Distributed resource time sequence scene generation method

Abstract

The invention discloses a distributed resource time sequence scene generation method, and belongs to the technical field of distributed energy access power grid analysis, and the method comprises the steps: obtaining historical operation data, meteorological historical data and meteorological prediction data of distributed photovoltaic and loads in a to-be-analyzed region; generating a historical sample day feature vector and a scene to-be-generated day feature vector according to the acquired data, and determining a similar day set; according to the similar day set, obtaining historical operation data of similar day distributed photovoltaic and load in a to-be-analyzed area, analyzing power fluctuation interval distribution characteristics and power probability distribution characteristics of the historical operation data, and obtaining fluctuation upper and lower bounds and a power probability density function of the distributed photovoltaic and load; and based on the fluctuation upper and lower bounds of the distributed photovoltaic and load and the power probability density function, obtaining a time sequence power scene set of the distributed photovoltaic and load considering fluctuation interval correction. According to the invention, historical power information and fluctuation information can be fully utilized to generate a high-precision uncertainty scene.

Description

Technical Field

[0001] The present invention relates to a scenario uncertainty analysis technology for distributed resources, belongs to the technical field of robust scheduling optimization of microgrids with high-proportion distributed energy access, and particularly relates to a distributed resource time series scenario generation method. Background Art

[0002] As a large number of distributed power sources are integrated into the distribution network and microgrid, their inherent randomness and fluctuations will intensify the uncertainty in the operation and regulation of microgrids. At the same time, due to the introduction of load regulation measures such as demand response, the traditional source-load structure and deterministic optimization methods can no longer reflect the role of distribution networks and microgrids in day-ahead scheduling, which has changed from traditional energy consumers to energy producers and consumers.

[0003] Therefore, using time series scenarios to quantify the typical operating conditions of loads and photovoltaics in microgrids can provide valuable boundary constraints for system operation. At the same time, the uncertainty information driven by scenarios should be effectively utilized to reduce the overly conservative tendency of traditional robust optimization while ensuring system stability, and make full use of the flexibility of distributed resources in the system. Summary of the invention

[0004] The purpose of the present invention is to provide a distributed resource time series scenario generation method, which can fully utilize historical power information and fluctuation information to generate high-precision uncertainty scenarios. The technical solution adopted by the present invention is as follows.

[0005] In one aspect, the present invention provides a method for generating a distributed resource time sequence scenario, comprising:

[0006] Obtain the historical operation data, historical meteorological data and meteorological forecast data of distributed photovoltaic and load in the area to be analyzed;

[0007] Generate historical sample day feature vectors and scenario day feature vectors to be generated based on the acquired data, and determine a similar day set;

[0008] According to the similar day set, the historical operation data of distributed photovoltaic and load on similar days in the area to be analyzed are obtained, the power fluctuation interval distribution characteristics and power probability distribution characteristics are analyzed, and the upper and lower bounds of power fluctuation and power probability density function of distributed photovoltaic and load are obtained;

[0009] Based on the power fluctuation upper and lower bounds and power probability density function of the distributed photovoltaic and load, a set of distributed photovoltaic and load time series power scenarios considering fluctuation interval correction is obtained.

[0010] Optionally, the meteorological characteristic data types in the meteorological historical data and the meteorological forecast data include: maximum temperature, minimum temperature, average temperature, precipitation, relative humidity, average irradiance, average wind speed, and average air pressure;

[0011] The forming of the historical sample day feature vector and the scene to-be-generated day feature vector according to the acquired data includes:

[0012] Calculate the daily average photovoltaic power and load power based on the historical operating data of distributed photovoltaic and load;

[0013] Calculate the Spearman correlation coefficients of the daily average photovoltaic power and load power and various types of historical meteorological characteristic data series respectively;

[0014] For each type of meteorological characteristic data, the corresponding Spearman correlation coefficient calculation result is compared with the correlation coefficient threshold, and the characteristic vector type is selected according to the comparison result;

[0015] According to the selected feature vector type, the feature vectors of each historical sample day and the day to be generated are obtained, which are expressed as:

[0016]

[0017] In the formula, Indicates the number of historical sample days, The feature vector representing the day of the scene to be generated, Indicates Historical daily sample feature vectors, Indicates The mth eigenvalue in the eigenvector of the historical daily samples.

[0018] In the above technical solution, the correlation coefficient threshold can be set or adjusted according to experience or needs, and the meteorological feature data type whose Spearman correlation coefficient calculation result is not less than the correlation coefficient threshold is selected to generate the feature vector.

[0019] Optionally, determining the similar day set includes:

[0020] Based on the feature vectors of each historical sample day and the day for the scenario to be generated, the comprehensive correlation coefficient between the meteorological characteristics of the day for the scenario to be generated and the meteorological characteristics of each historical sample is calculated;

[0021] The historical samples whose comprehensive correlation coefficient is higher than the comprehensive correlation coefficient threshold and the day for the scenario to be generated are combined into a similar day set.

[0022] Optionally, the calculation formula of the comprehensive correlation coefficient is:

[0023]

[0024] in, For the The comprehensive correlation coefficient between the historical samples and the meteorological characteristics of the day when the scene is to be generated, For the The geometric correlation between the historical samples and the day when the scene is to be generated, For the The distance correlation between the historical samples and the day when the scene is to be generated is expressed as:

[0025] ,

[0026] ;

[0027] Among them, the intermediate variable , .

[0028] Optionally, the historical operating data of distributed photovoltaic and load on similar days in the analysis area are analyzed to analyze the distribution characteristics of power fluctuation range, including:

[0029] For the distributed photovoltaic and load time series data of each historical sample day in the similar day set, the data fluctuation value of each period is calculated to obtain the fluctuation range of the distributed photovoltaic and load data corresponding to each period, which is expressed as:

[0030]

[0031] in, ,

[0032] In the formula, is the fluctuation value of distributed photovoltaic or load data in time period t, is the power fluctuation value from the start time to the end time of time period t on the dth day, T is the sampling time interval of adjacent time series data, and S is the set of similar days.

[0033] Optionally, the historical operating data of distributed photovoltaic and load on similar days in the analysis area are analyzed to analyze the power probability distribution characteristics, including:

[0034] Select the kernel function and its bandwidth;

[0035] For each data point on multiple similar days, the selected kernel function is used to estimate the local density of each data point;

[0036] Take a weighted average of the local densities of all data points to get a global density estimate;

[0037] Among them, for the distributed photovoltaic or load data set at any data point t on multiple similar days , the probability density function of the data set is estimated by the Gaussian kernel function:

[0038]

[0039] In the formula, It's on point The estimated density value at ; is the number of historical sample days; At each data point is the Gaussian kernel function applied at ; h is the kernel function bandwidth.

[0040] Optionally, based on the power fluctuation upper and lower bounds and power probability density function of the distributed photovoltaic and load, a set of distributed photovoltaic and load time series power scenarios taking into account fluctuation interval correction is obtained, including:

[0041] Based on the fluctuation range and power probability density function of distributed photovoltaic / load data corresponding to each time period, the probability density sampling correction space under each power scenario is calculated for each time period;

[0042] Simple random sampling is performed according to the probability density sampling correction space to obtain the photovoltaic / load power sampling value of each time period, and further obtain the time series power scenario set. The time series power scenario set includes multiple groups of power time series data subsets corresponding to different scenarios, and each power time series data subset includes power data corresponding to each time period.

[0043] Optionally, the calculating the probability density sampling correction space under each power scenario for each time period includes:

[0044] Determine the power value fluctuation range of time period t+1, expressed as:

[0045]

[0046] in, and Respectively represent the power sampling values ​​of time period t and time period t+1 under the sth power scenario;

[0047] Will Substitute the endpoint values ​​of the power fluctuation interval into the probability density function of the t+1 period to obtain the probability density sampling correction space of the t+1 period under the sth power scenario, which is expressed as:

[0048]

[0049] .

[0050] In the above technical solution, the sampling interval of each time series scenario will be corrected according to the fluctuation range of adjacent moments, and then random sampling will be performed within the corrected interval, thereby effectively simulating the fluctuation characteristics of photovoltaic load power within adjacent time intervals.

[0051] Optionally, the method further includes: taking the time series power scenario set as a time series original scenario set, forming a scenario-driven original uncertainty set according to the time series original scenario set, and using finite covering theory to solve several simplified uncertainty sets covering the original uncertain scenarios.

[0052] Optionally, the original uncertainty set driven by the scenario is expressed as:

[0053]

[0054] In the formula, and are the number of generation scenarios for distributed PV and load, respectively; , and , are the sampling value and average value of the distributed photovoltaic and load power scenarios in period t under the i-th time series scenario, is the original uncertainty set driven by the scenario, Γ is the budget parameter, indicating that there are at most Γ periods to reach the maximum uncertainty range, , Represent the sampling values ​​of distributed photovoltaic and load at time t respectively.

[0055] The finite covering theory is used to solve several simplified uncertainty sets covering the original uncertain scenario, including:

[0056] Based on the original uncertainty set driven by the scenario, the transformed optimization problem is obtained, which is expressed as:

[0057]

[0058] In the formula,

[0059] and are the upper and lower bounds of the power value of the scenario in the nth finite coverage set for the t-time period;

[0060] and is the upper and lower bounds of the scene at time t in the original scene set formed by the s-th sampling, that is: , ; is the balance factor; M is a very large number (big M method); It is a 0-1 variable that guides the mapping of the original scene set of the time series to the finite coverage scene set N; represents the balance factor; represents a finite covering set;

[0061] Solve the optimization problem to obtain an uncertain set that can cover the scenario drive , n box-like uncertain sets ,Right now: .

[0062] The objective function of the above optimization problem consists of two items. The first item is to minimize the total size of all uncertainty sets, and the second item is to minimize the size of the largest uncertainty set. The balance factor It is used to balance the weights of the first and second terms in the MILP objective function. The first term controls the complexity of the uncertainty set, and the second term can balance the size of a single uncertainty set, reducing the running cost of considering the uncertainty set in subsequent robust optimization.

[0063] In a second aspect, the present invention provides a computer-readable storage medium having a computer program stored thereon. When the computer program is executed by a processor, the distributed resource timing scenario generation method as described in the first aspect is implemented.

[0064] Beneficial Effects

[0065] The distributed resource time series scenario generation method of the present invention, on the one hand, takes into account the use of fluctuation information of historical data to reduce the width of the generated scenario, and proposes a scenario generation method based on fluctuation interval correction. By constructing a daily feature vector, historical data samples with a high correlation with the scene to be generated on the basis of a comprehensive correlation coefficient are screened out to form a similar day set, and based on the similar day set and distributed photovoltaic-load historical data, the source-load time series power probability density function is fitted and the historical fluctuation interval is calculated, and the upper and lower bounds of the fluctuation interval are used to correct the sampling interval of the probability density function based on the source-load power, thereby forming a day-ahead scenario set, thereby improving the quality of scenario generation.

[0066] On the other hand, based on the obtained day-ahead scenario set, the present invention defines a scenario-driven original uncertainty set, draws on the finite covering theory, covers a complex source-load scenario uncertainty set with a finite number of simpler uncertainty sets, and transforms the scenario-driven uncertainty set into multiple simple uncertainty sets for common coverage. This can reduce the conservatism of robust optimization modeling such as microgrid configuration and scheduling while ensuring the robustness of the system. BRIEF DESCRIPTION OF THE DRAWINGS

[0067] Figure 1 The figure is a schematic flow chart of a sampling method based on fluctuation interval correction in one embodiment of the method of the present invention;

[0068] Figure 2 Shown is a schematic flow chart of an embodiment of the method of the present invention. DETAILED DESCRIPTION

[0069] The invention is further described below with reference to the accompanying drawings and specific embodiments.

[0070] Example 1

[0071] This embodiment introduces a distributed resource time series scenario generation method suitable for day-ahead scheduling, which can generate a high-precision day-ahead scenario set based on the characteristics of historical similar days, and provide relatively accurate reference data for day-ahead scheduling. The method of this embodiment includes:

[0072] Obtain the historical operation data, historical meteorological data and meteorological forecast data of distributed photovoltaic and load in the area to be analyzed;

[0073] Generate historical sample day feature vectors and scenario day feature vectors to be generated based on the acquired data, and determine a similar day set;

[0074] According to the similar day set, the historical operation data of distributed photovoltaic and load on similar days in the area to be analyzed are obtained, the power fluctuation interval distribution characteristics and power probability distribution characteristics are analyzed, and the upper and lower bounds of power fluctuation and power probability density function of distributed photovoltaic and load are obtained;

[0075] Based on the power fluctuation upper and lower bounds and power probability density function of the distributed photovoltaic and load, a set of distributed photovoltaic and load time series power scenarios considering fluctuation interval correction is obtained.

[0076] The specific implementation of this embodiment includes the following parts.

[0077] 1. Formation of Similar Day Sets Based on Comprehensive Correlation Coefficient

[0078] This embodiment first obtains the historical operation data of distributed photovoltaic and load in the region, as well as the historical meteorological data and weather forecast, and then forms the historical sample day feature vector and the scene day feature vector based on the numerical weather forecast and historical sample meteorological data of the scene to be generated. By calculating the comprehensive meteorological correlation coefficient, samples with high comprehensive similarity are selected as the components of the similar day set to obtain the historical similar sample set.

[0079] Among them, the meteorological characteristic data types in the meteorological historical data and meteorological forecast data include: maximum temperature, minimum temperature, average temperature, precipitation, relative humidity, average irradiance, average wind speed, average air pressure, etc.

[0080] This section specifically includes the following steps:

[0081] Step 1: Construct the feature vector of the historical sample day and the feature vector of the scene to be generated day, including:

[0082] Calculate the daily average photovoltaic power / load power based on the historical operating data of distributed photovoltaic and load;

[0083] Calculate the Spearman correlation coefficient between the daily average photovoltaic power / load power and various types of historical meteorological characteristic data series;

[0084] For each type of meteorological characteristic data, the corresponding Spearman correlation coefficient calculation result is compared with the correlation coefficient threshold, and the characteristic vector type is selected according to the comparison result. The correlation coefficient threshold can be set or adjusted according to experience or needs, and the meteorological characteristic data type whose Spearman correlation coefficient calculation result is not less than the correlation coefficient threshold is selected to generate the characteristic vector;

[0085] According to the selected feature vector type, the feature vectors of each historical sample day and the day to be generated are obtained, which are expressed as:

[0086]

[0087] In the formula, Indicates the number of historical sample days, The feature vector representing the day of the scene to be generated, Indicates Historical daily sample feature vectors, Indicates The mth eigenvalue in the eigenvector of the historical daily samples.

[0088] Step 2, determining a set of similar days, including: based on the feature vectors of each historical sample day and the scene to be generated day, calculating the comprehensive correlation coefficient between the meteorological characteristics of the scene to be generated day and the meteorological characteristics of each historical sample; and combining the historical samples and the scene to be generated day whose comprehensive correlation coefficient is higher than the comprehensive correlation coefficient threshold into a similar day set. The calculation of the comprehensive correlation coefficient includes the following steps:

[0089] 1.2.1) Calculation of geometric association:

[0090] Intermediate variables

[0091]

[0092] In the formula, That is The geometric correlation between the historical samples and the day when the scene is to be generated.

[0093] 1.2.2) Distance correlation calculation:

[0094] Intermediate variables

[0095]

[0096] In the formula, That is the The distance correlation between the historical samples and the day when the scene is to be generated.

[0097] 1.2.3) Combine geometric correlation and distance correlation to calculate comprehensive correlation , which is used to describe the comprehensive correlation between the scene to be generated and the historical samples, and is expressed as:

[0098] .

[0099] When the comprehensive correlation coefficient between the feature vector of the i-th historical sample and the feature vector of the scene to be generated day is higher than a specific threshold, it is considered that there is a similarity effect, otherwise it is not similar. At this time, the historical samples similar to the scene to be generated day can be grouped into a similar day set.

[0100] 2. Volatility Analysis and Power Probability Density Modeling of Similar Days

[0101] According to the similar set of historical samples, the power fluctuation interval distribution characteristics and power probability distribution characteristics are analyzed to obtain the upper and lower bounds of power fluctuation of distributed photovoltaic and load. The kernel density estimation method is used to model the probability density fitting function of distributed photovoltaic and load power.

[0102] Among them, the analysis process of the upper and lower bounds of power fluctuations of distributed photovoltaics and loads is:

[0103] Referring to the day-ahead scheduling, a day is divided into 96 periods. In the set of similar days, the power fluctuation of sources and loads in adjacent periods can be expressed as:

[0104]

[0105]

[0106] In the formula, is the fluctuation value of distributed photovoltaic or load data in time period t, is the power fluctuation value from the start time to the end time of time period t on the dth day, T is the sampling time interval of adjacent time series data, and S is the set of similar days. At this time, the upper and lower bounds of the power fluctuation of distributed photovoltaic / load are obtained:

[0107] .

[0108] In the set of similar days, the probability density function modeling process of distributed photovoltaic power / load power in any period is:

[0109] 2.2.1) Select the kernel function, usually the Gaussian kernel function:

[0110]

[0111] The probability density function of the data set can be estimated by the Gaussian kernel function:

[0112]

[0113] In the formula, is the estimated density value at data point 𝑥, 𝑛 is the sample size, It is at data point 𝑥 i is the Gaussian kernel function applied at , and h is the bandwidth, which controls the width of each kernel function and the smoothness of the density estimate.

[0114] 2.2.2) Smoothing bandwidth. The width (bandwidth) of the kernel function controls the degree of smoothing. The larger the bandwidth, the smoother the estimated density curve; the smaller the bandwidth, the closer the curve will be to the local characteristics of the original data.

[0115] 2.2.3) Superimpose all data points. For each data point, use the kernel function to estimate the local density of the point, and then perform a weighted average of the local densities of all data points to form a global density estimate.

[0116] In this embodiment, for a plurality of distributed photovoltaic or load data sets at any data point t on similar days, , the probability density function of the data set is estimated by the Gaussian kernel function:

[0117]

[0118] In the formula, It's on point The estimated density value at ; is the number of historical sample days; At each data point is the Gaussian kernel function applied at ; h is the kernel function bandwidth.

[0119] 3. Generation of original scenarios of distributed photovoltaic and load considering fluctuation interval correction: Based on the historical probability distribution of power and historical fluctuation range of the distributed photovoltaic and load, the upper and lower bounds of the fluctuation interval at time t are used to correct the sampling interval of the distributed photovoltaic and load time series power at time t+1, and the original set of distributed photovoltaic and load time series power scenarios considering fluctuation interval correction is obtained.

[0120] This embodiment uses a simple random sampling method for each of the 96 time periods to generate a day-ahead scenario based on the probability modeling of photovoltaic / load power in each time period within a similar day set. In order to narrow the scene width, the sampling interval of each sampling time point is corrected based on the fluctuation amplitude of adjacent moments, and then random sampling is performed within the corrected power probability interval, which effectively simulates the fluctuation characteristics of photovoltaic load power in adjacent time periods. Specifically, when sampling the sth scene time period by time period, the adjacent power values ​​must satisfy:

[0121]

[0122] Will Substitution The power probability density sampling interval correction expression of time period t+1 is obtained:

[0123]

[0124]

[0125] At this point, according to the sampling results at the moment, the sampling interval of the power probability density function of the photovoltaic and load in the period t+1 is adjusted from [0,1] to [c1,c2], and the corrected photovoltaic / load power sampling value sequence can be further obtained by the simple random sampling method. The specific sampling process is as follows Figure 1 As shown. Figure 1 After the random sampling process shown, multiple sets of power time series data subsets corresponding to different scenarios can be obtained, and each power time series data subset includes power data corresponding to each time period. , The number of sampling periods can be set as needed. is 96, corresponding to the number of data points in the power time series data subset.

[0126] Example 2

[0127] In order to reduce the conservatism of robust optimization modeling such as microgrid configuration and scheduling while ensuring system robustness, this embodiment, based on the day-ahead scenario set obtained in Example 1, defines a scenario-driven original uncertainty set, and draws on the finite covering theory to cover a complex source-load scenario uncertainty set with a finite number of simpler uncertainty sets, and converts the scenario-driven uncertainty set into a plurality of simple uncertainty sets for common coverage. That is, the time series power scenario set is used as the time series original scenario set, and a scenario-driven original uncertainty set is formed according to the time series original scenario set. The finite covering theory is used to solve the corresponding mixed integer linear programming problem, and the number of box uncertainty sets and corresponding parameters of the finite covering original uncertain scenarios are obtained, completing the conversion from the complex original uncertainty set to the finite number of simplified box uncertainty sets.

[0128] Specifically, according to the sampling results, the original uncertainty set of scenario-driven distributed photovoltaic and load under constraints can be expressed as:

[0129]

[0130] In the formula, and are the number of generation scenarios for distributed PV and load, respectively; , and , are the sampling value and average value of the distributed photovoltaic and load power scenarios in period t under the i-th time series scenario, is the original uncertainty set driven by the scenario, Γ is the budget parameter, indicating that there are at most Γ periods to reach the maximum uncertainty range, , Represent the sampling values ​​of distributed photovoltaic and load at time t respectively.

[0131] Based on existing sampling scenarios , and its corresponding fluctuation upper and lower bounds , , forming a finite coverage uncertainty set generation method based on historical database and mixed integer linear programming (MILP) problem. The optimization problem of uncertainty set conversion can be expressed as:

[0132]

[0133] In the formula,

[0134] and are the upper and lower bounds of the scenario at time t in the nth finite covering set;

[0135] and is the upper and lower bounds of the scene at time t in the original scene set formed by the s-th sampling, that is: , ; is the balance factor; M is a very large number (big M method); It is a 0-1 variable that guides the mapping of the original scene set S to the finite coverage scene set N.

[0136] The objective function consists of two terms. The first term is to minimize the total size of all uncertainty sets, the second term is to minimize the size of the largest uncertainty set, and is a parameter to adjust the weights of the two terms. It is important to note that the second term can balance the size of a single uncertainty set. The reason for introducing the second term is that the required flexibility will be limited by the maximum fluctuation. This means that optimizing the largest uncertainty set is important to reduce the running cost of the uncertainty set considered. The subsequent three lines of constraints are used to ensure that all possible scenarios and their fluctuations can be contained in a finite covering set. M is a sufficiently large number (big M method). By solving the above mixed integer linear programming problem, the original uncertainty set U (S) can be transformed into a finite number (the specific number is the optimization solution) of simple box uncertainty sets, which can reduce the conservatism of robust optimization while ensuring the coverage of the original uncertainty set.

[0137] So far, the scene-driven uncertain set Can be represented by n box-like uncertain sets Coverage, that is: .

[0138] In the implementation process of the above method, the present invention not only considers the influencing factors of distributed photovoltaics and loads to construct a similar day set, but also takes into account the full use of the fluctuation information of historical data to compress the sampling interval, thereby reducing the scene width of photovoltaics and loads. It can effectively compress the uncertainty of scene generation through the sampling interval and limited coverage of the scene while ensuring that historical data information is fully considered, and fully explore and compress the uncertainty of distributed resource operation scenarios. The analysis method can be effectively applied to the uncertainty characterization of distributed resources at the microgrid level and distribution network level, and provide an accurate description of uncertainty for subsequent uncertain optimization models such as robust scheduling models, robust configuration models, and robust planning models.

[0139] Example 3

[0140] Based on the same inventive concept as Example 1 and Example 2, this embodiment introduces a computer-readable storage medium on which a computer program is stored. When the computer program is executed by a processor, the method for generating a distributed resource timing scenario and converting its uncertain set as described in Example 1 is implemented.

[0141] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, systems, or computer program products. Therefore, the present application may adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Moreover, the present application may adopt the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program codes.

[0142] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0143] These computer program instructions may also be stored in a computer readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer readable memory produce an article of manufacture including an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0144] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 A step that specifies a function in one or more boxes.

[0145] The embodiments of the present invention are described above in conjunction with the accompanying drawings, but the present invention is not limited to the above-mentioned specific implementation methods. The above-mentioned specific implementation methods are merely illustrative and not restrictive. Under the enlightenment of the present invention, ordinary technicians in this field can also make many forms without departing from the scope of protection of the purpose of the present invention and the claims, which all fall within the protection of the present invention.

Claims

1. A distributed resource time sequence scenario generation method, characterized in that: include: Obtain the historical operation data, historical meteorological data and meteorological forecast data of distributed photovoltaic and load in the area to be analyzed; Generate historical sample day feature vectors and scenario day feature vectors to be generated based on the acquired data, and determine a similar day set; According to the similar day set, the historical operation data of distributed photovoltaic and load on similar days in the area to be analyzed are obtained, the power fluctuation interval distribution characteristics and power probability distribution characteristics are analyzed, and the upper and lower bounds of power fluctuation and power probability density function of distributed photovoltaic and load are obtained; Based on the power fluctuation upper and lower bounds and power probability density function of the distributed photovoltaic and load, a set of distributed photovoltaic and load time series power scenarios considering fluctuation interval correction is obtained.

2. The method according to claim 1, characterized in that: The meteorological characteristic data types in the meteorological historical data and meteorological forecast data include: maximum temperature, minimum temperature, average temperature, precipitation, relative humidity, average irradiance, average wind speed, and average air pressure; The forming of the historical sample day feature vector and the scene to-be-generated day feature vector according to the acquired data includes: Calculate the daily average photovoltaic power and load power based on the historical operating data of distributed photovoltaic and load; Calculate the Spearman correlation coefficients of the daily average photovoltaic power and load power and various types of historical meteorological characteristic data series respectively; For each type of meteorological characteristic data, the corresponding Spearman correlation coefficient calculation result is compared with the correlation coefficient threshold, and the characteristic vector type is selected according to the comparison result; According to the selected feature vector type, the feature vectors of each historical sample day and the day to be generated are obtained, which are expressed as: , In the formula, Indicates the number of historical sample days, The feature vector representing the day when the scene is to be generated, Indicates Historical daily sample feature vectors, Indicates The mth eigenvalue in the eigenvector of the historical daily samples.

3. The method according to claim 2, characterized in that: Determining the similar day set includes: Based on the feature vectors of each historical sample day and the day for the scenario to be generated, the comprehensive correlation coefficient between the meteorological characteristics of the day for the scenario to be generated and the meteorological characteristics of each historical sample is calculated; The historical samples with comprehensive correlation coefficients higher than the comprehensive correlation coefficient threshold and the days to be generated for the scenarios are combined into a similar day set; The calculation formula of the comprehensive correlation coefficient is: , in, For the The comprehensive correlation coefficient between the historical samples and the meteorological characteristics of the day when the scene is to be generated, For the The geometric correlation between the historical samples and the day when the scene is to be generated, For the The distance correlation between the historical samples and the day when the scene is to be generated is expressed as: ， ； Among them, the intermediate variable , .

4. The method according to claim 1, characterized in that: The historical operation data of distributed photovoltaic and load on similar days in the analysis area are analyzed to analyze the distribution characteristics of power fluctuation range, including: For the distributed photovoltaic and load time series data of each historical sample day in the similar day set, the data fluctuation value of each period is calculated to obtain the fluctuation range of the distributed photovoltaic and load data corresponding to each period, which is expressed as: , in, , In the formula, is the fluctuation value of distributed photovoltaic or load data in time period t, is the power fluctuation value from the start time to the end time of time period t on the dth day, T is the sampling time interval of adjacent time series data, and S is the set of similar days.

5. The method according to claim 4, characterized in that: The historical operation data of distributed photovoltaic and load on similar days in the analysis area are analyzed to analyze the power probability distribution characteristics, including: Select the kernel function and its bandwidth; For each data point on multiple similar days, the selected kernel function is used to estimate the local density of each data point; Take a weighted average of the local densities of all data points to get a global density estimate; Among them, for the distributed photovoltaic or load data set at any data point t on multiple similar days , the probability density function of the data set is estimated by the Gaussian kernel function: , In the formula, It's on point The estimated density value at ; is the number of historical sample days; At each data point is the Gaussian kernel function applied at ; h is the kernel function bandwidth.

6. The method according to claim 5, characterized in that: Based on the power fluctuation upper and lower bounds and power probability density function of the distributed photovoltaic and load, a set of distributed photovoltaic and load time series power scenarios considering fluctuation interval correction is obtained, including: Based on the fluctuation range and power probability density function of distributed photovoltaic / load data corresponding to each time period, the probability density sampling correction space under each power scenario is calculated for each time period; Simple random sampling is performed according to the probability density sampling correction space to obtain the photovoltaic / load power sampling value of each time period, and further obtain the time series power scenario set.

7. The method according to claim 1, characterized in that: The method of calculating the probability density sampling correction space under each power scenario for each time period includes: Determine the power value fluctuation range of time period t+1, expressed as: , in, and Respectively represent the power sampling values ​​of time period t and time period t+1 under the sth power scenario; Will Substitute the endpoint values ​​of the power fluctuation interval into the probability density function of the t+1 period to obtain the probability density sampling correction space of the t+1 period under the sth power scenario, which is expressed as: , 。 8. The method according to claim 7, characterized in that include: The time series power scenario set is taken as a time series original scenario set, and a scenario-driven original uncertainty set is formed according to the time series original scenario set. The finite covering theory is used to solve several simplified uncertainty sets covering the original uncertain scenarios.

9. The method according to claim 8, characterized in that: The original uncertainty set driven by the scenario is expressed as: , In the formula, and are the number of generation scenarios for distributed PV and load, respectively; , and , are the sampling value and average value of the distributed photovoltaic and load power scenarios in period t under the i-th time series scenario, is the original uncertainty set driven by the scenario, Γ is the budget parameter, indicating that there are at most Γ periods to reach the maximum uncertainty range, , Respectively represent the sampling values ​​of distributed photovoltaic and load at time t; The finite covering theory is used to solve several simplified uncertainty sets covering the original uncertain scenario, including: Based on the original uncertainty set driven by the scenario, the transformed optimization problem is obtained, which is expressed as: , In the formula, and are the upper and lower bounds of the power value of the scenario in the nth finite coverage set for the t-time period; and is the upper and lower bounds of the scene at time t in the original scene set formed by the s-th sampling, that is: , ; is the balance factor; M is a very large number (big M method); It is a 0-1 variable that guides the mapping of the original scene set of the time series to the finite coverage scene set N; represents the balance factor; represents a finite covering set; Solve the optimization problem to obtain an uncertain set that can cover the scenario drive , n box-like uncertain sets ,Right now: .

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the distributed resource timing scenario generation method as described in any one of claims 1 to 9 is implemented.

Images