Intelligent power plant prediction and dispatching system and method based on big data technology

Abstract

The present application relates to the technical field of power management, in particular to a smart power plant prediction and scheduling system and method based on big data technology, the system comprises a prediction analysis module, the prediction analysis module combines the analysis results in the reporting data fluctuation characteristic analysis module, constructs a data reporting interval length adaptive model, and combines the historical power load data of the enterprise to be measured to obtain the prediction result corresponding to the latest submitted load reporting prediction information based on the current time. The present application considers that the power load of the enterprise to be measured at different times in a time period is fluctuating, and the influence of the fluctuation intensity on the subsequent prediction result, and realizes adaptive adjustment of the load data reporting time according to the fluctuation degree of the power load data, ensures the accuracy of the dynamic load prediction result, and realizes accurate control of the power reserve of the power plant.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of power management, in particular to a smart power plant prediction and scheduling system and method based on big data technology. BACKGROUND

[0002] With the development of social economy and the improvement of people's living standards, people's demand for electricity is getting stronger and stronger, whether it is production or life, it is inseparable from the use of electricity. However, the current storage of electric energy is difficult, and thus the power needs to be estimated in advance in the production process. User's power demand (load) is adjusted and distributed. Generally speaking, the power type is divided into two categories, residential power and enterprise power; for residential power, it maintains regularity; but for enterprise power, the power demand of each period is different. At present, power plants often store electricity by receiving the power demand reported by enterprises and residents on a regular basis. However, due to the irregularity of enterprise power demand and the passive situation of relying on the power demand reported by enterprises on a regular basis for power storage, it is impossible to actively predict enterprise power, resulting in a mismatch between power storage and power demand in power plants. Therefore, how to predict the power storage of power plants has become a problem to be solved in the current society. SUMMARY

[0003] The purpose of the present application is to provide a smart power plant prediction and scheduling system and method based on big data technology to solve the problems raised in the background.

[0004] In order to solve the above technical problems, the present application provides the following technical scheme: a smart power plant prediction and scheduling method based on big data technology, the method comprising the following steps:

[0005] S1, receiving the power demand prediction information reported by the power plant covered power consumption area, obtaining the historical power load data of the enterprise to be tested, and extracting the load reporting prediction information submitted by the enterprise to be tested each time in the historical data;

[0006] S2, obtaining the actual order deviation rate corresponding to the adjacent two load reporting prediction information of the enterprise to be tested, analyzing the relationship between the data deviation amount in the time period corresponding to the load reporting prediction information submitted by the enterprise to be tested each time and the data fluctuation characteristics of the actual power load information in the time period corresponding to the previous load reporting prediction information under the condition that the actual order deviation rate is the same, and obtaining the reporting fluctuation characteristic function of the enterprise to be tested under the condition of the corresponding order deviation rate;

[0007] S3, combining the analysis result in S2, constructing a data reporting interval length adaptive model, and combining the historical power load data of the enterprise to be tested to obtain the prediction result corresponding to the last submitted load reporting prediction information based on the current time.

[0008] S4, based on the prediction result obtained in S3, managing the electricity information of the enterprise to be tested in the corresponding time period, and warning of abnormal electricity consumption state.

[0009] Further, the historical electricity load data in S1 includes predicted electricity load and actual electricity load corresponding to different time points;

[0010] The load reporting prediction information includes prediction duration and average electricity load in the corresponding prediction duration.

[0011] Further, when obtaining the actual order deviation rate corresponding to the adjacent two times of load reporting prediction information in S2, the actual order deviation rate is denoted as A, A=A1-A2, and the obtained actual order deviation rate is bound with the last load reporting prediction information in the adjacent two times of load reporting prediction information,

[0012] A1 represents the ratio of the total amount of electricity load order not completed by the enterprise to be tested to the remaining order duration when submitting the first load reporting prediction information in the adjacent two times of load reporting prediction information,

[0013] A2 represents the ratio of the total amount of electricity load order not completed by the enterprise to be tested to the remaining order duration when submitting the last load reporting prediction information in the adjacent two times of load reporting prediction information;

[0014] When the corresponding order deviation rate is obtained in S2, the method for reporting fluctuation characteristic function of the enterprise to be tested includes the following steps:

[0015] S21, obtaining each load reporting prediction information with the same actual order deviation rate, and obtaining the corresponding feature correlation information data pair of each load reporting prediction information under the corresponding actual order deviation rate, denoted as (B1, C1),

[0016] The C1 represents the data deviation amount in the corresponding time period of the corresponding load reporting prediction information, and the B1 represents the data fluctuation characteristic of the actual electricity load information in the corresponding time period of the previous load reporting prediction information of the corresponding load reporting prediction information.

[0017] S22, for a plurality of feature correlation information data pairs under the same actual order deviation rate, determining the corresponding coordinate point of each feature correlation information data pair in a plane rectangular coordinate system, and the plane rectangular coordinate system is a coordinate system of data fluctuation characteristic and data deviation amount.

[0018] S23, based on the function model y=P1×(x-P2) in the database 2+P3 and P1, P2 and P3 are coefficients of a function model, the coordinate points in the plane rectangular coordinate system in S22 are fitted, the function corresponding to the fitting result is taken as the case of the order deviation rate, the reporting fluctuation characteristic function of the to-be-tested enterprise, when the order deviation rate is M, the reporting fluctuation characteristic function of the to-be-tested enterprise is denoted as GM(x), and the average value of the distances from each obtained coordinate point in the plane rectangular coordinate system to GM(x) before fitting is obtained, which is denoted as LGM;

[0019] The method for obtaining the B1 comprises the following steps:

[0020] S201, obtaining a prediction time t1 corresponding to the corresponding load reporting prediction information and an average power consumption E in the corresponding prediction time;

[0021] S202, obtaining actual power consumptions corresponding to different time points in a time period corresponding to the corresponding load reporting prediction information, and taking the actual power consumption at a time t in the time period of the prediction time t1 as Et;

[0022] S203, obtaining B1, wherein the B1=1 / t1×(∫ t=0 t1 Etdt-E×t1);

[0023] The method for obtaining the C1 comprises the following steps:

[0024] S211, obtaining actual power consumptions corresponding to different time points in a time period corresponding to a previous load reporting prediction information of the corresponding load reporting prediction information, taking the actual power consumption at a time point with a time interval t2 from the minimum time point in the obtained time period as ESt2, and taking the maximum interval time in the obtained time period as t3;

[0025] S212, obtaining C1, wherein the C1=F[1 / t3×∫ t2=0 t3 ESt2dt2, EP, {ESt2|0≤t2≤t3}] / t3,

[0026] Wherein, EP represents a preset load error bearing value in the database,

[0027] {ESt2|0≤t2≤t3} represents a set composed of each ESt2 corresponding to a different value of t2 when 0≤t2≤t3,

[0028] F[1 / t3×∫ t2=0 t3 ESt2dt2, EP, {ESt2|0≤t2≤t3}] represents that ESt2 in {ESt2|0≤t2≤t3} does not belong to [1 / t3×∫ t2=0 t3ESt2dt2-EP,1 / t3x∫ t2=0 t3 The interval length of the time interval corresponding to the time point of all elements of ESt2dt2+EP.

[0029] When the order deviation rate is M, the reporting fluctuation characteristic function GM(x) of the enterprise to be tested is used to obtain the influence of the fluctuation of the actual power load data corresponding to the previous load reporting prediction information on the deviation of the subsequent load reporting prediction information, and provide data support for the selection of the marked points in the process of constructing the data reporting interval length adaptive model in the subsequent step.

[0030] Further, the method for constructing the data reporting interval length adaptive model in S3 comprises the following steps:

[0031] S301, obtaining the reporting fluctuation characteristic function GM(x) and LGM of the enterprise to be tested when the order deviation rate is M;

[0032] S302, obtaining all feature correlation information data pairs of the enterprise to be tested corresponding to GM(x) whose distance is less than or equal to β×LGM when the order deviation rate is M, and marking the coordinate points corresponding to the obtained feature correlation information data pairs in the plane rectangular coordinate system, wherein β is a constant preset in the database;

[0033] In the present application, β is set considering that the reference values of the coordinate points in the process of fitting the curve are different for the subsequent analysis steps. In the fitting process, the greater the distance between the coordinate point and GM(x), the greater the deviation of the coordinate point from the fitting result (theoretical analysis result), and the greater the error caused by including the coordinate point in the analysis object in the subsequent step. To a certain extent, β limits the selection range of the coordinate points, and thus ensures the accuracy of the subsequent analysis result to a certain extent. When the number of coordinate points is sufficient, the smaller the value of β, the greater the analysis accuracy.

[0034] S303, taking W as a reference point, obtaining all marked points whose absolute value of the difference between the first value and W is less than or equal to a first preset value h,

[0035] When the number of obtained marked points is less than a second preset value k, the first preset value h is adaptively adjusted, the first preset value h is added by an adjustment step h1 to obtain a new first preset value, and S303 is re-executed, wherein h1 is a constant preset in the database,

[0036] When the number of obtained marked points is greater than or equal to the second preset value k, S304 is jumped to.

[0037] The first preset value is adaptively adjusted, so as to ensure that the number of analyzed mark points is sufficient (the number of samples is sufficient), and then the accuracy of the analysis result is ensured.

[0038] S304, a data reporting interval time length adaptive adjustment feature corresponding to the reference point W is obtained, denoted as RW, the RW={RW1, RW2}, wherein RW1 represents an average value of the data reporting interval time length adaptive adjustment coefficient corresponding to each mark point of W, and RW2 represents an upper limit of the data reporting interval time length adaptive adjustment of W,

[0039] The data reporting interval time length adaptive adjustment coefficient corresponding to each mark point is equal to the ratio of the prediction time length of the second data corresponding to the load reporting prediction information in the corresponding mark point to the maximum interval time length of the actual power load information corresponding to the load reporting prediction information of the first data in the mark point,

[0040] When RW2 is obtained, first, the respective mark points corresponding to W are obtained, the difference between the prediction time length of the second data corresponding to the load reporting prediction information in each mark point and the maximum interval time length of the actual power load information corresponding to the load reporting prediction information of the first data in the mark point is calculated, and RW2 is equal to the average value of the respective mark points corresponding to W.

[0041] S305, the prediction time length of the load reporting prediction information corresponding to W in the data reporting interval time length adaptive model is obtained, denoted as TW, TW=min{RW1×T, RW2+T}, wherein TW represents the prediction time length in the load reporting prediction information of the previous submission of the corresponding load reporting prediction information, and min{RW1×T, RW2+T} represents the minimum value of RW1×T and RW2+T.

[0042] Further, in S3, the prediction result corresponding to the load reporting prediction information of the nearest submission based on the current time is obtained, the actual order deviation rate before and after the load reporting prediction information of the nearest submission based on the current time is calculated, and the reporting fluctuation feature function of the corresponding enterprise to be tested is obtained according to the obtained actual order deviation rate,

[0043] The load reporting prediction information of the nearest submission based on the current time is denoted as U, the prediction time length in the prediction result corresponding to U is denoted as TZ, and the average power load amount in the corresponding prediction time length of U is denoted as Q, wherein Q represents the average value of the actual power load in the time period corresponding to the load reporting prediction information of the previous submission of U,

[0044] obtaining the maximum interval duration of the time period corresponding to the load reporting prediction information submitted by U last time, denoted as TU1, obtaining the data fluctuation characteristics of the actual power load information in the time period corresponding to the load reporting prediction information submitted by U last time, denoted as W1, and combining the constructed data reporting interval duration adaptive model to obtain the prediction duration TZ corresponding to the load reporting prediction information when W in the data reporting interval duration adaptive model is W1,

[0045] when TZ is greater than or equal to the preset value in the database, the value corresponding to TZ is kept unchanged,

[0046] when TZ is less than the preset value in the database, the value corresponding to TZ is changed to the preset value to obtain a new TZ.

[0047] Further, when the power consumption information of the enterprise to be tested in the corresponding time period is managed in S4, the abnormal power consumption state represents a state in which the data deviation amount in the time period corresponding to the corresponding load reporting prediction information is greater than the preset value in the database, and the corresponding prediction duration in the next submitted load reporting prediction information after the sending of the early warning information is manually set by the user.

[0048] In the present application, the corresponding prediction duration in the next submitted load reporting prediction information after the sending of the early warning information is manually set by the user, which is considering that when the early warning condition occurs, it means that the prediction duration obtained by the data reporting interval duration adaptive model has a certain degree of deviation. Without human intervention (manual adjustment) in this state, the adaptive generated data reporting interval duration adaptive adjustment result in the subsequent step will have a large error, which will affect the user management of the user to be tested in the subsequent process.

[0049] The intelligent power plant prediction and scheduling system based on big data technology, the system comprises the following modules:

[0050] The power consumption data acquisition module acquires the historical power load data of the enterprise to be tested, and extracts the load reporting prediction information submitted by the enterprise to be tested each time in the historical data;

[0051] The reporting data fluctuation characteristic analysis module obtains the actual order deviation rate corresponding to the adjacent two load reporting prediction information of the enterprise to be tested, analyzes the relationship between the data deviation amount in the time period corresponding to the load reporting prediction information submitted by the enterprise to be tested each time and the data fluctuation characteristics of the actual power load information in the time period corresponding to the previous load reporting prediction information under the condition that the actual order deviation rate is the same, and obtains the reporting fluctuation characteristic function of the enterprise to be tested under the corresponding order deviation rate;

[0052] A prediction analysis module combines the analysis result in the report data fluctuation feature analysis module, constructs a data report interval length adaptive model, and combines the historical power load data of the to-be-tested enterprise to obtain a prediction result corresponding to the nearest submitted load report prediction information based on the current time;

[0053] A power consumption early warning management module manages the power consumption information of the to-be-tested enterprise in a corresponding time period based on the prediction result obtained in the prediction analysis module, and gives a warning for abnormal power consumption state.

[0054] Further, the report data fluctuation feature analysis module comprises an order deviation acquisition module and a fluctuation feature analysis module,

[0055] The order deviation acquisition module is used to acquire the actual order deviation rate corresponding to the nearest load report prediction information of the to-be-tested enterprise.

[0056] The fluctuation feature analysis module analyzes the relationship between the data deviation amount of the to-be-tested enterprise in the time period corresponding to the submitted load report prediction information and the data fluctuation feature of the actual power load information in the time period corresponding to the previous load report prediction information under the condition that the actual order deviation rate is the same, and obtains the report fluctuation feature function of the to-be-tested enterprise under the corresponding order deviation rate.

[0057] Compared with the prior art, the present application has the beneficial effects that the present application takes into account that the power load of the to-be-tested enterprise at different times in a time period is fluctuant, and the influence of the fluctuation intensity on the subsequent prediction result, and realizes adaptive adjustment of the load data report time according to the fluctuation degree of the power load data, ensures the accuracy of the dynamic load prediction result, and realizes accurate control of the power reserve of the power plant. BRIEF DESCRIPTION OF DRAWINGS

[0058] The accompanying drawings are used to provide a further understanding of the present application, and constitute a part of the specification, and are used to explain the present application together with embodiments of the present application, and do not constitute a limitation of the present application. In the drawings:

[0059] Figure 1 is a flowchart of the intelligent power plant prediction and scheduling method based on big data technology of the present application;

[0060] Figure 2 is a structural schematic diagram of the intelligent power plant prediction and scheduling system based on big data technology of the present application. DETAILED DESCRIPTION

[0061] With reference to the drawings of the embodiments of the present application, the technical solutions in the embodiments of the present application will be described clearly and completely. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments of the present application, all other embodiments obtained by those skilled in the art without creative work fall within the scope of the present application.

[0062] Please refer to Figure 1 The present application provides a technical solution: a smart power plant prediction and scheduling method based on big data technology, which comprises the following steps:

[0063] S1, receiving power consumption demand prediction information reported by a power consumption area covered by a power plant, obtaining historical power consumption load data of a to-be-tested enterprise, and extracting load reporting prediction information submitted by the to-be-tested enterprise each time in the historical data;

[0064] The historical power consumption load data in S1 comprises predicted power consumption load and actual power consumption load corresponding to different time points;

[0065] The load reporting prediction information comprises a prediction duration and an average power consumption load in the corresponding prediction duration.

[0066] S2, obtaining actual order deviation rates corresponding to adjacent two times of load reporting prediction information of the to-be-tested enterprise, analyzing the relationship between data deviation amounts in a time period corresponding to each time of load reporting prediction information submitted by the to-be-tested enterprise and data fluctuation characteristics of actual power consumption load information in a time period corresponding to previous load reporting prediction information in a case that actual order deviation rates are the same, and obtaining a reporting fluctuation characteristic function of the to-be-tested enterprise in the case of the corresponding order deviation rate;

[0067] In S2, when obtaining actual order deviation rates corresponding to adjacent two times of load reporting prediction information of the to-be-tested enterprise, the actual order deviation rate is recorded as A, A=A1-A2, and the obtained actual order deviation rate is bound with the last time of load reporting prediction information in the adjacent two times of load reporting prediction information,

[0068] A1 represents a ratio of a total amount of uncompleted power load orders of the to-be-tested enterprise to a remaining order duration when the first time of load reporting prediction information in the adjacent two times of load reporting prediction information is submitted,

[0069] A2 represents a ratio of a total amount of uncompleted power load orders of the to-be-tested enterprise to a remaining order duration when the second time of load reporting prediction information in the adjacent two times of load reporting prediction information is submitted;

[0070] In the embodiment, if the adjacent two times of load reporting prediction information are recorded as A and B respectively,

[0071] If the submission A, the enterprise to be tested the amount of uncompleted orders is e1 and the remaining order duration is te1,

[0072] If the submission B, the enterprise to be tested the amount of uncompleted orders is e2 and the remaining order duration is te2,

[0073] The actual order deviation rate of B binding is e1 / te1-e2 / te2,

[0074] At this time, e1 / te1 reflects the order completion pressure of the enterprise to be tested to a certain extent when submitting A, e2 / te2 reflects the order completion pressure of the enterprise to be tested to a certain extent when submitting B; the difference of e1 / te1-e2 / te2 reflects the change of the order completion pressure of the enterprise to be tested to a certain extent, and when the change is different, the data fluctuation characteristics and data deviation amount of subsequent analysis are also different;

[0075] Under the condition of obtaining the corresponding order deviation rate in S2, the method for reporting fluctuation characteristics of the enterprise to be tested comprises the following steps:

[0076] S21, obtaining each load reporting prediction information with the same actual order deviation rate of binding, and obtaining the corresponding feature correlation information data pair of each load reporting prediction information under the corresponding actual order deviation rate, denoted as (B1, C1),

[0077] The C1 represents the data deviation amount in the corresponding time period of the corresponding load reporting prediction information, and the B1 represents the data fluctuation characteristics of the actual power load information in the corresponding time period of the previous load reporting prediction information of the corresponding load reporting prediction information.

[0078] S22, for multiple feature correlation information data pairs under the same actual order deviation rate, determining the corresponding coordinate point of each feature correlation information data pair in the plane rectangular coordinate system, and the plane rectangular coordinate system is the coordinate system of data fluctuation characteristics and data deviation amount.

[0079] S23, based on the function model y=P1×(x-P2) 2 +P3 and P1, P2 and P3 are all coefficients of the function model, fitting the coordinate points in the plane rectangular coordinate system in S22, taking the function corresponding to the fitting result as the reporting fluctuation characteristic function of the enterprise to be tested under the corresponding order deviation rate, and taking the reporting fluctuation characteristic function of the enterprise to be tested when the order deviation rate is M as GM(x), and obtaining the average value of the distance from each obtained coordinate point in the plane rectangular coordinate system before fitting to GM(x), denoted as LGM.

[0080] The method for obtaining B1 comprises the following steps:

[0081] S201, acquire a prediction time t1 corresponding to the prediction information reported by the corresponding load and an average electricity consumption load E in the corresponding prediction time;

[0082] S202, acquire actual electricity consumption loads corresponding to different time points in a time period corresponding to the prediction information reported by the corresponding load, and record the actual electricity consumption load at a time t in the time period of the prediction time t1 as Et;

[0083] S203, obtain B1, B1 = 1 / t1 × (∫ t=0 t1 Etdt-E×t1);

[0084] The method for obtaining C1 includes the following steps:

[0085] S211, acquire actual electricity consumption loads corresponding to different time points in a time period corresponding to the previous prediction information reported by the corresponding load, record the actual electricity consumption load at a time point with a time interval t2 from the minimum time point in the obtained time period as ESt2, and record the maximum interval time in the obtained time period as t3;

[0086] S212, obtain C1, C1 = F[1 / t3 × ∫ t2=0 t3 ESt2dt2, EP, {ESt2|0≤t2≤t3}] / t3,

[0087] Wherein, EP represents a preset load error tolerance value in the database,

[0088] {ESt2|0≤t2≤t3} represents a set of ESt2 corresponding to different values of t2 when 0≤t2≤t3,

[0089] F[1 / t3 × ∫ t2=0 t3 ESt2dt2, EP, {ESt2|0≤t2≤t3}] represents the interval length of a time interval formed by the time points corresponding to all elements in {ESt2|0≤t2≤t3} that do not belong to [1 / t3 × ∫ t2=0 t3 ESt2dt2-EP, 1 / t3 × ∫ t2=0 t3 ESt2dt2+EP],

[0090] In this embodiment, the interval length of the time interval formed by the time points corresponding to all elements in {ESt2|0≤t2≤t3} that do not belong to [1 / t3 × ∫ t2=0 t3 ESt2dt2-EP, 1 / t3 × ∫ t2= 0 t3The time interval of all elements of ESt2dt2+EP] corresponding to the time point can be a union of a plurality of different time intervals, in which case F[1 / t3×∫ t2=0 t3 ESt2dt2, EP, {ESt2|0≤t2≤t3}] is equal to the sum of the interval lengths of the corresponding time intervals of each component in the resulting union.

[0091] S3, in combination with the analysis result in S2, constructs a data reporting interval length adaptive model, and in combination with the historical power load data of the enterprise to be tested, obtains a prediction result corresponding to the prediction information of the last submitted load report based on the current time;

[0092] The method for constructing a data reporting interval length adaptive model in S3 includes the following steps:

[0093] S301, when the order deviation rate is M, obtaining the reporting fluctuation characteristic function GM(x) and LGM of the enterprise to be tested;

[0094] S302, obtaining all feature correlation information data pairs of GM(x) corresponding to the enterprise to be tested when the order deviation rate is M, the distance of which is less than or equal to β×LGM, and marking the coordinate points corresponding to the obtained feature correlation information data pairs in the plane rectangular coordinate system, wherein β is a constant preset in the database;

[0095] S303, taking W as a reference point, obtaining all marked points in S302, the absolute value of the difference between the first value and W being less than or equal to a first preset value h,

[0096] When the number of obtained marked points is less than a second preset value k, then the first preset value h is adaptively adjusted, the first preset value h is added by an adjustment step h1 to obtain a new first preset value, and S303 is re-executed, wherein h1 is a constant preset in the database,

[0097] When the number of obtained marked points is greater than or equal to the second preset value k, then jump to S304;

[0098] S304, obtaining the data reporting interval length adaptive adjustment feature corresponding to the reference point W, denoted as RW, wherein RW={RW1, RW2}, wherein RW1 represents the average value of the data reporting interval length adaptive adjustment coefficient corresponding to each marked point of W, and RW2 represents the upper limit of the data reporting interval length adaptive adjustment of W,

[0099] The adaptive adjustment coefficient of the data reporting interval length corresponding to each marker point is equal to the ratio of the prediction length of the prediction information of the second data corresponding to the load in the corresponding marker point to the maximum interval length of the actual power load information of the prediction information of the first data corresponding to the load in the marker point,

[0100] When RW2 is obtained, the respective marker points corresponding to W are obtained first, the difference between the prediction length of the prediction information of the second data corresponding to the load in each marker point and the maximum interval length of the actual power load information of the prediction information of the first data corresponding to the load in the marker point is calculated, and RW2 is equal to the average of the respective difference values corresponding to each marker point of W.

[0101] S305, the prediction length of the load reporting prediction information corresponding to W in the data reporting interval length adaptive model is obtained, denoted as TW, TW = min{RW1 x T, RW2 + T}, the TW represents the prediction length in the load reporting prediction information of the previous submission of the corresponding load reporting prediction information, and the min{RW1 x T, RW2 + T} represents the minimum value of RW1 x T and RW2 + T.

[0102] The S3 obtains the prediction result corresponding to the latest submitted load reporting prediction information based on the current time, calculates the actual order deviation rate before and after the latest submitted load reporting prediction information based on the current time, and obtains the reporting fluctuation characteristic function of the corresponding enterprise to be tested according to the obtained actual order deviation rate,

[0103] The latest submitted load reporting prediction information based on the current time is denoted as U, the prediction length in the prediction result corresponding to U is denoted as TZ, and the average power load amount in the corresponding prediction length of U is denoted as Q, the Q represents the average value of the actual power load in the time period corresponding to the load reporting prediction information of the previous submission of U,

[0104] The maximum interval length of the time period corresponding to the load reporting prediction information of the previous submission of U is obtained, denoted as TU1, the data fluctuation characteristic of the actual power load information in the time period corresponding to the load reporting prediction information of the previous submission of U is obtained, denoted as W1, and the prediction length TZ of the load reporting prediction information corresponding to W as W1 in the data reporting interval length adaptive model is obtained by combining the constructed data reporting interval length adaptive model.

[0105] When TZ is greater than or equal to the preset value in the database, the value corresponding to TZ is kept unchanged,

[0106] When TZ is less than the preset value in the database, the value corresponding to TZ is changed to the preset value, and a new TZ is obtained.

[0107] S4. Based on the prediction results obtained in S3, manage the electricity consumption information of the enterprises to be tested within the corresponding time period and issue early warnings for abnormal electricity consumption status.

[0108] In S4, when managing the electricity consumption information of the enterprise to be tested within the corresponding time period, the abnormal electricity consumption status indicates that the data deviation of the corresponding load reporting prediction information within the corresponding time period is greater than the preset value of the database. In the next load reporting prediction information submitted after sending the warning information, the corresponding prediction duration is manually set by the user.

[0109] like Figure 2 As shown, a smart power plant prediction and scheduling system based on big data technology includes the following modules:

[0110] The electricity data acquisition module acquires the historical electricity load data of the enterprise under test and extracts the load reporting forecast information submitted by the enterprise under test each time from the historical data.

[0111] The data fluctuation feature analysis module obtains the actual order deviation rate corresponding to two adjacent load reporting forecasts of the enterprise under test. When the actual order deviation rate is the same, it analyzes the relationship between the data deviation amount of each load reporting forecast submitted by the enterprise under test within the corresponding time period and the data fluctuation characteristics of the actual electricity load information within the corresponding time period of the previous load reporting forecast, and obtains the reporting fluctuation feature function of the enterprise under test under the corresponding order deviation rate.

[0112] The predictive analysis module combines the analysis results from the reported data fluctuation characteristic analysis module to construct an adaptive model for the data reporting interval duration, and combines it with the historical electricity load data of the enterprise under test to obtain the prediction result corresponding to the most recently submitted load reporting prediction information based on the current time.

[0113] The electricity consumption early warning management module manages the electricity consumption information of the enterprises under test within the corresponding time period based on the prediction results obtained from the prediction analysis module, and issues early warnings for abnormal electricity consumption status.

[0114] The reported data fluctuation characteristic analysis module includes an order deviation acquisition module and a fluctuation characteristic analysis module.

[0115] The order deviation acquisition module is used to acquire the actual order deviation rate corresponding to two consecutive load forecast information reports from the enterprise under test.

[0116] The fluctuation feature analysis module analyzes the relationship between the data deviation amount of the data of the time period corresponding to the load reporting prediction information submitted by the to-be-tested enterprise each time and the data fluctuation feature of the actual power load information of the time period corresponding to the previous load reporting prediction information under the condition that the actual order deviation rate is the same, to obtain the reporting fluctuation feature function of the to-be-tested enterprise under the corresponding order deviation rate.

[0117] It should be noted that the relational terms herein such as first and second and the like are used solely to distinguish one entity or action from another, without necessarily requiring or implying any such actual relationship or order between such entities or actions. Moreover, the terms "comprises", "comprising", or any other variations thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but can include other elements not expressly listed or inherent to such process, method, article, or apparatus.

[0118] Finally, it should be noted that the above only describes the preferred embodiments of the present application and is not intended to limit the present application. Although the present application has been described in detail with reference to the foregoing embodiments, those skilled in the art will still be able to modify the technical solutions recorded in the foregoing embodiments or make equivalent replacements to some technical features. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principles of the present application shall be included in the protection scope of the present application.

Claims

1. A smart power plant prediction and scheduling method based on big data technology, characterized in that, The method comprises the following steps: S1, receiving power demand prediction information reported by a power consumption area covered by a power plant, obtaining historical power consumption load data of the enterprise to be tested, and extracting load reporting prediction information submitted by the enterprise to be tested each time in the historical data; The historical power consumption load data in S1 comprises predicted power consumption load and actual power consumption load corresponding to different time points; The load reporting prediction information comprises a prediction time length and an average power consumption load in the corresponding prediction time length; S2, obtaining actual order deviation rates corresponding to adjacent two load reporting prediction information of the enterprise to be tested, analyzing the relationship between the data deviation amount in the time period corresponding to the load reporting prediction information submitted by the enterprise to be tested each time and the data fluctuation characteristics of the actual power consumption load information in the time period corresponding to the previous load reporting prediction information under the condition that the actual order deviation rates are the same, and obtaining a reporting fluctuation characteristic function of the enterprise to be tested under the condition of the corresponding order deviation rate; In S2, the actual order deviation rate is denoted as A, A=A1-A2, and the obtained actual order deviation rate is bound to the last load reporting prediction information in the adjacent two load reporting prediction information, A1 represents the ratio of the total amount of the power load order not completed by the enterprise to be tested to the remaining order time length when the previous load reporting prediction information in the adjacent two load reporting prediction information is submitted, A2 represents the ratio of the total amount of the power load order not completed by the enterprise to be tested to the remaining order time length when the last load reporting prediction information in the adjacent two load reporting prediction information is submitted; The method for obtaining the reporting fluctuation characteristic function of the enterprise to be tested under the condition of the corresponding order deviation rate in S2 comprises the following steps: S21, obtaining each load reporting prediction information with the same actual order deviation rate, and obtaining a feature correlation information data pair corresponding to each load reporting prediction information under the corresponding actual order deviation rate, denoted as (B1, C1), C1 represents the data deviation amount in the time period corresponding to the corresponding load reporting prediction information, and B1 represents the data fluctuation characteristics of the actual power consumption load information in the time period corresponding to the previous load reporting prediction information of the corresponding load reporting prediction information; S22, for a plurality of feature correlation information data pairs under the same actual order deviation rate, determining a coordinate point corresponding to each feature correlation information data pair in a plane rectangular coordinate system, and the plane rectangular coordinate system is a coordinate system of data fluctuation characteristics and data deviation amount; S23, fitting the coordinate points in the plane rectangular coordinate system in S22 based on the function model y = P1x (x - P2) + P3, and P1, P2 and P3 are coefficients of the function model 2 + P3, and P1, P2 and P3 are coefficients of the function model, fitting the coordinate points in the plane rectangular coordinate system in S22, taking the function corresponding to the fitting result as the case of the order deviation rate, taking the order deviation rate as M, the reported fluctuation characteristic function of the enterprise to be tested is recorded as GM (x), and the average value of the distance between each obtained coordinate point in the plane rectangular coordinate system before fitting and GM (x) is obtained, which is recorded as LGM. The method for obtaining B1 comprises the following steps: S201, obtaining a prediction time length t1 corresponding to the corresponding load reporting prediction information and an average power consumption load in the corresponding prediction time length; S202, obtaining actual power consumption loads corresponding to different time points in the time period corresponding to the corresponding load reporting prediction information, and denoting the actual power consumption load at time t in the time period of the prediction time length t1 as Et; S203, obtaining B1, which is B1 = 1 / t1 x (∫ t=0 t1 Etdt-E×t1); The method for obtaining C1 comprises the following steps: S211, acquire actual power consumption load corresponding to different time points in a time period before the corresponding load reporting prediction information is acquired, and record the actual power consumption load corresponding to a time point with a time interval t2 from the minimum time point in the time period as ESt2, and record the maximum interval t3 in the time period; S212, obtain C1, C1=F[1 / t3 x ∫ t2=0 t3 ESt2dt2, EP, {ESt2|0≤t2≤t3} / t3, wherein EP represents a preset load error bearing value in the database, {ESt2|0≤t2≤t3} represents a set of each ESt2 corresponding to different values of t2 when 0≤t2≤t3, F[1 / t3×∫ t2=0 t3 ESt2dt2, EP, {ESt2|0≤t2≤t3}] represent that in {ESt2|0≤t2≤t3}, the subset [1 / t3×∫] is not found in [1 / t3×∫]. t2=0 t3 ESt2dt2-EP,1 / t3×∫ t2=0 t3 The length of the time interval formed by the time points corresponding to all elements of ESt2dt2+EP; S3, combined with the analysis result in S2, an adaptive model of data reporting interval is constructed, and combined with the historical power consumption load data of the enterprise to be tested, a prediction result corresponding to the last submitted load reporting prediction information based on the current time is obtained; S4, based on the prediction result obtained in S3, the power consumption information of the enterprise to be tested in the corresponding time period is managed, and the abnormal power consumption state is warned. 2.The smart power plant prediction and scheduling method based on big data technology according to claim 1, characterized in that: The method for constructing the adaptive model of data reporting interval in S3 comprises the following steps: S301, when the order deviation rate is M, acquire the reporting fluctuation characteristic function GM(x) and LGM of the enterprise to be tested; S302, when the order deviation rate is M, acquire all the feature correlation information data pairs in which the distance of GM(x) corresponding to each feature correlation information data of the enterprise to be tested is less than or equal to β×LGM, and mark the coordinate points corresponding to the obtained feature correlation information data pairs in the plane rectangular coordinate system, wherein β is a constant preset in the database; S303, taking W as a reference point, acquire all the marked points in S302 in which the absolute value of the difference between the first value and W is less than or equal to a first preset value h, when the number of the obtained marked points is less than a second preset value k, then the first preset value h is adaptively adjusted, the first preset value h is added by an adjustment step h1 to obtain a new first preset value, and S303 is re-executed, wherein h1 is a constant preset in the database, when the number of the obtained marked points is greater than or equal to the second preset value k, then jump to S304; S304, acquire the adaptive adjustment feature of the data reporting interval corresponding to the reference point W, denoted as RW, wherein RW={RW1, RW2}, wherein RW1 represents the average value of the adaptive adjustment coefficient of the data reporting interval corresponding to each marked point of W, and RW2 represents the upper limit of the adaptive adjustment of the data reporting interval of W, the adaptive adjustment coefficient of the data reporting interval corresponding to each marked point is equal to the ratio of the prediction time length of the load reporting prediction information corresponding to the second data in the corresponding marked point to the maximum interval of the actual power consumption load information corresponding to the load reporting prediction information of the first data in the marked point, when RW2 is acquired, first acquire each marked point corresponding to W, calculate the difference between the prediction time length of the load reporting prediction information corresponding to the second data in each marked point and the maximum interval of the actual power consumption load information corresponding to the load reporting prediction information of the first data in the marked point, and RW2 is equal to the average value of the difference values corresponding to each marked point of W; S305, obtain the prediction time length corresponding to the load report prediction information corresponding to W in the data reporting interval length adaptive model, denoted as TW, TW = min{RW1 × T, RW2 + T}, the TW indicates the prediction time length in the load report prediction information of the previous submission corresponding to the load report prediction information, and the min{RW1 × T, RW2 + T} indicates the minimum value of RW1 × T and RW2 + T. 3.The smart power plant prediction and scheduling method based on big data technology according to claim 2, characterized in that: In S3, the prediction result based on the latest submitted load report prediction information corresponding to the current time is obtained, the actual order deviation rate before and after the latest submitted load report prediction information corresponding to the current time is calculated, and the reporting fluctuation characteristic function of the corresponding enterprise is obtained according to the obtained actual order deviation rate, The latest submitted load report prediction information corresponding to the current time is denoted as U, the prediction time length in the prediction result corresponding to U is denoted as TZ, the average power consumption load amount in the corresponding prediction time length of U is denoted as Q, the Q indicates the average value of the actual power consumption load in the time period corresponding to the load report prediction information of the previous submission of U, The maximum interval length of the time period corresponding to the load report prediction information of the previous submission of U is obtained, denoted as TU1, the data fluctuation characteristics of the actual power consumption load information in the time period corresponding to the load report prediction information of the previous submission of U are obtained, denoted as W1, and the prediction time length TZ corresponding to the load report prediction information corresponding to W as W1 in the data reporting interval length adaptive model is obtained by combining the constructed data reporting interval length adaptive model, When TZ is greater than or equal to the preset value in the database, the value corresponding to TZ is kept unchanged, When TZ is less than the preset value in the database, the value corresponding to TZ is changed to the preset value, and a new TZ is obtained. 4.The smart power plant prediction and scheduling method based on big data technology according to claim 1, characterized in that: In S4, when the power consumption information of the enterprise to be tested in the corresponding time period is managed, the abnormal power consumption state indicates that the data deviation amount in the time period corresponding to the corresponding load report prediction information is greater than the preset value in the database, and the corresponding prediction time length in the next submitted load report prediction information after sending the warning information is manually set by the user. 5.The smart power plant prediction and scheduling system based on big data technology using the smart power plant prediction and scheduling method based on big data technology according to any one of claims 1-4, characterized in that, The system comprises the following modules: The power consumption data acquisition module acquires the historical power consumption load data of the enterprise to be tested, and extracts the load report prediction information submitted by the enterprise to be tested each time in the historical data; The reporting data fluctuation characteristic analysis module obtains the actual order deviation rate corresponding to the adjacent two load report prediction information of the enterprise to be tested, analyzes the relationship between the data deviation amount in the time period corresponding to the load report prediction information submitted by the enterprise to be tested each time and the data fluctuation characteristics of the actual power consumption load information in the time period corresponding to the previous load report prediction information under the condition that the actual order deviation rate is the same, and obtains the reporting fluctuation characteristic function of the enterprise to be tested under the condition of the corresponding order deviation rate; A prediction analysis module combines the analysis results in the reporting data fluctuation feature analysis module, constructs a data reporting interval length adaptive model, and combines historical power load data of the enterprise to be tested to obtain a prediction result corresponding to the most recently submitted load reporting prediction information based on the current time; A power consumption early warning management module manages power consumption information of the enterprise to be tested in a corresponding time period based on the prediction result obtained in the prediction analysis module, and warns of abnormal power consumption states. 6.The smart power plant prediction and scheduling system based on big data technology according to claim 5, characterized in that: The reporting data fluctuation feature analysis module includes an order deviation acquisition module and a fluctuation feature analysis module, The order deviation acquisition module is used to acquire actual order deviation rates corresponding to adjacent twice load reporting prediction information of the enterprise to be tested. The fluctuation feature analysis module analyzes the relationship between data deviation amounts of the enterprise to be tested in a time period corresponding to each submitted load reporting prediction information and data fluctuation features of actual power load information in a time period corresponding to previous load reporting prediction information under the condition that actual order deviation rates are the same, and obtains a reporting fluctuation feature function of the enterprise to be tested under the condition of a corresponding order deviation rate.

Images