the cluster at the shortest place, center of distance, is formed a new K cluster;

3) for forming K new cluster, averaging method is utilized to recalculate the cluster centre of each cluster calculation;

4) repeat 2), 3) carry out iteration, until each cluster member no longer changes, then iteration terminates; Export the similar cluster period;

3rd step, the prediction adding up each period is exerted oneself deviation profile

According to the similitude clustering period in the needs of scheduling, current wind energy power output situation and date of enterprise, add up the deviation profile of the prediction of each period, concrete grammar is for a certain period T, and time range is [t _t0, t _tN], be provided with N number of time point, t _t0for the starting point of period T, t _tNfor the end point of period T, adjacent time point be spaced apart 15 minutes, the length of historical sample data is M days, then for this period T, total M*N output deviation data, deviation data Normal Distribution, utilizes Maximum Likelihood Estimation Method, can obtain the expectation and variance of deviation profile according to formula (5), (6), the expectation estimation value of the deviation profile of predicated error is:

$\widetilde{\mathrm{\&mu;}}=\frac{1}{M*N}\&CenterDot;\underset{i=1}{\overset{M*N}{\mathrm{\&Sigma;}}}{\mathrm{wp}}_{t,i}^{\mathrm{error}}---\left(5\right)$

The estimate of variance of the deviation profile of predicated error is:

${\widetilde{\mathrm{\&sigma;}}}^2=\frac{1}{M*N}\&CenterDot;\underset{i=1}{\overset{M*N}{\mathrm{\&Sigma;}}}{({\mathrm{wp}}_{t,i}^{\mathrm{error}}-\widetilde{\mathrm{\&mu;}})}^2---\left(6\right)$

Wherein in formula (2), (3) for wind power prediction error amount;

4th step, draws the probability density function of the reserve capacity of wind energy

With represent the probability density function of the spare capacity needs of the wind energy of corresponding period, with wind power prediction probability-distribution function f by mistake _t(wp ^error) identical, relation as formula (1),

$f_T^{\mathrm{reserve}}\left({\mathrm{wp}}^{\mathrm{error}}\right)=f_T\left({\mathrm{wp}}^{\mathrm{error}}\right)---\left(1\right)$

According to the expectation and variance obtaining deviation profile of the 3rd step, the expression (2) of the reserve capacity of wind energy is,

${\&Integral;}_{-\&infin;}^{R_{\mathrm{wind},h}}\frac{1}{\sqrt{2\mathrm{\&pi;}}\&CenterDot;{\mathrm{\&sigma;}}_{\mathrm{wind},h}}\&CenterDot;\exp (-\frac{{(\mathrm{\&xi;}-{\mathrm{\&mu;}}_{\mathrm{wind},h})}^2}{2\&CenterDot;{\mathrm{\&sigma;}}_{\mathrm{wind},h}^2})\&CenterDot;\mathrm{d\&xi;}=\mathrm{\&alpha;}---\left(2\right)$

Wherein R _{wind, h}for the wind energy reserve capacity of h period, μ _{wind, h}for the wind power prediction error distribution of h period is expected, σ _{wind, h}for the wind power prediction error distribution variance of h period, α is the level of confidence that the reserve capacity of wind energy meets system cloud gray model, and ξ is integration variable;

5th step, calculates the spinning reserve capacity of system under wind energy and conventional energy resource coordinated dispatching mode

After wind power integration, the spinning reserve capacity of system is divided into for subsequent use and wind energy two parts for subsequent use of conventional energy resource, the conventional energy resource employing experience for subsequent use estimation technique, percentage according to load is arranged, the reserve capacity of wind energy is that formula (2) represents, under wind energy and conventional energy resource coordinated dispatching mode, the spinning reserve capacity expression formula of system is as formula (3)

$\left\{\begin{array}{c}{R}_{\mathrm{total},h}=R_{\mathrm{cus},h}+R_{\mathrm{wind},h}\\ {\&Integral;}_{-\&infin;}^{R_{\mathrm{wind},h}}\frac{1}{\sqrt{2\mathrm{\&pi;}}\&CenterDot;{\mathrm{\&sigma;}}_{\mathrm{wind},h}}\&CenterDot;\exp (-\frac{{(\mathrm{\&xi;}-{\mathrm{\&mu;}}_{\mathrm{wind},h})}^2}{2\&CenterDot;{\mathrm{\&sigma;}}_{\mathrm{wind},h}^2})\&CenterDot;\mathrm{d\&xi;}=\mathrm{\&alpha;}\end{array}\right.---\left(3\right)$

Wherein R _{total, h}for the reserve capacity of the system of h period, R _{cus, h}for the reserve capacity of the conventional energy resource of h period, R _{wind, h}for the wind energy reserve capacity of h period.

The present invention is through considering the coordinated scheduling of wind energy and conventional energy resource, consider uncertainty and the fluctuation of Wind turbines in advance, ensure that the enforceability of wind energy and conventional energy resource coordinated scheduling, ensure wind fire coordinated scheduling power system security, economic and stable operation; According to historical forecast data and the history practical operation situation of wind-powered electricity generation, electric power system is load prediction data a few days ago, wind power prediction data etc., by carrying out predicated error analysis to wind power historical service data, form the plan of needs of spinning reserve capacity a few days ago, ensure the secure accessing of wind power, contribute to the safety and economic operation better instructing electric power system; On the basis of the fail safe of electric power system after ensureing wind energy access, rational reserve capacity is provided, for the coordinated scheduling of wind energy and conventional energy resource provides safety guarantee, contributes to the cost-effectiveness requirement realizing power system operation.

More than show and describe general principle of the present invention, principal character and advantage.The technical staff of the industry should understand; the present invention is not restricted to the described embodiments; what describe in above-described embodiment and specification just illustrates principle of the present invention; without departing from the spirit and scope of the present invention; the present invention also has various changes and modifications, and these changes and improvements all fall in the claimed scope of the invention.Application claims protection range is defined by appending claims and equivalent thereof.

Claims (4)

1. reserve capacity analytical method under wind energy and conventional energy resource coordinated dispatching mode, is characterized in that: comprise the following steps,

It is interval that step (1) calculates historical sample data for subsequent use

Adopt the wind power prediction error of method to electric power system of directly statistics to carry out statistical analysis, select the history wind power output power sequence similar to date of enterprise, interval as calculating historical sample data for subsequent use;

Step (2) obtains the similar cluster period in date of enterprise

Period cluster is carried out in the historical sample data interval that step (1) obtains, obtains the similitude clustering period in date of enterprise;

The prediction that step (3) adds up each period is exerted oneself deviation profile

According to the similitude clustering period in the needs of scheduling, current wind energy power output situation and date of enterprise, add up the deviation profile of the prediction of each period;

Step (4) draws the probability density function of the reserve capacity of wind energy

With represent the probability density function of the spare capacity needs of the wind energy of corresponding period, with wind power prediction probability-distribution function f by mistake _t(wp ^error) identical, relation as formula (1),

$f_T^{\mathrm{reserve}}\left({\mathrm{wp}}^{\mathrm{error}}\right)=f_T\left({\mathrm{wp}}^{\mathrm{error}}\right)---\left(1\right)$

Wherein, for the probability density function of the spare capacity needs of the wind energy of corresponding period, f _tfor wind power prediction probability-distribution function by mistake, reserve is the spare capacity needs of wind energy, and error is predicated error, wp ^errorbe respectively the probability density function of the spare capacity needs of the wind energy of corresponding period wind power prediction is probability-distribution function f by mistake _twind power prediction error variance;

Namely the expression (2) of the reserve capacity of wind energy is,

${\&Integral;}_{-\&infin;}^{R_{\mathrm{wind},h}}\frac{1}{\sqrt{2}\mathrm{\&pi;}\&CenterDot;{\mathrm{\&sigma;}}_{\mathrm{wind},h}}\&CenterDot;\exp (-\frac{{(\mathrm{\&xi;}-{\mathrm{\&mu;}}_{\mathrm{wind},h})}^2}{2\&CenterDot;{\mathrm{\&sigma;}}_{\mathrm{wind},h}^2})\&CenterDot;\mathrm{d\&xi;}=\mathrm{\&alpha;}---\left(2\right)$

Wherein R _{wind, h}for the wind energy reserve capacity of h period, μ _{wind, h}for the wind power prediction error distribution of h period is expected, σ _{wind, h}for the wind power prediction error distribution variance of h period, α is the level of confidence that the reserve capacity of wind energy meets system cloud gray model, and ξ is integration variable;

Step (5) calculates the spinning reserve capacity of system under wind energy and conventional energy resource coordinated dispatching mode

After wind power integration, the spinning reserve capacity of system is divided into for subsequent use and wind energy two parts for subsequent use of conventional energy resource, the conventional energy resource employing experience for subsequent use estimation technique, percentage according to load is arranged, the reserve capacity of wind energy is that formula (2) represents, under wind energy and conventional energy resource coordinated dispatching mode, the spinning reserve capacity expression formula of system is as formula (3)

$\left\{\begin{array}{c}{R}_{\mathrm{total},h}=R_{\mathrm{cus},h}+R_{\mathrm{wind},h}\\ {\&Integral;}_{-\&infin;}^{R_{\mathrm{wind},h}}\frac{1}{\sqrt{2}\mathrm{\&pi;}\&CenterDot;{\mathrm{\&sigma;}}_{\mathrm{wind},h}}\&CenterDot;\exp (-\frac{{(\mathrm{\&xi;}-{\mathrm{\&mu;}}_{\mathrm{wind},h})}^2}{2\&CenterDot;{\mathrm{\&sigma;}}_{\mathrm{wind},h}^2})\&CenterDot;\mathrm{d\&xi;}=\mathrm{\&alpha;}\end{array}\right.---\left(3\right)$

Wherein R _{total, h}for the reserve capacity of the system of h period, R _{cus, h}for the reserve capacity of the conventional energy resource of h period, R _{wind, h}for the wind energy reserve capacity of h period.

2. reserve capacity analytical method under wind energy according to claim 1 and conventional energy resource coordinated dispatching mode, it is characterized in that: described step (1) selects the history wind power output power sequence similar to date of enterprise, the method wherein selected, comprises the following steps

(1) as the foundation judging similarity, formula (4) is used to express similar Euclidean distance d _i,j:

$d_{i,j}=\sqrt{{|\frac{x_{i,1}}{{\mathrm{Cap}}_i}-\frac{x_{j,1}}{{\mathrm{Cap}}_j}|}^2+{|\frac{x_{i,2}}{{\mathrm{Cap}}_i}-\frac{x_{j,2}}{{\mathrm{Cap}}_j}|}^2+...{|\frac{x_{i,96}}{{\mathrm{Cap}}_i}-\frac{x_{j,96}}{{\mathrm{Cap}}_j}|}^2}---\left(4\right)$

Wherein set i as plan day, j is history day, then Xi (xi, 1, xi, 2 ... .xi, 96), Xj (xj, 1, xj, 2 ... .xj, 96) i is expressed as, the j wind power output sequence of two days, Capi, Capj represent i respectively, the installed capacity of j sampling unit on the two, criterion Euclidean distance d _i,jrepresent the similitude of ratio in mean geometrical distance of i, the j wind power output of two days and the installed capacity of the sampling unit of history;

(2) what provide similarity determines threshold value Z, determines threshold value Z at [0, max{d _i,j] in interval, i is plan day, j ∈ { all history dates };

(3) that lists according to (2) determines threshold value Z, if d _i,j<Z, then the power output sequence of history day and the prediction of corresponding plan day sequence of exerting oneself is set of metadata of similar data, interval as calculating historical sample data for subsequent use.

3. reserve capacity analytical method under wind energy according to claim 1 and conventional energy resource coordinated dispatching mode, it is characterized in that: the method that period cluster is carried out in the historical sample data interval that step (1) obtains by described step (2) is specific as follows

(1) carry out hierarchical clustering to the day part power output sequence in historical sample data interval, obtain K division, a calculating K center divided, respectively as the initial cluster centre of clustering algorithm;

(2) for the power output sequence of each period, the Euclidean distance of the K asking it a to correspond to cluster centre, is grouped into the cluster at the shortest place, center of distance, is formed a new K cluster;

(3) for forming K new cluster, averaging method is utilized to recalculate the cluster centre of each cluster calculation;

(4) iteration is carried out in repetition step (2), (3), until class members no longer changes, then iteration terminates, and exports the similar cluster period.

4. reserve capacity analytical method under wind energy according to claim 1 and conventional energy resource coordinated dispatching mode, it is characterized in that: the prediction of step (3) described statistics each period exerts oneself the method for deviation profile for for period T, there is N number of time point, t _t0for the starting point of period T, t _tNfor the end point of period T, adjacent time point be spaced apart 15 minutes, the length of historical sample data is M days, then for period T, total M*N output deviation data, deviation data Normal Distribution, utilizes Maximum Likelihood Estimation Method, can obtain the expectation and variance of deviation profile according to formula (5), (6), the expectation estimation value of the deviation profile of predicated error is:

$\widetilde{\mathrm{\&mu;}}=\frac{1}{M*N}\&CenterDot;\underset{i=1}{\overset{M*N}{\mathrm{\&Sigma;}}}{\mathrm{wp}}_{t,i}^{\mathrm{error}}---\left(5\right)$

The estimate of variance of the deviation profile of predicated error is:

${\widetilde{\mathrm{\&sigma;}}}^2=\frac{1}{M*N}\&CenterDot;\underset{i=1}{\overset{M*N}{\mathrm{\&Sigma;}}}{({\mathrm{wp}}_{t,i}^{\mathrm{error}}-\widetilde{\mathrm{\&mu;}})}^2---\left(6\right)$

Wherein in formula (5), (6) for wind power prediction error amount.