CN113807615A - Electric heating energy consumption prediction method and prediction system based on BP neural network - Google Patents

Electric heating energy consumption prediction method and prediction system based on BP neural network Download PDF

Info

Publication number
CN113807615A
CN113807615A CN202111195519.6A CN202111195519A CN113807615A CN 113807615 A CN113807615 A CN 113807615A CN 202111195519 A CN202111195519 A CN 202111195519A CN 113807615 A CN113807615 A CN 113807615A
Authority
CN
China
Prior art keywords
neural network
layer
energy consumption
output
network model
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
CN202111195519.6A
Other languages
Chinese (zh)
Inventor
石松林
朱烔名
张小梅
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.)
Beijing Jiajieneng Technology Co ltd
Original Assignee
Beijing Jiajieneng Technology 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 Beijing Jiajieneng Technology Co ltd filed Critical Beijing Jiajieneng Technology Co ltd
Priority to CN202111195519.6A priority Critical patent/CN113807615A/en
Publication of CN113807615A publication Critical patent/CN113807615A/en
Pending legal-status Critical Current

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q10/00—Administration; Management
    • G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/04—Architecture, e.g. interconnection topology
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/06—Physical realisation, i.e. hardware implementation of neural networks, neurons or parts of neurons
    • G06N3/061—Physical realisation, i.e. hardware implementation of neural networks, neurons or parts of neurons using biological neurons, e.g. biological neurons connected to an integrated circuit
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/08—Learning methods
    • G06N3/084—Backpropagation, e.g. using gradient descent

Abstract

The invention discloses an electric heating energy consumption prediction method and a prediction system based on a BP neural network, wherein the prediction method comprises the following steps: establishing a BP neural network model, dividing the BP neural network model into an input layer, a hidden layer and an output layer, determining an activation function of the BP neural network model according to an input vector of the input layer, and determining an error function of the BP neural network model according to an output vector of the output layer and an expected output vector; establishing a BP neural network model training algorithm based on sample data and an error function; training a BP neural network model by taking historical energy consumption data and weather data as sample data; and inputting weather data to be predicted into the trained BP neural network model to predict the electric heating energy consumption of the public building or the residential building, so as to obtain an energy consumption prediction result. The invention improves the accuracy of the heating controller on temperature control, provides guarantee for temperature data analysis, and reduces disputes caused by substandard heating due to temperature measurement.

Description

Electric heating energy consumption prediction method and prediction system based on BP neural network
Technical Field
The application relates to the field of energy consumption prediction of heating systems, in particular to an electric heating energy consumption prediction method and system based on a BP neural network.
Background
The carbon fiber electric heating prediction method has certain limitations and errors when the ordinary linear calculation function is used for predicting the energy consumption analysis in the future period by taking historical energy consumption data and weather data as samples, and the purpose of energy consumption prediction cannot be achieved. A new electric heating energy consumption prediction method and a new electric heating energy consumption prediction system need to be designed to improve the prediction accuracy of the carbon fiber electric heating energy consumption prediction system.
Disclosure of Invention
The invention provides an electric heating energy consumption prediction method and a prediction system based on a BP neural network aiming at the existing prediction error.
In a first aspect of the present invention, an electric heating energy consumption prediction method based on a BP neural network is provided, where the method includes the following steps:
establishing a BP neural network model, dividing the BP neural network model into an input layer, a hidden layer and an output layer, calculating the number of nodes of the hidden layer according to the number of the nodes of the input layer and the number of the nodes of the output layer, determining an activation function of the BP neural network model according to an input vector of the input layer, and determining an error function of the BP neural network model according to an output vector of the output layer and an expected output vector;
establishing a BP neural network model training algorithm based on sample data and an error function;
training a BP neural network model by taking historical energy consumption data and weather data as sample data until the training times exceed a preset value or the average error of model output reaches preset precision, and finishing training;
and inputting weather data to be predicted into the trained BP neural network model to predict the electric heating energy consumption of the public building or the residential building, so as to obtain an energy consumption prediction result.
Further, the specific expression for calculating the number of the hidden layer nodes according to the number of the input layer nodes and the number of the output layer nodes is as follows:
Figure BDA0003302789940000011
wherein h represents the number of hidden layer neurons, m represents the number of input layer neurons, n represents the number of output layer neurons, and λ ∈ [1,10 ]]Taking an integer.
Further, the activation function of the BP neural network model is:
Figure BDA0003302789940000012
where x is the input layer input vector.
Further, the error function is:
Figure BDA0003302789940000021
wherein n represents the number of neurons in the output layer, and the output vector yo of the output layer is { yo ═ yo1,yo2,yo3,...yonV ═ v, desired output vector1,v2,v3,...vn}。
Further, the establishing of the BP neural network model training algorithm based on the sample data and the error function includes optimizing a weight β from the hidden layer to the output layer and a weight γ from the input layer to the hidden layer, specifically:
randomly selecting k input samples and expected output from the sample data, wherein the input sample vector is as follows:
x(k)＝(x1(k),x2(k),...,xn(k))；
the desired output vector is:
v(k)＝(v1(k),v2(k),...,vn(k) where n represents the number of output layer neurons,
the inputs to the hidden layer h neurons are:
Figure BDA0003302789940000022
d 1,2, h, y denotes the weights of the input layer to the hidden layer, bdA threshold representing a neuron in the hidden layer;
the output of the hidden layer h neurons is: ho (electronic clock)d(k)＝f(hid(k))；
The inputs to the output layer are:
Figure BDA0003302789940000023
where t 1,2, n, β denote hidden-layer-to-output-layer weights, btA threshold value representing a neuron in the output layer;
the output of the output layer is: yot(k)＝f(yit(k))；
Partial derivative δ (k) of error function e to weight β from hidden layer to output layer:
Figure BDA0003302789940000024
using weight beta from hidden layer to output layer, partial derivative delta (k) of output layer and output ho of hidden layerd(k) Calculating partial derivatives for hidden layer neurons using an error function e, wherein the error function e calculates partial derivatives epsilon for hidden layer input neurons1(k) Comprises the following steps:
ε1(k)＝-δ(k)hod(k)
partial derivative epsilon of error function e to weight gamma of input layer to hidden layer2(k) Comprises the following steps:
Figure BDA0003302789940000025
optimizing weight beta from hidden layer to output layer, using epsilon1(k) And (3) calculating:
Figure BDA0003302789940000031
βN+1＝βN+με1(k)
wherein mu represents the learning rate of the BP neural network model, mu belongs to (0, 1), the weight gamma from the input layer to the hidden layer is optimized, and the partial derivative epsilon is calculated on the input neuron of the hidden layer by using an error function e2(k) And (3) calculating:
Figure BDA0003302789940000032
γN+1＝γN-με2(k)
further, the establishing of the BP neural network model training algorithm based on the sample data and the error function further includes calculating an output global error by using actual output and expected output, and a specific expression is as follows:
Figure BDA0003302789940000033
where m represents the number of input layer neurons, n represents the number of output layer neurons, v (k) represents the expected output value, yo (k) represents the actual output value.
In a second aspect of the present invention, there is provided an electric heating energy consumption prediction system based on a BP neural network, the system including:
the model building module is used for building a BP neural network model, dividing the BP neural network model into an input layer, a hidden layer and an output layer, calculating the number of nodes of the hidden layer according to the number of the nodes of the input layer and the number of the nodes of the output layer, determining an activation function of the BP neural network model according to an input vector of the input layer, and determining an error function of the BP neural network model according to an output vector of the output layer and an expected output vector;
a training model algorithm building module used for building a BP neural network model training algorithm based on sample data and an error function;
the training model module is used for training the BP neural network model by taking historical energy consumption data and weather data as sample data until the training times exceed a preset value or the average error of model output reaches preset precision, and then training is finished;
and the energy consumption prediction module is used for inputting the weather data to be predicted into the trained BP neural network model to predict the electric heating energy consumption of the public building or the residential building to obtain an energy consumption prediction result.
Further, in the above-mentioned case,the specific expression of calculating the number of the hidden layer nodes according to the number of the input layer nodes and the number of the output layer nodes in the model building module is as follows:
Figure BDA0003302789940000034
wherein h represents the number of hidden layer neurons, m represents the number of input layer neurons, n represents the number of output layer neurons, and λ ∈ [1,10 ]]Taking an integer.
Further, the activation function of the BP neural network model in the model building module is:
Figure BDA0003302789940000035
where x is the input layer input vector.
Further, the error function in the model building module is:
Figure BDA0003302789940000041
wherein n represents the number of neurons in the output layer, and the output vector yo of the output layer is { yo ═ yo1,yo2,yo3,...yonV ═ v, desired output vector1,v2,v3,...vn}。
In a third aspect of the present invention, a terminal is provided, where the terminal is equipped with a processor for implementing the above method for predicting electric heating energy consumption based on a BP neural network.
According to the electric heating energy consumption prediction method and the prediction system based on the BP neural network, the neural network model is established through the artificial intelligence BP neural network, training is carried out through a large number of data samples, the BP neural network model is used as a prediction core in combination with historical data, and a target prediction energy consumption curve is obtained, so that the prediction result is closer to the actual result, the purpose of energy consumption use prediction is achieved, and decision information is provided for a manager. The beneficial effects that finally reach: the accuracy of the heating controller on temperature control is improved, the controller can realize more accurate temperature control, guarantee is provided for temperature data analysis, and disputes caused by substandard heating due to temperature measurement problems are reduced.
Drawings
FIG. 1 is a flow chart of an electric heating energy consumption prediction method based on a BP neural network according to an embodiment of the invention;
FIG. 2 is a schematic diagram of a BP neural network model training process according to an embodiment of the present invention;
FIG. 3 is a schematic diagram of a sample data to test data fit according to an embodiment of the present invention;
fig. 4 is a schematic structural diagram of an electric heating energy consumption prediction system based on a BP neural network according to an embodiment of the present invention.
Detailed Description
In order to further describe the technical scheme of the present invention in detail, the present embodiment is implemented on the premise of the technical scheme of the present invention, and detailed implementation modes and specific steps are given.
As shown in fig. 1, the method for predicting electrical heating energy consumption based on a BP neural network according to the embodiment of the present invention includes establishing a BP neural network model, establishing a BP neural network model training algorithm based on sample data and an error function, training the BP neural network model by using historical energy consumption data and weather data as sample data, and inputting weather data to be predicted into the trained BP neural network model to predict electrical heating energy consumption of a public building or a residential building, and the specific implementation steps are:
s100, establishing a BP neural network model, dividing the BP neural network model into an input layer, a hidden layer and an output layer, calculating the number of nodes of the hidden layer according to the number of the nodes of the input layer and the number of the nodes of the output layer, determining an activation function of the BP neural network model according to an input vector of the input layer, and determining an error function of the BP neural network model according to an output vector of the output layer and an expected output vector;
in a specific embodiment of the invention, the prediction accuracy of the BP neural network model is related to the number of neurons of an implicit layer in a three-layer framework, an empirical formula is used for calculating the number of nodes of the implicit layer, and a specific expression for calculating the number of nodes of the implicit layer according to the number of nodes of an input layer and the number of nodes of an output layer is as follows:
Figure BDA0003302789940000051
wherein h represents the number of hidden layer neurons, m represents the number of input layer neurons, n represents the number of output layer neurons, and λ ∈ [1,10 ]]Taking an integer, λ can be a random constant within the interval.
In one embodiment of the present invention, the activation function of the BP neural network model is:
Figure BDA0003302789940000052
where x is the input layer input vector. The error function is:
Figure BDA0003302789940000053
wherein n represents the number of neurons in the output layer, and the output vector yo of the output layer is { yo ═ yo1,yo2,yo3,...yonV ═ v, desired output vector1,v2,v3,...vn}。
S200, establishing a BP neural network model training algorithm based on sample data and an error function;
in a specific embodiment of the invention, the specific implementation steps for establishing the BP neural network model training algorithm based on the sample data and the error function are as follows:
s2001, randomly selecting k input samples and an expected output from the sample data, where the input sample vector is:
x(k)＝(x1(k),x2(k),...,xn(k))；
the desired output vector is:
v(k)＝(v1(k),v2(k),...,vn(k) where n represents the number of output layer neurons,
s2002, calculating input and output values of each neuron:
the inputs to the hidden layer h neurons are:
Figure BDA0003302789940000054
where d 1,2, h, γ denotes the weight of the input layer to the hidden layer, γ ∈ (-1, 1), bdA threshold representing a neuron in the hidden layer;
the output of the hidden layer h neurons is: ho (electronic clock)d(k)＝f(hid(k))；
The inputs to the output layer are:
Figure BDA0003302789940000055
where t 1,2, n, β represents the weight from hidden layer to output layer, β ∈ (-1, 1), btA threshold value representing a neuron in the output layer;
the output of the output layer is: yot(k)＝f(yit(k))；
S2003, calculating partial derivatives delta (k) of neurons of an output layer by using an error function e:
Figure BDA0003302789940000056
s2004, utilizing weight beta from hidden layer to output layer, partial derivative delta (k) of output layer and output ho of hidden layerd(k) And calculating partial derivatives of the hidden layer neurons by using an error function e, wherein the error function e calculates the partial derivatives epsilon of the hidden layer input neurons1(k) Comprises the following steps:
ε1(k)＝-δ(k)hod(k)
error function e calculates partial derivative epsilon for hidden layer output neurons2(k) Comprises the following steps:
Figure BDA0003302789940000061
s2005, optimizing weight beta from hidden layer to output layer, using epsilon1(k) And (3) calculating:
Figure BDA0003302789940000062
βN+1＝βN+με1(k)
where μ represents the learning rate of the BP neural network model, μE (0, 1), optimizing the weight gamma from the input layer to the hidden layer, and calculating the partial derivative epsilon of the input neuron of the hidden layer by using an error function e2(k) And (3) calculating:
Figure BDA0003302789940000063
γN+1＝γN-με2(k)
s2006, calculating an output global error by using the actual output and the expected output:
Figure BDA0003302789940000064
s300, training a BP neural network model by taking historical energy consumption data and weather data as sample data until the training times exceed a preset value or the model output error reaches preset precision, and finishing training;
in a specific embodiment of the invention, model training is carried out by importing historical energy consumption data and weather data, and if the training times exceed a set value or the error meets the expected precision requirement, the model is successfully established; otherwise, continuing training the sample data of the next batch until the expected accuracy is met, wherein the specific training process is as shown in fig. 2, and the specific implementation process is as follows:
s3001, inputting a sample data into the established BP neural network model;
s3002, calculating the input and output of each neuron according to the step S2002;
s3003, calculating an output error according to the error function e;
s3004, according to the output error back propagation, updating and adjusting the weight according to the step S2005;
s3005, judging whether the training sample is used up, if not, returning to S3001, and if so, continuing to S3006;
s3006, calculating a BP neural network model output global error according to S2006;
s3007, judging whether the model output global error precision meets the requirement, if so, finishing the training, and if not, continuing to S3008;
and S3008, judging whether the training frequency reaches the upper limit, if so, finishing the training, otherwise, returning to S3001.
In a specific embodiment of the invention, the upper limit value of the training times is set to 10, and the expected precision requirement can be met when the global error of the model output is less than 5 kwh.
S400, inputting weather data to be predicted into the trained BP neural network model to predict electric heating energy consumption of public buildings or residential buildings, and obtaining an energy consumption prediction result.
In a specific embodiment of the invention, the trained BP neural network model is used as a prediction algorithm core of the carbon fiber electric heating prediction system to predict the electric heating energy consumption of public buildings or residential buildings. The method specifically comprises the following steps: selecting a public building or a residential building which needs to be subjected to electric heating energy consumption prediction, and acquiring basic information of the building, historical weather data and energy consumption data; configuring sample data, selecting the extensive historical weather data and energy consumption data as the sample data, inputting the sample data into a BP neural network model for training, and inputting the predicted weather data after training for prediction to obtain an energy consumption prediction curve.
A specific example is provided below, since the standard requirement of winter heating is that the indoor temperature reaches 18 ℃, the average indoor temperature of the selected sample data is above 18 ℃, and the average temperature is 9% earlier per natural day: average temperature from 00 to 18:00 PM, sample data as shown in the following table:
TABLE 1 sample data
Serial number Average indoor temperature Averaging chamberOutside air temperature Consumption data (approximately equal to)
1 23℃ 5℃ 40kwh
2 23℃ -1℃ 43kwh
3 23℃ 8℃ 36kwh
4 20℃ 5℃ 38kwh
5 20℃ -1℃ 40kwh
6 20℃ 8℃ 34kwh
7 18℃ 5℃ 30kwh
8 18℃ -1℃ 35kwh
9 18℃ 8℃ 38kwh
10 25℃ 5℃ 49kwh
11 25℃ -1℃ 45kwh
12 25℃ 8℃ 52kwh
Inputting the sample data in the table 1 into a BP neural network model for sample training, and then predicting by using the sample data in the table 1 to obtain predicted data of the sample, as shown in table 2:
TABLE 2 sample data and prediction data
Figure BDA0003302789940000071
Figure BDA0003302789940000081
As shown in fig. 3, a curve fitting graph of sample data and test data is shown, and although there is a certain prediction error, the overall prediction trend obtained by the method of the present invention is similar to the actual trend, and there is an intersection point, so that it can be said that the prediction is effective. In the field of electric heating energy consumption analysis, an energy consumption trend is a core, and in addition, in a heating season, due to the variability of outdoor temperature, an energy consumption error can be accepted within 5kwh, so that a corresponding energy consumption trend can be predicted, and managers can make a corresponding energy consumption regulation and control strategy.
Hereinafter, a system corresponding to the method shown in fig. 1 according to an embodiment of the present disclosure will be described with reference to fig. 4, and an electric heating energy consumption prediction system based on a BP neural network includes: the model establishing module 011 is used for establishing a BP neural network model, dividing the BP neural network model into an input layer, a hidden layer and an output layer, calculating the number of nodes of the hidden layer according to the number of the nodes of the input layer and the number of the nodes of the output layer, determining an activation function of the BP neural network model according to an input vector of the input layer, and determining an error function of the BP neural network model according to an output vector of the output layer and an expected output vector; a training model establishing algorithm module 012, configured to establish a BP neural network model training algorithm based on sample data and an error function; the training model module 013 trains the BP neural network model by taking historical energy consumption data and weather data as sample data until the training times exceed a preset value or the average error of model output reaches a preset precision, and then training is completed; and the energy consumption prediction module 014 is used for inputting the weather data to be predicted into the trained BP neural network model to predict the electric heating energy consumption of the public building or the residential building, so as to obtain an energy consumption prediction result. In addition to these 4 modules, the system 01 may include other components, however, since these components are not related to the content of the embodiments of the present disclosure, illustration and description thereof are omitted here.
The model establishing module 011 calculates the specific expression of the number of the hidden layer nodes according to the number of the input layer nodes and the number of the output layer nodes as follows:
Figure BDA0003302789940000082
wherein h represents the number of hidden layer neurons, m represents the number of input layer neurons, n represents the number of output layer neurons, and λ ∈ [1,10 ]]Taking an integer.
The activation function of the BP neural network model in the model module 011 is established as follows:
Figure BDA0003302789940000083
where x is the input layer input vector.
The error function in the build model module 011 is:
Figure BDA0003302789940000084
wherein n represents the number of neurons in the output layer, and the output vector yo of the output layer is { yo ═ yo1,yo2,yo3,...yonV ═ v, desired output vector1,v2,v3,...vn}。
The specific working process of the electric heating energy consumption prediction system 011 based on the BP neural network refers to the description of the electric heating energy consumption prediction method based on the BP neural network, and is not repeated.
In addition, the method for predicting electric heating energy consumption based on the BP neural network according to the embodiment of the present invention may also be implemented by using a terminal, where the terminal is equipped with a processor for implementing the method for predicting electric heating energy consumption based on the BP neural network, and the terminal includes, but is not limited to, an input/output component, a memory, and the like in addition to the processor, and when implementing different embodiments, one or more components are added according to actual needs.
According to the electric heating energy consumption prediction method and the prediction system based on the BP neural network, the neural network model is established through the artificial intelligence BP neural network, training is carried out through a large number of data samples, the BP neural network model is used as a prediction core in combination with historical data, and a target prediction energy consumption curve is obtained, so that the prediction result is closer to the actual result, the purpose of energy consumption use prediction is achieved, and decision information is provided for a manager. The beneficial effects that finally reach: the accuracy of the temperature measurement of the carbon fiber controller is improved, the controller can realize more accurate temperature control, guarantee is provided for temperature data analysis, and disputes caused by substandard heating due to the temperature measurement problem are reduced.
In this document, the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process or method that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process or method.
The foregoing is a more detailed description of the invention in connection with specific preferred embodiments and it is not intended that the invention be limited to these specific details. For those skilled in the art to which the invention pertains, several simple deductions or substitutions can be made without departing from the spirit of the invention, and all shall be considered as belonging to the protection scope of the invention.

Claims (10)

1. An electric heating energy consumption prediction method based on a BP neural network is characterized by comprising the following steps:
establishing a BP neural network model, dividing the BP neural network model into an input layer, a hidden layer and an output layer, calculating the number of nodes of the hidden layer according to the number of the nodes of the input layer and the number of the nodes of the output layer, determining an activation function of the BP neural network model according to an input vector of the input layer, and determining an error function of the BP neural network model according to an output vector of the output layer and an expected output vector;
establishing a BP neural network model training algorithm based on sample data and an error function;
training a BP neural network model by taking historical energy consumption data and weather data as sample data until the training times exceed a preset value or the average error of model output reaches preset precision, and finishing training;
and inputting weather data to be predicted into the trained BP neural network model to predict the electric heating energy consumption of the public building or the residential building, so as to obtain an energy consumption prediction result.
2. The base of claim 1The method for predicting the electric heating energy consumption of the BP neural network is characterized in that the specific expression of calculating the number of the hidden layer nodes according to the number of the input layer nodes and the number of the output layer nodes is as follows:
Figure FDA0003302789930000011
wherein h represents the number of hidden layer neurons, m represents the number of input layer neurons, n represents the number of output layer neurons, and λ ∈ [1,10 ]]Taking an integer.
3. The BP neural network-based electric heating energy consumption prediction method according to claim 2, wherein the activation function of the BP neural network model is:
Figure FDA0003302789930000012
where x is the input layer input vector.
4. The BP neural network-based electric heating energy consumption prediction method according to claim 3, wherein the error function is:
Figure FDA0003302789930000013
wherein n represents the number of neurons in the output layer, and the output vector yo of the output layer is { yo ═ yo1,yo2,yo3,...yonV ═ v, desired output vector1,v2,v3,...vn}。
5. The method for predicting electric heating energy consumption based on the BP neural network according to claim 4, wherein the establishing of the BP neural network model training algorithm based on the sample data and the error function comprises optimizing a weight β from a hidden layer to an output layer and a weight γ from an input layer to the hidden layer, and the specific expression is as follows:
Figure FDA0003302789930000014
βN+1＝βN+με1(k)
wherein mu represents the learning rate of the BP neural network model, mu belongs to (0, 1), epsilon1(k) Representing the partial derivative of the error function e to the weight beta from the hidden layer to the output layer, wherein k represents the number of input samples;
Figure FDA0003302789930000021
γN+1＝γN-με2(k)
wherein epsilon2(k) Representing the partial derivative of the error function e to the weights y of the input layer to the hidden layer.
6. The method for predicting electric heating energy consumption based on the BP neural network according to claim 5, wherein the establishing of the BP neural network model training algorithm based on the sample data and the error function further comprises calculating an output global error by using an actual output and an expected output, and the specific expression is as follows:
Figure FDA0003302789930000022
where m represents the number of input layer neurons, n represents the number of output layer neurons, v (k) represents the expected output value, yo (k) represents the actual output value.
7. An electric heating energy consumption prediction system based on a BP neural network, which is characterized by comprising:
the model building module is used for building a BP neural network model, dividing the BP neural network model into an input layer, a hidden layer and an output layer, calculating the number of nodes of the hidden layer according to the number of the nodes of the input layer and the number of the nodes of the output layer, determining an activation function of the BP neural network model according to an input vector of the input layer, and determining an error function of the BP neural network model according to an output vector of the output layer and an expected output vector;
a training model algorithm building module used for building a BP neural network model training algorithm based on sample data and an error function;
the training model module is used for training the BP neural network model by taking historical energy consumption data and weather data as sample data until the training times exceed a preset value or the average error of model output reaches preset precision, and then training is finished;
and the energy consumption prediction module is used for inputting the weather data to be predicted into the trained BP neural network model to predict the electric heating energy consumption of the public building or the residential building to obtain an energy consumption prediction result.
8. The electrical heating energy consumption prediction system based on the BP neural network of claim 7, wherein the specific expression for calculating the number of hidden layer nodes according to the number of input layer nodes and the number of output layer nodes in the model building module is as follows:
Figure FDA0003302789930000023
wherein h represents the number of hidden layer neurons, m represents the number of input layer neurons, n represents the number of output layer neurons, and λ ∈ [1,10 ]]Taking an integer.
9. The BP neural network-based electric heating energy consumption prediction system of claim 8, wherein the activation function of the BP neural network model in the model building module is:
Figure FDA0003302789930000024
where x is the input layer input vector.
10. A terminal, characterized in that the terminal is equipped with a processor for implementing the BP neural network-based electric heating energy consumption prediction method according to any one of claims 1 to 6.
CN202111195519.6A 2021-10-14 2021-10-14 Electric heating energy consumption prediction method and prediction system based on BP neural network Pending CN113807615A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202111195519.6A CN113807615A (en) 2021-10-14 2021-10-14 Electric heating energy consumption prediction method and prediction system based on BP neural network

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202111195519.6A CN113807615A (en) 2021-10-14 2021-10-14 Electric heating energy consumption prediction method and prediction system based on BP neural network

Publications (1)

Publication Number Publication Date
CN113807615A true CN113807615A (en) 2021-12-17

Family

ID=78897654

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202111195519.6A Pending CN113807615A (en) 2021-10-14 2021-10-14 Electric heating energy consumption prediction method and prediction system based on BP neural network

Country Status (1)

Country Link
CN (1) CN113807615A (en)

Cited By (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113723534A (en) * 2021-09-02 2021-11-30 南京润内克西信息科技有限公司 Urban raise dust on-line monitoring system based on BP neural network
CN114997692A (en) * 2022-06-17 2022-09-02 哈尔滨工业大学 Rotary table rotation reference angular pendulum error evaluation method based on BP neural network
CN115099129A (en) * 2022-05-31 2022-09-23 中海石油（中国）有限公司天津分公司 Natural gas well yield prediction method based on input characteristic error correction
CN116304968A (en) * 2023-01-06 2023-06-23 杭州山科智能科技股份有限公司 Ultrasonic water meter flow data fusion method and device based on BP neural network
CN116862077A (en) * 2023-08-31 2023-10-10 吉林电力交易中心有限公司 Electric heating operation cost prediction method and medium based on multi-mode combination model
CN117075549A (en) * 2023-08-17 2023-11-17 湖南源达智能科技有限公司 Plant control method and system based on artificial neural network

Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN105446139A (en) * 2015-12-18 2016-03-30 华南理工大学 Construction energy consumption analysis method and construction energy consumption analysis system based on BP neural network
CN105956683A (en) * 2016-04-19 2016-09-21 国网甘肃省电力公司电力科学研究院 Public institution electric power requirement side management system and electric power quality prediction method

Patent Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN105446139A (en) * 2015-12-18 2016-03-30 华南理工大学 Construction energy consumption analysis method and construction energy consumption analysis system based on BP neural network
CN105956683A (en) * 2016-04-19 2016-09-21 国网甘肃省电力公司电力科学研究院 Public institution electric power requirement side management system and electric power quality prediction method

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
戴永亮: ""基于负荷预测的既有建筑制冷机组运行节能研究"", 《中国优秀硕士学位论文全文数据库 工程科技II辑》 *

Cited By (9)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113723534A (en) * 2021-09-02 2021-11-30 南京润内克西信息科技有限公司 Urban raise dust on-line monitoring system based on BP neural network
CN115099129A (en) * 2022-05-31 2022-09-23 中海石油（中国）有限公司天津分公司 Natural gas well yield prediction method based on input characteristic error correction
CN115099129B (en) * 2022-05-31 2024-04-16 中海石油（中国）有限公司天津分公司 Natural gas well yield prediction method based on input characteristic error correction
CN114997692A (en) * 2022-06-17 2022-09-02 哈尔滨工业大学 Rotary table rotation reference angular pendulum error evaluation method based on BP neural network
CN116304968A (en) * 2023-01-06 2023-06-23 杭州山科智能科技股份有限公司 Ultrasonic water meter flow data fusion method and device based on BP neural network
CN116304968B (en) * 2023-01-06 2023-09-15 杭州山科智能科技股份有限公司 Ultrasonic water meter flow data fusion method and device based on BP neural network
CN117075549A (en) * 2023-08-17 2023-11-17 湖南源达智能科技有限公司 Plant control method and system based on artificial neural network
CN117075549B (en) * 2023-08-17 2024-03-22 湖南源达智能科技有限公司 Plant control method and system based on artificial neural network
CN116862077A (en) * 2023-08-31 2023-10-10 吉林电力交易中心有限公司 Electric heating operation cost prediction method and medium based on multi-mode combination model

Similar Documents

Publication Publication Date Title
CN113807615A (en) Electric heating energy consumption prediction method and prediction system based on BP neural network
Mustafaraj et al. Development of room temperature and relative humidity linear parametric models for an open office using BMS data
Bakker et al. Improving the performance of water demand forecasting models by using weather input
Paudel et al. Pseudo dynamic transitional modeling of building heating energy demand using artificial neural network
Nghiem et al. Data-driven demand response modeling and control of buildings with gaussian processes
CN106408223A (en) Short-term load prediction based on meteorological similar day and error correction
Knudsen et al. Experimental test of a black-box economic model predictive control for residential space heating
Chen et al. Energy cost and consumption reduction of an office building by Chaotic Satin Bowerbird Optimization Algorithm with model predictive control and artificial neural network: A case study
CN111191854A (en) Photovoltaic power generation prediction model and method based on linear regression and neural network
CN115965132A (en) Distributed photovoltaic digital twin system power prediction method based on GA-BP neural network
Killian et al. Short-term occupancy prediction and occupancy based constraints for MPC of smart homes
Reynolds et al. A smart heating set point scheduler using an artificial neural network and genetic algorithm
Hrehová et al. The fuzzy logic predictive model for remote increasing energy efficiency
Eguizabal et al. Simplified model for the short-term forecasting of heat loads in buildings
CN111815051B (en) GRNN photovoltaic power generation prediction method considering weather influence factors
Agouzoul et al. Using neural network in a model-based predictive control loop to enhance energy performance of buildings
Amasyali et al. A machine learning-based approach to predict the aggregate flexibility of hvac systems
CN114861555A (en) Regional comprehensive energy system short-term load prediction method based on Copula theory
Pandey et al. Mathematical modeling for short term indoor room temperature forecasting using Box-Jenkins models: An Indian evidence
CN115952888A (en) Multivariable grey model-based energy carbon emission prediction method and system
JP3707589B2 (en) Electricity demand forecast method
Palermo et al. Optimization of a feedforward neural network’s architecture for an HVAC system problem
Zuazo et al. Short-Term Load Forecasting of building electricity consumption using NARX Neural Networks model
Moftakhari et al. Thermal analysis of HVAC and solar panels using genetic optimization algorithm
Bogdanovs et al. Impact of Intelligence System of Building Management System for Energy Efficiency in the Test Facility

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
RJ01 Rejection of invention patent application after publication
RJ01 Rejection of invention patent application after publication

Application publication date: 20211217