CN111415037A - Electric power system short-term load prediction method based on similar day and artificial neural network - Google Patents

Electric power system short-term load prediction method based on similar day and artificial neural network Download PDF

Info

Publication number
CN111415037A
CN111415037A CN202010193485.6A CN202010193485A CN111415037A CN 111415037 A CN111415037 A CN 111415037A CN 202010193485 A CN202010193485 A CN 202010193485A CN 111415037 A CN111415037 A CN 111415037A
Authority
CN
China
Prior art keywords
day
similar
neural network
power system
artificial neural
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.)
Pending
Application number
CN202010193485.6A
Other languages
Chinese (zh)
Inventor
谢春泽
刘春明
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.)
North China Electric Power University
Original Assignee
North China Electric Power University
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 North China Electric Power University filed Critical North China Electric Power University
Priority to CN202010193485.6A priority Critical patent/CN111415037A/en
Publication of CN111415037A publication Critical patent/CN111415037A/en
Pending legal-status Critical Current

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
    • G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/04—Architecture, e.g. interconnection topology
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/08—Learning methods
    • 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

Landscapes

  • Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • Business, Economics & Management (AREA)
  • Economics (AREA)
  • Health & Medical Sciences (AREA)
  • General Physics & Mathematics (AREA)
  • General Health & Medical Sciences (AREA)
  • Human Resources & Organizations (AREA)
  • Strategic Management (AREA)
  • Marketing (AREA)
  • Computing Systems (AREA)
  • Software Systems (AREA)
  • General Business, Economics & Management (AREA)
  • Mathematical Physics (AREA)
  • General Engineering & Computer Science (AREA)
  • Tourism & Hospitality (AREA)
  • Molecular Biology (AREA)
  • Evolutionary Computation (AREA)
  • Data Mining & Analysis (AREA)
  • Computational Linguistics (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Artificial Intelligence (AREA)
  • Biomedical Technology (AREA)
  • Biophysics (AREA)
  • Primary Health Care (AREA)
  • Development Economics (AREA)
  • Water Supply & Treatment (AREA)
  • Public Health (AREA)
  • Game Theory and Decision Science (AREA)
  • Entrepreneurship & Innovation (AREA)
  • Operations Research (AREA)
  • Quality & Reliability (AREA)
  • Management, Administration, Business Operations System, And Electronic Commerce (AREA)
  • Supply And Distribution Of Alternating Current (AREA)

Abstract

The invention provides a power system short-term load prediction method based on a similar day and an artificial neural network, which comprises the following steps: acquiring historical data and forecast date data of the load; screening similar days in past historical days according to the predicted day data and the historical data; calculating the similarity degree of the similar day and the predicted day; and selecting a training sample, and predicting by adopting an artificial neural network to obtain the load of the prediction day. On the premise of considering meteorological factors, time factors and other factors, the method can accurately predict the short-term load, and effectively ensure the safe and economic operation of the power system.

Description

Electric power system short-term load prediction method based on similar day and artificial neural network
Technical Field
The invention relates to the technical field of power system load prediction, in particular to a power system short-term load prediction method based on a similar day and an artificial neural network.
Background
The short-term power load prediction is the basis of operation and analysis of a power system, and has important significance on unit combination, economic dispatching, safety check and the like. The method improves the load prediction precision and is an important means for guaranteeing the scientificity of the optimization decision of the power system. In modern power systems, the electrical loads constituting the power load are of various types, the duty ratio of the load affected by weather conditions, such as an air conditioner, is continuously increased, and the influence of weather factors (temperature, humidity, rainfall, and the like) on the power system load is more prominent. The consideration of meteorological factors becomes one of the main means for the dispatching center to further improve the load prediction accuracy.
If the short-term power load cannot be predicted accurately, great loss is caused to production safety and economic development, so that with further deepening and promotion of urbanization development and industrial kinetic energy, how to carry out scientific and reasonable short-term power load prediction becomes a problem to be solved urgently.
Disclosure of Invention
In view of the above, the present invention provides a method for predicting a short-term load of an electrical power system based on a similar day and an artificial neural network, which can accurately predict the short-term load on the premise of considering various factors such as weather and time, and effectively ensure safe and economic operation of the electrical power system.
In a first aspect, an embodiment of the present invention provides a power system short-term load prediction method based on a similar day and an artificial neural network, including:
acquiring historical data and forecast date data of the load;
according to the predicted date data and the historical data, similar days are screened in past historical days;
calculating the similarity degree of the similar day and the predicted day;
and selecting a training sample, and predicting by adopting an artificial neural network to obtain the load of the prediction day.
With reference to the first aspect, an embodiment of the present invention provides a first possible implementation manner of the first aspect, where the screening, according to the predicted day data and the historical data, similar days in past historical days includes:
reducing the search range according to the change cycle rule of the load;
similar days were screened in past history days.
With reference to the first possible implementation manner of the first aspect, an embodiment of the present invention provides a second possible implementation manner of the first aspect, wherein the calculating a similarity degree between the similar day and the predicted day includes:
measuring the first similarity degree of the similar day and the predicted day by using the trend similarity factor;
measuring a second similarity degree of the similar day and the predicted day by using the time similarity factor;
and screening similar days according to the first similarity degree and the second similarity degree.
With reference to the first aspect, an embodiment of the present invention provides a third possible implementation manner of the first aspect, where the training samples are similar day data.
With reference to the first possible implementation manner of the first aspect, an embodiment of the present invention provides a fourth possible implementation manner of the first aspect, where the historical data includes date and day types, and the day types include weekdays, weekends, holidays, and other special days.
With reference to the second possible implementation manner of the first aspect, an embodiment of the present invention provides a fifth possible implementation manner of the first aspect, where the trend similarity factor is a similarity degree of a load level change rule between a similar day and a predicted day, and the time similarity factor is a distance degree between the similar day and the predicted day in terms of time distance.
With reference to the fourth possible implementation manner of the first aspect, the embodiment of the present invention provides a sixth possible implementation manner of the first aspect, wherein the historical data further includes meteorological factors, and the meteorological factors include a maximum temperature, an average temperature, a minimum temperature, rainfall and relative humidity.
With reference to the sixth possible implementation manner of the first aspect, an embodiment of the present invention provides a seventh possible implementation manner of the first aspect, where the method further includes:
establishing a day eigenvector according to the highest temperature, the average temperature, the lowest temperature, the rainfall and the relative humidity, and calculating a meteorological similarity factor according to the day eigenvector;
and correcting the meteorological similarity factor according to the contribution degree of different indexes to the load change.
The invention provides a power system short-term load prediction method based on a similar day and an artificial neural network, which comprises the following steps: acquiring historical data and forecast date data of the load; screening similar days in past historical days according to the predicted day data and the historical data; calculating the similarity degree of the similar day and the predicted day; and selecting a training sample, and predicting by adopting an artificial neural network to obtain the load of the prediction day. On the premise of considering meteorological factors, time factors and other factors, the method can accurately predict the short-term load, and effectively ensure the safe and economic operation of the power system.
Additional features and advantages of the invention will be set forth in the description which follows, and in part will be obvious from the description, or may be learned by practice of the invention. The objectives and other advantages of the invention will be realized and attained by the structure particularly pointed out in the written description and claims hereof as well as the appended drawings.
In order to make the aforementioned and other objects, features and advantages of the present invention comprehensible, preferred embodiments accompanied with figures are described in detail below.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly described below, and it is obvious that the drawings in the following description are some embodiments of the present invention, and other drawings can be obtained by those skilled in the art without creative efforts.
Fig. 1 is a flowchart of a short-term load prediction method for a power system based on a similar day and artificial neural network according to an embodiment of the present invention.
Detailed Description
To make the objects, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings, and it is apparent that the described embodiments are some, but not all embodiments of the present invention. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
The short-term power load prediction is the basis of operation and analysis of a power system, and has important significance on unit combination, economic dispatching, safety check and the like. The method improves the load prediction precision and is an important means for guaranteeing the scientificity of the optimization decision of the power system. In modern power systems, the electrical loads constituting the power load are of various types, the duty ratio of the load affected by weather conditions, such as an air conditioner, is continuously increased, and the influence of weather factors (temperature, humidity, rainfall, and the like) on the power system load is more prominent. The consideration of meteorological factors becomes one of the main means for the dispatching center to further improve the load prediction accuracy.
If the short-term power load cannot be predicted accurately, great loss is caused to production safety and economic development, so that with further deepening and promotion of urbanization development and industrial kinetic energy, how to carry out scientific and reasonable short-term power load prediction becomes a problem to be solved urgently. Based on the method, the embodiment of the invention provides the power system short-term load prediction method based on the similar day and the artificial neural network, and on the premise of considering various factors such as weather and time, the short-term load can be accurately predicted, so that the safe and economic operation of the power system is effectively ensured.
Fig. 1 is a flowchart of a short-term load prediction method for a power system based on a similar day and artificial neural network according to an embodiment of the present invention.
Referring to fig. 1, a method for predicting short-term load of a power system based on a similar day and an artificial neural network comprises the following steps:
step S101, acquiring historical data and forecast date data of a load;
step S102, according to the forecast date data and the historical data, similar days are screened in past historical days;
step S103, calculating the similarity degree between the similar day and the prediction day;
and step S104, selecting training samples, and predicting by adopting an artificial neural network to obtain the load of the prediction day.
According to an exemplary embodiment of the present invention, step S102 includes:
reducing the search range according to the change cycle rule of the load;
similar days were screened in past history days.
Specifically, the similar day method is one of the basic methods for short-term load prediction, and the basic principle is that the loads are similar in two days with similar influence factors such as day type, date distance and the like, the loads are also similar, and the loads on the day to be predicted are corrected according to the historical similar days or the selected similar days are used as training samples to be applied to a prediction model, so that a good prediction effect can be achieved. Obviously, the quality of similar day selection directly affects the prediction precision, and the factors affecting similar days are as follows: (1) the day type. The change rule of the power load is mainly determined by the load composition and the social activities of people. During the working day, the industrial load accounts for a large proportion of the total load, and the plant schedules regularly during the working day, so that the load changes during the working day have similarities. During weekends, the load components are mainly residential life load and business load, so the weekend load is greatly different from the weekday load. These result in a cyclic nature of the load, i.e. the load exhibits a certain regularity in the variation over one week. In addition, the load change law is greatly different on special days, such as holidays, disastrous weather, and days when specific events are held. For this reason, the influence factors of the day type are introduced, namely working days or weekends, or the same special day. (2) The distance of the date. The date distance refers to the number of days the predicted day is from the historical day. The power load is continuously changed and is influenced by the requirement of safety and stability of the power system, and the power load generally does not have sudden change. For this reason, the smaller the date distance between the predicted day and the historical day, the more similar the load characteristics therebetween without ignoring other factors. Furthermore, the load has a growing trend, determined by economic development. The annual growth trend of the load is also taken into account when the historical day is far from the predicted day. In general, the date distance should embody the principle of "big and small. From the two influencing factors, the same date of different years or adjacent dates of the same year may be selected to approximate a range of similar days.
According to an exemplary embodiment of the present invention, step S103 includes:
measuring the first similarity degree of the similar day and the predicted day by using the trend similarity factor;
measuring a second similarity degree of the similar day and the predicted day by using the time similarity factor;
and screening similar days according to the first similarity degree and the second similarity degree.
According to an exemplary embodiment of the present invention, the training samples are similar day data.
According to an exemplary embodiment of the present invention, the historical data includes date and day types including weekdays, weekends, holidays, and other special days.
According to the exemplary embodiment of the invention, the trend similarity factor is the similarity degree of the load level change rule between the similar day and the prediction day, and the time similarity factor is the distance degree between the similar day and the prediction day in time distance.
According to an exemplary embodiment of the invention, the historical data further comprises meteorological factors including maximum temperature, average temperature, minimum temperature, rainfall and relative humidity.
According to an exemplary embodiment of the present invention, further comprising:
establishing a day eigenvector according to the highest temperature, the average temperature, the lowest temperature, the rainfall and the relative humidity, and calculating a meteorological similarity factor according to the day eigenvector;
and correcting the weather similarity factor according to the contribution degree of different indexes to the load change.
Specifically, since the meteorological factors are not subjected to weighted analysis before, and actually the influence strength of the meteorological factors on the power load is different, and the most important factor influencing the load change due to rainfall in summer is not the rainfall but the temperature change, the load is corrected according to the temperature change. When the temperature of the day to be predicted has a large change temperature mutation relative to the temperature of the previous day, a large deviation can be generated by adopting a conventional linear temperature correction model. For this purpose, the following two correction models are proposed here for different similar day types.
① weather load correction on days of similar temperature change
When the predicted day temperature has changed greatly from the previous day, a similar case, i.e., a similar temperature change day, is first searched for in the history day. If such similar days exist, the following model can be used for correction:
Figure BDA0002416752160000071
in the formula, △LweatherFor predicting weather load correction amount of day, △ T for predicting temperature variation of day before day, △L'weatherThe change amount of the meteorological load on the similar temperature change day is larger than that on the previous day, and △ T' is the change amount of the temperature on the similar temperature day is larger than that on the previous day.
② weather load correction on days of similar temperature
The similar temperature daily correction model is shown as follows:
Figure BDA0002416752160000072
of formula (II b), L'weatherL historical solar weather load with temperature close to that of the day to be measured in summerweatherFor weather load conventional prediction of the day to be measured, i.e. uncorrected load prediction Lw/L’wIs a load increase factor of LwIs a prediction of the average load of k days ahead of day, L'wThe average load was k days before the similar day.
The invention provides a power system short-term load prediction method based on a similar day and an artificial neural network, which comprises the following steps: acquiring historical data and forecast date data of the load; screening similar days in past historical days according to the predicted day data and the historical data; calculating the similarity degree of the similar day and the predicted day; and selecting a training sample, and predicting by adopting an artificial neural network to obtain the load of the prediction day. On the premise of considering meteorological factors, time factors and other factors, the method can accurately predict the short-term load, and effectively ensure the safe and economic operation of the power system.
The terms "first", "second", and "third" referred to in the embodiments of the present invention are used for descriptive purposes only and are not to be construed as indicating or implying relative importance.
Finally, it should be noted that: the above-mentioned embodiments are only specific embodiments of the present invention, which are used for illustrating the technical solutions of the present invention and not for limiting the same, and the protection scope of the present invention is not limited thereto, although the present invention is described in detail with reference to the foregoing embodiments, those skilled in the art should understand that: any person skilled in the art can modify or easily conceive the technical solutions described in the foregoing embodiments or equivalent substitutes for some technical features within the technical scope of the present disclosure; such modifications, changes or substitutions do not depart from the spirit and scope of the embodiments of the present invention, and they should be construed as being included therein. Therefore, the protection scope of the present invention shall be subject to the protection scope of the claims.

Claims (8)

1. A power system short-term load prediction method based on similar day and artificial neural network is characterized by comprising the following steps:
acquiring historical data and forecast date data of the load;
according to the predicted date data and the historical data, similar days are screened in past historical days;
calculating the similarity degree of the similar day and the predicted day;
and selecting a training sample, and predicting by adopting an artificial neural network to obtain the load of the prediction day.
2. The similar day and artificial neural network based power system short-term load prediction method of claim 1, wherein the screening similar days in past historical days according to the predicted day data and historical data comprises:
reducing the search range according to the change cycle rule of the load;
similar days were screened in past history days.
3. The power system short-term load prediction method based on the similar day and the artificial neural network as claimed in claim 2, wherein the calculating the similarity degree of the similar day and the prediction day comprises:
measuring the first similarity degree of the similar day and the predicted day by using the trend similarity factor;
measuring a second similarity degree of the similar day and the predicted day by using the time similarity factor;
and screening similar days according to the first similarity degree and the second similarity degree.
4. The similar day and artificial neural network-based power system short-term load prediction method according to claim 1, wherein the training samples are similar day data.
5. The similar day and artificial neural network based power system short term load prediction method of claim 2, wherein the historical data comprises date and day types, the day types comprising weekdays, weekends, holidays and special days.
6. The power system short-term load prediction method based on the similar day and artificial neural network as claimed in claim 3, wherein the trend similarity factor is a similarity degree of load level change rules between the similar day and the predicted day, and the time similarity factor is a distance degree between the similar day and the predicted day in time distance.
7. The similar day and artificial neural network-based power system short-term load prediction method of claim 5, wherein the historical data further comprises meteorological factors, the meteorological factors comprising maximum temperature, average temperature, minimum temperature, rainfall and relative humidity.
8. The similar day and artificial neural network-based power system short-term load prediction method of claim 7, further comprising:
establishing a day eigenvector according to the highest temperature, the average temperature, the lowest temperature, the rainfall and the relative humidity, and calculating a meteorological similarity factor according to the day eigenvector;
and correcting the meteorological similarity factor according to the contribution degree of different indexes to the load change.
CN202010193485.6A 2020-03-18 2020-03-18 Electric power system short-term load prediction method based on similar day and artificial neural network Pending CN111415037A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202010193485.6A CN111415037A (en) 2020-03-18 2020-03-18 Electric power system short-term load prediction method based on similar day and artificial neural network

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202010193485.6A CN111415037A (en) 2020-03-18 2020-03-18 Electric power system short-term load prediction method based on similar day and artificial neural network

Publications (1)

Publication Number Publication Date
CN111415037A true CN111415037A (en) 2020-07-14

Family

ID=71491218

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202010193485.6A Pending CN111415037A (en) 2020-03-18 2020-03-18 Electric power system short-term load prediction method based on similar day and artificial neural network

Country Status (1)

Country Link
CN (1) CN111415037A (en)

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN111985689A (en) * 2020-07-17 2020-11-24 江苏方天电力技术有限公司 Short-time load prediction method and system
CN112465196A (en) * 2020-11-12 2021-03-09 云南电网有限责任公司 System load prediction method, device, equipment and storage medium

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN111985689A (en) * 2020-07-17 2020-11-24 江苏方天电力技术有限公司 Short-time load prediction method and system
CN112465196A (en) * 2020-11-12 2021-03-09 云南电网有限责任公司 System load prediction method, device, equipment and storage medium

Similar Documents

Publication Publication Date Title
US7451017B2 (en) Energy and cost savings calculation system
US7228234B2 (en) Weather data quality control and ranking method
CN111985701B (en) Power consumption prediction method based on power supply enterprise big data model base
CN107423836B (en) Short-term load prediction method based on body sensing temperature
CN110097220B (en) Method for predicting monthly electric quantity of wind power generation
CN102426674A (en) Power system load prediction method based on Markov chain
CN104156786A (en) Non-working day maximum daily load forecasting system considering meteorological multi-factor influences
CN110909958A (en) Short-term load prediction method considering photovoltaic grid-connected power
CN113743673B (en) Power load prediction method during typhoon
CN111415037A (en) Electric power system short-term load prediction method based on similar day and artificial neural network
CN102509173A (en) Markov chain based method for accurately forecasting power system loads
CN117094475B (en) Fault analysis system for electric power distribution network
CN112884601B (en) Power system operation risk assessment method based on weather division strategy
CN111008727A (en) Power distribution station load prediction method and device
CN113205252A (en) Aggregated load scheduling method based on demand side load peak regulation potential parameter prediction
CN109299814B (en) Meteorological influence electric quantity decomposition prediction method
CN113326654A (en) Method and device for constructing gas load prediction model
CN111860985A (en) Day-ahead power load prediction method based on load decomposition
CN104252647B (en) Electro-load forecast method based on anti-distance weighting interpolation method
CN117853275B (en) Method and device for electricity utilization prediction
CN110991747A (en) Short-term load prediction method considering wind power plant power
CN116485443A (en) Electric power spot market price information prediction optimization method and system
CN115481918A (en) Active sensing and predictive analysis system for unit state based on source network load storage
CN108346009A (en) A kind of power generation configuration method and device based on user model self study
CN114676896A (en) Electric power operation monitoring method and system based on time decomposition improved algorithm

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
WD01 Invention patent application deemed withdrawn after publication

Application publication date: 20200714

WD01 Invention patent application deemed withdrawn after publication