Movatterモバイル変換

[0]ホーム
URL:

CN110909911B - Aggregation method of multidimensional time series data considering space-time correlation - Google Patents

Aggregation method of multidimensional time series data considering space-time correlationDownload PDF

Info

Publication number
CN110909911B
CN110909911BCN201910932946.4ACN201910932946ACN110909911BCN 110909911 BCN110909911 BCN 110909911BCN 201910932946 ACN201910932946 ACN 201910932946ACN 110909911 BCN110909911 BCN 110909911B
Authority
CN
China
Prior art keywords
load
photovoltaic
power
scene
wind power
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Active
Application number
CN201910932946.4A
Other languages
Chinese (zh)
Other versions
CN110909911A (en
Inventor
叶林
李镓辰
李延和
马明顺
李湃
黄越辉
张舒捷
肖明
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
China Agricultural University
China Electric Power Research Institute Co Ltd CEPRI
State Grid Qinghai Electric Power Co Ltd
Electric Power Research Institute of State Grid Qinghai Electric Power Co Ltd
Original Assignee
China Agricultural University
China Electric Power Research Institute Co Ltd CEPRI
State Grid Qinghai Electric Power Co Ltd
Electric Power Research Institute of State Grid Qinghai Electric Power Co Ltd
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by China Agricultural University, China Electric Power Research Institute Co Ltd CEPRI, State Grid Qinghai Electric Power Co Ltd, Electric Power Research Institute of State Grid Qinghai Electric Power Co LtdfiledCriticalChina Agricultural University
Priority to CN201910932946.4ApriorityCriticalpatent/CN110909911B/en
Publication of CN110909911ApublicationCriticalpatent/CN110909911A/en
Application grantedgrantedCritical
Publication of CN110909911BpublicationCriticalpatent/CN110909911B/en
Activelegal-statusCriticalCurrent
Anticipated expirationlegal-statusCritical

Links

Images

Classifications

Landscapes

Abstract

Translated fromChinese

本发明公开了考虑时空相关性的多维时序数据的聚合方法。本发明针对含风电、光伏发电的电力系统中多能源跨季互补年/月优化调度的问题，提出采用马尔可夫决策的方法优化出风电、光伏与负荷在不同初始状态组合下的最优动作策略。从而达到在该初始状态组合条件下所选的日场景组合在数值概率分布式上最接近原始时序数据。再采用马尔科夫蒙特卡洛方法抽样生成具有相关性的3×N的马尔可夫状态矩阵，将马尔科夫决策获得的最优策略与马尔可夫状态矩阵内的状态列向量组合相匹配，得到具有时空相关性的风电、光伏与负荷的聚合序列，得到符合目标地区风、光与负荷的典型功率场景，从而为系统优化调度上的年/月电量计划做指导。

Figure 201910932946

The invention discloses an aggregation method of multi-dimensional time series data considering space-time correlation. Aiming at the problem of multi-energy inter-seasonal complementary annual/monthly optimal scheduling in a power system including wind power and photovoltaic power generation, the invention proposes to use the Markov decision method to optimize the optimal actions of wind power, photovoltaic and load under different initial state combinations Strategy. Therefore, under the condition of the initial state combination, the selected daily scene combination is closest to the original time series data in terms of numerical probability distribution. Then use the Markov Monte Carlo method to sample and generate a 3×N Markov state matrix with correlation, and match the optimal strategy obtained by the Markov decision with the state column vector combination in the Markov state matrix, The aggregation sequence of wind power, photovoltaic and load with spatial and temporal correlation is obtained, and the typical power scenarios in line with the wind, light and load of the target area are obtained, so as to provide guidance for the annual/monthly power plan in the system optimal scheduling.

Figure 201910932946

Description

Translated fromChinese
考虑时空相关性的多维时序数据的聚合方法Aggregation method for multidimensional time series data considering spatiotemporal correlation

技术领域technical field

本发明涉及多能源电力系统互补优化调度领域，特别是考虑时空相关性的多维时序数据的聚合方法。The invention relates to the field of complementary optimal scheduling of multi-energy power systems, in particular to an aggregation method of multi-dimensional time series data considering space-time correlation.

背景技术Background technique

随着煤炭、石油等不可再生资源的日益枯竭以及能源困局的日益严峻，风能、太阳能、潮汐能及生物质能等可再生能源在世界范围内愈发备受关注。其中对于风、光自然资源的利用是可再生能源发电技术中技术最成熟、最具开发价值的两种可再生能源。发展风电、光电对于保障能源安全，调整能源结构，减轻环境污染，实现可持续发展等都具有非常重要的意义。With the increasing depletion of non-renewable resources such as coal and oil and the increasingly severe energy dilemma, renewable energy such as wind energy, solar energy, tidal energy and biomass energy has attracted more and more attention worldwide. Among them, the utilization of natural resources of wind and light is the two most mature and valuable renewable energy sources in the renewable energy power generation technology. The development of wind power and photovoltaics is of great significance for ensuring energy security, adjusting the energy structure, reducing environmental pollution, and achieving sustainable development.

自然界风、光可再生能源具有高不确定性，该特性决定了风电、光电的功率具有很强的波动性，随着大规模风电、光电接入电网，这种功率波动将给电网的安全经济运行带来巨大挑战。通过分析风、光等不确定性电源发电的广域时空相关性，研究考虑多能源跨季互补的年/月优化调度方法，充分利用可再生能源资源之间的跨季互补作用，通过分析目标地区内风、光与负荷电场的典型出力场景，实现年/月电量计划的合理科学的制定，有效提高电网对可再生能源的消纳能力。Natural wind and photovoltaic renewable energy have high uncertainty, which determines that the power of wind power and photovoltaic has strong fluctuation. Operation brings great challenges. By analyzing the wide-area temporal and spatial correlation of power generation with uncertain power sources such as wind and light, the annual/monthly optimal scheduling method considering multi-energy inter-seasonal complementarity is studied, and the inter-seasonal complementarity between renewable energy resources is fully utilized. The typical output scenarios of wind, light and load power fields in the region realize the rational and scientific formulation of annual/monthly power plans, and effectively improve the grid's ability to absorb renewable energy.

目前，国内外对于优化调度的研究越来越广泛和深入。在进行风电/光伏的年/月电量计划制定研究的过程中发现风电、光伏与负荷的出力时间序列数据过于庞大，数据冗杂，这导致了进行年/月优化计算时，时间断面多，处理时间过长不能快速求解、难以满足年/月优化调度时效性的问题，并且数据的冗杂在很大程度上不利于构建风电、光伏与负荷的典型出力场景。目前最基础的方法是通过在原出力时间序列的数据点上进行等间隔抽样，从而构成新的出力序列；此外，还有采用基于信息熵的分段聚合近似方法，将原出力时间序列计算信息熵的分布并进行分段聚合近似，从而构成新的出力序列，此外还有基于主成分分析方法、基于离散傅里叶变换方法等，在此基础上运用k-means聚类或者层次聚类等聚类算法进行典型场景的构建。但以上聚合方法的结果不能很好的反应原序列的趋势与波动性。并且传统的k-means聚类算法对初始类中心选择的敏感性强，多次聚类的稳定性较差，同时不能给出最优分类数。层次聚类的缺点是计算量比较大，另外，由于层次聚类使用的是贪心算法，得到的显然只是局域最优，不一定就是全局最优。所以以上方法所得到的典型场景均存在不足。At present, the research on optimal scheduling at home and abroad is more and more extensive and in-depth. In the process of researching the annual/monthly power plan formulation of wind power/photovoltaic, it is found that the output time series data of wind power, photovoltaics and loads is too large and the data is redundant, which leads to the annual/monthly optimization calculation, which requires many time sections and processing time. Too long cannot be solved quickly, it is difficult to meet the timeliness of annual/monthly optimal scheduling, and the complexity of data is not conducive to the construction of typical output scenarios of wind power, photovoltaics and loads to a large extent. The most basic method at present is to form a new output sequence by sampling at equal intervals on the data points of the original output time series; in addition, there is also a piecewise aggregation approximation method based on information entropy to calculate the information entropy of the original output time series. The distribution of the data is approximated by piecewise aggregation to form a new output sequence. In addition, there are methods based on principal component analysis and discrete Fourier transform methods. On this basis, clustering such as k-means clustering or hierarchical clustering is used. Class algorithm for the construction of typical scenarios. However, the results of the above polymerization methods cannot reflect the trend and volatility of the original sequence very well. In addition, the traditional k-means clustering algorithm is highly sensitive to the selection of the initial cluster center, the stability of multiple clustering is poor, and the optimal number of classifications cannot be given. The disadvantage of hierarchical clustering is that the amount of calculation is relatively large. In addition, because the hierarchical clustering uses a greedy algorithm, it is obviously only a local optimum, not necessarily a global optimum. Therefore, the typical scenarios obtained by the above methods have shortcomings.

发明内容SUMMARY OF THE INVENTION

针对现有技术中存在的缺陷，本发明的目的在于提供一种考虑时空相关性的多维时序数据的聚合方法。该方法采用马尔可夫决策的方法优化出风电、光伏与负荷在不同初始状态组合下的最优动作策略。从而达到在该初始状态组合条件下所选的日场景组合在数值概率分布式上最接近原始时序数据。再采用马尔科夫蒙特卡洛方法抽样生成具有相关性的3×N的马尔可夫状态矩阵，将马尔科夫决策获得的最优动作策略与马尔可夫状态矩阵内的状态列向量组合相匹配，得到具有时空相关性的风电、光伏与负荷的聚合序列。In view of the defects existing in the prior art, the purpose of the present invention is to provide a method for aggregating multi-dimensional time series data considering the spatiotemporal correlation. The method adopts the Markov decision method to optimize the optimal action strategy of wind power, photovoltaic and load under different initial state combinations. Therefore, under the condition of the initial state combination, the selected daily scene combination is closest to the original time series data in terms of numerical probability distribution. Then use the Markov Monte Carlo method to sample and generate a 3×N Markov state matrix with correlation, and match the optimal action strategy obtained by the Markov decision with the state column vector combination in the Markov state matrix. , to obtain the aggregation sequence of wind power, photovoltaic and load with spatial and temporal correlation.

为达到以上目的，本发明采取的技术方案是：In order to achieve the above purpose, the technical scheme adopted in the present invention is:

一种考虑时空相关性的多维时序数据的聚合方法，包括如下步骤：A method for aggregation of multidimensional time series data considering spatiotemporal correlation, comprising the following steps:

S1、分别获取同时期长度为m的风电、光伏与负荷中长期历史功率序列，对各序列进行极值归一化处理得到长度为m的风电中长期功率时间序列、光伏中长期功率时间序列、负荷中长期功率时间序列；S1. Obtain the mid- and long-term historical power series of wind power, photovoltaic and load at the same period of length m respectively, and perform extreme value normalization on each sequence to obtain the mid- and long-term power time series of wind power, photovoltaic mid- and long-term power time series, Long-term power time series of loads;

其中，长度为m的风电中长期功率时间序列表示为：Among them, the medium and long-term power time series of wind power with length m is expressed as:

Figure BDA0002220746900000031
序列中的
Figure BDA0002220746900000032
均代表一个风电功率历史值；
Figure BDA0002220746900000031
in sequence
Figure BDA0002220746900000032
Both represent a historical value of wind power;

长度为m的光伏中长期功率时间序列表示为：The PV medium and long-term power time series with length m is expressed as:

Figure BDA0002220746900000033
序列中的
Figure BDA0002220746900000034
均代表一个光伏功率历史值；
Figure BDA0002220746900000033
in sequence
Figure BDA0002220746900000034
Both represent a historical value of photovoltaic power;

长度为m的负荷中长期功率时间序列表示为：The medium and long-term power time series of a load of length m is expressed as:

Figure BDA0002220746900000035
序列中的
Figure BDA0002220746900000036
均代表一个负荷功率历史值；
Figure BDA0002220746900000035
in sequence
Figure BDA0002220746900000036
Both represent a historical value of load power;

然后对风电、光伏与负荷的中长期功率时间序列分别进行日场景分割获得相同数量的：Then the medium and long-term power time series of wind power, photovoltaic and load are divided into daily scenes to obtain the same number of:

风电日场景

Figure BDA0002220746900000037
Wind power day scene
Figure BDA0002220746900000037

其中

Figure BDA0002220746900000038
Figure BDA0002220746900000039
表示分割后的第k个风电日场景，1≤k≤dayn，in
Figure BDA0002220746900000038
Figure BDA0002220746900000039
represents the k-th wind power day scene after division, 1≤k≤dayn,

光伏日场景

Figure BDA00022207469000000310
Photovoltaic day scene
Figure BDA00022207469000000310

其中

Figure BDA00022207469000000311
Figure BDA00022207469000000312
表示分割后的第k个光伏日场景，1≤k≤dayn，in
Figure BDA00022207469000000311
Figure BDA00022207469000000312
Indicates the k-th PV day scene after division, 1≤k≤dayn,

负荷日场景

Figure BDA00022207469000000313
Load Day Scenario
Figure BDA00022207469000000313

其中

Figure BDA00022207469000000314
Figure BDA00022207469000000315
表示分割后的第k个负荷日场景，1≤k≤dayn，dayn是由中长期历史功率序列中所包含的总天数决定的，风电、光伏、负荷的中长期历史功率序列中所包含的总天数是一样的；in
Figure BDA00022207469000000314
Figure BDA00022207469000000315
Indicates the k-th load day scene after division, 1≤k≤dayn, dayn is determined by the total number of days included in the medium and long-term historical power series, and the total number of days included in the medium- and long-term historical power series of wind power, photovoltaic and load. The number of days is the same;

风电日场景SCwind、光伏日场景SCpv、负荷日场景SCload中所包含的日场景元素通过近邻传播聚类算法分别被分配到风电日场景集

Figure BDA0002220746900000041
光伏日场景集
Figure BDA0002220746900000042
负荷日场景集
Figure BDA0002220746900000043
内，通过Davies-Bouldin Index指标分析获得风电、光伏与负荷的日场景最佳聚类数分别为
Figure BDA0002220746900000044
Figure BDA0002220746900000045
同时获得在各自最佳聚类数下的聚类结果：The daily scene elements contained in the wind power daily scene SCwind , the photovoltaic daily scene SCpv , and the load daily scene SCload are respectively assigned to the wind power daily scene set through the neighbor propagation clustering algorithm.
Figure BDA0002220746900000041
Photovoltaic day scene set
Figure BDA0002220746900000042
load day scene set
Figure BDA0002220746900000043
The optimal clustering numbers of daily scenes of wind power, photovoltaic and load obtained by Davies-Bouldin Index analysis are as follows:
Figure BDA0002220746900000044
Figure BDA0002220746900000045
At the same time, the clustering results under the respective optimal number of clusters are obtained:

Figure BDA0002220746900000046
Figure BDA0002220746900000046

其中，

Figure BDA0002220746900000047
Figure BDA0002220746900000048
代表wi类风电日场景集，由nwi个风电日场景构成，in,
Figure BDA0002220746900000047
Figure BDA0002220746900000048
represents thewi class wind power day scene set, which consists of nwi wind power day scenes,

Figure BDA0002220746900000049
Figure BDA0002220746900000049

其中

Figure BDA00022207469000000410
Figure BDA00022207469000000411
代表第pi类光伏日场景集，由npi个光伏日场景构成，in
Figure BDA00022207469000000410
Figure BDA00022207469000000411
Represents thep i-th photovoltaic daily scene set, which consists of npi photovoltaic daily scenes,

Figure BDA00022207469000000412
Figure BDA00022207469000000412

其中

Figure BDA00022207469000000413
Figure BDA00022207469000000414
代表第li类负荷日场景集，由nli个光伏日场景构成，in
Figure BDA00022207469000000413
Figure BDA00022207469000000414
represents theli -th load day scene set, which is composed of nli photovoltaic day scenes,

其中nwi、npi与nli由近邻传播聚类算法通过计算获得的值；Among them, nwi , npi and nli are calculated by the nearest neighbor propagation clustering algorithm;

S2、在步骤S1的基础上，将采用近邻传播聚类算法获得的风电日场景、光伏日场景与负荷日场景的聚类结果作为状态量，得到风电、光伏与负荷的随机状态模型，由马尔可夫蒙特卡洛过程来表示，马尔可夫蒙特卡洛过程的高阶状态转移概率表示为：S2. On the basis of step S1, the clustering results of the wind power daily scene, photovoltaic daily scene and load daily scene obtained by using the neighbor propagation clustering algorithm are used as the state quantity, and the random state model of wind power, photovoltaic and load is obtained. The high-order state transition probability of the Markov Monte Carlo process is expressed as:

Figure BDA00022207469000000415
Figure BDA00022207469000000415

Figure BDA00022207469000000416
Figure BDA00022207469000000416

Figure BDA0002220746900000051
Figure BDA0002220746900000051

式中n表示马尔可夫聚合状态矩阵中的第n列状态且n∈[1,N]，where n represents the nth column state in the Markov aggregation state matrix and n∈[1,N],

Figure BDA0002220746900000052
Figure BDA0002220746900000052

Figure BDA0002220746900000053
Figure BDA0002220746900000053

Figure BDA0002220746900000054
Figure BDA0002220746900000054

通过统计历史数据获得马尔可夫蒙特卡洛过程的状态转移概率矩阵：Obtain the state transition probability matrix of the Markov Monte Carlo process from statistical historical data:

Figure BDA0002220746900000055
Figure BDA0002220746900000055

Figure BDA0002220746900000056
Figure BDA0002220746900000056

Figure BDA0002220746900000057
Figure BDA0002220746900000057

S3、马尔科夫决策过程由五部分组成(S,A,{PSA},γ,R)，S3, the Markov decision process consists of five parts (S, A, {PSA }, γ, R),

S是由聚类结果

Figure BDA0002220746900000058
Figure BDA0002220746900000059
组成的状态集合，S＝<Swind,Spv,Sload>；S is the result of clustering
Figure BDA0002220746900000058
Figure BDA0002220746900000059
The composed state set, S=<Swind , Spv , Sload >;

第n步的状态集表示为：The state set of the nth step is expressed as:

Figure BDA0002220746900000061
Figure BDA0002220746900000061

A表示动作集，由聚类结果中各日场景集内所包含的日场景元素组成

Figure BDA0002220746900000062
其中：A represents the action set, which consists of the daily scene elements contained in each daily scene set in the clustering result
Figure BDA0002220746900000062
in:

Figure BDA0002220746900000063
Figure BDA0002220746900000063

Figure BDA0002220746900000064
Figure BDA0002220746900000064

Figure BDA0002220746900000065
Figure BDA0002220746900000065

Figure BDA0002220746900000066
Figure BDA0002220746900000066

Figure BDA0002220746900000067
Figure BDA0002220746900000067

Figure BDA0002220746900000068
Figure BDA0002220746900000068

第n步的动作集可以表示为：The action set of the nth step can be expressed as:

Figure BDA0002220746900000069
Figure BDA0002220746900000069

{PSA}表示状态转移概率，表示为：{PSA } represents the state transition probability, which is expressed as:

Figure BDA00022207469000000610
Figure BDA00022207469000000610

γ∈[0，1)表示阻尼系数，表示随着时间的推移回报率的折扣，按照经验值一般可以选择γ＝0.5；γ∈[0, 1) represents the damping coefficient, which represents the discount of the rate of return over time. Generally, γ=0.5 can be selected according to the empirical value;

R代表误差函数：R stands for error function:

Figure BDA00022207469000000611
Figure BDA00022207469000000612
其中sort函数表示将时序数据按照数值从大到小排列，
Figure BDA00022207469000000613
表示将风电日场景
Figure BDA00022207469000000614
扩充(dayn)倍后将时序数据按照数值从大到小排列，使长度与原始序列长度一致，便于后续进行欧式距离的求解，
Figure BDA00022207469000000615
表示将光伏日场景
Figure BDA0002220746900000071
扩充(dayn)倍后将时序数据按照数值从大到小排列，使长度与原始序列长度一致，
Figure BDA0002220746900000072
表示将负荷日场景
Figure BDA0002220746900000073
扩充(dayn)倍后将时序数据按照数值从大到小排列，使长度与原始序列长度一致；
Figure BDA00022207469000000611
Figure BDA00022207469000000612
The sort function means to arrange the time series data in descending order of value.
Figure BDA00022207469000000613
Indicates the wind power day scene
Figure BDA00022207469000000614
After the expansion (dayn) times, the time series data are arranged in descending order of value, so that the length is consistent with the original sequence length, which is convenient for the subsequent solution of Euclidean distance.
Figure BDA00022207469000000615
Indicates the PV day scene
Figure BDA0002220746900000071
After the expansion (dayn) times, the time series data are arranged in descending order according to the value, so that the length is consistent with the original sequence length,
Figure BDA0002220746900000072
Indicates that the day scenario will be loaded
Figure BDA0002220746900000073
After the expansion (dayn) times, the time series data are arranged in descending order according to the value, so that the length is consistent with the original sequence length;

S4、定义值函数

Figure BDA0002220746900000074
S0是人为设定的初始状态，初始状态S0的数量为
Figure BDA0002220746900000075
定义最优值函数
Figure BDA0002220746900000076
S4, define the value function
Figure BDA0002220746900000074
S0 is an artificially set initial state, and the number of initial states S0 is
Figure BDA0002220746900000075
Define the optimal value function
Figure BDA0002220746900000076

利用贝尔曼等式求解

Figure BDA0002220746900000077
获得最优策略
Figure BDA0002220746900000078
Solve using Bellman's equation
Figure BDA0002220746900000077
get the best strategy
Figure BDA0002220746900000078

S5、将所有初始状态S0分别代入

Figure BDA0002220746900000079
等式中，共获得|S|个方程，从而获得在不同初始状态S0下的最优风电、光伏与负荷的日场景组合策略π*(S0)；S5. Substitute all initial states S0 into
Figure BDA0002220746900000079
In the equation, a total of |S| equations are obtained, so as to obtain the optimal daily scene combination strategy π* (S0 ) of wind power, photovoltaic and load under different initial states S0 ;

S6、运用马尔可夫蒙特卡洛方法，通过Gibbs抽样生成由风电、光伏与负荷组成的3×N阶的马尔可夫聚合状态矩阵：S6. Use the Markov Monte Carlo method to generate a 3×N-order Markov aggregation state matrix composed of wind power, photovoltaics and loads through Gibbs sampling:

Figure BDA00022207469000000710
Figure BDA00022207469000000710

其中

Figure BDA00022207469000000711
in
Figure BDA00022207469000000711

S7、将马尔可夫状态矩阵内的每一组状态向量与最优策略的映射相匹配，得到日场景矩阵

Figure BDA0002220746900000081
S7. Match each group of state vectors in the Markov state matrix with the mapping of the optimal strategy to obtain the daily scene matrix
Figure BDA0002220746900000081

其中

Figure BDA0002220746900000082
in
Figure BDA0002220746900000082

Figure BDA0002220746900000083
Figure BDA0002220746900000083

Figure BDA0002220746900000084
Figure BDA0002220746900000084

S8、判断日场景矩阵每行相邻的日场景间的首尾连接处差值是否大于中长期功率时间序列的一阶差分最大值，如果大于，则对首尾连接处进行小波滤波处理，直到差值小于一阶差分最大值，如果小于或等于，直接对日场景首尾相连，最后生成三条由N个日场景构成的聚合序列。S8. Determine whether the difference between the head and tail connections between adjacent daily scenes in each row of the daily scene matrix is greater than the maximum value of the first-order difference of the medium and long-term power time series. It is less than the maximum value of the first-order difference, and if it is less than or equal to, the daily scenes are directly connected end to end, and finally three aggregate sequences consisting of N daily scenes are generated.

本发明所述的方法考虑到了在优化调度层面进行年/月电量优化计算时，时间断面过多，数据量过大的问题，并且考虑到了在同一地区内的风、光与负荷因地形、纬度等自然因素的影响，很多风、光与负荷功率波动特性在一定程度上具有很大的相似性，需要提取出具有代表性的出力曲线的问题。该方法弥补了在优化调度层面进行年/月电量优化计算时计算量过大的缺陷，以及现有的风、光与负荷场景分析方法中时序仿真法存在计算效率低的问题，典型日法由于不能体现风、光与负荷的年、月、日出力特征，计算结果过于保守的问题。本方法兼顾计算效率和数据变化特性，为新能源发电系统的年/月优化调度提供了有效的指导，促进了风、光不确定性电源的消纳。The method described in the present invention takes into account the problems of too many time sections and too much data when the annual/monthly power optimization calculation is performed at the optimal scheduling level, and also takes into account that the wind, light and load in the same area are affected by terrain and latitude. Due to the influence of other natural factors, many wind, light and load power fluctuation characteristics have great similarity to a certain extent, and it is necessary to extract a representative output curve. This method makes up for the problem of excessive calculation amount in the annual/monthly power optimization calculation at the optimal scheduling level, and the problem of low calculation efficiency of the time series simulation method in the existing wind, light and load scene analysis methods. It cannot reflect the year, month, and sun force characteristics of wind, light and load, and the calculation results are too conservative. The method takes into account the characteristics of calculation efficiency and data change, provides effective guidance for the annual/monthly optimal scheduling of new energy power generation systems, and promotes the consumption of wind and light uncertain power sources.

附图说明Description of drawings

本发明有如下附图：The present invention has the following accompanying drawings:

图1为本发明一种考虑时空相关性的多维时序数据的聚合方法流程示意图。FIG. 1 is a schematic flowchart of a method for aggregating multi-dimensional time series data considering spatiotemporal correlation according to the present invention.

具体实施方式Detailed ways

以下结合附图1对本发明作进一步详细说明。The present invention will be further described in detail below in conjunction with FIG. 1 .

一种考虑时空相关性的多维时序数据的聚合方法，包括如下步骤：A method for aggregation of multidimensional time series data considering spatiotemporal correlation, comprising the following steps:

S1、分别获取同时期长度为m的风电、光伏与负荷中长期历史功率序列，对各序列进行极值归一化处理得到长度为m的风电中长期功率时间序列、光伏中长期功率时间序列、负荷中长期功率时间序列；S1. Obtain the mid- and long-term historical power series of wind power, photovoltaic and load at the same period of length m respectively, and perform extreme value normalization on each sequence to obtain the mid- and long-term power time series of wind power, photovoltaic mid- and long-term power time series, Long-term power time series of loads;

其中，长度为m的风电中长期功率时间序列表示为：Among them, the medium and long-term power time series of wind power with length m is expressed as:

Figure BDA0002220746900000091
序列中的
Figure BDA0002220746900000092
均代表一个风电功率历史值；
Figure BDA0002220746900000091
in sequence
Figure BDA0002220746900000092
Both represent a historical value of wind power;

长度为m的光伏中长期功率时间序列表示为：The PV medium and long-term power time series with length m is expressed as:

Figure BDA0002220746900000093
序列中的
Figure BDA0002220746900000094
均代表一个光伏功率历史值；
Figure BDA0002220746900000093
in sequence
Figure BDA0002220746900000094
Both represent a historical value of photovoltaic power;

长度为m的负荷中长期功率时间序列表示为：The medium and long-term power time series of a load of length m is expressed as:

Figure BDA0002220746900000095
序列中的
Figure BDA0002220746900000096
均代表一个负荷功率历史值；
Figure BDA0002220746900000095
in sequence
Figure BDA0002220746900000096
Both represent a historical value of load power;

然后对归一化处理后的风电、光伏与负荷的中长期功率时间序列分别进行日场景分割获得相同数量的：Then, the normalized medium and long-term power time series of wind power, photovoltaic and load are divided into daily scenes to obtain the same number of:

风电日场景

Figure BDA0002220746900000097
Wind power day scene
Figure BDA0002220746900000097

其中

Figure BDA0002220746900000098
Figure BDA0002220746900000099
表示分割后的第k个风电日场景，1≤k≤dayn，in
Figure BDA0002220746900000098
Figure BDA0002220746900000099
represents the k-th wind power day scene after division, 1≤k≤dayn,

光伏日场景

Figure BDA0002220746900000101
Photovoltaic day scene
Figure BDA0002220746900000101

其中

Figure BDA0002220746900000102
Figure BDA0002220746900000103
表示分割后的第k个光伏日场景，1≤k≤dayn，in
Figure BDA0002220746900000102
Figure BDA0002220746900000103
Indicates the k-th PV day scene after division, 1≤k≤dayn,

负荷日场景

Figure BDA0002220746900000104
Load Day Scenario
Figure BDA0002220746900000104

其中

Figure BDA0002220746900000105
Figure BDA0002220746900000106
分别表示分割后的第k个负荷日场景，1≤k≤dayn，dayn是由中长期历史功率序列中所包含的总天数决定的，风电、光伏、负荷的中长期历史功率序列中所包含的总天数是一样的；in
Figure BDA0002220746900000105
Figure BDA0002220746900000106
Respectively represent the k-th load day scenario after division, 1≤k≤dayn, dayn is determined by the total number of days included in the mid- and long-term historical power The total number of days is the same;

风电日场景SCwind、光伏日场景SCpv、负荷日场景SCload中所包含的日场景元素通过近邻传播聚类算法分别被分配到风电日场景集

Figure BDA0002220746900000107
光伏日场景集
Figure BDA0002220746900000108
负荷日场景集
Figure BDA0002220746900000109
内，通过Davies-Bouldin Index指标分析获得风电、光伏与负荷的日场景最佳聚类数分别为
Figure BDA00022207469000001010
Figure BDA00022207469000001011
同时获得在各自最佳聚类数下的聚类结果：The daily scene elements contained in the wind power daily scene SCwind , the photovoltaic daily scene SCpv , and the load daily scene SCload are respectively assigned to the wind power daily scene set through the neighbor propagation clustering algorithm.
Figure BDA0002220746900000107
Photovoltaic day scene set
Figure BDA0002220746900000108
load day scene set
Figure BDA0002220746900000109
The optimal clustering numbers of daily scenes of wind power, photovoltaic and load obtained by Davies-Bouldin Index analysis are as follows:
Figure BDA00022207469000001010
Figure BDA00022207469000001011
At the same time, the clustering results under the respective optimal number of clusters are obtained:

Figure BDA00022207469000001012
Figure BDA00022207469000001012

其中，

Figure BDA00022207469000001013
Figure BDA00022207469000001014
代表wi类风电日场景集，由nwi个风电日场景构成，in,
Figure BDA00022207469000001013
Figure BDA00022207469000001014
represents thewi class wind power day scene set, which consists of nwi wind power day scenes,

Figure BDA00022207469000001015
Figure BDA00022207469000001015

其中

Figure BDA00022207469000001016
Figure BDA00022207469000001017
代表第pi类光伏日场景集，由npi个光伏日场景构成，in
Figure BDA00022207469000001016
Figure BDA00022207469000001017
Represents thep i-th photovoltaic daily scene set, which consists of npi photovoltaic daily scenes,

Figure BDA00022207469000001018
Figure BDA00022207469000001018

其中

Figure BDA00022207469000001019
Figure BDA00022207469000001020
代表第li类负荷日场景集，由nli个光伏日场景构成，in
Figure BDA00022207469000001019
Figure BDA00022207469000001020
represents theli -th load day scene set, which is composed of nli photovoltaic day scenes,

其中nwi、npi与nli是由近邻传播聚类算法通过计算获得的值；Among them, nwi , npi and nli are the values obtained by the nearest neighbor propagation clustering algorithm;

S2、在步骤S1的基础上，将采用近邻传播聚类算法获得的风电日场景、光伏日场景与负荷日场景的聚类结果作为状态量，得到风电、光伏与负荷的随机状态模型，由马尔可夫蒙特卡洛过程来表示，马尔可夫蒙特卡洛过程的高阶状态转移概率表示为：S2. On the basis of step S1, the clustering results of the wind power daily scene, photovoltaic daily scene and load daily scene obtained by using the neighbor propagation clustering algorithm are used as the state quantity, and the random state model of wind power, photovoltaic and load is obtained. The high-order state transition probability of the Markov Monte Carlo process is expressed as:

Figure BDA0002220746900000111
Figure BDA0002220746900000111

Figure BDA0002220746900000112
Figure BDA0002220746900000112

Figure BDA0002220746900000113
Figure BDA0002220746900000113

式中n表示马尔可夫聚合状态矩阵中的第n列状态且n∈[1,N]，where n represents the state of the nth column in the Markov aggregation state matrix and n∈[1,N],

Figure BDA0002220746900000114
Figure BDA0002220746900000114

Figure BDA0002220746900000115
Figure BDA0002220746900000115

Figure BDA0002220746900000116
Figure BDA0002220746900000116

通过统计历史数据获得马尔可夫蒙特卡洛过程的状态转移概率矩阵：Obtain the state transition probability matrix of the Markov Monte Carlo process from statistical historical data:

Figure BDA0002220746900000117
Figure BDA0002220746900000117

Figure BDA0002220746900000118
Figure BDA0002220746900000118

Figure BDA0002220746900000121
Figure BDA0002220746900000121

S3、马尔科夫决策过程由五部分组成(S,A,{PSA},γ,R)，S3, the Markov decision process consists of five parts (S, A, {PSA }, γ, R),

S是由聚类结果

Figure BDA0002220746900000122
Figure BDA0002220746900000123
组成的状态集合，S＝<Swind,Spv,Sload>；S is the result of clustering
Figure BDA0002220746900000122
Figure BDA0002220746900000123
The composed state set, S=<Swind , Spv , Sload >;

第n步的状态集表示为：The state set of the nth step is expressed as:

Figure BDA0002220746900000124
Figure BDA0002220746900000124

A表示动作集，由聚类结果中各日场景集内所包含的日场景元素组成

Figure BDA0002220746900000125
其中：A represents the action set, which consists of the daily scene elements contained in each daily scene set in the clustering result
Figure BDA0002220746900000125
in:

Figure BDA0002220746900000126
Figure BDA0002220746900000126

Figure BDA0002220746900000127
Figure BDA0002220746900000127

Figure BDA0002220746900000128
Figure BDA0002220746900000128

Figure BDA0002220746900000129
Figure BDA0002220746900000129

Figure BDA00022207469000001210
Figure BDA00022207469000001210

Figure BDA00022207469000001211
Figure BDA00022207469000001211

第n步的动作集可以表示为：The action set of the nth step can be expressed as:

Figure BDA00022207469000001212
Figure BDA00022207469000001212

{PSA}表示状态转移概率，表示为：{PSA } represents the state transition probability, which is expressed as:

Figure BDA00022207469000001213
Figure BDA00022207469000001213

γ∈[0，1)表示阻尼系数，表示随着时间的推移回报率的折扣，按照经验值一般可以选择γ＝0.5；γ∈[0, 1) represents the damping coefficient, which represents the discount of the rate of return over time. Generally, γ=0.5 can be selected according to the empirical value;

R代表误差函数：R stands for error function:

Figure BDA0002220746900000131
Figure BDA0002220746900000132
其中sort函数表示将时序数据按照数值从大到小排列，
Figure BDA0002220746900000133
表示将风电日场景
Figure BDA0002220746900000134
扩充(dayn)倍后将时序数据按照数值从大到小排列，使长度与原始序列长度一致，便于后续进行欧式距离的求解，
Figure BDA0002220746900000135
表示将光伏日场景
Figure BDA0002220746900000136
扩充(dayn)倍后将时序数据按照数值从大到小排列，使长度与原始序列长度一致，
Figure BDA0002220746900000137
表示将负荷日场景
Figure BDA0002220746900000138
扩充(dayn)倍后将时序数据按照数值从大到小排列，使长度与原始序列长度一致；
Figure BDA0002220746900000131
Figure BDA0002220746900000132
The sort function means to arrange the time series data in descending order of value.
Figure BDA0002220746900000133
Indicates the wind power day scene
Figure BDA0002220746900000134
After the expansion (dayn) times, the time series data are arranged in descending order of value, so that the length is consistent with the original sequence length, which is convenient for the subsequent solution of Euclidean distance.
Figure BDA0002220746900000135
Indicates the PV day scene
Figure BDA0002220746900000136
After the expansion (dayn) times, the time series data are arranged in descending order according to the value, so that the length is consistent with the original sequence length,
Figure BDA0002220746900000137
Indicates that the day scenario will be loaded
Figure BDA0002220746900000138
After the expansion (dayn) times, the time series data are arranged in descending order according to the value, so that the length is consistent with the original sequence length;

S4、定义值函数

Figure BDA0002220746900000139
S0是人为设定的初始状态，初始状态S0的数量为
Figure BDA00022207469000001310
定义最优值函数
Figure BDA00022207469000001311
S4, define the value function
Figure BDA0002220746900000139
S0 is an artificially set initial state, and the number of initial states S0 is
Figure BDA00022207469000001310
Define the optimal value function
Figure BDA00022207469000001311

利用贝尔曼等式求解

Figure BDA00022207469000001312
获得最优策略
Figure BDA00022207469000001313
Solve using Bellman's equation
Figure BDA00022207469000001312
get the best strategy
Figure BDA00022207469000001313

S5、将所有初始状态S0分别代入

Figure BDA00022207469000001314
等式中，共获得|S|个方程，从而获得在不同初始状态S0下的最优风电、光伏与负荷的日场景组合策略π*(S0)；S5. Substitute all initial states S0 into
Figure BDA00022207469000001314
In the equation, a total of |S| equations are obtained, so as to obtain the optimal daily scene combination strategy π* (S0 ) of wind power, photovoltaic and load under different initial states S0 ;

S6、运用马尔可夫蒙特卡洛方法，通过Gibbs抽样生成由风电、光伏与负荷组成的3×N阶的马尔可夫聚合状态矩阵：S6. Use the Markov Monte Carlo method to generate a 3×N-order Markov aggregation state matrix composed of wind power, photovoltaics and loads through Gibbs sampling:

Figure BDA0002220746900000141
Figure BDA0002220746900000141

其中

Figure BDA0002220746900000142
in
Figure BDA0002220746900000142

S7、将马尔可夫状态矩阵内的每一组状态向量与最优策略的映射相匹配，得到日场景矩阵

Figure BDA0002220746900000143
S7. Match each group of state vectors in the Markov state matrix with the mapping of the optimal strategy to obtain the daily scene matrix
Figure BDA0002220746900000143

其中

Figure BDA0002220746900000144
in
Figure BDA0002220746900000144

Figure BDA0002220746900000145
Figure BDA0002220746900000145

Figure BDA0002220746900000146
Figure BDA0002220746900000146

S8、判断日场景矩阵每行相邻的日场景间的首尾连接处差值是否大于中长期功率时间序列的一阶差分最大值，如果大于，则对首尾连接处进行小波滤波处理，直到差值小于一阶差分最大值，如果小于或等于，直接对日场景首尾相连，最后生成三条由N个日场景构成的聚合序列。S8. Determine whether the difference between the head and tail connections between adjacent daily scenes in each row of the daily scene matrix is greater than the maximum value of the first-order difference of the medium and long-term power time series. It is less than the maximum value of the first-order difference, and if it is less than or equal to, the daily scenes are directly connected end to end, and finally three aggregate sequences consisting of N daily scenes are generated.

以上所述，仅是本发明的较佳实例而已，并非对本发明作任何形式上的限制，本领域技术人员利用上述揭示的技术内容做出些许简单修改、等同变化或装饰，均落在本发明的保护范围内。The above are only preferred examples of the present invention, and are not intended to limit the present invention in any form. Those skilled in the art make some simple modifications, equivalent changes or decorations using the technical contents disclosed above, all of which fall within the present invention. within the scope of protection.

本说明书中未作详细描述的内容属于本领域专业技术人员公知的现有技术。Contents not described in detail in this specification belong to the prior art known to those skilled in the art.

Claims (3)

1. A method for aggregating multidimensional time series data considering spatio-temporal correlation, comprising the steps of:
s1, respectively obtaining medium and long term historical power sequences of wind power, photovoltaic power and load with the length of m at the same time, and carrying out extremum normalization processing on the sequences to obtain medium and long term power time sequences of wind power, medium and long term power time sequences of photovoltaic power and medium and long term power time sequences of load with the length of m;
the wind power medium-long term power time sequence with the length of m is represented as follows:
Figure FDA0003478305040000011
in a sequence
Figure FDA0003478305040000012
All represent a historical value of wind power;
the photovoltaic medium and long term power time series of length m is represented as:
Figure FDA0003478305040000013
in sequence
Figure FDA0003478305040000014
Each represents a photovoltaic power historical value;
the load medium and long term power time series with length m is represented as:
Figure FDA0003478305040000015
in sequence
Figure FDA0003478305040000016
Each represents a load power history value;
then, the daily scene segmentation is respectively carried out on the medium-term and long-term power time sequences of wind power, photovoltaic power and load to obtain the same quantity:
wind power day scene
Figure FDA0003478305040000017
Wherein
Figure FDA0003478305040000018
Figure FDA0003478305040000019
Representing the k-th wind power day scene after segmentation, k is more than or equal to 1 and less than or equal to dayn,
photovoltaic day scene
Figure FDA00034783050400000110
Wherein
Figure FDA00034783050400000111
Figure FDA00034783050400000112
Representing the kth photovoltaic day scene after division, k is more than or equal to 1 and less than or equal to dayn,
load day scene
Figure FDA0003478305040000021
Wherein
Figure FDA0003478305040000022
Figure FDA0003478305040000023
Representing the k-th load day scene after division, k is more than or equal to 1 and less than or equal to dayn,
wind power day scene SCwindPhotovoltaic solar scene SCpvLoad day scene SCloadThe day scene elements contained in the method are respectively distributed to the wind power day scene set through a neighbor propagation clustering algorithm
Figure FDA0003478305040000024
Photovoltaic day scene set
Figure FDA0003478305040000025
Load day scene set
Figure FDA0003478305040000026
In the method, the optimal clustering numbers of the daily scenes of wind power, photovoltaic and load are obtained through Davies-Bouldin Index analysis
Figure FDA0003478305040000027
Figure FDA0003478305040000028
And simultaneously obtaining clustering results under respective optimal clustering numbers:
Figure FDA0003478305040000029
wherein,
Figure FDA00034783050400000210
Figure FDA00034783050400000211
representswiWind-like power day scene set composed of nwiThe solar energy power generation system is composed of a plurality of wind power day scenes,
Figure FDA00034783050400000212
wherein
Figure FDA00034783050400000213
Figure FDA00034783050400000214
Represents the firstpiSet of quasi-photovoltaic daily scenes, from npiThe solar photovoltaic solar scene is formed by a plurality of photovoltaic solar scenes,
Figure FDA00034783050400000215
wherein
Figure FDA00034783050400000216
Figure FDA00034783050400000217
Represents the firstliClass load daily scene set, from nliThe composition of the daily scene of each load,
s2, on the basis of the step S1, clustering results of the wind power daily scene, the photovoltaic daily scene and the load daily scene obtained by adopting a neighbor propagation clustering algorithm are used as state quantities to obtain a random state model of the wind power, the photovoltaic and the load, the random state model is represented by a Markov Monte Carlo process, and the high-order state transition probability of the Markov Monte Carlo process is represented as follows:
Figure FDA0003478305040000031
Figure FDA0003478305040000032
Figure FDA0003478305040000033
where N represents the nth column state in the Markov polymerization state matrix and N is [1, N ],
Figure FDA0003478305040000034
Figure FDA0003478305040000035
Figure FDA0003478305040000036
obtaining a state transition probability matrix of the Markov Monte Carlo process by counting historical data:
Figure FDA0003478305040000037
Figure FDA0003478305040000038
Figure FDA0003478305040000039
s3, Markov decision process is composed of five parts (S, A, { PSA},γ,R)，
S is the result of clustering
Figure FDA00034783050400000310
Figure FDA0003478305040000041
Set of constituent states, S ═<Swind,Spv,Sload>；
The state set for step n is represented as:
Figure FDA0003478305040000042
a represents action set, and is composed of daily scene elements contained in each daily scene set in clustering result
Figure FDA0003478305040000043
Wherein:
Figure FDA0003478305040000044
Figure FDA0003478305040000045
Figure FDA0003478305040000046
Figure FDA0003478305040000047
Figure FDA0003478305040000048
Figure FDA0003478305040000049
the set of actions in step n is represented as:
Figure FDA00034783050400000410
{PSAdenotes the state transition probability, expressed as:
Figure FDA00034783050400000411
Figure FDA00034783050400000412
γ ∈ [0, 1) denotes a damping coefficient;
r represents the error function:
Figure FDA00034783050400000413
Figure FDA00034783050400000414
wherein XwindRepresenting a wind power medium-long term power time sequence with the length of m; xpvRepresenting a photovoltaic medium and long term power time series with the length of m; xloadRepresenting a load medium and long term power time series with the length of m; the sort function represents sorting the time series data from large to small in value,
Figure FDA0003478305040000051
representing wind-solar scenes
Figure FDA0003478305040000052
After the dayn times are expanded, the time sequence data are arranged from large to small according to the numerical value, the length is consistent with the length of the original sequence,
Figure FDA0003478305040000053
representing a photovoltaic daily scene
Figure FDA0003478305040000054
After the dayn times are expanded, the time sequence data are arranged from large to small according to the numerical value, the length is consistent with the length of the original sequence,
Figure FDA0003478305040000055
indicating loading daily scenes
Figure FDA0003478305040000056
After the dayn times are expanded, arranging the time sequence data from large to small according to the numerical value, and enabling the length to be consistent with the length of the original sequence;
s4, defining a value function
Figure FDA0003478305040000057
S0Is an artificially set initial state, the initial state S0Is in an amount of
Figure FDA0003478305040000058
Defining an optimum function
Figure FDA0003478305040000059
Solving using Bellman equations
Figure FDA00034783050400000510
Obtaining an optimal strategy
Figure FDA00034783050400000511
S5, setting all initial states S0Respectively substitute for
Figure FDA00034783050400000512
In the equation, | S | equations are obtained in total, thereby obtaining S in different initial states0Day scene combination strategy pi of optimal wind power, photovoltaic and load*(S0)；
S6, generating a Markov aggregation state matrix of 3 XN order consisting of wind power, photovoltaic and load through Gibbs sampling by using a Markov Monte Carlo method:
Figure FDA00034783050400000513
wherein T represents a Markov polymerization state matrix of order 3 XN composed of wind power, photovoltaic and load,
Figure FDA00034783050400000514
s7, matching each group of state vectors in the Markov state matrix with the mapping of the optimal strategy to obtain a day scene matrix
Figure FDA0003478305040000061
Wherein
Figure FDA0003478305040000062
Figure FDA0003478305040000063
Figure FDA0003478305040000064
S8, judging whether the difference value of the head-tail connection position between the adjacent day scenes of each row of the day scene matrix is larger than the first-order difference maximum value of the medium-long term power time sequence, if so, performing wavelet filtering processing on the head-tail connection position until the difference value is smaller than the first-order difference maximum value, if not, directly connecting the head and the tail of the day scenes, and finally generating three aggregation sequences formed by N day scenes.
2. The method of claim 1, wherein the dayn is determined by a total number of days included in the medium and long term historical power sequences.
3. The method for aggregating multi-dimensional time-series data considering spatio-temporal correlation according to claim 1, wherein n iswi、npiAnd nliObtained by calculation through a neighbor propagation clustering algorithm.
CN201910932946.4A2019-09-292019-09-29Aggregation method of multidimensional time series data considering space-time correlationActiveCN110909911B (en)

Priority Applications (1)

Application NumberPriority DateFiling DateTitle
CN201910932946.4ACN110909911B (en)2019-09-292019-09-29Aggregation method of multidimensional time series data considering space-time correlation

Applications Claiming Priority (1)

Application NumberPriority DateFiling DateTitle
CN201910932946.4ACN110909911B (en)2019-09-292019-09-29Aggregation method of multidimensional time series data considering space-time correlation

Publications (2)

Publication NumberPublication Date
CN110909911A CN110909911A (en)2020-03-24
CN110909911Btrue CN110909911B (en)2022-05-20

Family

ID=69815220

Family Applications (1)

Application NumberTitlePriority DateFiling Date
CN201910932946.4AActiveCN110909911B (en)2019-09-292019-09-29Aggregation method of multidimensional time series data considering space-time correlation

Country Status (1)

CountryLink
CN (1)CN110909911B (en)

Families Citing this family (9)

* Cited by examiner, † Cited by third party
Publication numberPriority datePublication dateAssigneeTitle
CN111523230B (en)*2020-04-222023-05-26国网能源研究院有限公司Self-adaptive clustering method for wind-light load composite typical scene
CN112149346B (en)*2020-09-072024-04-26华中科技大学 A wind farm equivalent modeling method, device, electronic device and storage medium
CN113077166B (en)*2021-04-162024-06-28国网吉林省电力有限公司Community energy storage scheduling method based on Markov decision process
CN113128063B (en)*2021-04-272024-08-16国网北京市电力公司 A typical load scenario generation method and system for ice and snow sports venues
CN113128064B (en)*2021-04-272023-10-24国网北京市电力公司Thermoelectric data aggregation method, system, device and storage medium for simulation
CN113343491A (en)*2021-06-302021-09-03东南大学Random scene analysis method considering time sequence autocorrelation and cross correlation
CN113807019A (en)*2021-09-242021-12-17清华大学 MCMC wind power simulation method based on improved scene classification and de-coarse-graining
CN114358656A (en)*2022-02-252022-04-15华北电力大学 Method and system for extracting typical scenarios of new energy output considering space-time correlation
CN117370825B (en)*2023-10-112024-06-21国网经济技术研究院有限公司 Long-term scene generation method and system based on attention conditional generative adversarial network

Citations (2)

* Cited by examiner, † Cited by third party
Publication numberPriority datePublication dateAssigneeTitle
CN109524993A (en)*2018-12-192019-03-26中国农业大学The typical week power output scene generating method of wind-powered electricity generation photovoltaic for Mid-long Term Optimized Scheduling
CN109783841A (en)*2018-11-282019-05-21河海大学A kind of photovoltaic power output time series analogy method based on more scene state transfer matrixes and conditional probability sampling

Family Cites Families (1)

* Cited by examiner, † Cited by third party
Publication numberPriority datePublication dateAssigneeTitle
US20090228516A1 (en)*2008-03-102009-09-10The Hong Kong Polytechnic UniversityHarmful algae indexing (HaiDex) method

Patent Citations (2)

* Cited by examiner, † Cited by third party
Publication numberPriority datePublication dateAssigneeTitle
CN109783841A (en)*2018-11-282019-05-21河海大学A kind of photovoltaic power output time series analogy method based on more scene state transfer matrixes and conditional probability sampling
CN109524993A (en)*2018-12-192019-03-26中国农业大学The typical week power output scene generating method of wind-powered electricity generation photovoltaic for Mid-long Term Optimized Scheduling

Non-Patent Citations (2)

* Cited by examiner, † Cited by third party
Title
Quantifying parameter uncertainty in a coral reef model using Metropolis-Coupled Markov Chain Monte Carlo;Clancy,D 等;《ECOLOGICAL MODELLING》;20100524;第221卷(第10期);全文*
考虑气象因素的短期光伏出力预测的奇异谱分析方法;黎静华 等;《电力自动化设备》;20180504;全文*

Also Published As

Publication numberPublication date
CN110909911A (en)2020-03-24

Similar Documents

PublicationPublication DateTitle
CN110909911B (en)Aggregation method of multidimensional time series data considering space-time correlation
CN107301472B (en) Distributed photovoltaic planning method based on scene analysis method and voltage regulation strategy
CN106529814B (en) A distributed photovoltaic ultra-short-term forecast method based on Adaboost clustering and Markov chain
CN109524993A (en)The typical week power output scene generating method of wind-powered electricity generation photovoltaic for Mid-long Term Optimized Scheduling
CN105741193A (en)Multi-target distribution network reconstruction method considering distributed generation and load uncertainty
CN108075471B (en) Multi-objective constraint optimization grid dispatch strategy based on stochastic power output forecasting
CN111092451B (en)Photovoltaic consumption early warning method based on power distribution network grid
CN113270898A (en)Distributed photovoltaic optimization configuration method based on two-stage planning
CN107679687A (en)A kind of photovoltaic output modeling method and Generation System Reliability appraisal procedure
Jin et al.Wind and photovoltaic power time series data aggregation method based on an ensemble clustering and Markov chain
Liao et al.Two-layer optimization configuration method for distributed photovoltaic and energy storage systems based on IDEC-K clustering
CN110011358B (en)Distribution network load state adjustment controller
CN117132132A (en)Photovoltaic power generation power prediction method based on meteorological data
CN110084430A (en)A method of considering space-time characterisation design distributed photovoltaic power output prediction model
CN112052996B (en)Small hydropower station cluster power prediction method based on convolutional neural network technology
CN119988925A (en) A method, medium and system for generating extreme weather scenes based on CGAN algorithm
CN117592255A (en)Wind-light output sequence modeling method considering space-time correlation
CN117574197A (en) Data-driven photovoltaic power plant cluster division method
CN108321840A (en)The grid-connected logout selection method contributed based on photo-voltaic power generation station fining
CN117060407B (en) A wind power cluster power prediction method and system based on similar day classification
CN105914752A (en)Pilot node selection method based on clustering by fast search and density peaks
CN109149644B (en) An online strategy matching and collaborative optimization method for optical storage integration based on big data analysis
CN117748444A (en)Operation simulation method of power distribution system
Shi et al.Short-term photovoltaic power forecast based on long short-term memory network
CN108233373B (en) A Probabilistic Harmonic Analysis Method for Distributed Photovoltaic Access to Distribution Network Considering Weather Scenarios

Legal Events

DateCodeTitleDescription
PB01Publication
PB01Publication
SE01Entry into force of request for substantive examination
SE01Entry into force of request for substantive examination
GR01Patent grant
GR01Patent grant
[8]ページ先頭
©2009-2025 Movatter.jp