Power prediction and load prediction integrated method based on multi-source data

Abstract

The application discloses a power prediction and load prediction integrated method based on multi-source data, and relates to the technical field of communication.The historical power supply and power consumption data of a power grid company and its users are extracted from a cloud database, a backtracking period is determined, the ratio of power supply power to power consumption load in the period is calculated, and a power supply power-power consumption load change curve is generated.The dynamic matching of historical data is analyzed, the ratio of power supply power to power consumption load in a current time period is combined, and the power consumption load in a next time period is predicted by using the change curve, so as to provide decision support for power grid dispatching and load management.The method takes data driving as a core, realizes fine analysis and prediction of power supply and power consumption, and is suitable for power grid operation optimization and load management scenes.

Description

TECHNICAL FIELD

[0001] The present application belongs to the technical field of communication, and in particular relates to a power prediction and load prediction integrated method based on multi-source data. BACKGROUND

[0002] With the development of economy and the improvement of people's living standards, as one of the most important energy supply methods, the balance between demand and supply of electricity is crucial for the stability of society and the sustainable growth of economy.

[0003] In the traditional operation and management of power grid, power dispatching and load prediction mainly rely on experience and manual judgment. However, with the continuous expansion and increasing complexity of the power system, this traditional method has been difficult to meet the efficient operation and management needs of modern power grid; therefore, using advanced data analysis technology to deeply mine and analyze the historical power supply data of the power grid and the historical power consumption data of users has become the key to improving the operation efficiency and management level of the power grid.

[0004] Based on the above, the present application provides a power prediction and load prediction integrated method based on multi-source data. SUMMARY

[0005] In view of the deficiencies of the prior art, the present application provides a power prediction and load prediction integrated method based on multi-source data, which can find the rules and trends of power supply and user power consumption through data backtracking and period analysis, providing a scientific basis for power grid dispatching and planning; at the same time, based on the analysis results of historical data, the future power consumption load can be predicted, so as to optimize the power supply scheme of the power grid and improve the reliability and economy of power supply.

[0006] The purpose of the present application can be achieved by the following technical solutions:

[0007] The power prediction and load prediction integrated method based on multi-source data comprises the following steps:

[0008] S1, obtaining the historical power supply data of the current power grid company and the historical power consumption data of all power consumption users supplied by the current power grid company from the cloud database;

[0009] S2, determining a backtracking period, extracting a plurality of groups of historical power supply data in the backtracking period, locking any one group of period power supply data, extracting the power supply power in the period power supply data and the power consumption load in the period power consumption data associated with the period power supply data, and determining the ratio relationship between any one group of power supply power and power consumption load in the backtracking period;

[0010] S3, repeating step S2 to determine the ratio relationship between all power supply power and power consumption load in the backtracking period, and generating a power supply power-power consumption load change curve graph associated with the backtracking period;

[0011] S4, calculate the ratio between the power supply power and the electricity load in the time period in which the current time is located in the backtracking period, and combine the predicted electricity load in the next time period and the power supply power in the power supply power-electricity load change curve diagram.

[0012] As a further scheme of the present application, the historical power supply data includes the power supply power associated with the power grid company;

[0013] The historical electricity consumption data includes the electricity load associated with the electricity consumption user.

[0014] As a further scheme of the present application, in step S3, the specific way of determining the ratio between any one group of power supply power and electricity load in the backtracking period is:

[0015] S31, taking 0 o'clock and 24 o'clock every day as the start time and end time of the backtracking period;

[0016] S32, taking the current time as the reference to obtain the time interval t in the past, the time interval t contains m backtracking periods, the length of t is preset by the operator, and the value of m is obtained according to the length of t;

[0017] S33, extracting j groups of historical power supply data and j groups of historical electricity consumption data in any one backtracking period in t, wherein j is the number of segments segmented by the operator for the backtracking period;

[0018] S34, extracting j power supply powers in j groups of historical power supply data, and sorting them in time sequence as: P1, P2,..., P j , extracting j electricity loads in j groups of historical electricity consumption data, and recording them in time sequence as L1, L2,..., L j ;

[0019] Wherein, the power supply power P i corresponds to the electricity load L i , i is a count index, i starts from 1 and does not exceed j;

[0020] S35, extracting P i and L i in any one time period in the backtracking period, and calculating the ratio O i between P i and L i to obtain the ratio O i between P i and L i .

[0021] As a further scheme of the present application, in step S3, the specific way of generating the power supply power-electricity load change curve diagram associated with the backtracking period is:

[0022] S41, repeat the method in step S35 for the j time periods in the backtracking period, obtain j ratio relationships, sort them in time order, and obtain a ratio relationship sequence: O1, O2,..., Oj. j ;

[0023] S42, construct a two-dimensional coordinate system with the time line as the horizontal axis and the numerical value of the ratio relationship as the vertical axis, mark the obtained ratio relationship sequence in the two-dimensional coordinate system, and fit a curve, to obtain a power supply power-electricity load change curve graph associated with the backtracking period.

[0024] As a further scheme of the present application, based on the determined m backtracking periods in the time interval t, the m backtracking periods in the time interval t are processed according to the method described in steps S41 to S42, to obtain m power supply power-electricity load change curve graphs, each corresponding to a backtracking period.

[0025] As a further scheme of the present application, in step S4, the specific way of predicting the electricity load and the power supply power in the next time period is:

[0026] From the first time period of each of the m backtracking periods in the time interval t, m ratio relationships are extracted, and the average is taken as the standard ratio relationship of the first time period, denoted as

[0027] Repeat the above steps for the j time periods of each of the m backtracking periods, obtain j standard ratio relationships, and sort them in time order, denoted as a standard ratio relationship sequence Construct a two-dimensional coordinate system with the time line as the horizontal axis and the numerical value of the standard ratio relationship as the vertical axis, mark the standard ratio relationship sequence in the two-dimensional coordinate system, and fit a curve to obtain a power supply power-electricity load change curve graph S associated with the time interval t;

[0028] Obtain x time intervals determined by the operator to cover different conditions, denoted as t1, t2,... t x , process the x time intervals according to the method of obtaining S for the time interval t, to obtain x power supply power-electricity load change curve graphs associated with the x time intervals, respectively denoted as S1, S2,..., Sx x , the value of x is determined by the operator;

[0029] Obtain the segment number a of the current time period located in the backtracking period of the current time, denoted as a, a starts from 1 and does not exceed j;

[0030] Extract the ratio relationship Oa in the time period a nowand the ratio relationship in the n time periods to construct a two-dimensional coordinate system with the timeline as the horizontal axis and the value of the ratio relationship as the vertical axis, and the n ratio relationships together with the ratio relationship O now is fitted into a power supply-power load change curve S now , n is a preset value of the operator, and n is less than a;

[0031] Based on the determined time period a and the number of time periods of the previous n time periods of time period a, the time period interval [a-n, a] is obtained;

[0032] A time period interval with the same number of time periods as the time period interval [a-n, a] is obtained from S1, S2,..., S x ;

[0033] And the part of each of the x time period intervals in S1, S2,..., S x is cut off and extracted, and the first segment curve sequence S ′ 1, S ′ 2,..., S ′ x is obtained;

[0034] S ′ 1, S ′ 2,..., S ′ x is matched with S now , and S ′ 1, S ′ 2,..., S ′ x is sorted in descending order of matching degree, and the second segment curve sequence S ′ 1 ′ , S ′ 2 ′ ,..., S ′ x ′ is obtained;

[0035] The n+1 groups of power supply power and power load corresponding to S ′ 1 ′ are extracted, and the n+1 groups of power supply power and power load corresponding to S now are obtained;

[0036] The similarity of the n+1 groups of power supply power and power load corresponding to S ′ 1 ′ and the n+1 groups of power supply power and power load corresponding to S now is calculated;

[0037] If similar, S ′ 1 ′ is extracted as a preferred segment curve graph, and from S ′ 1 ′ , the power supply power and electricity load corresponding to the a+1th time period are extracted as the predicted power supply power and predicted electricity load of the next time period of the time period in which the current time is located, and the predicted results are informed to the operator.

[0038] If not similar, continue to obtain S ′ 2 ′ , and repeat the above processing S ′ 1 ′ .

[0039] As a further scheme of the present application, S ′ 1, S ′ 2,...,S ′ x The specific way of matching degree calculation with S now is:

[0040] Extract any one S ′ 1, S ′ 2,...,S ′ x from S ′ y , where y is the count index, taking values from 1 to x;

[0041] Extract the ratio relationship of all time periods from S ′ y , a total of n+1, denoted as

[0042] Obtain the ratio relationship of all time periods in S now , denoted as O a-n ,...,O a , extract and O a-n , use to calculate the difference absolute value C a-n of the two ratio relationships, and continue to execute backward until the difference absolute value C a is calculated, the n+1 difference absolute values are summarized and the average is taken, the average difference absolute value is obtained, the greater the value of the average difference absolute value, the lower the matching degree of S now and S ′ y .

[0043] As a further aspect of the present invention, the specific method for calculating the matching degree also includes the following:

[0044] For S ′ 1,S ′ 2,...,S ′ x Each segment of the curve is processed to obtain x absolute values ​​of the average difference, and then the curves are sorted in ascending order of the absolute values ​​of the average difference. ′ 1,S ′ 2,...,S ′ x The sequences are reordered, and the resulting sequence is denoted as the second segment curve sequence S. ′ 1 ′ ,S ′ 2 ′ ,...,S ′ x ′ .

[0045] As a further aspect of the present invention, S is calculated. ′ 1 ′ The corresponding n+1 groups of power supply and power load and S now The specific method for determining the similarity between the power supply and electrical load of the corresponding n+1 groups is as follows:

[0046] Extract S ′ 1 ′ The associated n+1 power supplies, denoted in chronological order as: P a-n ,...,P a Similarly, extract S now The associated n+1 power supplies, denoted in chronological order as: P ′ a-n ,...,P ′ a ;

[0047] Construct a two-dimensional coordinate system with the timeline as the horizontal axis and the power supply value as the vertical axis, and plot P... a-n ,...,P a Plotting this data point yields n+1 data points, which are then connected by short lines to obtain information about S. ′ 1 ′ The power supply line graph Z1;

[0048] Then P ′ a-n ,...,P ′ a Plotting this in a two-dimensional coordinate system, we obtain n+1 more data points. Connecting these points with short lines yields S. now The power supply line graph Z2;

[0049] P a-n and P ′ a-n A straight line H1 is constructed which is perpendicular to the horizontal axis and passes through P a and P ′ a A straight line H2 is constructed which is perpendicular to the horizontal axis and passes through Z1, Z2, H1 and H2, and the area of the closed region formed thereby is calculated and denoted as F1;

[0050] If F1≥F 阈 , then it is determined that S ′ 1 ′ is not similar to S 阈 , wherein F 阈 is a closed region area threshold value preset by an operator;

[0051] If F1<F 阈 , then the next step is performed.

[0052] As a further scheme of the present application, if F1<F ′ , then the n+1 electric loads associated with S ′ 1 a-n are extracted and denoted in time sequence as L a ,...,L now , and the n+1 electric loads associated with S ′ are extracted and denoted as L a-n ′ ,...,L a ;

[0053] L a-n ,...,L a and L ′ a-n ,...,L ′ a are processed in the same way as P a-n ,...,P a and P ′ a-n ,...,P ′ a ;

[0054] The area of the closed region formed thereby is calculated and denoted as F2;

[0055] If F2≥F 阈 , then it is determined that S ′ 1 ′ is not similar to S now ;

[0056] If F2<F阈 If S ′ 1 ′ Similarity degree of S now is similar.

[0057] The beneficial effects of the present application are:

[0058] (1) The present application proposes a power prediction and load prediction integrated method based on multi-source data. By obtaining the historical power supply and power consumption data of power grid companies and their users, combining backtracking period analysis, the ratio relationship between power supply power and power consumption load is calculated to generate a change curve graph, realizing dynamic matching analysis of power supply and power consumption. Then, by using periodic analysis driven by historical data, the power grid operation law is accurately captured and combined with real-time data of the current time period, which can effectively predict the power consumption load of the next time period, optimize the power supply scheme of the power grid, and improve the power supply reliability and economy. This method is suitable for power grid operation optimization and load management scenarios, and provides decision support for power resource allocation and demand side management.

[0059] (2) The present application extracts the ratio relationship of the corresponding time period from the power supply power-power consumption load change curve graph of multiple backtracking periods, which can quantify and simplify complex historical data, and present the relative relationship between power supply power and power consumption load of different time periods in the form of numerical values and graphs. It can more intuitively display the dynamic change relationship between power supply power and power consumption load, provide visual reference for subsequent prediction, help operators quickly understand the change trend and law of historical data, and provide strong support for prediction.

[0060] (3) The present application draws the power supply power and power consumption load data in a two-dimensional coordinate system to form a line graph, quantifies the similarity degree between the segment curve graph and the power supply power-power consumption load change curve graph by constructing and calculating the area of the closed region, provides an objective and quantitative evaluation standard, and avoids the uncertainty of subjective judgment. The similarity of power supply power and power consumption load is verified step by step, the judgment range is gradually narrowed, and the accuracy and efficiency of judging the similarity degree are improved. BRIEF DESCRIPTION OF DRAWINGS

[0061] The present application will be further described below in conjunction with the drawings.

[0062] Figure 1 is a flowchart of the method of the present application;

[0063] Figure 2 is a flowchart of the method described in Example 2 of the present application;

[0064] Figure 3 is a flowchart of the method described in Example 3 of the present application. DETAILED DESCRIPTION

[0065] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0066] Example 1

[0067] Integrated methods for power forecasting and load forecasting based on multi-source data, such as Figure 1 As shown, the specific steps include the following:

[0068] Step 1: Obtain the historical power supply data of the current power grid company and the historical electricity consumption data of all electricity users supplied by the current power grid company from the cloud database. Specifically, the historical power supply data of the power grid company includes various power supply data. In this method, the power supply power is selected as the research object. The value of the power supply power comes from the electricity load in the historical electricity consumption data of the electricity users of the power grid company. The total power supply power should be greater than the total electricity load to avoid the daily electricity consumption of users being affected by circuit losses.

[0069] Step 2: Determine the backtracking period and identify several sets of historical power supply data within the backtracking period. Lock any set of periodic power supply data within the backtracking period from the set of historical power supply data. Extract the power supply power and associated periodic power consumption data from the periodic power supply data. Extract the power consumption load from the periodic power consumption data. Determine the ratio between any set of power supply power and power consumption load within the backtracking period. Specifically, the time of a backtracking period defined in this method is one day (with 0:00 and 24:00 of each day as the start and end times of the backtracking period, respectively). Then, by dividing the time within a backtracking period into segments, several time periods are obtained. Each time period corresponds to a set of periodic power supply data and a set of periodic power consumption data. Extract any set of power supply power from the periodic power supply data and extract any set of power consumption load from the periodic power consumption data. The arbitrary set of power supply power and arbitrary set of power consumption load are within the same time period.

[0070] Further determine the ratio between power supply and power load within the same time period, and proceed to the next step of data processing.

[0071] Step three, repeat step two to determine all cycle power supply data and its associated cycle power consumption data in the backtracking period, and calculate the ratio of power supply power and power consumption load, generate the power supply data and power consumption data associated with the power supply power-power consumption load change curve in the backtracking period. Specifically, based on a plurality of time periods in a determined backtracking period, the same processing steps are performed on the power supply power and power consumption load in the plurality of time periods according to the method of determining the ratio between the power supply power and the power consumption load in the same time period in step two, a plurality of ratio relationships are obtained, and the sorting operation is performed according to the time sequence;

[0072] Then select a plurality of time intervals under different conditions;

[0073] The different conditions include season, temperature, humidity;

[0074] The one time interval is composed of a plurality of backtracking periods, and the conditions of the plurality of backtracking periods are the same;

[0075] Get a plurality of ratio relationships corresponding to a plurality of backtracking periods in a time interval, and average process a plurality of ratio relationships in each time period to obtain a standard ratio relationship of the time period. Repeat a plurality of times to obtain a plurality of standard ratio relationships corresponding to a plurality of time periods, and summarize as a standard ratio relationship sequence of the time interval, and fit the power supply power-power consumption load change curve associated with the time interval.

[0076] Step four, calculate the ratio between the power supply power of the power grid company and the power consumption load of the user in the time period in the backtracking period where the current time is located, and predict the power consumption load of the user in the next time period in the power supply power-power consumption load change curve. The power supply power of the next time period, specifically, the time period in the backtracking period where the current time is located is obtained, as well as the number of time periods, and the power supply power and power consumption load in the time period where the current time is located are obtained. First, calculate the ratio, get a plurality of time periods, and determine a plurality of ratio relationships. The obtained plurality of ratio relationships are fitted into a power supply power-power consumption load change curve together with the ratio relationship of the current time period according to the time line, and then one with the highest similarity to the power supply power-power consumption load change curve determined this time is extracted from the power supply power-power consumption load change curves associated with a plurality of time intervals under different conditions. The power supply power and power consumption load corresponding to the next time period of the number of time periods in it are extracted as the predicted power supply power and predicted power consumption load.

[0077] Embodiment 2

[0078] The embodiment discloses a method for generating a plurality of power supply-power consumption load change graphs based on a plurality of backtracking periods in a time interval, as shown in Figure 2 The method comprises the following steps:

[0079] Based on a time interval described in embodiment 1, now recorded as t, and m backtracking periods are determined, the length of the time interval t is determined by the operator in combination with the actual situation, and the value of m is determined in combination with the length of the time interval t;

[0080] Determine any backtracking period in the time interval t, and j time periods in the backtracking period, the j time periods are the number of segments segmented by the operator for the backtracking period, and each time period corresponds to a set of historical power supply data and a set of historical power consumption data;

[0081] Further obtain j sets of historical power supply data corresponding to j time periods and j sets of historical power consumption data associated with j sets of historical power supply data, and further obtain j power supply powers in j sets of historical power supply data, and sort the obtained j power supply powers in time sequence, and the sorted result is represented as: P1, P2,..., P j ;

[0082] Similarly, j power consumption loads in j sets of historical power consumption data are further obtained, and the obtained j power consumption loads are sorted in time sequence, and the sorted result is represented as: L1, L2,..., L j , wherein P i and L i correspond, in the same time period i in the backtracking period, i is a count index, and i starts from 1, i is not more than j;

[0083] For j power supply powers P1, P2,..., P j and j power consumption loads L1, L2,..., L j in the backtracking period, any corresponding power supply power P i and power consumption load L i in the group, the ratio O i =P i / L i between the power supply power and the power consumption load in the group is calculated; i ;

[0084] The processing power supply power P i and power consumption load L i, the same processing is performed on the j time periods in the backtracking period, and finally j ratio relationships can be obtained. The j ratio relationships obtained are sorted according to the sorting order of the time periods, and the sorted result is recorded as a ratio relationship sequence, denoted as: O1, O2,..., O j ;

[0085] A two-dimensional coordinate system is constructed with the time line as the horizontal axis of the two-dimensional coordinate system and the value of the ratio relationship as the vertical axis of the two-dimensional coordinate system, the determined ratio relationship sequence O1, O2,..., O j is sequentially marked on the constructed two-dimensional coordinate system, and the marked ratio relationship sequence is fitted by image fitting technology to obtain the power supply power-electricity load change curve graph associated with the backtracking period.

[0086] For the remaining m-1 backtracking periods of the time interval t, the above processing steps are performed to obtain the power supply power-electricity load change curve graph corresponding to each of the m-1 backtracking periods.

[0087] Embodiment 3

[0088] The embodiment discloses a method for predicting the electricity load and power supply power of the next time period, as shown in Figure 3 , specifically comprising the following steps:

[0089] Based on the m power supply power-electricity load change curve graphs associated with the m backtracking periods obtained in embodiment 2, the first time period of each backtracking period is determined from the m backtracking periods to obtain m first time periods, and the ratio relationship of the m first time periods is further extracted and processed to obtain the common ratio relationship of the m first time periods as a standard ratio relationship

[0090] The second time period of each of the m backtracking periods is determined again, and m second time periods are obtained, and the m second time periods are processed according to the method of processing the first time period to obtain a standard ratio relationship By analogy, until the standard ratio relationship of the last time period

[0091] The j standard ratio relationships obtained are sorted according to the time sequence, and the sorted result is recorded as a standard ratio relationship sequence, denoted as:

[0092] The two-dimensional coordinate system is constructed again with the time line as the horizontal axis of the two-dimensional coordinate system and the value of the standard ratio relationship as the vertical axis of the two-dimensional coordinate system, and the obtained standard ratio relationship sequence: In the constructed two-dimensional coordinate system, the standard ratio sequence is correlated with the power load curve S, which is also the power load curve correlated with the time interval t.

[0093] The x time intervals determined by the operator are obtained, and each time interval has different conditions, which are expressed as t1, t2,...t x The same processing is performed on the x time intervals t1, t2,...t x The x power load curves associated with the x time intervals are obtained, and are expressed as S1, S2,...,S x The value of x is determined by the operator according to the actual situation.

[0094] The number of time intervals in which the current time is located in the current time is obtained, which is expressed as a, and the a starts from 1 and does not exceed j.

[0095] The ratio between the power and the load in the time interval in which the current time is located is obtained, which is expressed as O now The n time intervals in the past time of the current time are obtained, and the ratio between the power and the load in the n time intervals is extracted, which is combined with O now There are n+1 ratios in total.

[0096] A two-dimensional coordinate system is constructed again with the time line as the horizontal axis and the value of the ratio as the vertical axis, and the n+1 ratios are marked in the constructed two-dimensional coordinate system and fitted into the power load curve S now , wherein n is a preset value of the operator, and n is less than a.

[0097] The number of time intervals a in which the current time is located is extracted, and the number of time intervals in the n time intervals obtained in the past time of the current time is extracted, and the time interval [a-n, a] is obtained.

[0098] From the power load curves S1, S2,...,S x correlated with the x time intervals under different conditions, S1 is extracted, and the part corresponding to the time interval [a-n, a] in S1 is extracted, which is expressed as S ′ 1, and the remaining S2,...,S x are processed in the same way, and the part corresponding to the time interval [a-n, a] in each power load curve is obtained, and is combined with S ′1Commonly recorded as the first segment graph sequence: S ′ 1,S ′ 2,...,S ′ x ;

[0099] Based on the determined first segment graph sequence: S ′ 1,S ′ 2,...,S ′ x and the power supply power-electricity load change graph S now associated with the current time period ; 1

[0100] Extract any one segment graph from the first segment graph sequence: S ′ 1,S ′ 2,...,S ′ x , recorded as S ′ y , where y is the count index, taking the value of 1 to x;

[0101] Extract the ratio relationship between the power supply power and the electricity load associated with the first time period a-n from the segment graph S ′ y , and record it as And repeat the step of extracting the ratio relationship for the subsequent time period until the ratio relationship of time period a is extracted n+1, listed in turn as:

[0102] Get the n+1 ratio relationships of time period a-n to time period a in the power supply power-electricity load change graph S now , listed as O a-n ,...,O a , extract the ratio relationship and the ratio relationship O a-n , use to calculate the difference absolute value C a-n of the two ratio relationships, and continue to execute the step of calculating the difference absolute value of the ratio relationship until the difference absolute value C a is calculated, and the n+1 difference absolute values calculated are averaged to obtain the average difference absolute value. The larger the value of the average difference absolute value, the lower the matching degree of the power supply power-electricity load change graph S now and the segment graph S ′ y ;

[0103] If the value of the average difference absolute value is smaller, it means that the power supply power-electricity load change graph Snow The higher the matching degree of the segment curve graph S ′ y is, the higher the matching degree of the segment curve graph S

[0104] The above-mentioned processing segment curve graph S ′ y is repeated, and the same processing operation is performed on each segment curve graph in the first segment curve graph sequence: S ′ 1, S ′ 2,...,S ′ x ;

[0105] x average absolute differences can be obtained, and the first segment curve graph sequence: S ′ 1, S ′ 2,...,S ′ x is sorted in ascending order of the values of the average absolute differences, and a second segment curve graph sequence: S ′ 1 ′ , S ′ 2 ′ ,...,S ′ x ′ is obtained.

[0106] The segment curve graph with the highest matching degree S now from the second segment curve graph sequence, i.e., the segment curve graph at the first position in the sorting, is extracted, and is denoted as S ′ 1 ′ .

[0107] The corresponding n+1 power supply powers and power consumption loads in the segment curve graph S ′ 1 ′ are obtained, and the corresponding n+1 power supply powers and power consumption loads in the power supply power-power consumption load change curve graph S now are also obtained.

[0108] The n+1 power supply powers associated with the segment curve graph S ′ 1 ′ are sequentially denoted as P a-n ,...,P a in time sequence, and the n+1 power supply powers associated with the power supply power-power consumption load change curve graph S now are sequentially denoted as P ′ a-n ,...,P ′ a .

[0109] Then, using the timeline as the horizontal axis and the power supply value as the vertical axis, construct a two-dimensional coordinate system and plot the segment curve S. ′ 1 ′ n+1 power supply P a-n ,...,P a Plotting in the constructed two-dimensional coordinate system yields n+1 data points. Connecting adjacent data points with short lines creates a curve S about the segment. ′ 1 ′ The power supply power is plotted as Z1;

[0110] Then, plot the obtained power supply versus power load variation curve S. now n+1 power supply P ′ a-n ,...,P ′ a Plotting in the constructed two-dimensional coordinate system yields n+1 more data points. Connecting adjacent data points with short lines again creates the curve S representing the change in power supply versus power load. now The power supply line graph is denoted as Z2;

[0111] Construct a path through P a-n and P ′ a-n And a straight line H1 perpendicular to the horizontal axis, then construct a line passing through P a and P ′ a And perpendicular to the horizontal axis, H2, calculate the area of ​​one or more closed regions formed by the power supply power broken line graph Z1, the power supply power broken line graph Z2, and the lines H1 and H2 (these one or more closed regions will be referred to as closed regions from now on), denoted as F1;

[0112] Extract the closed area threshold F preset by the operator based on actual needs. 阈 The calculated power supply line graphs Z1 and Z2, along with the areas of one or more closed regions formed by lines H1 and H2, are denoted as F1, and compared with the closed region area threshold F preset by the operator. 阈 Perform a comparison operation;

[0113] If the area of ​​the closed region F1 is greater than or equal to the threshold F of the closed region 阈 Then determine the segment curve S ′ 1 ′ The corresponding n+1 groups of power supply and power consumption load are shown in the power supply-power load variation curve S. now The similarity between the power supply and electrical load of the corresponding n+1 groups is dissimilar;

[0114] If the calculated closed area F1 is less than the operator's preset closed area threshold F 阈 If so, proceed to the next step;

[0115] Extract the segment curve S again ′ 1 ′ The associated n+1 electrical loads, in chronological order, are denoted as: L a-n ,...,L a Similarly, extract the power supply-load variation curve S. now The n+1 associated electrical loads are denoted as: L ′ a-n ,...,L ′ a ;

[0116] Construct a two-dimensional coordinate system with the timeline as the horizontal axis and the power supply value as the vertical axis, and plot the resulting segment curve S. ′ 1 ′ n+1 electrical loads L a-n ,...,L a Plotting in the constructed two-dimensional coordinate system yields n+1 data points. Connecting adjacent data points with short lines creates a curve S about the segment. ′ 1 ′ The line graph of the electricity load is denoted as Z3;

[0117] Then, plot the obtained power supply versus power load variation curve S. now n+1 electrical loads L ′ a-n ,...,L ′ a Plotting in the constructed two-dimensional coordinate system yields n+1 more data points. Connecting adjacent data points with short lines creates the curve S representing the change in power supply versus power load. now The line graph of the electricity load is denoted as Z4;

[0118] After L a-n and L ′ a-n Construct a straight line H3 perpendicular to the horizontal axis, and then pass through L. a and L ′ a Construct a straight line H4 perpendicular to the horizontal axis, and calculate the area of ​​the closed region formed by the power load polygon Z3, the power load polygon Z4, and the straight line H3 and the straight line H4, denoted as F2;

[0119] If the calculated closed area F2 is greater than or equal to the operator's preset closed area threshold F 阈 Then determine the segment curve S′ 1 ′ The n+1 groups of power supply power and power consumption corresponding to the segment curve S now The similarity of the n+1 groups of power supply power and power consumption corresponding to the segment curve S is dissimilar, and the process continues to the next group. ′ 2 ′ The above process is repeated. ′ 1 ′ The above process is repeated.

[0120] If the calculated closed area F2 is less than the closed area threshold F 阈 , the segment curve S ′ 1 ′ corresponding to the n+1 groups of power supply power and power consumption is determined to be the segment curve S now 1 now corresponding to the n+1 groups of power supply power and power consumption is similar.

[0121] Based on the determined similarity of the n+1 groups of power supply power and power consumption corresponding to the segment curve S ′ 1 ′ corresponding to the n+1 groups of power supply power and power consumption, the segment curve S ′ 1 ′ is determined to be the preferred segment curve.

[0122] The segment curve S ′ 1 ′ is extracted. ′ 1 ′ The segment curve S ′ 1 ′ corresponds to the portion of the original power supply power and power consumption change curve from the a-nth time period to the a-th time period.

[0123] The curve portion corresponding to the a+1-th time period is extracted from the original power supply power and power consumption change curve in which the segment curve S ′ 1 ′ is located, and the power supply power and power consumption corresponding to this portion are further extracted as the predicted power supply power and predicted power consumption of the next time period of the time period in which the current time is located, and the predicted power supply power and predicted power consumption are notified to the operator.

[0124] Some data in the above formulas are dimensionless numerical calculations, and the contents not described in detail in the specification are all prior art known to those skilled in the art.

[0125] The above merely illustrates and describes the present application, and those skilled in the art can make various modifications or supplements to the specific embodiments described or replace them with similar ways without departing from the application or exceeding the scope defined by the claims, which should belong to the protection scope of the present application.

[0126] It needs to be declared that all the user data collected in this application is collected with the consent and authorization of the user. And the purpose of the user data is legal and compliant, and the use and processing of the user data comply with the relevant laws, regulations and standards of the relevant region.

Claims

1. An integrated method for power forecasting and load forecasting based on multi-source data, characterized in that, This method includes the following steps: S1. Obtain historical power supply data of the current power grid company and historical electricity consumption data of all electricity users supplied by the current power grid company from the cloud database; S2. Determine the backtracking period, extract several sets of historical power supply data within the backtracking period, lock any set of periodic power supply data, extract the power supply in the periodic power supply data and the power load in the periodic power consumption data associated with the periodic power supply data, and determine the ratio relationship between any set of power supply and power load within the backtracking period. S3. Use 0:00 and 24:00 each day as the start and end times of the backtracking cycle; Using the current time as a reference, a time interval t is obtained from the past. The time interval t contains m backtracking cycles. The duration of t is preset by the operator, and the value of m is obtained based on the duration of t. Extract j sets of historical power supply data and j sets of historical power consumption data within any backtracking period within the time interval t, where j is the number of segments into which the backtracking period is divided by the operator. Repeat step S2 to determine the ratio between all power supply and power consumption within the traceback period, and generate a power supply-power consumption change curve associated with the traceback period. S4. Calculate the ratio between power supply and power consumption within the time period of the current backtracking cycle, and predict the power consumption and power supply in the next time period based on the power supply-power consumption change curve. The specific method is as follows: From the first time period of each of the m backtracking cycles within the time interval t, extract m ratio relationships, and take the average as the standard ratio relationship for the first time period, denoted as . ; Repeat the above steps for each of the m backtracking periods and the j time periods to obtain j standard ratio relationships, and sort them in chronological order, denoted as the standard ratio relationship sequence. A two-dimensional coordinate system is constructed with the time line as the horizontal axis and the standard ratio relationship as the vertical axis. The standard ratio relationship sequence is marked in the two-dimensional coordinate system and fitted into a power supply-load change curve S associated with the time interval t. Obtain x time intervals, each covering different conditions, as determined by the operator, denoted as... The method of obtaining S based on the processing time interval t is used to process x time intervals to obtain the power supply-load variation curves associated with the x time intervals, denoted as . x is a preset value; Get the segment number of the current time period within the backtracking period of the current time, denoted as a, where a starts from 1 and does not exceed j; Extract the ratio relationship within time period a It then continuously obtains n time intervals and the ratio relationships within those n time intervals from the past. A two-dimensional coordinate system is constructed with the timeline as the horizontal axis and the ratio values ​​as the vertical axis. The obtained n ratio relationships are then linked to the ratio... The graph is plotted in a two-dimensional coordinate system and fitted to form a curve showing the change in power supply versus power load. n is a preset value for the operator, and n is less than a; Based on the determined time period 'a' and the number of segments in the n preceding time periods of 'a', the time interval is obtained. ; from Each time interval is extracted from the time interval [an, a] and has the same number of segments, resulting in x time intervals; And each of the x time intervals in The portion of the graph is extracted and summarized, and recorded as the first segment of the curve sequence. ; Will and Perform a matching score calculation and sort the results in descending order of matching score. Sort the data to obtain the second segment of the curve sequence. ; extract The corresponding n+1 groups of power supply and power load, then obtain The corresponding n+1 groups of power supply and electrical load; calculate The corresponding n+1 groups of power supply and power load are The degree of similarity between the power supply and electrical load of the corresponding n+1 groups; If they are similar, As a preferred segment curve, and from Extract the power supply and power load corresponding to the (a+1)th time period from the original power supply and power load change curve, and use them as the predicted power supply and power load for the next time period of the current time period. Then notify the operator of the prediction results. If they are dissimilar, continue obtaining the results. And repeat the above process. The steps.

2. The integrated method for power forecasting and load forecasting based on multi-source data according to claim 1, characterized in that, The historical power supply data includes the power supply capacity associated with the power grid company; The historical electricity consumption data includes the electricity load associated with each electricity user.

3. The integrated method for power forecasting and load forecasting based on multi-source data according to claim 1, characterized in that, In step S3, the specific method for determining the ratio between any set of power supply and electrical load within the traceback period is as follows: S31. Extract j power values ​​from j sets of historical power supply data and sort them in chronological order as follows: Extract j electrical loads from j sets of historical electricity consumption data, and categorize them in chronological order as follows: ; Among them, power supply With electrical load Correspondingly, i is the counting index, starting from 1 and not exceeding j; S32. Extract any time period within the backtracking cycle. and ,use Calculated and ratio relationship .

4. The integrated method for power forecasting and load forecasting based on multi-source data according to claim 1, characterized in that, In step S3, the specific method for generating the power supply-load change curve associated with the traceback period is as follows: S41. Repeat step S32 for j time periods in the backtracking period to obtain j ratio relationships. Sort them by time to obtain the ratio relationship sequence: ; S42. Construct a two-dimensional coordinate system with the time line as the horizontal axis and the numerical values ​​of the ratio relationship as the vertical axis. Mark the obtained ratio relationship sequence in the two-dimensional coordinate system and fit it with a curve to obtain the power supply-load change curve of the backtracking period.

5. The integrated method for power forecasting and load forecasting based on multi-source data according to claim 4, characterized in that, Based on the m backtracking cycles in the determined time interval t, the m backtracking cycles in the time interval t are processed according to the methods described in steps S41 to S42 to obtain m power supply-load change curves, each corresponding to one of the m backtracking cycles.

6. The integrated method for power forecasting and load forecasting based on multi-source data according to claim 1, characterized in that, Will The specific method for calculating the matching degree is as follows: from Extract any one , where y is the counting index, with a value from 1 to x; from Extract the ratio relationships across all time periods, totaling n+1, denoted as . ; Get again The ratio relationship of all time periods is denoted as ,extract ,use Calculate the absolute value of the difference between two ratios. And continue executing until the absolute value of the difference is calculated. Summarize the absolute values ​​of the n+1 differences and take the average to obtain the average absolute value of the differences. The larger the value of the average absolute value of the differences, the greater the difference. The lower the match, the higher the match.

7. The integrated method for power forecasting and load forecasting based on multi-source data according to claim 6, characterized in that, The specific methods for calculating the matching degree also include the following: right Each segment of the curve is processed to obtain x absolute values ​​of the average difference, and then sorted in ascending order of the absolute values ​​of the average difference. The sequences are reordered, and the resulting sequence is denoted as the second segment of the graph. .

8. The integrated method for power forecasting and load forecasting based on multi-source data according to claim 1, characterized in that, calculate The corresponding n+1 groups of power supply and power load are The specific method for determining the similarity between the power supply and electrical load of the corresponding n+1 groups is as follows: extract The associated n+1 power supplies, denoted in chronological order, are: Similarly, extract The associated n+1 power supplies, denoted in chronological order, are: ; Construct a two-dimensional coordinate system with the timeline as the horizontal axis and the power supply value as the vertical axis. Plotting this data into a graph yields n+1 data points, which are then connected by short lines to obtain information about... Power supply line graph ; Then Plotting in a two-dimensional coordinate system, we obtain n+1 more data points, which are then connected by short lines to obtain... Power supply line graph ; through Construct a straight line perpendicular to the horizontal axis. After Construct a straight line perpendicular to the horizontal axis. ,calculate as well as The area of ​​one or more enclosed regions formed by these regions is denoted as ; like Then determine The similarity between and is dissimilar, where The preset threshold for the area of ​​the enclosed zone for operators; like If so, proceed to the next step.

9. The integrated method for power forecasting and load forecasting based on multi-source data according to claim 8, characterized in that, like Then extract The associated n+1 electrical loads, in chronological order, are denoted as follows: Similarly, extract The n+1 associated electrical loads are denoted as follows: ; To process and Method of processing and ; The area of ​​one or more enclosed regions is obtained, denoted as . ; like Then determine and The similarity is not similar; like Then determine and The degree of similarity is similar.

Images