CN114254801A - Industrial user power consumption long-term prediction method and device based on similarity measurement - Google Patents

Industrial user power consumption long-term prediction method and device based on similarity measurement Download PDF

Info

Publication number
CN114254801A
CN114254801A CN202111320362.5A CN202111320362A CN114254801A CN 114254801 A CN114254801 A CN 114254801A CN 202111320362 A CN202111320362 A CN 202111320362A CN 114254801 A CN114254801 A CN 114254801A
Authority
CN
China
Prior art keywords
power consumption
samples
prediction
predicted
similarity
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Pending
Application number
CN202111320362.5A
Other languages
Chinese (zh)
Inventor
褚晓泉
仇瑜
唐杰
李亚坤
王朝亮
胡若云
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Tsinghua University
Marketing Service Center of State Grid Zhejiang Electric Power Co Ltd
Original Assignee
Tsinghua University
Marketing Service Center of State Grid Zhejiang Electric Power Co Ltd
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Tsinghua University, Marketing Service Center of State Grid Zhejiang Electric Power Co Ltd filed Critical Tsinghua University
Priority to CN202111320362.5A priority Critical patent/CN114254801A/en
Publication of CN114254801A publication Critical patent/CN114254801A/en
Pending legal-status Critical Current

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q10/00—Administration; Management
    • G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00—Pattern recognition
    • G06F18/20—Analysing
    • G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
    • G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00—Pattern recognition
    • G06F18/20—Analysing
    • G06F18/22—Matching criteria, e.g. proximity measures
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q50/00—Systems or methods specially adapted for specific business sectors, e.g. utilities or tourism
    • G06Q50/06—Electricity, gas or water supply

Abstract

The application provides a similarity measurement-based long-term prediction method for power consumption of industrial users, and relates to the technical field of long-term prediction of power consumption, wherein the method comprises the following steps: preprocessing monthly power consumption time sequence data to generate a sample; carrying out standardization processing on a sample, and constructing a data set, wherein the data set is divided into a training set and a testing set; for the samples to be predicted in the test set, introducing dynamic time warping to measure the similarity between the samples to be predicted and the samples in the training set, and selecting the samples with higher similarity degree as reference samples for implementing prediction; and taking the screened samples as a basis, constructing a time sequence prediction model to predict the power consumption of the target user, and obtaining a predicted value of the monthly power consumption of the target user. By adopting the scheme, the method and the device solve the difficult problem of prediction of large power consumption characteristic difference of industrial users in different industries, realize long-term prediction of power consumption for industrial users, and provide reference for overall planning and planning of regional power consumption.

Description

Industrial user power consumption long-term prediction method and device based on similarity measurement
Technical Field
The application relates to the technical field of long-term prediction of power consumption, in particular to a method and a device for long-term prediction of power consumption of an industrial user based on similarity measurement.
Background
Electric power is considered to be the most flexible energy source and an important secondary energy source, and is an important infrastructure component for socioeconomic development. At present, the rapid growth of domestic economy and population is accelerating the consumption of electric energy, and predicting the future change trend of the electric energy consumption is vital to promoting the sustainable development of the economy and the society of China and realizing the effective management of an electric power system.
Industrial enterprises take a leading position in power consumption, and the scientific prediction and mastering of the power consumption change trend of industrial users have important significance for upgrading the industrial structure in China and realizing the high-quality development of industrial departments. Long-term power usage prediction aims at estimating the power demand of a user over a long-term future range to achieve a dynamic balance between demand and resources. The long-term power consumption prediction can provide reference information for building new power generation facilities, purchasing existing generator sets, perfecting power transmission and distribution systems and the like, and plays an important guiding role in national power structure planning and energy development. Therefore, a long-term prediction method for the power consumption of industrial users is constructed, and scientific and accurate prediction and analysis are realized, so that the method has important practical significance on the development of the economic society.
The power consumption time series data type of the industrial user has the characteristics of nonlinearity and non-stability, and a great challenge is faced to the realization of long-term prediction of the power consumption time series data type, so that different technologies are proposed to realize a high-precision robust model. Most models are based on a sliding window mechanism, original data are processed into an isometric sequence and then are modeled, future numerical values are directly predicted, and the method has high requirements on data integrity and cannot guarantee the application condition in engineering practice. The similarity measurement method based on the sequence morphology not only screens candidate samples from historical trends, but also creates conditions for the prediction of sequences with different lengths which are ubiquitous in practical application.
Studies have shown that if two sequence data have similar morphology, then both may have similar values in a future time period, and thus the idea can be applied to prediction of time series data. Similarity metric algorithms can be divided into two categories: a lock step metric and a resiliency metric. The lock step measurement method is simple in calculation, but only the equivalent long-time sequence data has better performance and lower adaptability; the elastic measurement method represented by Dynamic Time Warping (DTW) maps one time sequence to another time sequence for measuring the similarity between the two time sequences and the minimum distance between the two time sequences, can measure the similarity between the sequences with different lengths, and has higher precision and robustness.
Disclosure of Invention
The present application is directed to solving, at least to some extent, one of the technical problems in the related art.
Therefore, a first purpose of the application is to provide a similarity measurement-based method for predicting the power consumption of the industrial users for a long time, so that the problems that the existing method has high requirements on data integrity and cannot guarantee the application condition in engineering practice are solved, the difficult problem of prediction of large power consumption characteristic difference of the industrial users in different industries is solved, a similarity measurement and least square support vector machine-based monthly power consumption prediction method is constructed from engineering practical data, the long-term prediction of the power consumption for the industrial users is realized, and a reference is provided for overall planning and planning of regional power consumption.
The second purpose of the application is to provide a device for predicting the electricity consumption of the industrial users for a long time based on the similarity measurement.
A third object of the present application is to propose a non-transitory computer-readable storage medium.
In order to achieve the above object, an embodiment of a first aspect of the present application provides a method for predicting a power consumption of an industrial user for a long time based on a similarity metric, including: preprocessing monthly power consumption time sequence data to generate a sample; carrying out standardization processing on a sample, and constructing a data set, wherein the data set is divided into a training set and a testing set; for the samples to be predicted in the test set, introducing dynamic time warping to measure the similarity between the samples to be predicted and the samples in the training set, and selecting the samples with higher similarity degree as reference samples for implementing prediction; and taking the screened samples as a basis, constructing a time sequence prediction model to predict the power consumption of the target user, and obtaining a predicted value of the monthly power consumption of the target user.
Optionally, in an embodiment of the present application, the sample is normalized, specifically:
Figure BDA0003345363500000021
wherein o ismax、omin、o、onormRespectively representing the maximum value, the minimum value, the current value and the standardized result value of the monthly electricity consumption time sequence data of the same industrial user.
Optionally, in an embodiment of the present application, a dynamic time warping metric is introduced to measure similarity between a sample to be predicted and a sample in a training set, where the dynamic time metric is expressed as:
Figure BDA0003345363500000022
wherein, Otest(t)、Otrain(t) respectively representing a sample to be predicted and data to be compared in a training set, w represents a DTW shortest path lower degree corresponding point element, and the ith element w in wl＝(Otest(i),Otrain(j))l。
Optionally, in an embodiment of the present application, a time sequence prediction model is constructed based on a least squares support vector machine LSSVR to predict the power consumption of the target user, so as to obtain a predicted value of the monthly power consumption of the target user, where the LSSVR is expressed as:
Figure BDA0003345363500000023
wherein alpha isiRepresenting a series of Lagrange multipliers, xiRepresenting an input variable, b representing a deviation, K () representing a kernel function, using RBF as the kernel function, the K () expression being
Figure BDA0003345363500000024
In order to achieve the above object, a second aspect of the present application provides an apparatus for predicting long-term power consumption of an industrial user based on a similarity metric, including: a preprocessing module, a data set generating module, a screening module and a model predicting module, wherein,
the preprocessing module is used for preprocessing monthly power consumption time sequence data to generate a sample;
the data set generating module is used for carrying out standardization processing on the samples and constructing a data set, wherein the data set is divided into a training set and a testing set;
the screening module is used for introducing dynamic time warping to the samples to be predicted in the test set, measuring the similarity between the samples to be predicted and the samples in the training set, and selecting the samples with higher similarity degree as reference samples for implementing prediction;
and the model prediction module is used for constructing a time sequence prediction model to predict the power consumption of the target user by taking the screened samples as the basis so as to obtain a predicted value of the monthly power consumption of the target user.
To achieve the above object, a non-transitory computer readable storage medium is provided in a third aspect of the present application, and when executed by a processor, the instructions in the storage medium can perform a method for long-term prediction of power consumption of an industrial user based on a similarity metric.
The similarity measurement-based industrial user power consumption long-term prediction method, the similarity measurement-based industrial user power consumption long-term prediction device and the non-transitory computer-readable storage medium solve the problems that the existing method has high requirements for data integrity and cannot guarantee the application condition in engineering practice, solve the prediction problem that the power consumption characteristics of industrial users in different industries have large differences, and construct a similarity measurement and least square support vector machine-based monthly power consumption prediction method from engineering actual data, thereby realizing the long-term prediction of industrial users and providing reference for regional power consumption planning and planning.
Additional aspects and advantages of the present application will be set forth in part in the description which follows and, in part, will be obvious from the description, or may be learned by practice of the present application.
Drawings
The foregoing and/or additional aspects and advantages of the present application will become apparent and readily appreciated from the following description of the embodiments, taken in conjunction with the accompanying drawings of which:
fig. 1 is a flowchart of a method for predicting power consumption of an industrial user for a long time based on similarity measurement according to an embodiment of the present disclosure;
FIG. 2 is another flow chart of a method for long-term prediction of power usage by an industrial user based on similarity measurements according to an embodiment of the present disclosure;
FIG. 3 is a diagram illustrating monthly power consumption data of an industrial user in a method for long-term prediction of power consumption of the industrial user based on similarity measurement according to an embodiment of the present disclosure;
fig. 4 is a schematic diagram of a dynamic time warping DTW of a similarity measurement-based method for long-term prediction of power consumption of an industrial user according to an embodiment of the present application;
fig. 5 is a schematic structural diagram of a device for predicting electricity consumption of an industrial user based on similarity measurement according to a second embodiment of the present application.
Detailed Description
Reference will now be made in detail to embodiments of the present application, examples of which are illustrated in the accompanying drawings, wherein like or similar reference numerals refer to the same or similar elements or elements having the same or similar function throughout. The embodiments described below with reference to the drawings are exemplary and intended to be used for explaining the present application and should not be construed as limiting the present application.
The method and the device for predicting the electricity consumption of the industrial users based on the similarity measurement are described below with reference to the accompanying drawings.
Fig. 1 is a flowchart of a method for predicting power consumption of an industrial user based on similarity measurement according to an embodiment of the present disclosure.
As shown in fig. 1, the method for predicting the electricity consumption of the industrial user based on the similarity measure includes the following steps:
101, preprocessing monthly power consumption time sequence data to generate a sample;
step 102, carrying out standardization processing on a sample, and constructing a data set, wherein the data set is divided into a training set and a testing set;
103, introducing dynamic time warping to the samples to be predicted in the test set, measuring the similarity between the samples to be predicted and the samples in the training set, and selecting the samples with higher similarity as reference samples for prediction;
and 104, taking the screened samples as a basis, constructing a time sequence prediction model to predict the power consumption of the target user, and obtaining a predicted value of the monthly power consumption of the target user.
According to the method for predicting the power consumption of the industrial user based on the similarity measurement, a sample is generated by preprocessing monthly power consumption time sequence data; carrying out standardization processing on a sample, and constructing a data set, wherein the data set is divided into a training set and a testing set; for the samples to be predicted in the test set, introducing dynamic time warping to measure the similarity between the samples to be predicted and the samples in the training set, and selecting the samples with higher similarity degree as reference samples for implementing prediction; and taking the screened samples as a basis, constructing a time sequence prediction model to predict the power consumption of the target user, and obtaining a predicted value of the monthly power consumption of the target user. Therefore, the problems that the existing method has high requirement on data integrity and cannot guarantee the application condition in engineering practice can be solved, the prediction difficulty that the difference of the power consumption characteristics of industrial users in different industries is large is solved, the monthly power consumption prediction method based on similarity measurement and a least square support vector machine is constructed from the actual engineering data, the long-term prediction of the power consumption of industrial users is realized, and reference is provided for overall planning and planning of regional power consumption.
The application designs a data processing and predicting method based on similarity measurement and machine learning. Firstly, for a user to be predicted, screening similar users as a basis for building a prediction model according to historical power consumption data of the user through a Dynamic Time Warping (DTW) of a similarity measurement algorithm, and training and building the model through the screened samples to ensure that the samples participating in model training and a target user have similar power consumption characteristics; then, a regression prediction model is constructed by adopting a Least Square Support Vector Machine (LSSVM), and prediction of future monthly power consumption is realized. The method and the device ensure the similarity between the data of the model training in the sample screening stage and the target user, solve the prediction problem of large power consumption characteristic difference of industrial users in different industries, and provide reference for overall planning and planning of regional power consumption.
Counting the electricity consumption data of the industrial users according to the months to obtain monthly electricity consumption time sequence data of a single user, and taking the monthly electricity consumption time sequence data as a sample Oi(t) of (d). Electricity consumption sequence O for original industrial useri(T), wherein T is 1, 2, the. The long-term prediction method aims at analyzing OiAnd (T) searching for similar users as a prediction basis according to the historical change rule, and constructing a model to predict the electricity consumption of the T month. The expression is as follows:
Figure BDA0003345363500000051
wherein the content of the first and second substances,
Figure BDA0003345363500000052
to predict the value of the target, F (-) represents the prediction model constructed in this application.
In the embodiment of the application, in order to research a monthly power consumption prediction model F (-) of an industrial user, power consumption data of different industrial users are collected at first, and time sequence data of 13 months in total in a certain period of time is used as a sample. The electricity consumption data of 12 months was used as a prediction input variable, and the electricity consumption data of 13 th month was used as a prediction target. According to definition, Oi(t), wherein t 1, 2, 13, the prediction window size is 12.
In order to avoid prediction deviation caused by difference of orders of magnitude of different samples, the samples are standardized.
Further, in the embodiment of the present application, the sample is normalized, specifically:
Figure BDA0003345363500000053
wherein o ismax、omin、o、onormRespectively representing the maximum value, the minimum value, the current value and the standardized result value of the monthly electricity consumption time sequence data of the same industrial user.
And dividing different industrial user samples into a training set and a testing set, and preparing for further similarity measurement and power consumption prediction.
And for the samples to be predicted in the test set, introducing Dynamic Time Warping (DTW) to measure the similarity between the samples to be predicted and the samples in the training set, and selecting the samples with higher similarity degree as prediction reference samples.
The Dynamic Time Warping (DTW) process aims to determine user power consumption data with similarity to a sample to be predicted through screening, and the data serve as a basis for model construction. Since in the time series prediction task, it is generally considered that if two sequence data have similar morphology, they may have similar values in a future period of time. Preliminary sample screening helps make model construction more reliable.
The specific implementation method of the dynamic time warping DTW process is as follows:
for the sample O to be predictedtest(t) and data O to be compared in training settrain(t), in the present embodiment, t 1, 2. To measure Otest(t) and OtrainAnd (t) the similarity distance between the two sequences needs to be aligned by operations such as lengthening or shortening, and the like.
In this application, for Otest(t) and Otrain(t) constructing a 13 matrix using d (o)test(i)，otrain(j) To represent a point Otest(t) points i and O in the sequencetrain(t) the distance between j points in the sequence, the distance between two points being calculated as the euclidean distance:
Figure BDA0003345363500000061
in the process of aligning the key elements, an optimal path is searched in the grid in a dynamic planning mode, and the sequence value corresponding to the grid where the points in the path are located is the point pair where the two sequences are to be aligned. The path W may be defined as W ═ (W)1，w2，...，wl，...，wL) Wherein the 1 st element W in Wl＝(otest(i)，otrain(j))1. The selection of points in the path follows the following condition: boundary conditions, for any two sequences, the path starts from the lower left corner of the grid and ends at the upper right corner; continuity, the path can only be matched with adjacent points during selection, and can not cross the adjacent points to be matched with other points; monotonicity, points in the path must be monotonously proceeding to the right or to the upper right. And when the conditions are met, the DTW selects the minimum path for the electricity consumption data of the industrial users.
Further, in the embodiment of the present application, a dynamic time warping metric is introduced to measure the similarity between the sample to be predicted and the sample in the training set, where the dynamic time metric is expressed as:
Figure BDA0003345363500000062
wherein, Otest(t)、Otrain(t) respectively representing a sample to be predicted and data to be compared in a training set, w represents a DTW shortest path lower degree corresponding point element, and the ith element w in wl＝(Otest(i)，Otrain(j))l。
Further, in the embodiment of the present application, a time sequence prediction model is constructed based on a least squares support vector machine LSSVR to predict the power consumption of the target user, so as to obtain a predicted value of the monthly power consumption of the target user, where the LSSVR is expressed as:
Figure BDA0003345363500000063
wherein alpha isiRepresenting a series of Lagrange multipliers, xiRepresenting an input variable, b representing a deviation, K () representing a kernel function, using RBF as the kernel function, the K () expression being
Figure BDA0003345363500000064
According to the method, a Least Square Support Vector Machine (LSSVM) is selected for prediction, and a Particle Swarm Optimization (PSO) is introduced as an Optimization algorithm for parameter selection.
The LSSVM is a variant of the support vector machine, solves a linear equation set by using equation constraint, replaces a quadratic programming mode in an original support vector machine, and expands the application of the support vector machine under the condition of increasing data quantity. For a given data set
Figure BDA0003345363500000065
xi represents an input variable, yi is a prediction target corresponding thereto, and N is the number of samples. According to the structural risk minimization criterion, the optimization problem of LSSVR can be summarized as:
Figure BDA0003345363500000066
s.t.yi＝wTφ(xi)+b+ξi，i＝1，2，...，N
where w is the weight vector, b is the offset, φ (·) represents the mapping between the input space and the output space, and the relaxation variable ξiNot less than 0, the penalty coefficient gamma is more than 0,
solving the Lagrange function by using a Lagrange method to obtain a Lagrange function corresponding to the above formula as follows:
Figure BDA0003345363500000071
wherein, { aiN is a series of Lagrange multipliers. By taking partial derivatives of w, b, ξ, α, respectively, the optimization condition of the Lagrange function can be obtained as follows:
Figure BDA0003345363500000072
the above optimization condition equation can be converted into the following equation form:
Figure BDA0003345363500000073
wherein α ═ α1，α2，...，αN]，Y＝[y1，y2，...，yN]Where I is the identity matrix, Ψkj＝K(xk，xj) N is a kernel function that satisfies the Mercer criterion.
And predicting through the model to obtain a predicted value of the T-month electricity consumption of the target user, designing and realizing a platform and a carrier which are realized by taking the electricity consumption prediction system of the industrial user as the model, and applying the platform and the carrier to the reality.
FIG. 2 is another flowchart of a method for predicting long-term power consumption of an industrial user based on similarity measurement according to an embodiment of the present disclosure.
As shown in fig. 2, in the similarity measurement-based method for long-term prediction of power consumption of an industrial user, considering that the power consumption change characteristics of the industrial user are obviously different due to different fields and industries of the industrial user, samples close to the historical trend of a sample to be predicted are screened out through time series similarity measurement before prediction is implemented, and the samples are used as candidate samples constructed by a prediction model, so that the prediction basis is stronger; a screening threshold value is set according to actual prediction requirements, a sample with the similarity ranking at the front is used as a prediction basis, a prediction model is built, corresponding devices and systems are realized, and long-term prediction of power consumption for industrial users is realized. According to the method and the device, a similarity measurement process is added in a traditional model building process, so that the similarity between a sample participating in prediction and a target sample is higher, and the model performance defect caused by the difference of power consumption time sequence data of different industrial users can be avoided.
Fig. 3 is a schematic diagram of monthly power consumption data of an industrial user in a method for long-term prediction of power consumption of the industrial user based on similarity measurement according to an embodiment of the present application.
As shown in fig. 3, monthly electricity usage data for 13 months in a certain period of time for some industrial users is shown.
Fig. 4 is a schematic diagram of a dynamic time warping DTW of a method for long-term prediction of power consumption of an industrial user based on similarity measurement according to an embodiment of the present application.
As shown in fig. 4, in order to select a warping path that minimizes the distance between two time series, matching is performed from the grid point at the bottom left corner of the matrix, when the next grid point is reached, the distances calculated by all previous grids are accumulated, and finally the end point at the top right corner is reached, and the points in the grids are sequentially connected from left to right and from bottom to top, so that the dynamic warping distance between the two time series can be obtained.
Fig. 5 is a schematic structural diagram of a device for predicting electricity consumption of an industrial user based on similarity measurement according to a second embodiment of the present application.
As shown in fig. 5, the apparatus for predicting long-term electricity consumption of industrial users based on similarity measure includes: a preprocessing module, a data set generating module, a screening module and a model predicting module, wherein,
the preprocessing module 10 is used for preprocessing monthly electricity consumption time sequence data to generate a sample;
the data set generating module 20 is configured to perform standardization processing on the samples to construct a data set, where the data set is divided into a training set and a test set;
the screening module 30 is configured to introduce dynamic time warping to the samples to be predicted in the test set to measure the similarity between the samples to be predicted and the samples in the training set, and select the sample with a higher degree of similarity as a reference sample for performing prediction;
and the model prediction module 40 is used for constructing a time sequence prediction model to predict the power consumption of the target user by taking the screened samples as the basis, so as to obtain a predicted value of the monthly power consumption of the target user.
The device for predicting the electricity consumption of the industrial user based on the similarity measurement comprises the following components: the device comprises a preprocessing module, a data set generating module, a screening module and a model predicting module, wherein the preprocessing module is used for preprocessing monthly power consumption time sequence data to generate a sample; the data set generating module is used for carrying out standardization processing on the samples and constructing a data set, wherein the data set is divided into a training set and a testing set; the screening module is used for introducing dynamic time warping to the samples to be predicted in the test set, measuring the similarity between the samples to be predicted and the samples in the training set, and selecting the samples with higher similarity degree as reference samples for implementing prediction; and the model prediction module is used for constructing a time sequence prediction model to predict the power consumption of the target user by taking the screened samples as the basis so as to obtain a predicted value of the monthly power consumption of the target user. Therefore, the problems that the existing method has high requirement on data integrity and cannot guarantee the application condition in engineering practice can be solved, the prediction difficulty that the difference of the power consumption characteristics of industrial users in different industries is large is solved, the monthly power consumption prediction method based on similarity measurement and a least square support vector machine is constructed from the actual engineering data, the long-term prediction of the power consumption of industrial users is realized, and reference is provided for overall planning and planning of regional power consumption.
In order to achieve the above embodiments, the present application further proposes a non-transitory computer readable storage medium, on which a computer program is stored, and the computer program, when executed by a processor, implements the method for long-term prediction of power consumption of an industrial user based on a similarity measure of the above embodiments.
In the description herein, reference to the description of the term "one embodiment," "some embodiments," "an example," "a specific example," or "some examples," etc., means that a particular feature, structure, material, or characteristic described in connection with the embodiment or example is included in at least one embodiment or example of the application. In this specification, the schematic representations of the terms used above are not necessarily intended to refer to the same embodiment or example. Furthermore, the particular features, structures, materials, or characteristics described may be combined in any suitable manner in any one or more embodiments or examples. Furthermore, various embodiments or examples and features of different embodiments or examples described in this specification can be combined and combined by one skilled in the art without contradiction.
Furthermore, the terms "first", "second" and "first" are used for descriptive purposes only and are not to be construed as indicating or implying relative importance or implicitly indicating the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include at least one such feature. In the description of the present application, "plurality" means at least two, e.g., two, three, etc., unless specifically limited otherwise.
Any process or method descriptions in flow charts or otherwise described herein may be understood as representing modules, segments, or portions of code which include one or more executable instructions for implementing steps of a custom logic function or process, and alternate implementations are included within the scope of the preferred embodiment of the present application in which functions may be executed out of order from that shown or discussed, including substantially concurrently or in reverse order, depending on the functionality involved, as would be understood by those reasonably skilled in the art of the present application.
The logic and/or steps represented in the flowcharts or otherwise described herein, e.g., an ordered listing of executable instructions that can be considered to implement logical functions, can be embodied in any computer-readable medium for use by or in connection with an instruction execution system, apparatus, or device, such as a computer-based system, processor-containing system, or other system that can fetch the instructions from the instruction execution system, apparatus, or device and execute the instructions. For the purposes of this description, a "computer-readable medium" can be any means that can contain, store, communicate, propagate, or transport the program for use by or in connection with the instruction execution system, apparatus, or device. More specific examples (a non-exhaustive list) of the computer-readable medium would include the following: an electrical connection (electronic device) having one or more wires, a portable computer diskette (magnetic device), a Random Access Memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber device, and a portable compact disc read-only memory (CDROM). Additionally, the computer-readable medium could even be paper or another suitable medium upon which the program is printed, as the program can be electronically captured, via for instance optical scanning of the paper or other medium, then compiled, interpreted or otherwise processed in a suitable manner if necessary, and then stored in a computer memory.
It should be understood that portions of the present application may be implemented in hardware, software, firmware, or a combination thereof. In the above embodiments, the various steps or methods may be implemented in software or firmware stored in memory and executed by a suitable instruction execution system. If implemented in hardware, as in another embodiment, any one or combination of the following techniques, which are known in the art, may be used: a discrete logic circuit having a logic gate circuit for implementing a logic function on a data signal, an application specific integrated circuit having an appropriate combinational logic gate circuit, a Programmable Gate Array (PGA), a Field Programmable Gate Array (FPGA), or the like.
It will be understood by those skilled in the art that all or part of the steps carried by the method for implementing the above embodiments may be implemented by hardware related to instructions of a program, which may be stored in a computer readable storage medium, and when the program is executed, the program includes one or a combination of the steps of the method embodiments.
In addition, functional units in the embodiments of the present application may be integrated into one processing module, or each unit may exist alone physically, or two or more units are integrated into one module. The integrated module can be realized in a hardware mode, and can also be realized in a software functional module mode. The integrated module, if implemented in the form of a software functional module and sold or used as a stand-alone product, may also be stored in a computer readable storage medium.
The storage medium mentioned above may be a read-only memory, a magnetic or optical disk, etc. Although embodiments of the present application have been shown and described above, it is understood that the above embodiments are exemplary and should not be construed as limiting the present application, and that variations, modifications, substitutions and alterations may be made to the above embodiments by those of ordinary skill in the art within the scope of the present application.

Claims (6)

1. A method for predicting electricity consumption of an industrial user for a long time based on similarity measurement is characterized by comprising the following steps:
preprocessing monthly power consumption time sequence data to generate a sample;
carrying out standardization processing on the sample to construct a data set, wherein the data set is divided into a training set and a testing set;
for the samples to be predicted in the test set, introducing dynamic time warping to measure the similarity between the samples to be predicted and the samples in the training set, and selecting the samples with higher similarity degree as reference samples for implementing prediction;
and taking the screened samples as a basis, constructing a time sequence prediction model to predict the power consumption of the target user, and obtaining a predicted value of the monthly power consumption of the target user.
2. The method according to claim 1, wherein the normalization of the sample is performed by:
Figure FDA0003345363490000011
wherein o ismax、omin、o、onormRespectively representing the maximum value, the minimum value, the current value and the standardized result value of the monthly electricity consumption time sequence data of the same industrial user.
3. The method of claim 1, wherein a dynamic time warping metric is introduced to measure the similarity of the samples to be predicted to the samples in the training set, wherein the dynamic time metric is expressed as:
Figure FDA0003345363490000012
wherein, Otest(t)、Otrain(t) respectively representing a sample to be predicted and data to be compared in a training set, w represents a DTW shortest path lower degree corresponding point element, and the ith element w in wl＝(Otest(i),Otrain(j))l。
4. The method as claimed in claim 1, wherein the time sequence prediction model is constructed based on the least squares support vector machine LSSVR to predict the power consumption of the target user, and the predicted value of the monthly power consumption of the target user is obtained, where the LSSVR is expressed as:
Figure FDA0003345363490000013
wherein alpha isiRepresenting a series of Lagrange multipliers, xiRepresenting an input variable, b representing a deviation, K () representing a kernel function, using RBF as the kernel function, the K () expression being
Figure FDA0003345363490000014
5. The long-term prediction device of the electricity consumption of the industrial users based on the similarity measurement is characterized by comprising a preprocessing module, a data set generating module, a screening module and a model prediction module, wherein,
the preprocessing module is used for preprocessing monthly power consumption time sequence data to generate a sample;
the data set generating module is used for carrying out standardization processing on the sample and constructing a data set, wherein the data set is divided into a training set and a testing set;
the screening module is used for introducing dynamic time warping measurement to the samples to be predicted in the test set and the similarity between the samples to be predicted and the samples in the training set, and selecting the samples with higher similarity degree as reference samples for implementing prediction;
and the model prediction module is used for constructing a time sequence prediction model to predict the power consumption of the target user by taking the screened samples as the basis so as to obtain a predicted value of the monthly power consumption of the target user.
6. A non-transitory computer-readable storage medium having stored thereon a computer program, wherein the computer program, when executed by a processor, implements the method of any one of claims 1-4.
CN202111320362.5A 2021-11-09 2021-11-09 Industrial user power consumption long-term prediction method and device based on similarity measurement Pending CN114254801A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202111320362.5A CN114254801A (en) 2021-11-09 2021-11-09 Industrial user power consumption long-term prediction method and device based on similarity measurement

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202111320362.5A CN114254801A (en) 2021-11-09 2021-11-09 Industrial user power consumption long-term prediction method and device based on similarity measurement

Publications (1)

Publication Number Publication Date
CN114254801A true CN114254801A (en) 2022-03-29

Family

ID=80792332

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202111320362.5A Pending CN114254801A (en) 2021-11-09 2021-11-09 Industrial user power consumption long-term prediction method and device based on similarity measurement

Country Status (1)

Country Link
CN (1) CN114254801A (en)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN116205554A (en) * 2023-04-26 2023-06-02 浙江天柜科技有限公司 Mobile self-service vending equipment and vending control method

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN116205554A (en) * 2023-04-26 2023-06-02 浙江天柜科技有限公司 Mobile self-service vending equipment and vending control method
CN116205554B (en) * 2023-04-26 2024-02-09 浙江天柜科技有限公司 Mobile self-service vending equipment and vending control method

Similar Documents

Publication Publication Date Title
Xiao et al. Parameter optimization for nonlinear grey Bernoulli model on biomass energy consumption prediction
Li et al. Remaining useful life prediction for lithium-ion batteries based on Gaussian processes mixture
Barbosa et al. Using performance profiles to analyze the results of the 2006 CEC constrained optimization competition
Strange et al. Prediction of future capacity and internal resistance of Li-ion cells from one cycle of input data
CN107506868B (en) Method and device for predicting short-time power load
WO2021208079A1 (en) Method and apparatus for obtaining power battery life data, computer device, and medium
JP2001502831A (en) A method for classifying the statistical dependence of measurable time series
Huang et al. A framework for scalable bilevel optimization: Identifying and utilizing the interactions between upper-level and lower-level variables
CN103729530B (en) Device and method for processing sequence
Carstens et al. Low-cost energy meter calibration method for measurement and verification
Schnorr et al. Feature tracking by two-step optimization
CN114254801A (en) Industrial user power consumption long-term prediction method and device based on similarity measurement
Anjum et al. Parallel operated hybrid Arithmetic-Salp swarm optimizer for optimal allocation of multiple distributed generation units in distribution networks
KR20190072292A (en) Apparatus and method for body growth prediction modeling
CN114330834A (en) Charging pile power consumption prediction method based on self-updating cubic exponential smoothing method
CN114881343A (en) Short-term load prediction method and device of power system based on feature selection
Franz et al. Prediction in trend-renewal processes for repairable systems
CN115017970A (en) Migration learning-based gas consumption behavior anomaly detection method and system
CN108694472B (en) Prediction error extreme value analysis method, device, computer equipment and storage medium
Jiménez et al. A closed-loop data-driven optimization framework for the unit commitment problem: A Q-learning approach under real-time operation
CN110990135B (en) Spark job time prediction method and device based on deep migration learning
Guo et al. Accelerating parameter estimation for multivariate self-exciting point processes
Long et al. Assessing lifecycle value using object-based modeling by incorporating excess and changeability
CN115577856A (en) Method and system for predicting construction cost and controlling balance of power transformation project
CN115423159A (en) Photovoltaic power generation prediction method and device and terminal equipment

Legal Events

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