CN111461377B - Prediction method/system, storage medium and apparatus for total energy demand - Google Patents
Prediction method/system, storage medium and apparatus for total energy demand Download PDFInfo
- Publication number
- CN111461377B CN111461377B CN201910046186.7A CN201910046186A CN111461377B CN 111461377 B CN111461377 B CN 111461377B CN 201910046186 A CN201910046186 A CN 201910046186A CN 111461377 B CN111461377 B CN 111461377B
- Authority
- CN
- China
- Prior art keywords
- sequence
- matrix
- user
- neural network
- correlation coefficient
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 238000000034 method Methods 0.000 title claims abstract description 77
- 238000003860 storage Methods 0.000 title claims abstract description 12
- 239000011159 matrix material Substances 0.000 claims abstract description 133
- 238000013528 artificial neural network Methods 0.000 claims abstract description 72
- 238000012549 training Methods 0.000 claims abstract description 31
- 238000005265 energy consumption Methods 0.000 claims abstract description 9
- 230000015654 memory Effects 0.000 claims description 16
- 230000000306 recurrent effect Effects 0.000 claims description 15
- 238000013527 convolutional neural network Methods 0.000 claims description 14
- 238000004590 computer program Methods 0.000 claims description 13
- 238000004422 calculation algorithm Methods 0.000 claims description 12
- 125000004122 cyclic group Chemical group 0.000 claims description 12
- 238000009826 distribution Methods 0.000 claims description 10
- 238000012216 screening Methods 0.000 claims description 10
- 238000004364 calculation method Methods 0.000 claims description 9
- 230000009466 transformation Effects 0.000 claims description 6
- 230000006870 function Effects 0.000 claims description 4
- 210000002569 neuron Anatomy 0.000 claims description 3
- 238000005457 optimization Methods 0.000 claims description 3
- 238000003384 imaging method Methods 0.000 claims 1
- 230000007787 long-term memory Effects 0.000 claims 1
- 210000005036 nerve Anatomy 0.000 claims 1
- 230000008569 process Effects 0.000 description 9
- 238000012545 processing Methods 0.000 description 6
- 238000004891 communication Methods 0.000 description 5
- 238000004519 manufacturing process Methods 0.000 description 5
- 230000001537 neural effect Effects 0.000 description 4
- 238000010586 diagram Methods 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 230000006403 short-term memory Effects 0.000 description 2
- 239000002699 waste material Substances 0.000 description 2
- 238000003491 array Methods 0.000 description 1
- 230000002860 competitive effect Effects 0.000 description 1
- 238000010276 construction Methods 0.000 description 1
- 230000002349 favourable effect Effects 0.000 description 1
- 238000012423 maintenance Methods 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 238000002360 preparation method Methods 0.000 description 1
- 238000011144 upstream manufacturing Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/045—Combinations of networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
- G06Q50/06—Energy or water supply
Landscapes
- Engineering & Computer Science (AREA)
- Business, Economics & Management (AREA)
- Economics (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Health & Medical Sciences (AREA)
- Strategic Management (AREA)
- Human Resources & Organizations (AREA)
- General Physics & Mathematics (AREA)
- General Business, Economics & Management (AREA)
- Tourism & Hospitality (AREA)
- Marketing (AREA)
- General Health & Medical Sciences (AREA)
- Software Systems (AREA)
- Entrepreneurship & Innovation (AREA)
- Mathematical Physics (AREA)
- Computing Systems (AREA)
- Molecular Biology (AREA)
- Evolutionary Computation (AREA)
- Development Economics (AREA)
- Data Mining & Analysis (AREA)
- Game Theory and Decision Science (AREA)
- Computational Linguistics (AREA)
- Biophysics (AREA)
- General Engineering & Computer Science (AREA)
- Biomedical Technology (AREA)
- Operations Research (AREA)
- Quality & Reliability (AREA)
- Artificial Intelligence (AREA)
- Life Sciences & Earth Sciences (AREA)
- Public Health (AREA)
- Water Supply & Treatment (AREA)
- Primary Health Care (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention discloses a prediction method/system, a storage medium and equipment for total energy demand, which are characterized in that user energy consumption data are arranged according to association degrees by using a maximization matrix with maximized correlation coefficients, and the characteristics of the inside, the association among and the like of energy consumption sequences of all users are extracted by using a maximization matrix training neural network, so that the constructed prediction model can more accurately predict the total energy demand of all users.
Description
Technical Field
The present invention relates to the field of data processing, and in particular, to a method/computing system, storage medium and apparatus for predicting total energy demand.
Background
The energy producer generally calculates the energy demand of each energy using user to obtain the total energy demand, in the statistical process, each energy using user is in need of preparation, the declared energy demand is often larger than the actual demand, when the energy producer supplies energy for more energy using users, the declared total energy of each user can far exceed the final actual total energy consumption, so that great resource waste is caused, if the energy producer does not have the basis to reduce the production quantity, the energy producing user needs to bear the liability risk of energy outage.
Disclosure of Invention
In view of the above-mentioned drawbacks of the prior art, a primary object of the present invention is to provide a prediction method/system, a storage medium and a device for total energy demand, so as to improve the accuracy of the prediction result.
To achieve the above object and other related objects, the present invention provides the following technical solutions:
a method for predicting total energy demand, comprising:
acquiring existing data and outputting a maximized matrix;
inputting the maximization matrix into a pre-constructed neural network, and outputting the energy total demand prediction result of each user;
wherein the existing data includes historical usage of each user, and the method for outputting the maximizing matrix includes: the historical usage amount of each user is arranged and configured into a historical usage amount sequence according to a time sequence, the historical usage amount sequence of each user is arranged and configured into a plurality of sequence matrixes according to a random sequence in a direction different from that of a single historical usage amount sequence, and the maximization matrix with the largest correlation coefficient in the sequence matrixes is calculated and screened.
Optionally, the method for calculating and screening the maximization matrix with the largest correlation coefficient in the sequence matrix includes: and calculating the correlation coefficient between the historical use sequences of any two users, and then calculating the correlation coefficient of each sequence matrix according to the correlation coefficient between the historical use sequences of any two users.
Optionally, a calculation formula of a correlation coefficient between the historical usage sequences of any two users is as follows:
wherein i represents the current day, n represents the total day, x represents any one of the users,represents the average usage of user x over 1 to n days, y represents any user other than x, < ->Represents the average usage amount of the user y in 1 to n days, r x,y Representing the correlation coefficients of user x and user y. If i is any one of 1 to n, the historical usage sequence of user x during 1 to n days is (x) 1 ,x 2 ,····,x n ) User(s) y The historical usage sequence over a period of 1 to n days is (y 1 ,y 2 ,····,y n )。
Optionally, the method for calculating the correlation coefficient of the sequence matrix includes: ignoring correlation coefficients between said historically used sequences in said sequence matrix at distances greater than 3.
Optionally, the correlation coefficient calculation formula of the sequence matrix is:
where N denotes the order of arrangement of individual users in the sequence matrix.
Optionally, if the correlation coefficients between the historical use sequences with the distance greater than 3 in the sequence matrix are ignored, and then the correlation coefficients between two or more sequence matrices are equal, the correlation coefficients of each sequence matrix are recalculated, and the correlation coefficients between the historical use sequences with the distance greater than 3 in the sequence matrix are sequentially increased during recalculation.
Optionally, the method for computing and screening the maximizing matrix includes a violent exhaustion algorithm.
Optionally, the method for computationally screening the maximization matrix includes a greedy algorithm.
Optionally, before configuring the sequence matrix, verifying whether the historical usage sequence of each user meets the normal distribution, and if not, performing data adjustment to enable the historical usage sequence to meet the normal distribution.
Optionally, the method of making data adjustments such that the historical usage sequence satisfies a normal distribution includes a logarithmic transformation and/or a square root transformation.
Optionally, the sequence matrix comprises a two-dimensional matrix or a multi-dimensional matrix.
Optionally, the sequence matrix is a multidimensional matrix, the existing data further includes one or more of historical energy reservation amount data of each user, overhaul day data of each user and holiday data of each user, the method for configuring and forming the sequence matrix further includes adding one or more of a reservation amount sequence of each user, an overhaul day sequence and a holiday sequence, one reservation amount sequence is formed by arranging historical energy reservation amounts of individual users in time sequence, one overhaul day sequence is formed by arranging overhaul schedules of individual users in time sequence, and further includes adding a holiday sequence of each user, and one holiday sequence is formed by arranging holiday schedules of individual users in time sequence.
Optionally, the data in the historical usage sequence is greater than 50.
Optionally, the method for predicting total energy demand further includes training the neural network, and the training method includes:
the parameters of the neural network are set,
in the maximization matrix, randomly intercepting the historical energy use amount of each user in a continuous time period as a training sample;
and sequentially inputting the training samples into the neural network, and adjusting parameters according to the output result of the neural network based on the training samples until the error degree of the output result is lower than an expected threshold value.
Optionally, the neural network includes a convolutional neural network and a cyclic neural network, and the method for inputting the training samples into the neural network in sequence for training includes:
inputting the training sample into the convolutional neural network to extract local features;
and the local features are input into the cyclic neural network to extract time sequence features, and the result is output.
Optionally, the parameters of the neural network include:
days of the continuous period;
one or more of the layer number, convolution size and network architecture of the convolution neural network; and
One or more of the number of hidden layers, the number of neurons and the network structure in the recurrent neural network.
Optionally, the cyclic neural network includes one of a long-short-term memory network, a gated cyclic neural network, and a neural turing machine.
Optionally, the method for inputting the maximization matrix into the pre-constructed neural network for training further comprises: the neural network is optimized using a loss function and/or a gradient optimization algorithm.
The present invention also provides a prediction system for total energy demand, comprising:
a maximizing matrix calculating unit for acquiring existing data and outputting the maximizing matrix,
the neural network is used for acquiring the maximization matrix and outputting the energy total demand prediction result of each user;
wherein the existing data includes a historical usage amount of each user, and the operation of outputting the maximizing matrix by the maximizing matrix calculating unit includes: the historical usage amount of each user is arranged and configured into a historical usage amount sequence according to a time sequence, the historical usage amount sequence of each user is arranged and configured into a plurality of sequence matrixes according to a random sequence in a direction different from that of a single historical usage amount sequence, and the maximization matrix with the largest correlation coefficient in the sequence matrixes is calculated and screened.
Optionally, the neural network includes:
an input matrix generation unit for intercepting historical energy consumption data of each user in a continuous time period in the maximization matrix to form an input matrix;
a convolutional neural network for extracting local features of the input matrix; and
And the cyclic neural network is used for extracting time sequence characteristics from the output result of the convolutional neural network and outputting a prediction result.
A storage medium having stored thereon a computer program which, when executed by a processor, implements the prediction method for total energy demand of any one of claims 3 to 20.
An apparatus comprising a processor and a memory, the memory for storing a computer program, the processor for executing the computer program stored by the memory to cause the apparatus to perform the prediction method for total energy demand of any one of claims 3 to 20.
According to the prediction method/system, the storage medium and the device for the total energy demand, disclosed by the invention, the user energy consumption data are arranged according to the association degree by utilizing the maximization matrix with the maximized correlation coefficient to identify the association of the energy consumption among all users, and the characteristics such as the association between the inside and the outside of the energy consumption sequences of all users are extracted by utilizing the maximization matrix to train the neural network, so that the constructed prediction model can more accurately predict the total energy demand of all users.
Drawings
FIG. 1 is a block diagram showing the construction of an energy total demand prediction system according to the present invention;
FIG. 2 is a diagram showing a network configuration of a neural network according to the present invention;
FIG. 3 shows a flow chart for energy total demand prediction according to the present invention;
fig. 4 is a flow chart of neural network training in the present invention.
Detailed Description
In order to make the objects, technical solutions and advantages of the present invention more apparent, the following detailed description of specific embodiments of the present invention is given with reference to the accompanying drawings. It is to be understood that the specific embodiments described herein are merely illustrative of the invention and are not limiting thereof.
It should be further noted that, for convenience of description, only some, but not all of the matters related to the present invention are shown in the accompanying drawings. Before discussing exemplary embodiments in more detail, it should be mentioned that some exemplary embodiments are described as processes or methods depicted as flowcharts. Although a flowchart depicts operations (or steps) as a sequential process, many of the operations can be performed in parallel, concurrently, or at the same time. Furthermore, the order of the operations may be rearranged. The process may be terminated when its operations are completed, but may have additional steps not included in the figures. The processes may correspond to methods, functions, procedures, subroutines, and the like.
For a clear and obvious description of the embodiments of the present invention, the inventive concept of the present invention will be described with emphasis on the following:
the energy production users generally obtain the total energy demand by adding the energy demand reported by each energy use user, and take the total energy demand as a production plan, but most of the energy use users have a phenomenon of multiple reporting in the reporting process, so that the total reported demand far exceeds the final actual use amount, and energy waste is caused. If the traditional algorithm is used for predicting the demand, a part of algorithms are too complex, and the prediction result of the part of algorithms is not accurate enough, the method mainly uses the historical use data of each user to obtain the maximized matrix with the maximum correlation coefficient, and takes the maximized matrix as the input item of the total energy demand, so that the prediction result can consider the relevance of the energy consumption among the users, and the prediction result is more accurate. The correlation between the users may be the correlation of the energy usage amounts between the users of the upstream and downstream enterprises, or the correlation of the energy usage amounts between the users having a competitive relationship, and the correlation is not limited to the two kinds of correlations, but may be other kinds of correlations.
Referring to fig. 3, the present invention provides a prediction method for total energy demand, the method comprising:
210. acquiring, by one or more computing devices, existing data and outputting a maximization matrix;
220. inputting, by one or more computing devices, the maximization matrix into a pre-built neural network and outputting a prediction of total energy demand for each user (e.g., business or home);
wherein the existing data includes a historical usage amount of each user, and in step 210, the method for outputting the maximizing matrix includes: the historical usage amount of each user is arranged and configured into a historical usage amount sequence according to a time sequence, the historical usage amount sequence of each user is arranged and configured into a plurality of sequence matrixes according to a random sequence in a direction different from that of a single historical usage amount sequence, and the maximization matrix with the largest correlation coefficient in the sequence matrixes is calculated and screened.
The present embodiment also provides a storage medium having stored thereon the computer program corresponding to the above-described method for predicting the total demand of energy, which when executed by a processor, implements any one of the above-described and below-described methods for predicting the total demand of energy.
The storage medium in this embodiment, as will be understood by those of ordinary skill in the art: all or part of the steps for carrying out the method embodiments of the present description may be accomplished by computer program related hardware. The aforementioned computer program may be stored in a computer readable storage medium. The program, when executed, performs steps including the method embodiments in the present specification; and the aforementioned storage medium includes: various media that can store program code, such as ROM, RAM, magnetic or optical disks.
The historical usage amount sequence may be a row vector or a column vector, and if the historical usage amount sequence is a row vector, the historical usage amounts of the users may be ranked up and down when the sequence matrix is created, and if the historical usage amount sequence is a column vector, the historical usage amounts of the users may be ranked left and right when the sequence matrix is created.
For example, the energy use users include user a, user b, and user c … …User p, n refers to the total number of days, then the historical usage sequence for user a may be (a 1 ,a 2 ····a i ····a n ) The created sequence matrix includes matrix one:
(a 1 ,a 2 ····a i ····a n )
(b 1 ,b 2 ····b i ····b n )
(c 1 ,c 2 ····c i ····c n )
....
(p 1 ,p 2 ····p i ····p n )
the created sequence matrix also comprises other sequences after the historical usage sequence of each user is randomly ordered.
In this embodiment, before the sequence matrix is configured, it is required to verify whether the historical usage sequence of each user satisfies the normal distribution, and if not, data adjustment is required to make the historical usage sequence satisfy the normal distribution. In particular, the method of making data adjustments such that the historical usage sequence satisfies a normal distribution may include logarithmic transformation and/or square root transformation.
In some embodiments, the method for calculating and screening the maximization matrix with the largest correlation coefficient in the sequence matrix comprises the following steps: and calculating the correlation coefficient between the historical use sequences of any two users, and then calculating the correlation coefficient of each sequence matrix according to the correlation coefficient between the historical use sequences of any two users.
In some embodiments, the correlation coefficient between the historical usage sequences of any two users may be calculated by a pearson correlation coefficient calculation formula:
wherein i represents the current day, n represents the total day, x represents any one of the users,represents the average usage of user x over 1 to n days, y represents any user other than x, < ->Represents the average usage amount of the user y in 1 to n days, r x,y Representing the correlation coefficients of user x and user y. If i is any one of 1 to n, the historical usage sequence of user x during 1 to n days is (x) 1 ,x 2 ,····,x n ) The historical usage sequence of user y over a period of 1-n days is (y 1 ,y 2 ,····,y n )。
In some embodiments, the method for calculating the correlation coefficient of the sequence matrix includes: and neglecting correlation coefficients between historical use sequences with distances greater than 3 in the sequence matrix. The data processing amount can be greatly reduced, and the accuracy of the final prediction result is little affected.
Specifically, after ignoring the correlation coefficient between the history use sequences with the distance greater than 3 in the sequence matrix, the calculation formula of the correlation coefficient of the sequence matrix is as follows:
where N refers to the order of arrangement of individual users in the sequence matrix.
In some embodiments, if the correlation coefficients between the historical usage sequences with a distance greater than 3 in the sequence matrix are ignored, and then the correlation coefficients between two or more sequence matrices are equal, the correlation coefficients of each sequence matrix are recalculated, and the correlation coefficients between the historical usage sequences with a distance greater than 3 in the sequence matrix are sequentially increased during the recalculation. If the distance between the two history use sequences is defined as d, the calculation formula of the correlation coefficient between each history use sequence with the distance d is as follows:
for example, when the correlation coefficients between the history-use sequences having a distance greater than 3 in the sequence matrix are ignored, and the correlation coefficients of the two sequence matrices are equal, the correlation coefficient between the history-use sequences having a distance of 4 can be increasedThe correlation coefficients of the matrices are recalculated, and the obtained correlation coefficient calculating company is as follows:
if there is still a sequence matrix with equal correlation coefficient after the correlation coefficient between the history used sequences with a distance of 4 is increased, the correlation coefficient of each matrix is calculated again from the new, and the correlation coefficient between the history used sequences with a distance of 5 is increased.
In some embodiments, the method for computing and screening the maximizing matrix includes a violent exhaustion algorithm, namely computing correlation matrices of all sequence matrices, and selecting the sequence matrix with the largest correlation coefficient as the maximizing matrix.
In some embodiments, the method for computing and screening the maximized matrix includes a greedy algorithm, namely, calculating a correlation coefficient of a current sequence matrix, randomly exchanging sequences of any two users, calculating the correlation coefficient of the exchanged sequence matrix, if the correlation coefficient of the exchanged sequence matrix is larger than the correlation coefficient of the sequence play matrix before exchanging, reserving the sequences after exchanging, otherwise, reserving the sequences before exchanging, and then iterating based on the principle until the value of the correlation coefficient after exchanging the sequences is not increased any more and the number of exchanges reaches a preset number threshold value, and stopping iterating.
In some embodiments, the sequence matrix may be a two-dimensional matrix, i.e., containing only each user and a chronological historical sequence of usage for each user. In other embodiments, the sequence matrix may be a multidimensional matrix, the existing data further including one or more of historical energy reserve data for each user, date of service data for each user, and holiday data for each user, the method of configuring the sequence matrix further including adding one or more of a reserve sequence for each user, date of service sequence, and holiday sequence, one of the reserve sequences being formed by chronologically ordering historical energy reserve for a single user, one of the date of service sequences being formed by chronologically ordering a service schedule for a single user, and further including adding a holiday sequence for each user, one of the holiday sequences being formed by chronologically ordering a holiday schedule for a single user. . In the prediction process, factors which possibly influence the use amount, such as the reservation amount, the maintenance day, the holiday and the like, are increased, and the prediction result is more accurate.
In some embodiments, the number of data in the historical usage sequence is greater than 50, that is, n is greater than 50, in the actual implementation process, the number of data in one historical usage sequence may be less than or equal to 50, or may be any number greater than 50, such as 60, 70, 80, 82, etc., when the number of data in a single historical usage sequence is greater than 50, it can be ensured that the data for constructing the prediction model is sufficient, so that the historical usage sequence can meet the normal distribution requirement after the data processing is performed.
Referring to fig. 4, in some embodiments, the prediction method for total energy demand of the present invention further includes training the neural network, and the training method includes:
221. the parameters of the neural network are set,
222. in the maximization matrix, randomly intercepting the historical energy use amount of each user in a continuous time period as a training sample;
223. sequentially inputting the training samples into the neural network, adjusting parameters according to the output result of the neural network based on the training samples,
224. steps 221-223 are performed in a loop until the error degree of the output result is lower than the expected threshold.
Specifically, when the historical usage sequence is a row vector, a column vector with a length of c is randomly intercepted to be used as a training sample, and if the output result is defined as the output result of the t day, the training sample is the historical usage sequence of each user on the t-c day and the t-1 day.
In some embodiments, referring to fig. 4, the neural network includes a convolutional neural network and a cyclic neural network, and the method for sequentially inputting the training samples into the neural network for training includes:
223a, inputting the maximization matrix into the convolutional neural network to extract local features;
223b, inputting the local characteristics into the cyclic neural network to extract time sequence characteristics, and outputting a result.
The convolutional neural network is favorable for preserving the correlation among all data in the maximized matrix when extracting features, the cyclic neural network can extract time sequence features, and a more excellent prediction model, namely a more accurate prediction result, can be obtained compared with other types of neural networks after the neural network consisting of the convolutional neural network and the cyclic neural network is subjected to combined training.
In some embodiments, the parameters of the neural network include:
days of the continuous period;
one or more of the layer number, convolution size and network architecture of the convolution neural network; and
One or more of the number of hidden layers, the number of neurons and the network structure in the recurrent neural network. In some embodiments, the recurrent neural network may include any one of Long Short-Term Memory (LSTM), gated recurrent neural network (Gated Recurrent Unit, GRU), neuro-graphic machine (Neural Turing Machines, NTM). In this embodiment, the recurrent neural network is a gated recurrent neural network.
In some embodiments, the method of inputting the maximization matrix into the pre-constructed neural network for training further comprises: the neural network is optimized using a loss function and/or a gradient optimization algorithm. The prediction accuracy of the prediction model can be improved.
Referring to fig. 1, there is shown a predictive system for total energy demand, comprising:
a maximizing matrix calculating unit for outputting the maximizing matrix according to the existing data;
the neural network is used for acquiring the maximization matrix and outputting the energy total demand prediction result of each user;
wherein the existing data includes a historical usage amount of each user, and the operation of outputting the maximizing matrix by the maximizing matrix calculating unit includes: the historical usage amount of each user is arranged and configured into a historical usage amount sequence according to a time sequence, the historical usage amount sequence of each user is arranged and configured into a plurality of sequence matrixes according to a random sequence in a direction different from that of a single historical usage amount sequence, and the maximization matrix with the largest correlation coefficient in the sequence matrixes is calculated and screened.
In some embodiments, referring to fig. 2, the neural network comprises:
an input matrix generation unit 21 for intercepting historical energy usage data of each user in a continuous period of time in the maximization matrix to form an input matrix;
a convolutional neural network 22 for extracting local features of the input matrix; and
And a recurrent neural network 23 for extracting a time sequence feature from the output result of the convolutional neural network and outputting a prediction result.
In practical implementation, the composition of the neural network is not limited to the above-described configuration, and a type of configuration may be adopted as long as the total energy demand can be predicted by using the maximization matrix.
In this embodiment, the recurrent neural network is a gated recurrent neural network (Gated Recurrent Unit, GRU), and in the practical implementation process, the recurrent neural network may be any one of Long Short-Term Memory (LSTM), gated recurrent neural network (Gated Recurrent Unit, GRU), and neural turing machine (Neural Turing Machines, NTM).
The invention also provides a device comprising a processor and a memory, the memory being for storing a computer program, the processor being for executing the computer program stored by the memory to cause the device to perform any one of the above described prediction methods for total energy demand.
The device provided in this embodiment includes a processor, a memory, a transceiver, and a communication interface, where the memory and the communication interface are connected to the processor and the transceiver and perform communication therebetween, the memory is used to store a computer program, and the communication interface is used to perform communication, and the processor and the transceiver are used to run the computer program, so that the device performs each step of the above method for predicting total energy demand.
In this embodiment, the memory may include a random access memory (Random Access Memory, abbreviated as RAM), and may further include a non-volatile memory (non-volatile memory), such as at least one magnetic disk memory.
The processor may be a general-purpose processor, including a central processing unit (Central Processing Unit, CPU for short), a network processor (Network Processor, NP for short), etc.; but also digital signal processors (Digital Signal Processing, DSP for short), application specific integrated circuits (Application Specific Integrated Circuit, ASIC for short), field-programmable gate arrays (Field-Programmable Gate Array, FPGA for short) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components.
The above embodiments are merely illustrative of the principles of the present invention and its effectiveness, and are not intended to limit the invention. Modifications and variations may be made to the above-described embodiments by those skilled in the art without departing from the spirit and scope of the invention. Accordingly, it is intended that all equivalent modifications and variations of the invention be covered by the claims, which are within the ordinary skill of the art, be within the spirit and scope of the present disclosure.
Claims (19)
1. A method for predicting total energy demand, comprising
Acquiring existing data and outputting a maximized matrix;
inputting the maximization matrix into a pre-constructed neural network, and outputting the energy total demand prediction result of each user;
wherein the existing data includes historical usage of each user, and the method for outputting the maximizing matrix includes: the historical usage amount of each user is arranged and configured into a historical usage amount sequence according to a time sequence, the historical usage amount sequence of each user is arranged and configured into a plurality of sequence matrixes according to a random sequence in a direction different from that of a single historical usage amount sequence, and the maximization matrix with the largest correlation coefficient in the sequence matrixes is calculated and screened; the method for calculating and screening the maximization matrix with the largest correlation coefficient in the sequence matrix comprises the following steps: calculating the correlation coefficient between the history use sequences of any two users, and calculating the correlation coefficient of each sequence matrix according to the correlation coefficient between the history use sequences of any two users;
the correlation coefficient calculation formula between the history use sequences of any two users is as follows:
wherein i represents the current day, n represents the total day, x represents any one of the users,represents the average usage of user x over 1 to n days, y represents any user other than x, < ->Represents the average usage amount of the user y in 1 to n days, r x,y Representing the correlation coefficients of user x and user y; if i is any one of 1 to n, user x is 1 to nThe historical usage sequence over the n days is (x) 1 ,x 2 ,····,x n ) The historical usage sequence of user y over a period of 1-n days is (y 1 ,y 2 ,····,y n )；
The calculation formula of the correlation coefficient of the sequence matrix is as follows:
where N denotes the order of arrangement of individual users in the sequence matrix.
2. The method for predicting total energy demand of claim 1, wherein the method for calculating the correlation coefficient of the sequence matrix comprises: ignoring correlation coefficients between said historically used sequences in said sequence matrix at distances greater than 3.
3. The prediction method for total energy demand according to claim 2, characterized in that: if the correlation coefficient between the historical use sequences with the distance larger than 3 in the sequence matrix is ignored, and then the correlation coefficient between two or more than two sequence matrices is equal, the correlation coefficient of each sequence matrix is recalculated, and the correlation coefficient between the historical use sequences with the distance larger than 3 in the sequence matrix is sequentially increased during recalculation.
4. The prediction method for total energy demand according to claim 1, characterized in that: the method for computing and screening the maximization matrix comprises a violence exhaustion algorithm.
5. The prediction method for total energy demand according to claim 1, characterized in that: the method for computing and screening the maximization matrix comprises a greedy algorithm.
6. The prediction method for total energy demand according to claim 1, characterized in that: before configuring the sequence matrix, verifying whether the historical usage sequence of each user meets normal distribution, and if not, carrying out data adjustment to enable the historical usage sequence to meet the normal distribution.
7. The prediction method for total energy demand according to claim 6, characterized in that: methods of making data adjustments such that the historical usage sequence satisfies a normal distribution include logarithmic transformation and/or square root transformation.
8. The prediction method for total energy demand according to claim 1, characterized in that: the sequence matrix comprises a two-dimensional matrix or a multi-dimensional matrix.
9. The prediction method for total energy demand according to claim 1, characterized in that: the sequence matrix is a multidimensional matrix, the existing data further comprises one or more of historical energy reservation amount data of each user, overhaul day data of each user and holiday data of each user, the method for configuring and forming the sequence matrix further comprises one or more of reservation amount sequences of each user, overhaul day sequences and holiday sequences, one reservation amount sequence is formed by arranging historical energy reservation amounts of single users in time sequence, one overhaul day sequence is formed by arranging overhaul schedules of single users in time sequence, and further comprises the addition of holiday sequences of each user, and one holiday sequence is formed by arranging holidays of single users in time sequence.
10. The prediction method for total energy demand according to claim 1, characterized in that: the data in the historical usage sequence is greater than 50.
11. The method for predicting total energy demand of claim 1, further comprising training the neural network, the training method comprising:
the parameters of the neural network are set,
in the maximization matrix, randomly intercepting the historical energy use amount of each user in a continuous time period as a training sample;
and sequentially inputting the training samples into the neural network, and adjusting parameters according to the output result of the neural network based on the training samples until the error degree of the output result is lower than an expected threshold value.
12. The method for predicting total energy demand of claim 11, wherein the neural network comprises a convolutional neural network and a cyclic neural network, and wherein the method for sequentially inputting the training samples into the neural network for training comprises: inputting the training sample into the convolutional neural network to extract local features;
and the local features are input into the cyclic neural network to extract time sequence features, and the result is output.
13. The method for predicting total energy demand of claim 12, wherein the parameters of the neural network comprise:
days of the continuous period;
one or more of the layer number, convolution size and network architecture of the convolution neural network; and
One or more of the number of hidden layers, the number of neurons and the network structure in the recurrent neural network.
14. The prediction method for total energy demand according to claim 13, characterized in that: the circulating neural network comprises any one of a long-term memory network, a gate-controlled circulating neural network and a nerve imaging machine.
15. The method for predicting total energy demand of claim 1, wherein inputting the maximizing matrix into the neural network for training comprises: the neural network is optimized using a loss function and/or a gradient optimization algorithm.
16. A predictive system for total energy demand, comprising:
a maximizing matrix calculating unit for acquiring existing data and outputting the maximizing matrix,
the neural network is used for acquiring the maximization matrix and outputting the energy total demand prediction result of each user;
wherein the existing data includes a historical usage amount of each user, and the operation of outputting the maximizing matrix by the maximizing matrix calculating unit includes: the historical usage amount of each user is arranged and configured into a historical usage amount sequence according to a time sequence, the historical usage amount sequence of each user is arranged and configured into a plurality of sequence matrixes according to a random sequence in a direction different from that of a single historical usage amount sequence, and the maximization matrix with the largest correlation coefficient in the sequence matrixes is calculated and screened; the method for calculating and screening the maximization matrix with the largest correlation coefficient in the sequence matrix comprises the following steps: calculating the correlation coefficient between the history use sequences of any two users, and calculating the correlation coefficient of each sequence matrix according to the correlation coefficient between the history use sequences of any two users;
the correlation coefficient calculation formula between the history use sequences of any two users is as follows:
wherein i represents the current day, n represents the total day, x represents any one of the users,represents the average usage of user x over 1 to n days, y represents any user other than x, < ->Represents the average usage amount of the user y in 1 to n days, r x,y Representing the correlation coefficients of user x and user y; if i is any one of 1 to n, the historical usage sequence of user x during 1 to n days is (x) 1 ,x 2 ,····,x n ) The historical usage sequence of user y over a period of 1-n days is (y 1 ,y 2 ,····,y n )；
The calculation formula of the correlation coefficient of the sequence matrix is as follows:
where N denotes the order of arrangement of individual users in the sequence matrix.
17. The prediction system for the total demand of energy of claim 16, wherein the neural network comprises: an input matrix generation unit for intercepting historical energy consumption data of each user in a continuous time period in the maximization matrix to form an input matrix;
a convolutional neural network for extracting local features of the input matrix; and
And the cyclic neural network is used for extracting time sequence characteristics from the output result of the convolutional neural network and outputting a prediction result.
18. A storage medium having a computer program stored thereon, characterized by: the program, when executed by a processor, implements the prediction method for total energy demand according to any one of claims 1 to 15.
19. An apparatus, characterized in that: comprising a processor and a memory for storing a computer program, the processor being adapted to execute the computer program stored by the memory to cause the apparatus to perform the prediction method for total energy demand according to any one of claims 1 to 15.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910046186.7A CN111461377B (en) | 2019-01-18 | 2019-01-18 | Prediction method/system, storage medium and apparatus for total energy demand |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910046186.7A CN111461377B (en) | 2019-01-18 | 2019-01-18 | Prediction method/system, storage medium and apparatus for total energy demand |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN111461377A CN111461377A (en) | 2020-07-28 |
| CN111461377B true CN111461377B (en) | 2024-02-02 |
Family
ID=71682115
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910046186.7A Active CN111461377B (en) | 2019-01-18 | 2019-01-18 | Prediction method/system, storage medium and apparatus for total energy demand |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111461377B (en) |
Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2007199862A (en) * | 2006-01-24 | 2007-08-09 | Nippon Telegr & Teleph Corp <Ntt> | Energy demand predicting method, predicting device, program and recording medium |
| JP2012247998A (en) * | 2011-05-27 | 2012-12-13 | Toshiba Corp | Time series data prediction apparatus, prediction method, prediction program and memory medium |
| CN102938742A (en) * | 2012-11-08 | 2013-02-20 | 广东电网公司电力调度控制中心 | Communication bandwidth forecasting method and device based on power business requirements |
| CN104134096A (en) * | 2014-06-12 | 2014-11-05 | 国网安徽省电力公司 | Intelligent load prediction method based on province and prefecture integration |
| CN105069519A (en) * | 2015-07-16 | 2015-11-18 | 国网天津市电力公司 | Intelligent power grid park terminal user energy demand condition dynamic prediction system and method |
| CN106779129A (en) * | 2015-11-19 | 2017-05-31 | 华北电力大学（保定） | A kind of Short-Term Load Forecasting Method for considering meteorologic factor |
| CN107730054A (en) * | 2017-11-15 | 2018-02-23 | 西南石油大学 | A kind of Gas Load combination forecasting method based on support vector regression |
| CN108573323A (en) * | 2018-03-19 | 2018-09-25 | 国网浙江省电力有限公司嘉兴供电公司 | A kind of energy Internet user electricity demand forecasting method and system |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2012034444A (en) * | 2010-07-28 | 2012-02-16 | Toshiba Corp | Power supply-demand planning device and method thereof |
-
2019
- 2019-01-18 CN CN201910046186.7A patent/CN111461377B/en active Active
Patent Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2007199862A (en) * | 2006-01-24 | 2007-08-09 | Nippon Telegr & Teleph Corp <Ntt> | Energy demand predicting method, predicting device, program and recording medium |
| JP2012247998A (en) * | 2011-05-27 | 2012-12-13 | Toshiba Corp | Time series data prediction apparatus, prediction method, prediction program and memory medium |
| CN102938742A (en) * | 2012-11-08 | 2013-02-20 | 广东电网公司电力调度控制中心 | Communication bandwidth forecasting method and device based on power business requirements |
| CN104134096A (en) * | 2014-06-12 | 2014-11-05 | 国网安徽省电力公司 | Intelligent load prediction method based on province and prefecture integration |
| CN105069519A (en) * | 2015-07-16 | 2015-11-18 | 国网天津市电力公司 | Intelligent power grid park terminal user energy demand condition dynamic prediction system and method |
| CN106779129A (en) * | 2015-11-19 | 2017-05-31 | 华北电力大学（保定） | A kind of Short-Term Load Forecasting Method for considering meteorologic factor |
| CN107730054A (en) * | 2017-11-15 | 2018-02-23 | 西南石油大学 | A kind of Gas Load combination forecasting method based on support vector regression |
| CN108573323A (en) * | 2018-03-19 | 2018-09-25 | 国网浙江省电力有限公司嘉兴供电公司 | A kind of energy Internet user electricity demand forecasting method and system |
Non-Patent Citations (2)
| Title |
|---|
| A Data Oriented Analysis and Design Method for Smart Complex Software Systems of IoT;D.J. Ning;《2018 International Symposium in Sensing and Instrumentation in IoT Era (ISSI)》;全文 * |
| 基于指标数据的供电可靠性预测方法;杨江;刘鹏飞;;云南电力技术(04);全文 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN111461377A (en) | 2020-07-28 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Pourahmadi et al. | Distributionally robust chance-constrained generation expansion planning | |
| Kaytez et al. | Forecasting electricity consumption: A comparison of regression analysis, neural networks and least squares support vector machines | |
| Aytac et al. | Characterization of demand for short life-cycle technology products | |
| CN107544251B (en) | Single machine scheduling method for minimizing total stall period based on distributed robust model | |
| CN104468413B (en) | A kind of network service method and system | |
| CN110910004A (en) | Reservoir dispatching rule extraction method and system with multiple uncertainties | |
| CN112001556B (en) | Reservoir downstream water level prediction method based on deep learning model | |
| CN115600774B (en) | Multi-target production scheduling optimization method for assembly type building component production line | |
| CN111045907B (en) | System capacity prediction method based on traffic | |
| CN110197307B (en) | Regional sea surface temperature prediction method combined with attention mechanism | |
| CN112949913A (en) | Photovoltaic output prediction method and terminal equipment | |
| US7672815B2 (en) | Method and apparatus for evaluating a proposed solution to a constraint problem | |
| Huh et al. | A general strategic capacity planning model under demand uncertainty | |
| Nielsen et al. | Inference on the dimension of the nonstationary subspace in functional time series | |
| CN103649942A (en) | Retail forecasting using parameter estimation | |
| CN108241864A (en) | Server performance Forecasting Methodology based on multivariable grouping | |
| CN110007371A (en) | Wind speed forecasting method and device | |
| CN111461377B (en) | Prediction method/system, storage medium and apparatus for total energy demand | |
| Yang et al. | Host load prediction based on PSR and EA-GMDH for cloud computing system | |
| Wang et al. | Bayesian sequential data collection for stochastic simulation calibration | |
| CN115629576A (en) | Non-invasive flexible load aggregation characteristic identification and optimization method, device and equipment | |
| CN113837782B (en) | Periodic term parameter optimization method and device of time sequence model and computer equipment | |
| Nhu et al. | Optimizing resource scaling in network slicing | |
| CN114722490A (en) | Agent model global optimization method based on mixed increase and interval reduction | |
| CN111091242A (en) | Optimal nonparametric interval prediction method for power load |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |