CN111861259B - Load modeling method, system and storage medium considering time sequence - Google Patents

Load modeling method, system and storage medium considering time sequence Download PDF

Info

Publication number
CN111861259B
CN111861259B CN202010751650.5A CN202010751650A CN111861259B CN 111861259 B CN111861259 B CN 111861259B CN 202010751650 A CN202010751650 A CN 202010751650A CN 111861259 B CN111861259 B CN 111861259B
Authority
CN
China
Prior art keywords
load
data
time point
daily
days
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
CN202010751650.5A
Other languages
Chinese (zh)
Other versions
CN111861259A (en
Inventor
王琨玥
戴申华
纪胜强
刘尧
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.)
Datang Boiler Pressure Vessel Examination Center Co Ltd
East China Electric Power Test Institute of China Datang Corp Science and Technology Research Institute Co Ltd
Original Assignee
Datang Boiler Pressure Vessel Examination Center Co Ltd
East China Electric Power Test Institute of China Datang Corp Science and Technology Research Institute 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 Datang Boiler Pressure Vessel Examination Center Co Ltd, East China Electric Power Test Institute of China Datang Corp Science and Technology Research Institute Co Ltd filed Critical Datang Boiler Pressure Vessel Examination Center Co Ltd
Priority to CN202010751650.5A priority Critical patent/CN111861259B/en
Publication of CN111861259A publication Critical patent/CN111861259A/en
Application granted granted Critical
Publication of CN111861259B publication Critical patent/CN111861259B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q10/00—Administration; Management
    • G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
    • G06Q10/063—Operations research, analysis or management
    • G06Q10/0639—Performance analysis of employees; Performance analysis of enterprise or organisation operations
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F17/00—Digital computing or data processing equipment or methods, specially adapted for specific functions
    • G06F17/10—Complex mathematical operations
    • G06F17/16—Matrix or vector computation, e.g. matrix-matrix or matrix-vector multiplication, matrix factorization
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F17/00—Digital computing or data processing equipment or methods, specially adapted for specific functions
    • G06F17/10—Complex mathematical operations
    • G06F17/18—Complex mathematical operations for evaluating statistical data, e.g. average values, frequency distributions, probability functions, regression analysis
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F30/00—Computer-aided design [CAD]
    • G06F30/20—Design optimisation, verification or simulation
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
    • G06Q50/06—Energy or water supply
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F2111/00—Details relating to CAD techniques
    • G06F2111/08—Probabilistic or stochastic CAD

Landscapes

  • Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • Business, Economics & Management (AREA)
  • Data Mining & Analysis (AREA)
  • Mathematical Physics (AREA)
  • Human Resources & Organizations (AREA)
  • Mathematical Analysis (AREA)
  • Computational Mathematics (AREA)
  • Mathematical Optimization (AREA)
  • Pure & Applied Mathematics (AREA)
  • Economics (AREA)
  • Strategic Management (AREA)
  • General Engineering & Computer Science (AREA)
  • Development Economics (AREA)
  • Educational Administration (AREA)
  • Algebra (AREA)
  • Health & Medical Sciences (AREA)
  • Entrepreneurship & Innovation (AREA)
  • Marketing (AREA)
  • Databases & Information Systems (AREA)
  • General Business, Economics & Management (AREA)
  • Tourism & Hospitality (AREA)
  • Operations Research (AREA)
  • Software Systems (AREA)
  • Management, Administration, Business Operations System, And Electronic Commerce (AREA)
  • Evolutionary Computation (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Public Health (AREA)
  • Probability & Statistics with Applications (AREA)
  • Water Supply & Treatment (AREA)
  • Computer Hardware Design (AREA)
  • Bioinformatics & Cheminformatics (AREA)
  • Geometry (AREA)
  • General Health & Medical Sciences (AREA)
  • Evolutionary Biology (AREA)
  • Primary Health Care (AREA)
  • Game Theory and Decision Science (AREA)
  • Quality & Reliability (AREA)

Abstract

The invention relates to a load modeling method considering time sequence, which comprises the steps of 1, obtaining load historical data of the same type to form a load data matrix L, wherein the load at the jth moment of the ith day is LijThe data volume is m days, and the data time scale is n time points every day; step 2, acquiring load data of all days at each moment as a set, and recording the set as T1，T2，…,Tj,…,Tn(ii) a Then TjI.e., set { L }1j,L2j,…,Lij,…,Lmj}; step 3, respectively processing the data of each set to obtain the data of each setCumulative distribution function, noted FTj(x) And its inverse function, denoted KTj(x) Then n total number of FTj(x) And n number of KTj(x) (ii) a Step 4, sampling a number between 0 and 1 in a 0-1 uniform distribution way, recording the number as u, and respectively substituting the u into a function KTj(x) In order to obtain KT1(u),KT2(u),…,KTj(u),…,KTnAnd (u) is a reference value for simulating each time point of the daily load. After the inverse function of the cumulative distribution function at each time point is obtained, only one random number between 0 and 1 is needed to be directly substituted to obtain the load simulation data of one day, and the calculation time and the cost are both greatly reduced.

Description

Load modeling method, system and storage medium considering time sequence
Technical Field
The invention relates to the technical field of load simulation, in particular to a load modeling method, a load modeling system and a storage medium considering time sequence.
Background
Load modeling occupies an extremely important position in power system planning. Since the historical data of the load is fixed, all possible load conditions are difficult to consider in simulation operation by only relying on the historical data. Therefore, the method can accurately grasp the load characteristics of one region and has great significance on the reliability of the power system planning.
The traditional load modeling has two directions, a probability model and a time sequence model. The probability model mainly uses a Monte Carlo method to obtain load data by random sampling, has no time sequence and is mainly suitable for calculation of risk assessment. The time sequence model is mainly suitable for the problems of power grid planning and the like, a peak-to-load ratio method is used for mainstream time sequence load modeling, three groups of coefficients of year, week and day are extracted from load historical data, and the three groups of coefficients are multiplied by the maximum load to obtain simulation data. Although the peak-to-load ratio method can reproduce the time sequence characteristics of the load, the daily load change of the method is fixed and is essentially the copy of the typical daily load, so the method cannot reproduce the whole probability distribution characteristics of the load, does not have the random fluctuation of the load, and has poor accuracy in the power system simulation.
The short-term load prediction method and the short-term load prediction model training method and device disclosed as 202010341288.4 disclose that learning is performed through a neural network by relying on a database, and then the load at the next moment is presumed from the current load characteristics, and the method is suitable for load prediction limited to a few minutes or a few moments because of unsupervised learning and no correction function; so if the method is used for long-term evolution, accurate load simulation cannot be obtained.
With the continuous increase of new energy grid-connected proportion, the time sequence characteristics of the power system simulation are more and more emphasized, and the traditional peak-to-load ratio method is gradually difficult to adapt to the requirements. A new time sequence load simulation method is needed, which can not only reproduce the time sequence characteristics of the load in the area, but also combine the probability distribution and random fluctuation of the load.
Disclosure of Invention
The invention aims to provide a daily load simulation method with small calculation amount and low cost.
The invention solves the technical problems through the following technical means:
a load modeling method considering time sequence includes the following steps:
step 1, obtaining load historical data of the same type to form a load data matrix L, wherein the load at the jth moment of the ith day is LijThe data volume is m days, and the data time scale is n time points every day;
Figure BDA0002610188660000021
step 2, acquiring load data of all days at each moment as a set, and recording the set as T1，T2，…,Tj,…,Tn(ii) a Then TjI.e. set { L1j,L2j,…,Lij,…,Lmj}；
Tj＝{L1j,L2j,...,Lij,....,Lmj} (j＝1,2,...,n) (2)
Step 3, respectively processing the data of each set to obtain the cumulative distribution function of each set, and marking as FTj(x) And its inverse function, denoted KTj(x) Then n total number of FTj(x) And n number of KTj(x)；
FTj(x)＝CDF(Tj) (j＝1,2,...,n) (3)
KTj(x)＝FTj -1(x) (j＝1,2,...,n) (4)
Step 4, sampling a number between 0 and 1 in a 0-1 uniform distribution way, recording the number as u, and respectively substituting the u into a function KTj(x) In order to obtain KT1(u),KT2(u),…,KTj(u),…,KTnAnd (u) is a reference value for simulating each time point of the daily load.
The method uses the historical data of the single type load to count the distribution rule, respectively calculates the inverse function of the cumulative distribution function of each time point according to the time points, counts the error distribution rule, and then uses only the uniform sampling between 0 and 1 to substitute the inverse function, so as to obtain the simulated load with high accuracy and daily change rule of the load. Compared with the traditional load simulation method, after modeling is completed, the method has the advantages of simple sampling, low calculation cost, high accuracy and wide applicability, has a daily load time sequence change rule and a load probability distribution condition, and has better effects in various application scenes such as power supply planning, risk evaluation and the like.
Further, the method also comprises an error eliminating step, which specifically comprises the following steps:
step 5, all load historical data LijRespectively substituted into corresponding FTj(x) In (1), obtain data YijSaid Y isijRepresenting the load of a certain time point, and representing the magnitude relation among all load values of the time point all the year round; total of m x n Yij；
Yij＝FTj(Lij) (i＝1,2,...,m；j＝1,2,...,n) (5)
Step 6, obtaining daily YijAverage value of AiSubstituted into KTj(x) In (1), a daily load reference value is obtained and is marked as LA ijObtaining a load standard value matrix LA；
Figure BDA0002610188660000031
Figure BDA0002610188660000032
Figure BDA0002610188660000033
7, subtracting a load standard value matrix L from the load historical data matrix LAObtaining a simulation error, which is marked as C, and counting the error matrix C to obtain a probability density function of the error, which is marked as H (x);
Figure BDA0002610188660000034
H(x)＝CDF(Cij) (i＝1,2,...,m；j＝1,2,...,n) (10)
step 8, sampling the number between 0 and 1 in the same uniform distribution of 0 to 1, sampling n times in total, and recording as p1,p2,…,pj,…,pnSubstitution into H (x) to give H (p)1),H(p2),…,H(pj),…,H(pn) (ii) a The random fluctuation value of each time point of the simulated daily load is obtained;
step 9, adding the reference value of each time point of the simulated daily load and the random fluctuation value to obtain KT1(u)+H(p1),KT2(u)+H(p2),…,KTj(u)+H(pj),…,KTn(u)+H(pn) Is the sampledThe results of the daily load simulation of (a);
further, if a load simulation result of multiple days is needed, repeating the steps 4, 8 and 9 according to the needed days, and obtaining the simulation daily load data of the number of repeated times.
The invention also provides a load modeling system considering time sequence, which comprises
A load data matrix construction module for obtaining the same type of load historical data to form a load data matrix L, wherein the load at the jth moment of the ith day is LijThe data volume is m days, and the data time scale is n time points every day;
Figure BDA0002610188660000041
a load data set construction module for all days at each moment, wherein the load data of all days at each moment is acquired as a set which is marked as T1，T2，…,Tj,…,Tn(ii) a Then TjI.e., set { L }1j,L2j,…,Lij,…,Lmj}；
Tj＝{L1j,L2j,...,Lij,....,Lmj} (j＝1,2,...,n) (2)
An inverse function construction module of the cumulative distribution function, which respectively processes the data of each set to obtain the cumulative distribution function of each set, and records the cumulative distribution function as FTj(x) And its inverse function, denoted KTj(x) Then n total number of FTj(x) And n number of KTj(x)；
FTj(x)＝CDF(Tj) (j＝1,2,...,n) (3)
KTj(x)＝FTj -1(x) (j＝1,2,...,n) (4)
A reference value calculation module for simulating each time point of daily load, which samples a number between 0 and 1 in a 0-1 uniform distribution way and is recorded as u, and the u is respectively substituted into a function KTj(x) In (b) to obtain KT1(u),KT2(u),…,KTj(u),…,KTnAnd (u) is a reference value for simulating each time point of the daily load.
Further, also comprises
A calculation module for calculating the size relationship of all loads at a certain moment in the whole period of the momentijRespectively substituted into corresponding FTj(x) In (1), obtain data YijSaid Y isijRepresenting the load at a certain time point, and the magnitude relation among all load values at the time point in the whole year is m x n Yij；
Yij＝FTj(Lij) (i＝1,2,...,m；j＝1,2,...,n) (5)
A daily load standard value calculation module for calculating Y of each dayijAverage value of AiSubstituted into KTj(x) In (1), a daily load reference value is obtained and is marked as LA ijObtaining a load standard value matrix LA；
Figure BDA0002610188660000042
Figure BDA0002610188660000043
Figure BDA0002610188660000051
An error probability density function calculation module for subtracting a load standard value matrix L from the load historical data matrix LAObtaining a simulation error, which is marked as C, and counting the error matrix C to obtain a probability density function of the error, which is marked as H (x);
Figure BDA0002610188660000052
H(x)＝CDF(Cij) (i＝1,2,...,m；j＝1,2,...,n) (10)
the random fluctuation value calculation module samples the number between 0 and 1 in the same 0 to 1 uniform distribution for n times, and the number is marked as p1,p2,…,pj,…,pnSubstitution into H (x) to give H (p)1),H(p2),…,H(pj),…,H(pn) (ii) a The random fluctuation value of each time point of the simulated daily load is obtained;
the daily load simulation structure calculation module adds the reference value and the random fluctuation value of each time point of the simulated daily load to obtain KT1(u)+H(p1),KT2(u)+H(p2),…,KTj(u)+H(pj),…,KTn(u)+H(pn) The sampled load simulation result of the day is obtained;
if the load simulation result of multiple days is needed, the steps of the reference value calculation module, the random fluctuation value calculation module and the daily load simulation structure calculation module of each time point of the simulated daily load are repeatedly executed according to the needed days, and the simulated daily load data with the number of the repeated times can be obtained.
The present invention also provides a storage medium having stored therein a plurality of instructions adapted to be loaded and executed by a processor, the plurality of instructions comprising:
constructing a load data matrix, acquiring the historical load data of the same type to form a load data matrix L, wherein the load at the jth moment of the ith day is LijThe data volume is m days, and the data time scale is n time points every day;
Figure BDA0002610188660000061
constructing a load data set of all days at each moment, acquiring the load data of all days at each moment as a set, and recording the set as T1，T2，…,Tj,…,Tn(ii) a Then TjI.e. set { L1j,L2j,…,Lij,…,Lmj}；
Tj＝{L1j,L2j,...,Lij,....,Lmj} (j＝1,2,...,n) (2)
Constructing an inverse function of the cumulative distribution function, and respectively processing the data of each set to obtain the cumulative distribution function of each set, and recording the cumulative distribution function as FTj(x) And its inverse function, denoted KTj(x) Then n total number of FTj(x) And n number of KTj(x)；
FTj(x)＝CDF(Tj) (j＝1,2,...,n) (3)
KTj(x)＝FTj -1(x) (j＝1,2,...,n) (4)
Calculating the reference value of each time point of the simulated daily load, sampling a number between 0 and 1 in a 0-1 uniform distribution mode, recording the number as u, and respectively substituting the u into a function KTj(x) In order to obtain KT1(u),KT2(u),…,KTj(u),…,KTnAnd (u) is a reference value for simulating each time point of the daily load.
Further, the instructions may also include
Calculating the magnitude relation of all loads of a certain moment in the year, and storing all load historical data LijRespectively substituted into corresponding FTj(x) In (1), obtain data YijSaid Y isijRepresenting the load at a certain time point, and the magnitude relation among all the load values at the time point in the whole year totals m x n Yij；
Yij＝FTj(Lij) (i＝1,2,...,m；j＝1,2,...,n) (5)
A daily load standard value calculation module for calculating Y of each dayijAverage value of AiSubstituted into KTj(x) In (1), a daily load reference value is obtained and is marked as LA ijObtaining a load standard value matrix LA；
Figure BDA0002610188660000062
Figure BDA0002610188660000063
Figure BDA0002610188660000071
Calculating error probability density function by subtracting load standard value matrix L from load historical data matrix LAObtaining a simulation error, which is marked as C, and counting an error matrix C to obtain a probability density function of the error, which is marked as H (x);
Figure BDA0002610188660000072
H(x)＝CDF(Cij) (i＝1,2,...,m；j＝1,2,...,n) (10)
calculating random fluctuation value, uniformly distributing and sampling the number between 0 and 1 in the same way by 0 to 1, sampling n times, and recording as p1,p2,…,pj,…,pnSubstitution into H (x) to give H (p)1),H(p2),…,H(pj),…,H(pn) (ii) a The random fluctuation value of each time point of the simulated daily load is obtained;
calculating a daily load simulation structure, and adding the reference value of each time point of the simulated daily load and the random fluctuation value to obtain KT1(u)+H(p1),KT2(u)+H(p2),…,KTj(u)+H(pj),…,KTn(u)+H(pn) The sampled load simulation result of the day is obtained;
further, the instructions further comprise
And if the multi-day load simulation result is needed, repeatedly executing the reference value calculation, the random fluctuation value calculation and the daily load simulation structure calculation instruction of each time point of the simulated daily load according to the needed days, and obtaining the simulated daily load data with the number of times of repetition.
The invention has the advantages that:
according to the method, the distribution rule of the single-type load is counted by using historical data of the single-type load, the inverse function of the cumulative distribution function of each time point is obtained according to the time points, the error distribution rule of the single-type load is counted, and then the single-type load can be simulated by only uniformly sampling between 0 and 1 and substituting the uniform sampling into the obtained inverse function, so that the high-accuracy and daily change rule of the load can be considered. Compared with the traditional load simulation method, after modeling is completed, the method has the advantages of simple sampling, low calculation cost, high accuracy and wide applicability, has a daily load time sequence change rule and a load probability distribution condition, and has better effects in various application scenes such as power supply planning, risk evaluation and the like.
The method provided by the invention has wide application range, can sample the load data of one day, one year and many years, and can also sample point by point to obtain the data of one moment. The method can be used for simulation of load flow calculation and calculation of regional power grid capacity planning.
Drawings
FIG. 1 is an opaque scatter plot of a one-year load of raw data for a case in an embodiment of the present invention;
FIG. 2 is an opaque scatter diagram of a year load of data simulated by a load modeling method considering time sequence in the embodiment of the present invention;
FIG. 3 is a one-year load opacity scattergram of case data simulated by a conventional method according to an embodiment of the present invention;
fig. 4 is a probability density function comparison graph of the load modeling method of case-considered time sequence, the conventional method and the original data in the embodiment of the present invention.
Detailed Description
In order to make the objects, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the embodiments of the present invention, and it is obvious that the described embodiments are some embodiments of the present invention, but not all embodiments. All other embodiments, which can be obtained by a person skilled in the art without inventive step based on the embodiments of the present invention, are within the scope of protection of the present invention.
A load modeling method considering time sequence comprises the following steps:
step 1, acquiring load historical data of the same type in the same region, wherein the historical data needs to be in units of years and needs load historical data of at least one year. The historical data is required to ensure that the load is a single type of load in the same region, such as commercial load, industrial load, residential load and the like. If the load is composed of complicated types and is the case with various types of loads, the method is applicable but the accuracy is reduced. The load point-taking time interval can be determined according to the requirement, the smaller the time interval, the larger the data volume, the initial modeling time and cost are increased, but the accuracy is improved. Conversely, the larger the time interval, the easier the modeling and the lower the accuracy. Generally, 15min can be taken, and a load data matrix L is constructed by taking load data of one year as an example. And 365 days are taken one year, 96 time points are taken one day at intervals of 15min, and then L is a matrix of 365 rows and 96 columns, which is shown in a formula (1).
Figure BDA0002610188660000091
In the formula, LijLoad history data of i days and j time; m is the total days of the taken historical data, 365 days are taken in one year, and m is equal to 365; n is the total time of the taken historical data every day, namely the number of sample points every day, in the embodiment, the time interval is 15min, and then n is 96;
step 2, acquiring load data of all days at each moment as a set, and recording the set as T1，T2，…,Tj,…,TnIn the embodiment, the interval is 15min, so that 96 time points are counted in one day and 96 sets are counted. Set of time j times TjI.e. set { L1j,L2j,…,Lij,…,Lmj}。
Tj＝{L1j,L2j,...,Lij,....,Lmj} (j＝1,2,...,n) (2)
Step 3, respectively processing the data of each set to obtain the cumulative distribution function of each set, and marking as FTj(x) And its inverse function, denoted KTj(x) Then the total number of the embodiments is 96FTj(x) And 96KTj(x)。
FTj(x)＝CDF(Tj) (j＝1,2,...,n) (3)
KTj(x)＝FTj -1(x) (j＝1,2,...,n) (4)
Step 4, sampling a number between 0 and 1 in a 0-1 uniform distribution way, recording the number as u, and respectively substituting the u into a function KTj(x) In (b) to obtain KT1(u),KT2(u),…,KTj(u),…,KTn(u) of the formula (I). The standard value of each time point of the simulated daily load is obtained.
The method uses the historical data of the single type load to count the distribution rule, respectively calculates the inverse function of the cumulative distribution function of each time point according to the time points, counts the error distribution rule, and then uses only the uniform sampling between 0 and 1 to substitute the inverse function, so as to obtain the simulated load with high accuracy and daily change rule of the load. Compared with the traditional load simulation method, after modeling is completed, the method has the advantages of simple sampling, low calculation cost, high accuracy and wide applicability, has a daily load time sequence change rule and a load probability distribution condition, and has better effects in various application scenes such as power supply planning, risk evaluation and the like.
To improve the accuracy, the present embodiment also eliminates the error by:
step 5, all load historical data LijRespectively substituted into corresponding FTj(x) In (1), obtain data YijYIj represents the load at a certain moment, and the magnitude relation among all load values at the moment in the whole year; total of m x n YijIn the embodiment, 96 time points per day for 365 days are 365 by 96, namely 350404Yij；
Yij＝FTj(Lij) (i＝1,2,...,m；j＝1,2,...,n) (5)
Step 6, finding out daily YijAverage value of AiAs shown in formula (6). Then A is mixediSubstituted into corresponding KTj(x) In (1), a daily load reference value is obtained and is marked as LA ijAs shown in formula (7). Obtaining a load standard value matrix LAMoment of embodimentThe array has 365 rows and 96 columns, as shown in equation (8).
Figure BDA0002610188660000101
Figure BDA0002610188660000102
Figure BDA0002610188660000103
7, subtracting a load standard value matrix L from the load historical data matrix LAAnd obtaining a simulation error which is marked as C, counting the error matrix C to obtain a probability density function of the error which is marked as H (x), and generally adopting Gaussian distribution.
Figure BDA0002610188660000104
H(x)＝CDF(Cij) (i＝1,2,...,m；j＝1,2,...,n) (10)
Step 8, sampling the number between 0 and 1 in the same uniform distribution of 0 to 1, sampling n times in total, and recording as p1,p2,…,pj,…,pnSubstitution into H (x) to give H (p)1),H(p2),…,H(pj),…,H(pn). Namely the random fluctuation value of each time point of the simulated daily load.
Step 9, adding the standard value of each time point of the simulated daily load and the random fluctuation value to obtain KT1(u)+H(p1),KT2(u)+H(p2),…,KTj(u)+H(pj),…,KTn(u)+H(pn) I.e. the sampled load simulation results for the day.
And 10, if the load simulation result of multiple days is needed, repeating the steps 4, 8 and 9 according to the needed days, and obtaining the simulation daily load data with the number of repeated times.
Case(s)
Fig. 1, 2, and 3 are opaque scattergram diagrams, in which the abscissa is time, the ordinate is load data, the color depth represents the occurrence frequency of the load data, the color depth represents more occurrence frequency, and the color depth represents less occurrence frequency, and the time sequence characteristics and distribution of the data can be seen from the opaque scattergram diagrams. Comparing fig. 1, fig. 2, and fig. 3, it can be seen that the simulated load result of the proposed method is closer to the original data, while the conventional method has a time-series variation trend, but the load probability distribution situation does not reappear. FIG. 4 is a comparison graph of probability density functions of simulated annual load data and original data of the proposed method, which is basically consistent with the probability density of the original data, and the conventional method, which has a large gap and cannot reproduce the probability distribution of the load original data.
The embodiment also provides a load modeling system considering time sequence, which comprises
And the load data matrix construction module is used for acquiring the load historical data of the same type in the same region, wherein the historical data needs to be in units of years and needs to be at least one year of load historical data. The historical data is required to ensure that the load is a single type of load in the same region, such as a commercial load, an industrial load, a residential load and the like. If the load is composed of complicated types and is the case with various types of loads, the method is applicable but the accuracy is reduced. The load point-taking time interval can be determined according to the requirement, the smaller the time interval, the larger the data volume, the initial modeling time and cost are increased, but the accuracy is improved. Whereas the larger the time interval, the easier the modeling and the lower the accuracy. Generally, 15min can be taken, and a load data matrix L is constructed by taking load data of one year as an example. And (3) taking 365 days in one year, taking 15min as an interval, and taking 96 time points every day, wherein L is a matrix of 365 rows and 96 columns, and is shown in a formula (1).
Figure BDA0002610188660000111
In the formula, LijLoad history data of i days and j time; m is total days of the taken historical data, 365 days are taken in one year, and m is equal to 365; n is takenThe total time number of the historical data in each day is the number of sample points in each day, in the embodiment, the time interval is 15min, and then n is 96;
a load data set construction module for all days at each moment, wherein the load data of all days at each moment is acquired as a set which is marked as T1，T2，…,Tj,…,TnIn the embodiment, the interval is 15min, so that 96 time points are counted in one day and 96 sets are counted. Set of times j TjI.e. set { L1j,L2j,…,Lij,…,Lmj}。
Tj＝{L1j,L2j,...,Lij,....,Lmj} (j＝1,2,...,n) (2)
And the inverse function construction module of the cumulative distribution function respectively processes the data of each set to obtain the cumulative distribution function of each set, and the cumulative distribution function is marked as FTj(x) And its inverse function, denoted KTj(x) Then the total number of the embodiments is 96FTj(x) And 96KTj(x)。
FTj(x)＝CDF(Tj) (j＝1,2,...,n) (3)
KTj(x)＝FTj -1(x) (j＝1,2,...,n) (4)
A reference value calculation module for simulating each time point of daily load, uniformly distributing and sampling a number between 0 and 1, which is recorded as u, by 0 to 1, and respectively substituting the u into a function KTj(x) In (b) to obtain KT1(u),KT2(u),…,KTj(u),…,KTn(u) of the formula (I). Namely the reference value of each time point of the simulated daily load.
The method uses the historical data of the single type load to count the distribution rule, respectively calculates the inverse function of the cumulative distribution function of each time point according to the time points, counts the error distribution rule, and then uses only the uniform sampling between 0 and 1 to substitute the inverse function, so as to obtain the simulated load with high accuracy and daily change rule of the load. Compared with the traditional load simulation method, after modeling is completed, the method has the advantages of simple sampling, low calculation cost, high accuracy, wide applicability, daily load time sequence change rule and load probability distribution condition, and better effect in various application scenes such as power supply planning, risk evaluation and the like.
To improve the accuracy, the present embodiment also eliminates errors by the following modules:
a calculation module for calculating the magnitude relation of all loads of a certain moment in the whole period of the momentijRespectively substituted into corresponding FTj(x) In (1), obtain data YijYIj represents the load at a certain time point, and the magnitude relation among all load values at the time point in the whole year; total of m x n YijIn the embodiment, 96 time points per day for 365 days are 365 x 96, that is, 350404Yij；
Yij＝FTj(Lij) (i＝1,2,...,m；j＝1,2,...,n) (5)
A daily load standard value calculation module for calculating daily YijAverage value of AiAs shown in formula (6). Then A is mixediSubstituted into corresponding KTj(x) In (1), a daily load reference value is obtained and recorded as LA ijAs shown in formula (7). Obtaining a load standard value matrix LAIn the embodiment, the number of rows in the matrix is 365, and the number of columns in the matrix is 96, as shown in formula (8).
Figure BDA0002610188660000131
Figure BDA0002610188660000132
Figure BDA0002610188660000133
An error probability density function calculation module for subtracting a load standard value matrix L from the load historical data matrix LAAnd obtaining a simulation error, which is marked as C, and counting the error matrix C to obtain a probability density function of the error, which is marked as H (x), and the probability density function is generally in Gaussian distribution.
Figure BDA0002610188660000134
H(x)＝CDF(Cij) (i＝1,2,...,m；j＝1,2,...,n) (10)
The random fluctuation value calculation module samples the number between 0 and 1 in the same 0 to 1 uniform distribution for n times, and the number is marked as p1,p2,…,pj,…,pnSubstitution into H (x) to give H (p)1),H(p2),…,H(pj),…,H(pn). Namely the random fluctuation value of each time point of the simulated daily load.
The daily load simulation structure calculation module is used for adding the standard value and the random fluctuation value of each time point of the simulated daily load to obtain KT1(u)+H(p1),KT2(u)+H(p2),…,KTj(u)+H(pj),…,KTn(u)+H(pn) I.e. the sampled load simulation results for the day.
And if the multi-day load simulation result is needed, the multi-day load simulation result module repeatedly executes the reference value calculation module, the random fluctuation value calculation module and the daily load simulation structure calculation module of each time point of the simulated daily load according to the needed days, so that the simulated daily load data with the number of repeated times can be obtained.
The present embodiment further provides a storage medium, where a plurality of instructions are stored, where the instructions are suitable for being loaded and executed by a processor, and the plurality of instructions are:
constructing a load data matrix, acquiring the historical load data of the same type to form a load data matrix L, wherein the load at the jth moment of the ith day is LijThe data volume is m days, and the data time scale is n time points every day;
Figure BDA0002610188660000141
constructing a load data set of all days at each moment, acquiring the load data of all days at each moment as a set, and recording the set as T1，T2，…,Tj,…,Tn(ii) a Then TjI.e. set { L1j,L2j,…,Lij,…,Lmj}；
Tj＝{L1j,L2j,...,Lij,....,Lmj} (j＝1,2,...,n) (2)
Constructing an inverse function of the cumulative distribution function, and respectively processing the data of each set to obtain the cumulative distribution function of each set, which is marked as FTj(x) And its inverse function, denoted KTj(x) Then n total number of FTj(x) And n number of KTj(x)；
FTj(x)＝CDF(Tj) (j＝1,2,...,n) (3)
KTj(x)＝FTj -1(x) (j＝1,2,...,n) (4)
Calculating the reference value of each time point of the simulated daily load, sampling a number between 0 and 1 in a 0-1 uniform distribution way, recording the number as u, and respectively substituting the u into a function KTj(x) In (b) to obtain KT1(u),KT2(u),…,KTj(u),…,KTnAnd (u) is a reference value for simulating each time point of the daily load.
Calculating the magnitude relation of all loads of a certain moment in the year, and storing all load historical data LijRespectively substituted into corresponding FTj(x) In (1), obtain data YijSaid Y isijRepresenting the load at a certain time point, and the magnitude relation among all the load values at the time point in the whole year totals m x n Yij；
Yij＝FTj(Lij) (i＝1,2,...,m；j＝1,2,...,n) (5)
A daily load standard value calculation module for calculating Y of each dayijAverage value of AiSubstituted into KTj(x) In (1), a daily load reference value is obtained and is marked as LA ijObtaining a load standard value matrix LA；
Figure BDA0002610188660000151
Figure BDA0002610188660000152
Figure BDA0002610188660000153
Calculating error probability density function by subtracting load standard value matrix L from load historical data matrix LAObtaining a simulation error, which is marked as C, and counting the error matrix C to obtain a probability density function of the error, which is marked as H (x);
Figure BDA0002610188660000154
H(x)＝CDF(Cij) (i＝1,2,...,m；j＝1,2,...,n) (10)
calculating random fluctuation value, uniformly distributing and sampling the number between 0 and 1 in the same way by 0 to 1, sampling n times, and recording as p1,p2,…,pj,…,pnSubstitution into H (x) to give H (p)1),H(p2),…,H(pj),…,H(pn) (ii) a The random fluctuation value of each time point of the simulated daily load is obtained;
calculating a daily load simulation structure, and adding the reference value and the random fluctuation value of each time point of the simulated daily load to obtain KT1(u)+H(p1),KT2(u)+H(p2),…,KTj(u)+H(pj),…,KTn(u)+H(pn) The sampled load simulation result of the day is obtained;
and if the multi-day load simulation result is needed, repeatedly executing the reference value calculation, the random fluctuation value calculation and the daily load simulation structure calculation instruction of each time point of the simulated daily load according to the needed days, and obtaining the simulated daily load data with the number of repeated times.
The above examples are only intended to illustrate the technical solution of the present invention, and not to limit it; although the present invention has been described in detail with reference to the foregoing embodiments, it will be understood by those of ordinary skill in the art that: the technical solutions described in the foregoing embodiments may still be modified, or some technical features may be equivalently replaced; and such modifications or substitutions do not depart from the spirit and scope of the corresponding technical solutions of the embodiments of the present invention.

Claims (6)

1. A load modeling method considering time sequence is characterized in that: the method comprises the following steps:
step 1, obtaining the load historical data of the same type to form a load data matrix L, wherein the load at the jth moment of the ith day is LijThe data volume is m days, and the data time scale is n time points every day;
Figure FDA0003565044020000011
step 2, acquiring load data of all days at each moment as a set, and recording the set as T1，T2，…,Tj,…,Tn；
Tj＝{L1j,L2j,...,Lij,....,Lmj}j＝1,2,...,n (2)
Step 3, respectively processing the data of each set to obtain the cumulative distribution function of each set, and marking as FTj(x) And its inverse function, denoted KTj(x) Then n total number of FTj(x) And n number of KTj(x)；
FTj(x)＝CDF(Tj)j＝1,2,...,n (3)
KTj(x)＝FTj -1(x)j＝1,2,...,n (4)
Step 4, sampling a number between 0 and 1 in a 0-1 uniform distribution way, recording the number as u, and respectively substituting the u into a function KTj(x) In (b) to obtain KT1(u),KT2(u),…,KTj(u),…,KTn(u), namely the reference value of each time point of the simulated daily load;
step 5, all load historical data LijRespectively substituted into corresponding FTj(x) In (1), obtain data YijSaid Y isijRepresenting the load of a certain time point, and representing the magnitude relation among all load values of the time point in a full period; total of m x n Yij；
Yij＝FTj(Lij)i＝1,2,...,m；j＝1,2,...,n (5)
Step 6, obtaining daily YijAverage value of AiSubstituted into KTj(x) In (1), a daily load reference value is obtained and recorded as LA ijObtaining a load standard value matrix LA；
Figure FDA0003565044020000012
Figure FDA0003565044020000013
Figure FDA0003565044020000021
7, subtracting a load standard value matrix L from the load historical data matrix LAObtaining a simulation error, which is marked as C, and counting the error matrix C to obtain a probability density function of the error, which is marked as H (x);
Figure FDA0003565044020000022
H(x)＝CDF(Cij)i＝1,2,...,m；j＝1,2,...,n (10)
step 8, sampling the number between 0 and 1 in the same uniform distribution of 0 to 1, sampling n times in total, and recording as p1,p2,…,pj,…,pnSubstitution into H (x) to give H (p)1),H(p2),…,H(pj),…,H(pn) (ii) a The random fluctuation value of each time point of the simulated daily load is obtained;
step 9, adding the reference value of each time point of the simulated daily load and the random fluctuation value to obtain KT1(u)+H(p1),KT2(u)+H(p2),…,KTj(u)+H(pj),…,KTn(u)+H(pn) I.e. the sampled load simulation results for the day.
2. The load modeling method in consideration of temporal sequence according to claim 1, wherein: if the load simulation result of multiple days is needed, repeating the steps 4, 8 and 9 according to the needed days to obtain the simulation daily load data with the number of the repeated times.
3. A load modeling system that considers temporal ordering, characterized by: comprises that
The load data matrix construction module is used for acquiring the historical load data of the same type to form a load data matrix L, wherein the load at the jth moment of the ith day is LijThe data volume is m days, and the data time scale is n time points every day;
Figure FDA0003565044020000031
a load data set construction module for all days at each moment, wherein the load data of all days at each moment is acquired as a set which is marked as T1，T2，…,Tj,…,Tn；
Tj＝{L1j,L2j,...,Lij,....,Lmj}j＝1,2,...,n (2)
And the inverse function construction module of the cumulative distribution function respectively processes the data of each set to obtain the cumulative distribution function of each set, and the cumulative distribution function is marked as FTj(x) And its inverse function, denoted KTj(x) Then n total number of FTj(x) And n number of KTj(x)；
FTj(x)＝CDF(Tj)j＝1,2,...,n (3)
KTj(x)＝FTj -1(x)j＝1,2,...,n (4)
A reference value calculation module for simulating each time point of daily load, which samples a number between 0 and 1 in a 0-1 uniform distribution way and is recorded as u, and the u is respectively substituted into a function KTj(x) In (b) to obtain KT1(u),KT2(u),…,KTj(u),…,KTn(u), namely the reference value of each time point of the simulated daily load;
a calculation module for calculating the size relationship of all loads at a certain moment in the whole period of the momentijRespectively substituted into corresponding FTj(x) In (1), obtain data YijSaid Y isijRepresenting the load of a certain time point, and the magnitude relation of all load values of the time point in the whole period totals m × n Yij；
Yij＝FTj(Lij)i＝1,2,...,m；j＝1,2,...,n (5)
A daily load standard value calculation module for calculating daily YijAverage value of AiSubstituted into KTj(x) In (1), a daily load reference value is obtained and recorded as LA ijObtaining a load standard value matrix LA；
Figure FDA0003565044020000032
Figure FDA0003565044020000033
Figure FDA0003565044020000041
An error probability density function calculation module for subtracting a load standard value matrix L from a load historical data matrix LAObtaining a simulation error, recording as C, and counting the probability of the error obtained by the error matrix CDensity function, denoted as h (x);
Figure FDA0003565044020000042
H(x)＝CDF(Cij)i＝1,2,...,m；j＝1,2,...,n(10)
the random fluctuation value calculation module samples the number between 0 and 1 in the same 0 to 1 uniform distribution for n times, and the number is marked as p1,p2,…,pj,…,pnSubstitution into H (x) to give H (p)1),H(p2),…,H(pj),…,H(pn) (ii) a The random fluctuation value of each time point of the simulated daily load is obtained;
a daily load simulation structure calculation module for adding the reference value and the random fluctuation value of each time point of the simulated daily load to obtain KT1(u)+H(p1),KT2(u)+H(p2),…,KTj(u)+H(pj),…,KTn(u)+H(pn) I.e. the sampled load simulation results for the day.
4. A load modeling system in accordance with claim 3 in which said load modeling system takes into account temporal sequencing: if the load simulation result of multiple days is needed, the steps of simulating the daily load data of the number of times of repetition can be obtained by repeatedly executing the reference value calculation module, the random fluctuation value calculation module and the daily load simulation structure calculation module of each time point of the simulated daily load according to the needed days.
5. A storage medium having stored therein a plurality of instructions adapted to be loaded and executed by a processor, characterized in that: the plurality of instructions are:
constructing a load data matrix, acquiring the historical load data of the same type to form a load data matrix L, wherein the load at the jth moment of the ith day is LijThe data volume is m days, and the data time scale is n time points every day;
Figure FDA0003565044020000051
constructing a load data set of all days at each moment, acquiring the load data of all days at each moment as a set, and recording the set as T1，T2，…,Tj,…,Tn(ii) a Then TjI.e. set { L1j,L2j,…,Lij,…,Lmj}；
Tj＝{L1j,L2j,...,Lij,....,Lmj}j＝1,2,...,n (2)
Constructing an inverse function of the cumulative distribution function, and respectively processing the data of each set to obtain the cumulative distribution function of each set, which is marked as FTj(x) And its inverse function, denoted KTj(x) Then n total FTj(x) And n number of KTj(x)；
FTj(x)＝CDF(Tj)j＝1,2,...,n (3)
KTj(x)＝FTj -1(x)j＝1,2,...,n (4)
Calculating the reference value of each time point of the simulated daily load, sampling a number between 0 and 1 in a 0-1 uniform distribution way, recording the number as u, and respectively substituting the u into a function KTj(x) In order to obtain KT1(u),KT2(u),…,KTj(u),…,KTn(u), namely, the reference value of each time point of the simulated daily load;
calculating the magnitude relation of all loads at a certain moment in a full period, and storing all load historical data LijRespectively substituted into corresponding FTj(x) In (1), obtain data YijSaid Y isijRepresenting the load of a certain time point, and the magnitude relation of all load values of the time point in the whole period totals m × n Yij；
Yij＝FTj(Lij)i＝1,2,...,m；j＝1,2,...,n (5)
A daily load standard value calculation module for calculating daily YijAverage value of AiSubstituted into KTj(x) In (1), a daily load standard is obtainedValue, is noted as LA ijObtaining a load standard value matrix LA；
Figure FDA0003565044020000052
Figure FDA0003565044020000053
Figure FDA0003565044020000061
Calculating error probability density function by subtracting load standard value matrix L from load historical data matrix LAObtaining a simulation error, which is marked as C, and counting an error matrix C to obtain a probability density function of the error, which is marked as H (x);
Figure FDA0003565044020000062
H(x)＝CDF(Cij)i＝1,2,...,m；j＝1,2,...,n (10)
calculating random fluctuation value, uniformly sampling the number between 0 and 1 in the same way by 0 to 1, sampling n times, and recording as p1,p2,…,pj,…,pnSubstitution into H (x) to give H (p)1),H(p2),…,H(pj),…,H(pn) (ii) a The random fluctuation value of each time point of the simulated daily load is obtained;
calculating a daily load simulation structure, and adding the reference value of each time point of the simulated daily load and the random fluctuation value to obtain KT1(u)+H(p1),KT2(u)+H(p2),…,KTj(u)+H(pj),…,KTn(u)+H(pn) I.e. the sampled load simulation results for the day.
6. A storage medium according to claim 5, wherein: the instructions further comprise
And if the multi-day load simulation result is needed, repeatedly executing the reference value calculation, the random fluctuation value calculation and the daily load simulation structure calculation instruction of each time point of the simulated daily load according to the needed days, and obtaining the simulated daily load data with the number of times of repetition.
CN202010751650.5A 2020-07-30 2020-07-30 Load modeling method, system and storage medium considering time sequence Active CN111861259B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202010751650.5A CN111861259B (en) 2020-07-30 2020-07-30 Load modeling method, system and storage medium considering time sequence

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202010751650.5A CN111861259B (en) 2020-07-30 2020-07-30 Load modeling method, system and storage medium considering time sequence

Publications (2)

Publication Number Publication Date
CN111861259A CN111861259A (en) 2020-10-30
CN111861259B true CN111861259B (en) 2022-07-15

Family

ID=72945689

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202010751650.5A Active CN111861259B (en) 2020-07-30 2020-07-30 Load modeling method, system and storage medium considering time sequence

Country Status (1)

Country Link
CN (1) CN111861259B (en)

Families Citing this family (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113296549B (en) * 2021-04-29 2022-07-26 江苏核电有限公司 Deaerator liquid level control system and method based on time sequence control and inverse function operation

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN106557828A (en) * 2015-09-30 2017-04-05 中国电力科学研究院 A kind of long time scale photovoltaic is exerted oneself time series modeling method and apparatus
CN107506331A (en) * 2017-08-25 2017-12-22 国网新疆电力公司经济技术研究院 Micro-capacitance sensor reliability calculation method based on temporal associativity and element run time
CN109767353A (en) * 2019-01-14 2019-05-17 国网江苏省电力有限公司苏州供电分公司 A kind of photovoltaic power generation power prediction method based on probability-distribution function
CN111401757A (en) * 2020-03-19 2020-07-10 西安交通大学 Multi-region power load time sequence scene modeling method considering time-space correlation

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN106557828A (en) * 2015-09-30 2017-04-05 中国电力科学研究院 A kind of long time scale photovoltaic is exerted oneself time series modeling method and apparatus
CN107506331A (en) * 2017-08-25 2017-12-22 国网新疆电力公司经济技术研究院 Micro-capacitance sensor reliability calculation method based on temporal associativity and element run time
CN109767353A (en) * 2019-01-14 2019-05-17 国网江苏省电力有限公司苏州供电分公司 A kind of photovoltaic power generation power prediction method based on probability-distribution function
CN111401757A (en) * 2020-03-19 2020-07-10 西安交通大学 Multi-region power load time sequence scene modeling method considering time-space correlation

Non-Patent Citations (6)

* Cited by examiner, † Cited by third party
Title
Optimization method for multi-access capacity of wind power based on copula;Xiaoxiao Huang etc;《IOP Conference Series: Earth and Environmental Science》;20190228;第227卷(第3期);全文 *
Research on distributed photovoltaic power prediction based on centralized photovoltaic output;Xiaoxiao Huang etc;《IOP Conference Series: Earth and Environmental Science》;20191130;第467卷;全文 *
基于时序生产模拟的配电网光伏装机容量规划;宋士瞻等;《山东大学学报(工学版)》;20181031;第48卷(第5期);全文 *
基于空间相关性的分布式光伏出力预测;张家安等;《电力建设》;20200331;第41卷(第3期);全文 *
考虑光伏不确定性的配电网负荷特性概率评估;黄伟等;《广东电力》;20180709(第06期);全文 *
考虑分布式电源接入的电网源荷时序随机波动特性概率潮流计算;袁晓冬等;《水电能源科学》;20160225(第02期);全文 *

Also Published As

Publication number Publication date
CN111861259A (en) 2020-10-30

Similar Documents

Publication Publication Date Title
Tabone et al. Modeling variability and uncertainty of photovoltaic generation: A hidden state spatial statistical approach
CN113159102B (en) Multi-time-scale photovoltaic power prediction method and system
CN111861259B (en) Load modeling method, system and storage medium considering time sequence
CN115689055A (en) Short-term solar irradiance prediction method and device
CN116128141A (en) Storm surge prediction method and device, storage medium and electronic equipment
CN110728008A (en) Method and device for determining expected life of smart electric meter
CN116316615B (en) Data enhancement-based distributed light Fu Qun short-term power prediction method and system
CN116565861A (en) Power distribution network reliability assessment method, system, equipment and medium
CN112581311B (en) Method and system for predicting long-term output fluctuation characteristics of aggregated multiple wind power plants
CN116316600A (en) Wind power day-ahead prediction self-adaptive filtering method based on seasonal typing
JP4350062B2 (en) Random number generation method according to non-normal distribution and its parameter estimation method
CN116307139A (en) Wind power ultra-short-term prediction method for optimizing and improving extreme learning machine
CN113591399B (en) Short-term wind power prediction method
CN114925931A (en) Platform area load prediction method and system
CN111612289A (en) Electric power system risk assessment method based on new energy multi-scene risk feature guidance
JP3708928B2 (en) Random number generation method according to multivariate non-normal distribution and its parameter estimation method
Mbasso et al. Demand-supply forecasting based on deep learning for electricity balance in Cameroon
CN117495195A (en) Wind power plant group capacity value assessment method based on Johnson distribution system
Pinson et al. Methods for the estimation of the uncertainty of wind power forecasts
De Felice Hydropower information for power system modelling: the JRC-EFAS-Hydropower dataset
Pizarro-Vallejos Use Of The Fission Matrix For Steady State And Quasi-static Kinetic Modeling Of The Treat Reactor With Temperature Feedback
CN113656953B (en) Wind power sequence modeling method based on state number optimal decision model
CN113191561B (en) Runoff random simulation method and system based on Gaussian mixture model
CN115907178B (en) Clean ecosystem CO 2 Exchange amount prediction method
Kleibergen et al. Priors, posteriors and Bayes factors for a Bayesian analysis of cointegration

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
GR01 Patent grant
GR01 Patent grant