CN113205203A - CNN-LSTM-based building energy consumption prediction method and system - Google Patents
CNN-LSTM-based building energy consumption prediction method and system Download PDFInfo
- Publication number
- CN113205203A CN113205203A CN202110335276.5A CN202110335276A CN113205203A CN 113205203 A CN113205203 A CN 113205203A CN 202110335276 A CN202110335276 A CN 202110335276A CN 113205203 A CN113205203 A CN 113205203A
- Authority
- CN
- China
- Prior art keywords
- data
- energy consumption
- building energy
- cnn
- 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
Links
- 238000005265 energy consumption Methods 0.000 title claims abstract description 92
- 238000000034 method Methods 0.000 title claims abstract description 43
- 238000013527 convolutional neural network Methods 0.000 claims abstract description 25
- 238000012549 training Methods 0.000 claims abstract description 22
- 238000004422 calculation algorithm Methods 0.000 claims abstract description 14
- 238000000605 extraction Methods 0.000 claims abstract description 10
- 238000007781 pre-processing Methods 0.000 claims description 11
- 210000002569 neuron Anatomy 0.000 claims description 10
- 230000006403 short-term memory Effects 0.000 claims description 10
- 230000006870 function Effects 0.000 claims description 8
- 238000010606 normalization Methods 0.000 claims description 8
- 230000002159 abnormal effect Effects 0.000 claims description 7
- 238000012544 monitoring process Methods 0.000 claims description 7
- 230000004913 activation Effects 0.000 claims description 6
- 238000012545 processing Methods 0.000 claims description 6
- 238000004140 cleaning Methods 0.000 claims description 4
- 230000008521 reorganization Effects 0.000 claims description 3
- 238000005457 optimization Methods 0.000 abstract description 2
- 238000013135 deep learning Methods 0.000 description 5
- 230000008569 process Effects 0.000 description 4
- 238000012360 testing method Methods 0.000 description 4
- 238000004458 analytical method Methods 0.000 description 3
- 238000013528 artificial neural network Methods 0.000 description 3
- 238000010586 diagram Methods 0.000 description 3
- 238000010801 machine learning Methods 0.000 description 3
- 238000012986 modification Methods 0.000 description 3
- 230000004048 modification Effects 0.000 description 3
- 230000008901 benefit Effects 0.000 description 2
- 230000005540 biological transmission Effects 0.000 description 2
- 230000008859 change Effects 0.000 description 2
- 238000004134 energy conservation Methods 0.000 description 2
- 230000002068 genetic effect Effects 0.000 description 2
- 238000007726 management method Methods 0.000 description 2
- 230000005856 abnormality Effects 0.000 description 1
- 230000006978 adaptation Effects 0.000 description 1
- 238000013473 artificial intelligence Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 238000010276 construction Methods 0.000 description 1
- 238000013499 data model Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000012217 deletion Methods 0.000 description 1
- 230000037430 deletion Effects 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 238000005286 illumination Methods 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 230000007787 long-term memory Effects 0.000 description 1
- 230000007774 longterm Effects 0.000 description 1
- 230000008520 organization Effects 0.000 description 1
- 230000001737 promoting effect Effects 0.000 description 1
- 230000009467 reduction Effects 0.000 description 1
- 238000012216 screening Methods 0.000 description 1
- 238000012546 transfer Methods 0.000 description 1
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
- G06N3/044—Recurrent networks, e.g. Hopfield networks
-
- 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
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
-
- 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
- G06Q50/08—Construction
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Business, Economics & Management (AREA)
- General Physics & Mathematics (AREA)
- General Health & Medical Sciences (AREA)
- Health & Medical Sciences (AREA)
- Software Systems (AREA)
- Economics (AREA)
- Evolutionary Computation (AREA)
- Computational Linguistics (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- General Engineering & Computer Science (AREA)
- Biophysics (AREA)
- Mathematical Physics (AREA)
- Biomedical Technology (AREA)
- Artificial Intelligence (AREA)
- Life Sciences & Earth Sciences (AREA)
- Strategic Management (AREA)
- Data Mining & Analysis (AREA)
- Human Resources & Organizations (AREA)
- Marketing (AREA)
- Tourism & Hospitality (AREA)
- General Business, Economics & Management (AREA)
- Game Theory and Decision Science (AREA)
- Development Economics (AREA)
- Entrepreneurship & Innovation (AREA)
- Operations Research (AREA)
- Quality & Reliability (AREA)
- Primary Health Care (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention provides a CNN-LSTM-based building energy consumption prediction method and a CNN-LSTM-based building energy consumption prediction system, wherein building energy consumption sequence data acquired by a building energy consumption acquisition module and data of relevant influence factors are used as data bases, and then the data are preprocessed and recombined into a data format required by a training algorithm; setting a Convolutional Neural Network (CNN) as a feature extraction layer, adopting a structure of performing convolution in a feature direction by adopting one-dimensional convolution, and then modeling the time sequence of data by using a long-short term memory network (LSTM); and obtaining a predicted value of the building energy consumption through the model and the new input data. By utilizing the algorithm provided by the invention, the high-precision prediction of the building energy consumption can be realized, and a reference value is provided for the optimization of the building operation.
Description
Technical Field
The invention relates to the technical field of building energy intelligentization, in particular to a CNN-LSTM-based building energy consumption prediction method and system, and particularly relates to a method for modeling and predicting building energy consumption by utilizing a deep learning method in machine learning.
Background
Energy conservation and emission reduction in the building field are necessary rings for promoting ecological civilization construction, building energy consumption prediction is taken as an important means for mastering the operation characteristics of the building, and the method has positive significance for building energy conservation management and building energy utilization rate improvement. In recent years, development of big data and artificial intelligence technology provides a data base and a modeling analysis algorithm for prediction of building energy consumption. The deep learning algorithm in machine learning is applied to different fields because the deep learning algorithm has better performance in processing complex data learning problems.
The traditional building energy consumption prediction analysis method is a physical modeling method based on heat transfer analysis, energy consumption is obtained by simulating building operation by virtue of simulation software, and large errors are caused by difficulty in obtaining accurate building parameters. The building is predicted based on the data driving mode of machine learning, and the method has the advantages of high modeling speed and high prediction precision.
The Convolutional Neural Network (CNN) has wide application in the field of image processing, and can process the local relevance of data; long and short term memory models (LSTM) are used to process natural language and time series data to learn the sequence variation rules in the data. Due to the difference of application fields, a reasonable input sample form and a reasonable model structure need to be designed to realize the organic combination of the two algorithms.
The prior art related to the present application is patent document CN 110046743 a, which discloses a method and a system for predicting energy consumption of public buildings based on GA-ANN. The invention relates to a public building energy consumption prediction method and a public building energy consumption prediction system based on GA-ANN, which are used for collecting time-by-time energy consumption of public buildings and data of influence factors of the time-by-time energy consumption, sorting the data and preprocessing the data; dividing a training set and a test set, and screening input variables by a correlation coefficient method; inputting test data, optimizing relevant parameters of an Artificial Neural Network (ANN) model through a Genetic Algorithm (GA), and then training the model by using training set data; predicting the energy consumption of the public building by inputting the input variable of the predicted period; and finally, evaluating the prediction effect on the test set through the error index, and giving out an allowable error range. The invention provides a flow and a method for predicting energy consumption of public buildings by using a genetic algorithm and an artificial neural network. The method realizes high-precision prediction for public buildings and provides basis for monitoring, managing and diagnosing the energy consumption of the public buildings.
Disclosure of Invention
Aiming at the defects in the prior art, the invention aims to provide a CNN-LSTM-based building energy consumption prediction method and system.
The building energy consumption prediction method and system based on CNN-LSTM provided by the invention comprises the following steps:
step S1: acquiring historical energy consumption sequence data of a building by using a building energy consumption monitoring module, and synchronously acquiring influence factor data;
step S2: preprocessing historical sequence data, wherein the data preprocessing comprises abnormal data cleaning, linear interpolation completion and data normalization, and the historical sequence data comprises historical energy consumption data of the whole year of the building; then, recombining the data into a two-dimensional data table format according to two dimensions of time sequence and characteristics, wherein the two-dimensional data table format is used as a single sample of data for training, and the time sequence is data acquisition time points arranged according to a time step;
step S3: using a Convolutional Neural Network (CNN) as a feature extraction layer, using a long short-term memory Layer (LSTM) after the CNN to realize modeling of the time sequence, and finally using a full-connection layer network to establish a building energy consumption prediction model by using the network structure;
step S4: training a model in a server by using the recombined historical energy consumption data set, and storing the model obtained by training by a model structure weighted value method, wherein the server is a computer resource for a prediction system to call, and the weighted value is the weight of network neuron connection;
step S5: and recombining the new influence factor data, inputting the recombined data into a model, dividing the output of the model by the scaling ratio during normalization, and calculating to obtain a building energy consumption predicted value.
Preferably, the influencing factor data includes three types: outdoor weather parameters, building operation rules and building internal use conditions; the meteorological parameters comprise outdoor temperature and outdoor relative humidity, the building operation rule comprises changes of time, working day, holiday and holiday, and the internal use condition comprises indoor set temperature and the number of indoor personnel;
preferably, the data reorganization mode is to organize the data into an input sample of a two-dimensional table, the data table is organized according to two dimensions of time and characteristics, the data are arranged in two adjacent days according to step in time, and the data of all influencing factors are arranged in characteristic dimensions;
preferably, the model adopts a specific network structure, a Convolutional Neural Network (CNN) is used as a feature extraction layer, a structure that a one-dimensional convolutional layer is convolved in a feature direction is adopted, and 3 layers of convolutional layers are arranged to improve the feature extraction capability; after CNN, using long short-term memory Layer (LSTM) to realize modeling of time sequence, setting 3 layers of long short-term memory layer; finally, the 3-layer full-connection layer network is used for realizing the nonlinear learning of the building operation; and preferably, specific parameters suitable for building energy consumption prediction in the network structure are given, wherein the specific parameters comprise the number of neurons and activation function selection;
the invention also provides a CNN-LSTM-based building energy consumption prediction system, which comprises:
module M1: the energy consumption monitoring module is used for measuring the building energy consumption data and the influence factor data in real time and storing the data according to the acquisition time point;
module M2: the data processing module is used for preprocessing the data, recombining the data and storing the recombined sample data into a computer for calling by the model training module M3;
module M3: the algorithm modeling module is used for realizing the model structure through computer language by computer programming to form a model to be trained; then calling a data sample training model processed by an M2 module, and storing the model after the model is iterated and stabilized;
module M4: and the model predicted value output module is used for obtaining the predicted value of the building energy consumption through the model obtained by the module M3 after the influence factor data of the predicted time period is processed by the data module M2.
Compared with the prior art, the invention has the following beneficial effects:
1. the prediction speed is high, the building energy consumption is predicted through CNN-LSTM modeling, the building historical energy consumption data is used for training and storing the model, the building energy consumption prediction can be directly called and rapidly provided subsequently, and the problems of long physical modeling period and large engineering quantity are solved.
2. The prediction precision is high, the learning capability of a deep learning algorithm on complex problems is utilized, the model structure can be used for well learning and expressing the change characteristic of the building energy consumption, and the prediction precision of the building energy consumption is obviously improved.
3. The method is high in practicability, can be widely applied to energy consumption prediction of public buildings and residential buildings, can optimize building energy management strategies by taking prediction results as reference, and improves the operation energy efficiency of the buildings.
Drawings
Other features, objects and advantages of the invention will become more apparent upon reading of the detailed description of non-limiting embodiments with reference to the following drawings:
FIG. 1 is a schematic flow diagram of a system for predicting building energy consumption according to the present invention;
FIG. 2 is a schematic diagram of data organization;
FIG. 3 is a diagram illustrating model structure and parameters.
Detailed Description
The present invention will be described in detail with reference to specific examples. The following examples will assist those skilled in the art in further understanding the invention, but are not intended to limit the invention in any way. It should be noted that it would be obvious to those skilled in the art that various changes and modifications can be made without departing from the spirit of the invention. All falling within the scope of the present invention.
The invention relates to a CNN-LSTM-based building energy consumption prediction method and a CNN-LSTM-based building energy consumption prediction system.
As shown in fig. 1, the method for predicting building energy consumption based on CNN-LSTM provided by the present invention includes:
step S1: acquiring historical energy consumption sequence data of a building by using a building energy consumption monitoring module, and synchronously acquiring influence factor data;
step S2: preprocessing historical sequence data, wherein the data preprocessing comprises abnormal data cleaning, linear interpolation completion and data normalization, and the historical sequence data comprises historical energy consumption data of the whole year of the building; then, recombining the data into a two-dimensional data table format according to two dimensions of time sequence and characteristics, wherein the two-dimensional data table format is used as a single sample of data for training, and the time sequence is data acquisition time points arranged according to a time step;
step S3: using a Convolutional Neural Network (CNN) as a feature extraction layer, using a long short-term memory Layer (LSTM) after the CNN to realize modeling of the time sequence, and finally using a full-connection layer network to establish a building energy consumption prediction model by using the network structure;
step S4: training a model in a server by using the recombined historical energy consumption data set, and storing the model obtained by training by using a model structure and weight, wherein the server is a computer resource for a prediction system to call, and the weight is the weight of network neuron connection;
step S5: and recombining the new influence factor data, inputting the recombined data into a model, dividing the output of the model by the scaling ratio during normalization, and calculating to obtain a building energy consumption predicted value.
Specifically, the collected building energy consumption sequence data comprises total building energy consumption, socket illumination subentries and air conditioner energy consumption subentries, and aiming at the building energy consumption, the method provided by the patent can be used for realizing high-precision prediction. The frequency of acquisition is time-by-time, and the data with higher frequency of acquisition is added to obtain time-by-time energy consumption sequence data. Taking the total building energy consumption as an example, sequence data of the total building energy consumption of one year in history is taken.
Specifically, the influence factor data includes three types: outdoor weather parameters, building operation rules and building internal use conditions; the meteorological parameters comprise outdoor temperature and outdoor relative humidity and come from data collected by a current meteorological station; the building operation rule comprises changes of time, working day, holiday and internal use conditions comprise indoor set temperature and indoor personnel number. All influencing factors are aligned with the building energy consumption data by taking time as a reference. The time law converts continuous numerical values into category variables in a One-hot coding mode, and the working day is arranged from calendar data. The indoor usage data is provided by building managers. Taking a certain building in the sea area as an example, the real-time meteorological parameters are obtained through the internet, an operation rule data file of a future month is preset, the indoor set temperature is uniformly managed by a building manager, and the indoor data of personnel is obtained through door access card data statistics.
Specifically, the data preprocessing process comprises abnormal data cleaning, linear interpolation completion and data normalization. The abnormal data comprises negative values and transmission abnormality, the transmission abnormal data is screened by setting a threshold value through the installed capacity of the building, and the total energy consumption value of the building in the example is 10 orders of magnitude3Setting the upper threshold value to 104. The linear interpolation completion aims at a small amount of breakpoints in the data, and the calculation formula is as follows:
n (N is less than or equal to 5) is the number of deletion points, Yn(N is 1,2, …, N) is the data value of the nth missing point, Y0And YN+1The data are respectively the data which are close to the front and the back of the missing sequence, the data are marked to be abnormal under the condition of long-term missing, and the corresponding data are not used in the subsequent training. Data normalization for each feature, divided by the maximum value Y in the corresponding feature history datamaxGet, processed data falls in [0,1 ]]Within the interval (c).
Specifically, the data reorganization method is that all historical data are stored in a data table in a manner that the acquisition time is used as a row label and the data category is used as a column label. Then, for each time node, the data table contents of the energy consumption sequence in two days and 48 hours are taken and stored as separate samples, and the sample form is shown in fig. 2. And processing all the nodes according to the time sequence to obtain a data sample set for training the model.
In particular, the particular network structure used by the model is built, as shown in FIG. 3. The size of an input layer is consistent with that of an input data sample, a Convolutional Neural Network (CNN) is used as a feature extraction layer, a structure that a one-dimensional convolutional layer is convoluted in a feature direction is adopted, and 3 layers of convolutional layers are arranged to improve the feature extraction capability; after CNN, using long short-term memory Layer (LSTM) to realize modeling of time sequence, setting 3 layers of long short-term memory layer; and finally, realizing nonlinear learning of building operation by using a 3-layer full-connection layer network, wherein the size of an output layer is 1, and outputting the energy consumption predicted value of the next moment point at a time.
Specifically, the network structure needs to adopt specific parameters suitable for building energy consumption prediction, including neuron number and activation function. The number of neurons is obtained through case data testing, the number of CNN neurons is reduced by half layer by layer, the number of LSTM neurons is 10-100, and the number of neurons in a full connecting layer is 10-100. In this example, the CNN first layer is 128, the LSTM first layer is 64, the fully connected layer is 20, and then the layer by layer is halved. The ReLU function is chosen as the activation function:
in the formula: y isoutIs the activation function output; x is the number ofinIs the activation function input.
Based on the method, a building energy consumption prediction system based on the CNN-LSTM is formed, and a person skilled in the art can understand the building energy consumption prediction method based on the CNN-LSTM as a preferred example of the building energy consumption prediction system based on the CNN-LSTM.
Specifically, the method comprises the following modules:
module M1: the energy consumption monitoring module is used for measuring the building energy consumption data and the influence factor data in real time and storing the data according to the acquisition time point;
module M2: the data processing module is used for preprocessing the data, recombining the data and storing the recombined sample data into a computer for calling by the model training module M3;
module M3: the algorithm modeling module is used for realizing the model structure through computer language by computer programming to form a model to be trained; then calling a data sample training model processed by an M2 module, and storing the model after the model is iterated and stabilized;
module M4: and the model predicted value output module is used for obtaining the predicted value of the building energy consumption through the model obtained by the module M3 after the influence factor data of the predicted time period is processed by the data module M2.
According to the method and the system for predicting the building energy consumption, disclosed by the invention, a deep learning algorithm is used for modeling and predicting the building energy consumption, and aiming at the adaptation of the building energy consumption data characteristics and the algorithm structure to the energy consumption prediction problem, a method and a system capable of predicting the building energy consumption with high precision are provided, so that a reference model is provided for the energy-saving optimization of a building energy system. The prediction accuracy of the data model on the building energy consumption is effectively improved by respectively utilizing the learning of the CNN on the data local relevance and the learning of the LSTM on the sequence data change rule.
In a specific embodiment, the method is implemented on the premise of the technical scheme of the invention, and a detailed implementation mode and a specific operation process are given and specific applicable parameters are given by depending on historical data and an open source Python library of a certain building in the Shanghai region. It should be noted that, for a person skilled in the art, several variations and modifications can be made without departing from the inventive concept, which falls within the scope of the present invention.
Those skilled in the art will appreciate that, in addition to implementing the systems, apparatus, and various modules thereof provided by the present invention in purely computer readable program code, the same procedures can be implemented entirely by logically programming method steps such that the systems, apparatus, and various modules thereof are provided in the form of logic gates, switches, application specific integrated circuits, programmable logic controllers, embedded microcontrollers and the like. Therefore, the system, the device and the modules thereof provided by the present invention can be considered as a hardware component, and the modules included in the system, the device and the modules thereof for implementing various programs can also be considered as structures in the hardware component; modules for performing various functions may also be considered to be both software programs for performing the methods and structures within hardware components.
The foregoing description of specific embodiments of the present invention has been presented. It is to be understood that the present invention is not limited to the specific embodiments described above, and that various changes or modifications may be made by one skilled in the art within the scope of the appended claims without departing from the spirit of the invention. The embodiments and features of the embodiments of the present application may be combined with each other arbitrarily without conflict.
Claims (7)
1. A building energy consumption prediction method based on CNN-LSTM is characterized by comprising the following steps:
step S1: acquiring historical energy consumption sequence data of a building by using a building energy consumption monitoring module, and synchronously acquiring influence factor data;
step S2: preprocessing historical sequence data, wherein the data preprocessing comprises abnormal data cleaning, linear interpolation completion and data normalization, and the historical sequence data comprises historical energy consumption data of the whole year of the building; then, recombining the data into a two-dimensional data table format according to two dimensions of time sequence and characteristics, wherein the two-dimensional data table format is used as a single sample of data for training, and the time sequence is data acquisition time points arranged according to a time step;
step S3: using a Convolutional Neural Network (CNN) as a feature extraction layer, using a long short-term memory Layer (LSTM) after the CNN to realize modeling of the time sequence, and finally using a full-connection layer network to establish a building energy consumption prediction model by using the network structure;
step S4: training a model in a server by using the recombined historical energy consumption data set, and storing the model obtained by training by a model structure weighted value method, wherein the server is a computer resource for a prediction system to call, and the weighted value is the weight of network neuron connection;
step S5: and recombining the new influence factor data, inputting the recombined data into a model, dividing the output of the model by the scaling ratio during normalization, and calculating to obtain a building energy consumption predicted value.
2. The CNN-LSTM based building energy consumption prediction method of claim 1, wherein the influencing factor data includes three types: outdoor weather parameters, building operation rules and building internal use conditions.
3. The CNN-LSTM-based building energy consumption prediction method of claim 1, wherein the data is organized into input samples of a two-dimensional table by the data reorganization method.
4. The CNN-LSTM-based building energy consumption prediction method of claim 3, wherein the data table is organized according to two dimensions of time and characteristic, the time comprises data arranged in two adjacent days of history according to step, and the characteristic dimension arranges data of all influencing factors.
5. The building energy consumption prediction method based on CNN-LSTM according to claim 1, characterized in that, establishing a specific network structure used by the model, using Convolutional Neural Network (CNN) as a feature extraction layer, adopting a structure that one-dimensional convolutional layers are convolved in the feature direction, and setting 3 layers of convolutional layers to improve the feature extraction capability; after CNN, using long short-term memory Layer (LSTM) to realize modeling of time sequence, setting 3 layers of long short-term memory layer; and finally, learning the nonlinearity of the building operation by using a 3-layer full-connection layer network.
6. The CNN-LSTM-based building energy consumption prediction method of claim 5, wherein the network structure comprises parameters suitable for building energy consumption prediction, such as neuron number and activation function selection.
7. A CNN-LSTM-based building energy consumption prediction system is characterized by comprising:
module M1: the energy consumption monitoring module is used for measuring the building energy consumption data and the influence factor data in real time and storing the data according to the acquisition time point;
module M2: the data processing module is used for preprocessing the data, recombining the data and storing the recombined sample data into a computer for calling by the model training module M3;
module M3: the algorithm modeling module is used for realizing the model structure through computer language by computer programming to form a model to be trained; then calling a data sample training model processed by an M2 module, and storing the model after the model is iterated and stabilized;
module M4: and the model predicted value output module is used for obtaining the predicted value of the building energy consumption through the model obtained by the module M3 after the influence factor data of the predicted time period is processed by the data module M2.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110335276.5A CN113205203A (en) | 2021-03-29 | 2021-03-29 | CNN-LSTM-based building energy consumption prediction method and system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110335276.5A CN113205203A (en) | 2021-03-29 | 2021-03-29 | CNN-LSTM-based building energy consumption prediction method and system |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN113205203A true CN113205203A (en) | 2021-08-03 |
Family
ID=77025791
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110335276.5A Pending CN113205203A (en) | 2021-03-29 | 2021-03-29 | CNN-LSTM-based building energy consumption prediction method and system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113205203A (en) |
Cited By (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113554361A (en) * | 2021-09-22 | 2021-10-26 | 广东电网有限责任公司东莞供电局 | Comprehensive energy system data processing and calculating method and processing system |
| CN113786205A (en) * | 2021-11-17 | 2021-12-14 | 华南理工大学 | Portable early warning system and method based on convolutional network and long-short term memory network |
| CN113835341A (en) * | 2021-09-18 | 2021-12-24 | 中邮科通信技术股份有限公司 | Energy consumption analysis and diagnosis method based on intelligent building AI |
| CN113837488A (en) * | 2021-10-14 | 2021-12-24 | 南京国铁电气有限责任公司 | Energy consumption data prediction method, system and equipment |
| CN114237049A (en) * | 2021-12-14 | 2022-03-25 | 西安建筑科技大学 | Intelligent building system prediction control parameter setting method based on LSTM |
| CN114648152A (en) * | 2022-01-28 | 2022-06-21 | 江苏鼋博群智能技术有限公司 | Building energy consumption prediction method and system based on state constraint and time-frequency characteristics |
| CN114693064A (en) * | 2022-02-25 | 2022-07-01 | 山东建筑大学 | Building group scheme generation performance evaluation method |
| CN116227899A (en) * | 2023-05-09 | 2023-06-06 | 深圳市明源云科技有限公司 | Cell house energy consumption prediction method and device, electronic equipment and readable storage medium |
| CN116734918A (en) * | 2023-06-05 | 2023-09-12 | 宁夏中昊银晨能源技术服务有限公司 | Indoor environment monitoring system suitable for near zero energy consumption building |
-
2021
- 2021-03-29 CN CN202110335276.5A patent/CN113205203A/en active Pending
Cited By (14)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113835341A (en) * | 2021-09-18 | 2021-12-24 | 中邮科通信技术股份有限公司 | Energy consumption analysis and diagnosis method based on intelligent building AI |
| CN113835341B (en) * | 2021-09-18 | 2024-05-17 | 中邮科通信技术股份有限公司 | Intelligent building AI-based energy consumption analysis and diagnosis method |
| CN113554361A (en) * | 2021-09-22 | 2021-10-26 | 广东电网有限责任公司东莞供电局 | Comprehensive energy system data processing and calculating method and processing system |
| CN113837488B (en) * | 2021-10-14 | 2023-12-26 | 南京国铁电气有限责任公司 | Method, system and equipment for predicting energy consumption data |
| CN113837488A (en) * | 2021-10-14 | 2021-12-24 | 南京国铁电气有限责任公司 | Energy consumption data prediction method, system and equipment |
| CN113786205A (en) * | 2021-11-17 | 2021-12-14 | 华南理工大学 | Portable early warning system and method based on convolutional network and long-short term memory network |
| CN114237049A (en) * | 2021-12-14 | 2022-03-25 | 西安建筑科技大学 | Intelligent building system prediction control parameter setting method based on LSTM |
| CN114648152A (en) * | 2022-01-28 | 2022-06-21 | 江苏鼋博群智能技术有限公司 | Building energy consumption prediction method and system based on state constraint and time-frequency characteristics |
| CN114648152B (en) * | 2022-01-28 | 2024-02-02 | 江苏鼋博群智能技术有限公司 | Building energy consumption prediction method and system based on state constraint and time-frequency characteristics |
| CN114693064A (en) * | 2022-02-25 | 2022-07-01 | 山东建筑大学 | Building group scheme generation performance evaluation method |
| CN114693064B (en) * | 2022-02-25 | 2023-03-10 | 山东建筑大学 | Building group scheme generation performance evaluation method |
| CN116227899A (en) * | 2023-05-09 | 2023-06-06 | 深圳市明源云科技有限公司 | Cell house energy consumption prediction method and device, electronic equipment and readable storage medium |
| CN116734918A (en) * | 2023-06-05 | 2023-09-12 | 宁夏中昊银晨能源技术服务有限公司 | Indoor environment monitoring system suitable for near zero energy consumption building |
| CN116734918B (en) * | 2023-06-05 | 2024-05-14 | 宁夏中昊银晨能源技术服务有限公司 | Indoor environment monitoring system suitable for near zero energy consumption building |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN113205203A (en) | CNN-LSTM-based building energy consumption prediction method and system | |
| CN109376960A (en) | Load Forecasting based on LSTM neural network | |
| CN110046743A (en) | Energy Consumption of Public Buildings prediction technique and system based on GA-ANN | |
| CN109242049B (en) | Water supply pipe network multipoint leakage positioning method and device based on convolutional neural network | |
| CN107315884A (en) | A kind of building energy consumption modeling method based on linear regression | |
| CN111460728A (en) | Method and device for predicting residual life of industrial equipment, storage medium and equipment | |
| CN112651665A (en) | Surface water quality index prediction method and device based on graph neural network | |
| CN111008726B (en) | Class picture conversion method in power load prediction | |
| CN112381673B (en) | Park electricity utilization information analysis method and device based on digital twin | |
| CN111882157A (en) | Demand prediction method and system based on deep space-time neural network and computer readable storage medium | |
| CN110852496A (en) | Natural gas load prediction method based on LSTM recurrent neural network | |
| CN114862032B (en) | XGBoost-LSTM-based power grid load prediction method and device | |
| CN115204491A (en) | Production line working condition prediction method and system based on digital twinning and LSTM | |
| CN115936546A (en) | Assembly type building green construction management method and device based on digital twinning | |
| CN113449919A (en) | Power consumption prediction method and system based on feature and trend perception | |
| CN116579447A (en) | Time sequence prediction method based on decomposition mechanism and attention mechanism | |
| CN116706992A (en) | Self-adaptive power prediction method, device and equipment for distributed photovoltaic cluster | |
| CN116523187A (en) | Engineering progress monitoring method and system based on BIM | |
| CN113849052B (en) | Machine room temperature prediction method and system based on artificial intelligence | |
| CN114692981A (en) | Medium-and-long-term runoff forecasting method and system based on Seq2Seq model | |
| Wang et al. | A transformer-based multi-entity load forecasting method for integrated energy systems | |
| CN114091776A (en) | K-means-based multi-branch AGCNN short-term power load prediction method | |
| CN113902181A (en) | Short-term prediction method and equipment for common variable heavy overload | |
| CN113111419A (en) | Method and system for establishing and predicting air-conditioning load prediction model in office building | |
| CN108520311B (en) | Haze prediction model establishing method and system combining SOFM (software on a programmable) network and BP (back propagation) neural network |
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 |