A typical wind and solar output scene recognition method considering the comprehensive characteristics of source and load

Abstract

The present invention provides a method for identifying typical wind and solar power output scenarios that considers the comprehensive characteristics of source and load, including: obtaining the original load fluctuation sequence, wind and solar power output fluctuation sequence, and net load fluctuation sequence of each daily analysis condition; determining the daily electricity index and the daily regulation index, and then obtaining the daily source-load comprehensive analysis index; performing multi-level classification on the wind and solar power output sequence, and extracting wind and solar power output scenarios under different guarantee rates. The comprehensive source-load characteristic index that takes into account the characteristics of the source end and the load end proposed in the present invention can fully reflect the changing law and internal relationship between the source and the load; the typical scenarios obtained under each guarantee rate can quantitatively evaluate the supporting and regulating role of each wind and solar power output scenario in the power system. Based on the scene extraction results, the formulation of future horizontal annual power supply planning plans can be more accurately guided to ensure the safe, reliable and green economy of future power system development.

Description

Technical Field

[0001] The present invention belongs to the technical field of power planning, and in particular relates to a method for identifying typical wind and solar power output scenarios taking into account comprehensive source-load characteristics. Background Art

[0002] In recent years, as the scale and proportion of wind and solar power generation in the power system continue to rise, the power system has gradually shifted from a structure dominated by water and fire to a structure dominated by new energy. However, in the existing technology, due to the large uncertainty of wind power and photovoltaic output, which is directly affected by meteorological factors, large-scale wind power and photovoltaic grid connection brings great pressure to the dispatching and operation of the power system. When planning power supply for future horizontal years, since it is impossible to accurately predict the wind and solar output scene, a large number of wind and solar output scenes are generally generated through statistical probability methods such as sampling and Markov chain, and then the future horizontal annual power supply planning is guided based on the wind and solar output scene. This method has the following problems: affected by the value of statistical probability parameters and the basic sample size, the obtained wind and solar output scene may deviate from the law of wind and solar power generation; starting only from the law and characteristics of wind and solar output, the correlation between load and wind and solar in the power system is ignored, so the extracted wind and solar scene is only a typical source characteristic, not a typical scene in the system, which in turn affects the safety, reliability and low-carbon economy of future horizontal annual power supply planning. Summary of the invention

[0003] In view of the defects of the prior art, the present invention provides a method for identifying typical wind and solar power output scenarios taking into account the comprehensive characteristics of source and load, which can effectively solve the above problems.

[0004] The technical solution adopted by the present invention is as follows:

[0005] The present invention provides a method for identifying typical wind and solar power output scenes taking into account the comprehensive characteristics of sources and loads, comprising the following steps:

[0006] Step S1, determine K daily analysis conditions; obtain the original load fluctuation sequence of each daily analysis condition k Wind and solar power output fluctuation sequence and net load fluctuation series Where k = 1, 2, ..., K;

[0007] Step S2: Analyze the original load fluctuation sequence of working condition k on each day Net load fluctuation sequence and wind and solar output fluctuation sequence Comprehensive analysis is performed to obtain the daily power index ED reflecting the original load and wind and solar output characteristics of the daily analysis condition k. k and daily adjustment index FD k ;

[0008] Step S3: according to the daily power indicator EDk From large to small, for K wind and solar power output fluctuation sequences Perform the first sorting to obtain the first sorting sequence, and obtain each wind and solar power output fluctuation sequence The first sorting number Num_ED k ;

[0009] According to the daily adjustment index FD k From large to small, for K wind and solar power output fluctuation sequences Perform the second sorting to obtain the second sorting sequence, and obtain each wind and solar power output fluctuation sequence The second sorting number Num_FD k ;

[0010] Using formula (1), we can get the daily source-load comprehensive analysis index AD representing the daily analysis condition k characteristics: k :

[0011] AD k =Num_ED k *Num_FD k (1)

[0012] Step S4, using daily power indicator ED k As a first-level classification indicator, the daily adjustment indicator FD k As a secondary classification indicator, for K wind and solar power output fluctuation sequences Classify the same type of wind and solar power output fluctuation sequence is divided into a wind-solar output set to form an element of the wind-solar output set, thereby obtaining X wind-solar output sets, which are respectively expressed as: wind-solar output set C 1 ,C 2 ,...,C X ;

[0013] Step S5, setting the wind and solar output characteristics; taking the wind and solar output characteristics similarity as the goal, for each wind and solar output set C j Perform cluster analysis on the elements in , j = 1, 2, ..., X, and aggregate the elements with similar wind and solar output characteristics into one cluster, thus forming Y clusters;

[0014] Since there are X wind and solar power output sets, a total of X*Y clusters are obtained, which are represented as follows: Cluster G 1 ,G 2 ,...,G X*Y ;

[0015] Step S6: for each cluster G i , i=1,2,...,X*Y, which has multiple wind and solar output fluctuation sequences with similar source-load comprehensive characteristics Represents a power output scenario, for cluster G i All wind and solar output fluctuation sequences Taking the average value, the wind and solar power output fluctuation sequence is clustered as G i Corresponding scene output scene i ;

[0016] For cluster G i All wind and solar power output fluctuation sequences Corresponding daily source load comprehensive analysis index AD k Take the average value to get the scene source load comprehensive analysis index CP i , representing the scene of the landscape output i Comprehensive analysis characteristics of source and load;

[0017] Step S7, thereby identifying X*Y wind and solar output scenes, which are: Scene 1 ,Scene 2 ,...,Scene X*Y , the corresponding scene source load comprehensive analysis index is: CP 1 ,CP 2 ,...,CP X*Y ;

[0018] According to the scene source-load comprehensive analysis index from small to large, X*Y wind and solar output scenes are analyzed. 1 ,Scene 2 ,...,Scene X*Y Sort by scene i , let its sorting number be SN(Scene i ), which has a series of wind and solar power fluctuations The number of scenes is Num(Scene i ), then the calculated sorting sequence number is 1 to SN (Scene i ) has a sequence of wind and solar output fluctuations in all wind and solar output scenarios The total number is expressed as: Nt; using formula (2), we get its guaranteed rate Prob(Scene i ):

[0019]

[0020] In this way, a variety of wind and solar power output scenarios are identified, and the guarantee rate of each wind and solar power output scenario is obtained.

[0021] Preferably, step S1 specifically comprises:

[0022] Step S1.1, determine a horizontal year and N historical years; determine that the number of days in each of the horizontal year and the N historical years is m; and the number of time periods t in each day is w;

[0023] Step S1.2: Determine the power plant equipment parameters of the study area in the horizontal year and the original load fluctuation sequence on the dth day of the horizontal year Where, d = 1, 2, ..., m;

[0024] Step S1.3: Collect the meteorological data fluctuation sequence of the study area on the dth day of each historical year u Where, u=1,2,...,N;

[0025] Step S1.4, based on the meteorological data fluctuation sequence of the dth day of historical year u and the power plant equipment parameters of the horizontal year, the wind power output fluctuation sequence corresponding to the dth day of the historical year u is obtained and photovoltaic output fluctuation series

[0026] Step S1.5: calculate the wind power output fluctuation sequence corresponding to the dth day of historical year u and photovoltaic output fluctuation series Superposition is performed to obtain the wind and solar power output fluctuation sequence

[0027] Step S1.6: The original load fluctuation sequence of the study area on the dth day of the horizontal year is The wind and solar power output fluctuation sequence on the dth day of historical year u Subtract them to get the net load fluctuation sequence corresponding to the dth day of historical year u

[0028] The original load fluctuation sequence corresponding to the d-th day of historical year u is the original load fluctuation sequence of the d-th day of the horizontal year.

[0029] Since there are N historical years, each year has m days, there are N*m days in total, and each day has a corresponding daily analysis condition, including: Wind and solar output fluctuation sequence Net load fluctuation sequence and the original load fluctuation series

[0030] Therefore, N*m is K in step S1; thereby, K daily analysis conditions are obtained; the K daily analysis conditions are arbitrarily sorted and numbered starting from 1 to obtain the K daily analysis conditions determined in step S1.

[0031] Preferably, the original load fluctuation sequence Net load fluctuation sequence and wind and solar output fluctuation sequence They are all fluctuation sequences of time period t on the d-th day of the corresponding historical year u.

[0032] Preferably, step S1.4 is specifically:

[0033] Wind power output fluctuation sequence corresponding to the dth day of historical year u Specifically, it is the wind power output value of each period t on the dth day of historical year u The formed sequence is expressed as:

[0034] Photovoltaic output fluctuation sequence corresponding to the dth day of historical year u Specifically, it is the photovoltaic output value of each period t on the dth day of historical year u The formed sequence is expressed as:

[0035] and We can obtain the following equations using formula (3) and (4) respectively:

[0036]

[0037]

[0038] Where: ρ represents the air density; v t represents the wind speed value of time period t on the dth day of historical year u; T c represents the PV array panel temperature at time t on the dth day of historical year u; I T represents the total irradiance intensity of the photovoltaic array slope at time t on the dth day of historical year u; ρ, v t , T c and I T , forming meteorological data;

[0039] S represents the swept area of ​​the wind turbine; η s represents the photoelectric conversion efficiency of the photovoltaic array under standard test conditions; α represents the temperature coefficient; K 1 Represents the photovoltaic array aging loss coefficient; K 2 Represents the photovoltaic array mismatch loss coefficient; K 3 represents the dust shielding loss coefficient; K 4 Represents the DC circuit line loss coefficient; S, η s ,α,K 1 , K 2 , K 3 and K 4 , constituting the power station equipment parameters.

[0040] Preferably, in step S2, the daily electricity indicator ED k The calculation method is:

[0041] (1) Original load fluctuation sequence of daily analysis condition k Specifically, there are w original loads in each time period t Composition; where w is the number of time periods t per day; the wind and solar output fluctuation sequence of daily analysis condition k Specifically, it consists of w wind and solar power output values ​​in each period t composition;

[0042] (2) Using formula (5), we can get the daily electricity consumption index ED k :

[0043]

[0044] Where: T is the length of time for each period t.

[0045] Preferably, the daily adjustment index FD k The calculation method is:

[0046] (1) Original load fluctuation sequence of daily analysis condition k Specifically, there are w original loads in each time period t Composition; where w is the number of time periods t per day; take w original loads The maximum value in and minimum value Using formula (6), we can get the original load peak load demand FD_load of daily analysis condition k: k :

[0047]

[0048] (2) Net load fluctuation sequence of daily analysis condition k Specifically, there are w net loads in each time period t Composition; take w net loads The maximum value in and minimum value Using formula (7), we can get the net load peak load demand FD_net of daily analysis condition k: k :

[0049]

[0050] (3) Using formula (8), we can get the daily adjustment index FD k :

[0051] FD k =FD_load k-FD_net k (8)

[0052] The daily adjustment index FD is obtained k .

[0053] Preferably, step S4 is specifically:

[0054] Step S4.1, for K wind and solar power output fluctuation sequences There are K daily electricity indicators ED corresponding to k , select K daily electricity indicators ED k The maximum and minimum values ​​in the data are taken as the two end points, and five daily power index intervals are obtained by using the equal distance binning method; then the wind and solar power output fluctuation sequence is into the corresponding daily power index interval, thus converting the K wind and solar power output fluctuation sequences Divided into five sets;

[0055] Step S4.2: for each of the five sets, obtain the wind and solar power output fluctuation sequence contained in the set Daily adjustment index FD k Then, each set is divided into five subsets by using equal distance binning. A total of 25 subsets are obtained, each of which is a wind and solar power output set C in step S4. j .

[0056] Preferably, in step S5, each wind and solar power output fluctuation sequence The wind and solar power output characteristics, specifically the wind and solar power output fluctuation sequence The maximum output of each period t in the daily analysis condition k and minimum effort This constitutes the wind and solar output characteristic array

[0057] Taking the wind and solar power output feature array as the clustering feature, the k-means clustering method in the machine learning library is used, and the number of clustering targets is set to Y. For each wind and solar power output set C j The elements in the cluster analysis are clustered and the wind and solar power output set C j The elements in are divided into Y clusters.

[0058] The method for identifying typical wind and solar power output scenes taking into account the comprehensive characteristics of sources and loads provided by the present invention has the following advantages:

[0059] The source-load comprehensive characteristic index that takes into account the characteristics of both the source end and the load end proposed in the present invention can fully reflect the changing rules and internal relationships of the source end and the load end, and can quantitatively and accurately reflect the regulatory support role of different wind and solar output scenarios on the power system through the guarantee rate. The larger the guarantee rate, the higher the dispatching pressure and accident risk of the power system under the wind and solar output scenario, and the power system planned for the wind and solar output scenario has higher robustness. Taking the guarantee rate of each wind and solar output scenario as the evaluation basis, the supporting and regulating role of each wind and solar output scenario in the power system can be quantitatively evaluated, thereby more accurately guiding the formulation of future horizontal annual power supply planning plans, and ensuring the safety, reliability and green economy of future power system development. BRIEF DESCRIPTION OF THE DRAWINGS

[0060] Figure 1 The overall flow chart of a typical wind and solar output scene identification method considering the comprehensive characteristics of source and load provided by the present invention;

[0061] Figure 2 It is a diagram of the generation process of the daily original load fluctuation sequence, the net load fluctuation sequence and the wind and solar output fluctuation sequence in the daily analysis working condition provided by the present invention;

[0062] Figure 3 It is a schematic diagram of the wind and solar power output scene extraction results provided by the present invention. DETAILED DESCRIPTION

[0063] In order to make the technical problems, technical solutions and beneficial effects solved by the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0064] See also Figure 1 The present invention provides a method for identifying typical wind and solar power output scenes taking into account the comprehensive characteristics of source and load, comprising the following steps:

[0065] Step S1, determine K daily analysis conditions; obtain the original load fluctuation sequence of each daily analysis condition k Wind and solar power output fluctuation sequence and net load fluctuation series Where k = 1, 2, ..., K;

[0066] See also Figure 2 , step S1 is specifically as follows:

[0067] Step S1.1, determine a horizontal year and N historical years; determine that the number of days in each of the horizontal year and the N historical years is m; the number of time periods t in each day is w; in actual application, the number of days m in each year is 365 days; the time period t is 1 hour, therefore, the number of time periods t in each day is w is 24; therefore, the number of time periods t in each year is 8760. The number of historical years N is set according to actual conditions, for example, 10 to 20 years, and the present invention does not limit the number of historical years N.

[0068] Step S1.2: Determine the power plant equipment parameters of the study area in the horizontal year and the original load fluctuation sequence on the dth day of the horizontal year Where, d = 1, 2, ..., m; original load fluctuation sequence Specifically, it is the fluctuation sequence of time period t on the dth day.

[0069] Step S1.3: Collect the meteorological data fluctuation sequence of the study area on the dth day of each historical year u Where, u=1,2,...,N;

[0070] Step S1.4, based on the meteorological data fluctuation sequence of the dth day of historical year u and the power plant equipment parameters of the horizontal year, the wind power output fluctuation sequence corresponding to the dth day of the historical year u is obtained and photovoltaic output fluctuation series

[0071] Among them, the wind power output fluctuation sequence corresponding to the dth day of historical year u is Specifically, it is the wind power output value of each period t on the dth day of historical year u The formed sequence is expressed as:

[0072] Photovoltaic output fluctuation sequence corresponding to the dth day of historical year u Specifically, it is the photovoltaic output value of each period t on the dth day of historical year u The formed sequence is expressed as:

[0073] and The wind energy conversion formula of formula (3) and the photoelectric conversion formula of formula (4) are used to obtain:

[0074]

[0075]

[0076] Where: ρ represents the air density, kg / m 3 ;vt represents the wind speed value of time period t on the dth day of historical year u; T c represents the PV array panel temperature at time t on the dth day of historical year u, ℃; I T Represents the total irradiance intensity of the photovoltaic array slope at time t on the dth day of historical year u, W / m 2 ; ρ, v t , T c and I T , forming meteorological data;

[0077] S represents the wind swept area of ​​the wind turbine, m 2 ; η s Represents the photoelectric conversion efficiency of the photovoltaic array under standard test conditions; α represents the temperature coefficient, ℃ -1 ; K 1 Represents the photovoltaic array aging loss coefficient; K 2 Represents the photovoltaic array mismatch loss coefficient; K 3 represents the dust shielding loss coefficient; K 4 Represents the DC circuit line loss coefficient; S, η s ,α,K 1 , K 2 , K 3 and K 4 , constituting the power station equipment parameters.

[0078] Step S1.5: calculate the wind power output fluctuation sequence corresponding to the dth day of historical year u and photovoltaic output fluctuation series Superposition is performed to obtain the wind and solar power output fluctuation sequence Wind and solar power output fluctuation sequence Specifically, it is the fluctuation sequence of time period t on the dth day of historical year u;

[0079] Step S1.6: The original load fluctuation sequence of the study area on the dth day of the horizontal year is The wind and solar power output fluctuation sequence on the dth day of historical year u Subtract them to get the net load fluctuation sequence corresponding to the dth day of historical year u Therefore, the net load fluctuation sequence Specifically, it is the fluctuation sequence of time period t on the d-th day of the corresponding historical year u.

[0080] The original load fluctuation sequence corresponding to the d-th day of historical year u is the original load fluctuation sequence of the d-th day of the horizontal year.

[0081] Since there are N historical years, each year has m days, there are N*m days in total, and each day has a corresponding daily analysis condition, including: Wind and solar output fluctuation sequence Net load fluctuation sequence and the original load fluctuation series

[0082] Therefore, N*m is K in step S1; thereby, K daily analysis conditions are obtained; the K daily analysis conditions are arbitrarily sorted and numbered starting from 1 to obtain the K daily analysis conditions determined in step S1.

[0083] Step S2: Analyze the original load fluctuation sequence of working condition k on each day Net load fluctuation sequence and wind and solar output fluctuation sequence Comprehensive analysis is performed to obtain the daily power index ED reflecting the original load and wind and solar output characteristics of the daily analysis condition k. k and daily adjustment index FD k ;

[0084] Daily power indicator ED k The calculation method is:

[0085] (1) Original load fluctuation sequence of daily analysis condition k Specifically, there are w original loads in each time period t Composition; where w is the number of time periods t per day; the wind and solar output fluctuation sequence of daily analysis condition k Specifically, it consists of w wind and solar power output values ​​in each period t composition;

[0086] (2) Using formula (5), we can get the daily electricity consumption index ED k :

[0087]

[0088] Where: T is the length of time for each period t.

[0089] Therefore, in the present invention, the daily electricity index ED k Represents the ratio of the total daily wind and solar output to the total daily original load, reflecting the ability of supply to meet demand. The daily electricity index ED k The larger it is, the greater the proportion of wind and solar power in the total day, which means the better the wind and solar power support effect on that day.

[0090] Daily Adjustment Index FD k The calculation method is:

[0091] (1) Original load fluctuation sequence of daily analysis condition k Specifically, there are w original loads in each time period t Composition; where w is the number of time periods t per day; take w original loads The maximum value in and minimum value Using formula (6), we can get the original load peak load demand FD_load of daily analysis condition k: k :

[0092]

[0093] (2) Net load fluctuation sequence of daily analysis condition k Specifically, there are w net loads in each time period t Composition; take w net loads The maximum value in and minimum value Using formula (7), we can get the net load peak load demand FD_net of daily analysis condition k: k :

[0094]

[0095] (3) Using formula (8), we can get the daily adjustment index FD k :

[0096] FD k =FD_load k -FD_net k (8)

[0097] The daily adjustment index FD is obtained k .

[0098] Therefore, in the present invention, the daily adjustment index FD k The difference between the original load peak demand and the net load peak demand of the day reflects the degree of change in the regulation capacity of the power system caused by wind and solar. The daily regulation index FD k The larger the value, the better the indicator. The daily adjustment indicator FD k When it is positive, it means that it has a supplementary effect on the regulation demand of the power system; the daily regulation index FD k When it is negative, it means that it further increases the regulation demand of the power system.

[0099] Step S3: according to the daily power indicator ED k From large to small, for K wind and solar power output fluctuation sequences Perform the first sorting to obtain the first sorting sequence, and obtain each wind and solar power output fluctuation sequence The first sorting number Num_ED k ;

[0100] According to the daily adjustment index FD k From large to small, for K wind and solar power output fluctuation sequences Perform the second sorting to obtain the second sorting sequence, and obtain each wind and solar power output fluctuation sequence The second sorting number Num_FD k ;

[0101] Using formula (1), we can get the daily source-load comprehensive analysis index AD representing the daily analysis condition k characteristics: k :

[0102] AD k =Num_ED k *Num_FD k (1)

[0103] Specifically, due to the daily adjustment index FD k and daily electricity indicator ED k , which is positively correlated with the support and regulation effect of the wind and solar output on that day. Its daily power index ED k The sorting number and daily adjustment index FD k The sorting numbers of can reflect the ranking of their support and regulation effects in all K wind and solar power output fluctuation sequences. k The sorting number and daily adjustment index FD k The product of the sorting numbers is used as the daily source load comprehensive analysis indicator AD k , which can reflect the ranking of its supporting and regulating effects in all K wind and solar power output fluctuation sequences.

[0104] Since the present invention is based on the daily electricity index ED k and daily adjustment index FD k From large to small, the order of each wind and solar output fluctuation sequence So, for a certain wind and solar power output fluctuation sequence Its daily power index ED k The sorting number and daily adjustment index FD k The smaller the sorting number, the higher the daily source load comprehensive analysis index AD k The smaller the value is, the better the support and regulation effect is in all K wind and solar power output fluctuation sequences.

[0105] That is to say, in the present invention, for a certain wind and solar power output fluctuation sequence Its daily source load comprehensive analysis index AD k It is negatively correlated with the supporting and regulating effect of wind and solar power output.

[0106] In addition, in the present invention, for each of the K wind and solar power output fluctuation sequences Due to its daily source load comprehensive analysis index AD k In the process, the interaction between the original load and the wind and solar power output is considered at the same time. Therefore, it is a comprehensive source-load analysis indicator, which can fully reflect the role of wind and solar power in the power system, and is conducive to the accurate extraction of wind and solar power output scenarios in subsequent steps.

[0107] Step S4, using daily power indicator ED k As a first-level classification indicator, the daily adjustment indicator FD k As a secondary classification indicator, for K wind and solar power output fluctuation sequences Classify the same type of wind and solar power output fluctuation sequence is divided into a wind-solar output set to form an element of the wind-solar output set, thereby obtaining X wind-solar output sets, which are respectively expressed as: wind-solar output set C 1 ,C 2 ,...,C X ;

[0108] Step S4 is specifically as follows:

[0109] Step S4.1, for K wind and solar power output fluctuation sequences There are K daily electricity indicators ED corresponding to k , select K daily electricity indicators ED k The maximum and minimum values ​​in the data are taken as the two end points, and five daily power index intervals are obtained by using the equal distance binning method; then the wind and solar power output fluctuation sequence is into the corresponding daily power index interval, thus converting the K wind and solar power output fluctuation sequences Divided into five sets;

[0110] Step S4.2: for each of the five sets, obtain the wind and solar power output fluctuation sequence contained in the set Daily adjustment index FD k Then, each set is divided into five subsets by using equal distance binning. A total of 25 subsets are obtained, each of which is a wind and solar power output set C in step S4. j .

[0111] Step S5, setting the wind and solar output characteristics; taking the wind and solar output characteristics similarity as the goal, for each wind and solar output set C j Perform cluster analysis on the elements in , j = 1, 2, ..., X, and aggregate the elements with similar wind and solar output characteristics into one cluster, thus forming Y clusters;

[0112] Since there are X wind and solar power output sets, a total of X*Y clusters are obtained, which are represented as follows: Cluster G 1 ,G 2 ,...,G X*Y ;

[0113] Specifically, each wind and solar power output fluctuation sequence The wind and solar power output characteristics, specifically the wind and solar power output fluctuation sequence The maximum output of each period t in the daily analysis condition k and minimum effort This constitutes the wind and solar output characteristic array

[0114] Taking the wind and solar power output feature array as the clustering feature, the k-means clustering method in the machine learning library is used, and the number of clustering targets is set to Y. For each wind and solar power output set C j The elements in the cluster analysis are clustered and the wind and solar power output set C j The elements in are divided into Y clusters.

[0115] Step S6: for each cluster G i , i=1,2,...,X*Y, which has multiple wind and solar output fluctuation sequences with similar source-load comprehensive characteristics Represents a power output scenario, for cluster G i All wind and solar output fluctuation sequences Taking the average value, the wind and solar power output fluctuation sequence is clustered as G i Corresponding scene output scene i ;

[0116] For cluster G i All wind and solar power output fluctuation sequences Corresponding daily source load comprehensive analysis index AD k Take the average value to get the scene source load comprehensive analysis index CP i , representing the scene of the landscape output i Comprehensive analysis characteristics of source and load;

[0117] Step S7, thereby identifying X*Y wind and solar output scenes, which are: Scene 1 ,Scene 2 ,...,Scene X*Y , the corresponding scene source load comprehensive analysis index is: CP 1 ,CP 2 ,...,CP X*Y ;

[0118] According to the scene source-load comprehensive analysis index from small to large, X*Y wind and solar output scenes are analyzed. 1 ,Scene 2 ,...,Scene X*Y Sort by scene i , let its sorting number be SN(Scene i ), which has a series of wind and solar power fluctuations The number of scenes is Num(Scene i ), then the calculated sorting sequence number is 1 to SN (Scene i ) has a sequence of wind and solar output fluctuations in all wind and solar output scenarios The total number is expressed as: Nt; using formula (2), we get its guaranteed rate Prob(Scene i ):

[0119]

[0120] Thus, multiple wind and solar power output scenarios are identified, and the guaranteed rate of each wind and solar power output scenario is obtained. Figure 3 As shown, a corresponding wind and solar output scenario provided by the present invention is obtained when the guarantee rate is 50%. Figure 3 In the figure, 1 represents the photovoltaic output fluctuation sequence 2 represents the wind power output fluctuation sequence 3 represents the wind and solar power output fluctuation sequence 4 represents the original load fluctuation sequence 5 represents the net load fluctuation sequence

[0121] Therefore, in the present invention, for K wind and solar power output fluctuation sequences Daily power indicator ED k First-level classification, daily adjustment index FD k After secondary classification and clustering of wind and solar output feature arrays, the wind and solar output fluctuation sequences with very similar wind and solar output effects can be classified into The clusters are grouped into a cluster, which represents a wind-solar power output scenario, which is a typical wind-solar power output scenario obtained by combining the horizontal annual power supply planning and historical annual meteorological data.

[0122] And because of the daily source load comprehensive analysis index AD kIt is negatively correlated with the wind-solar output support and regulation effect. Therefore, the scene source-load comprehensive analysis index is negatively correlated with the wind-solar output support and regulation effect of the wind-solar output scene. In the present invention, each wind-solar output scene is sorted in order from small to large according to the scene source-load comprehensive analysis index. Therefore, the wind-solar output scene with a higher ranking has a better support and regulation effect and a lower guarantee rate. The source-load comprehensive characteristic index that takes into account the characteristics of the source end and the load end proposed in the present invention can fully reflect the change law and internal relationship of the source end and the load end, and can quantitatively and accurately reflect the regulatory support effect of different wind-solar output scenes on the power system through the guarantee rate. The larger the guarantee rate, the higher the dispatching pressure and accident risk of the power system under the wind-solar output scene, and the power system planned on the wind-solar output scene has higher robustness. Based on the guarantee rate of each wind-solar output scene as the evaluation basis, the support and regulation effect of each wind-solar output scene in the power system can be quantitatively evaluated, thereby more accurately guiding the formulation of future horizontal annual power supply planning plans, and ensuring the safety, reliability and green economy of future power system development.

[0123] The above is only a preferred embodiment of the present invention. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principle of the present invention. These improvements and modifications should also be considered as the scope of protection of the present invention.

Claims

1. A method for identifying typical wind and solar power output scenarios considering the comprehensive characteristics of source and load, characterized in that: The following steps are involved: Step S1, determine K daily analysis conditions; obtain the original load fluctuation sequence of each daily analysis condition k Wind and solar power output fluctuation sequence and net load fluctuation series Where k = 1, 2, ..., K; Step S2: Analyze the original load fluctuation sequence of working condition k on each day Net load fluctuation sequence and wind and solar output fluctuation sequence Comprehensive analysis is performed to obtain the daily power index ED reflecting the original load and wind and solar output characteristics of the daily analysis condition k. k and daily adjustment index FD k ; Step S3: according to the daily power indicator ED k From large to small, for K wind and solar power output fluctuation sequences Perform the first sorting to obtain the first sorting sequence, and obtain each wind and solar power output fluctuation sequence The first sorting number Num_ED k ; According to the daily adjustment index FD k From large to small, for K wind and solar power output fluctuation sequences Perform the second sorting to obtain the second sorting sequence, and obtain each wind and solar power output fluctuation sequence The second sorting number Num_FD k ; Using formula (1), we can get the daily source-load comprehensive analysis index AD representing the daily analysis condition k characteristics: k : TO k =Num_ED k *Num_FD k (1) Step S4, using daily power indicator ED k As a first-level classification indicator, the daily adjustment indicator FD k As a secondary classification indicator, for K wind and solar power output fluctuation sequences Classify the same type of wind and solar power output fluctuation sequence is divided into a wind-solar output set to form an element of the wind-solar output set, thereby obtaining X wind-solar output sets, which are respectively represented as: wind-solar output sets C1, C2, ..., C X ; Step S5, setting the wind and solar output characteristics; taking the wind and solar output characteristics similarity as the goal, for each wind and solar output set C j Perform cluster analysis on the elements in , j = 1, 2, ..., X, and aggregate the elements with similar wind and solar output characteristics into one cluster, thus forming Y clusters; Since there are X wind and solar power output sets, a total of X*Y clusters are obtained, which are represented as: clusters G1, G2, ..., G X*Y ; Step S6: for each cluster G i , i=1,2,...,X*Y, which has multiple wind and solar output fluctuation sequences with similar source-load comprehensive characteristics Represents a power output scenario, for cluster G i All wind and solar output fluctuation sequences Taking the average value, the wind and solar power output fluctuation sequence is clustered as G i Corresponding scene output scene i ; For cluster G i All wind and solar power output fluctuation sequences Corresponding daily source load comprehensive analysis index AD k Take the average value to get the scene source load comprehensive analysis index CP i , representing the scene of the landscape output i Comprehensive analysis characteristics of source and load; Step S7, thereby identifying X*Y wind and solar power output scenes, which are: Scene 1, Scene 2, ..., Scene X*Y , the corresponding scene source load comprehensive analysis indicators are: CP1, CP2,..., CP X*Y ; According to the scene source-load comprehensive analysis index from small to large, X*Y wind and solar output scenes Scene1, Scene2,..., Scene X*Y Sort by scene i , let its sorting number be SN(Scene i ), which has a series of wind and solar power fluctuations The number of scenes is Num(Scene i ), then the calculated sorting sequence number is 1 to SN (Scene i ) has a sequence of wind and solar output fluctuations in all wind and solar output scenarios The total number is expressed as: Nt; using formula (2), we get its guaranteed rate Prob(Scene i ): In this way, a variety of wind and solar power output scenarios are identified, and the guarantee rate of each wind and solar power output scenario is obtained.

2. According to claim 1, a method for identifying typical wind and solar output scenes considering the comprehensive characteristics of source and load, characterized in that: Step S1 is specifically as follows: Step S1.1, determine a horizontal year and N historical years; determine that the number of days in each of the horizontal year and the N historical years is m; and the number of time periods t in each day is w; Step S1.2: Determine the power plant equipment parameters of the study area in the horizontal year and the original load fluctuation sequence on the dth day of the horizontal year Where, d = 1, 2, ..., m; Step S1.3: Collect the meteorological data fluctuation sequence of the study area on the dth day of each historical year u Where, u=1,2,...,N; Step S1.4, based on the meteorological data fluctuation sequence of the dth day of historical year u and the power plant equipment parameters of the horizontal year, the wind power output fluctuation sequence corresponding to the dth day of the historical year u is obtained and photovoltaic output fluctuation series Step S1.5: calculate the wind power output fluctuation sequence corresponding to the dth day of historical year u and photovoltaic output fluctuation series Superposition is performed to obtain the wind and solar power output fluctuation sequence Step S1.6: The original load fluctuation sequence of the study area on the dth day of the horizontal year is The wind and solar power output fluctuation sequence on the dth day of historical year u Subtract them to get the net load fluctuation sequence corresponding to the dth day of historical year u The original load fluctuation sequence corresponding to the d-th day of historical year u is the original load fluctuation sequence of the d-th day of the horizontal year. Since there are N historical years, each year has m days, there are N*m days in total, and each day has a corresponding daily analysis condition, including: Wind and solar output fluctuation sequence Net load fluctuation sequence and the original load fluctuation series Therefore, N*m is K in step S1; thereby, K daily analysis conditions are obtained; the K daily analysis conditions are arbitrarily sorted and numbered starting from 1 to obtain the K daily analysis conditions determined in step S1.

3. The method for identifying typical wind and solar power output scenes considering the comprehensive characteristics of source and load according to claim 2 is characterized in that: Original load fluctuation sequence Net load fluctuation sequence and wind and solar output fluctuation sequence They are all fluctuation sequences of time period t on the d-th day of the corresponding historical year u.

4. The method for identifying typical wind and solar power output scenes considering the comprehensive characteristics of source and load according to claim 3 is characterized in that: Step S1.4 is specifically as follows: Wind power output fluctuation sequence corresponding to the dth day of historical year u Specifically, it is the wind power output value of each period t on the dth day of historical year u The formed sequence is expressed as: Photovoltaic output fluctuation sequence corresponding to the dth day of historical year u Specifically, it is the photovoltaic output value of each period t on the dth day of historical year u The formed sequence is expressed as: and We can obtain the following equations using formula (3) and (4) respectively: Where: ρ represents the air density; v t represents the wind speed value of time period t on the dth day of historical year u; T c represents the PV array panel temperature at time t on the dth day of historical year u; I T represents the total irradiance intensity of the photovoltaic array slope at time t on the dth day of historical year u; ρ, v t , T c and I T , forming meteorological data; S represents the swept area of ​​the wind turbine; η s represents the photovoltaic conversion efficiency of the photovoltaic array under standard test conditions; α represents the temperature coefficient; K1 represents the photovoltaic array aging loss coefficient; K2 represents the photovoltaic array mismatch loss coefficient; K3 represents the dust shielding loss coefficient; K4 represents the DC circuit line loss coefficient; S, η s , α, K1, K2, K3 and K4 constitute the power station equipment parameters.

5. The method for identifying typical wind and solar power output scenes considering the comprehensive characteristics of source and load according to claim 4 is characterized in that: In step S2, the daily power indicator ED k The calculation method is: (1) Original load fluctuation sequence of daily analysis condition k Specifically, there are w original loads in each time period t Composition; where w is the number of time periods t per day; the wind and solar output fluctuation sequence of daily analysis condition k Specifically, it consists of w wind and solar power output values ​​in each period t composition; (2) Using formula (5), we can get the daily electricity consumption index ED k : Where: T is the length of time for each period t.

6. The method for identifying typical wind and solar power output scenes considering the comprehensive characteristics of source and load according to claim 4, characterized in that: Daily Adjustment Index FD k The calculation method is: (1) Original load fluctuation sequence of daily analysis condition k Specifically, there are w original loads in each time period t Composition; where w is the number of time periods t per day; take w original loads The maximum value in and minimum value Using formula (6), we can get the original load peak load demand FD_load of daily analysis condition k: k : (2) Net load fluctuation sequence of daily analysis condition k Specifically, there are w net loads in each time period t Composition; take w net loads The maximum value in and minimum value Using formula (7), we can get the net load peak load demand FD_net of daily analysis condition k: k : (3) Using formula (8), we can get the daily adjustment index FD k : FD k =FD_load k -FD_net k (8) The daily adjustment index FD is obtained k .

7. The method for identifying typical wind and solar power output scenes considering the comprehensive characteristics of source and load according to claim 1, characterized in that: Step S4 is specifically as follows: Step S4.1, for K wind and solar power output fluctuation sequences There are K daily electricity indicators ED corresponding to k , select K daily electricity indicators ED k The maximum and minimum values ​​in the data are taken as the two end points, and five daily power index intervals are obtained by using the method of equal distance binning; then the fluctuation sequence of each wind and solar power output is into the corresponding daily power index interval, thus converting the K wind and solar power output fluctuation sequences Divided into five sets; Step S4.2: for each of the five sets, obtain the wind and solar power output fluctuation sequence contained in the set Daily adjustment index FD k Then, each set is divided into five subsets by using equal distance binning. A total of 25 subsets are obtained, each of which is a wind and solar power output set C in step S4. j .

8. The method for identifying typical wind and solar power output scenes considering the comprehensive characteristics of source and load according to claim 1, characterized in that: In step S5, each wind and solar power output fluctuation sequence The wind and solar power output characteristics, specifically the wind and solar power output fluctuation sequence The maximum output of each period t in the daily analysis condition k and minimum effort This constitutes the wind and solar output characteristic array Taking the wind and solar power output feature array as the clustering feature, the k-means clustering method in the machine learning library is used, and the number of clustering targets is set to Y. For each wind and solar power output set C j The elements in the cluster analysis are clustered and the wind and solar power output set C j The elements in are divided into Y clusters.

Images