CN117495119A - Intelligent prediction method for carbon emission in building operation stage of neural network optimization algorithm - Google Patents
Intelligent prediction method for carbon emission in building operation stage of neural network optimization algorithm Download PDFInfo
- Publication number
- CN117495119A CN117495119A CN202311306117.8A CN202311306117A CN117495119A CN 117495119 A CN117495119 A CN 117495119A CN 202311306117 A CN202311306117 A CN 202311306117A CN 117495119 A CN117495119 A CN 117495119A
- Authority
- CN
- China
- Prior art keywords
- carbon emission
- building
- data
- neural network
- module
- 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
- 229910052799 carbon Inorganic materials 0.000 title claims abstract description 232
- OKTJSMMVPCPJKN-UHFFFAOYSA-N Carbon Chemical compound [C] OKTJSMMVPCPJKN-UHFFFAOYSA-N 0.000 title claims abstract description 226
- 238000013528 artificial neural network Methods 0.000 title claims abstract description 72
- 238000005457 optimization Methods 0.000 title claims abstract description 62
- 238000004422 calculation algorithm Methods 0.000 title claims abstract description 59
- 238000000034 method Methods 0.000 title claims abstract description 23
- 238000005265 energy consumption Methods 0.000 claims abstract description 30
- 239000000463 material Substances 0.000 claims abstract description 15
- 238000012544 monitoring process Methods 0.000 claims abstract description 13
- 230000000694 effects Effects 0.000 claims abstract description 12
- 238000012545 processing Methods 0.000 claims abstract description 5
- 230000006870 function Effects 0.000 claims description 36
- 238000004364 calculation method Methods 0.000 claims description 24
- 238000010276 construction Methods 0.000 claims description 22
- 238000003062 neural network model Methods 0.000 claims description 11
- 239000011159 matrix material Substances 0.000 claims description 9
- 238000013480 data collection Methods 0.000 claims description 6
- 238000009826 distribution Methods 0.000 claims description 6
- 210000002569 neuron Anatomy 0.000 claims description 6
- 238000013486 operation strategy Methods 0.000 claims description 6
- 238000007781 pre-processing Methods 0.000 claims description 6
- 238000011160 research Methods 0.000 claims description 6
- 238000012216 screening Methods 0.000 claims description 6
- 238000007726 management method Methods 0.000 claims description 5
- 238000012549 training Methods 0.000 claims description 5
- 230000004913 activation Effects 0.000 claims description 4
- 238000004378 air conditioning Methods 0.000 claims description 4
- 238000013523 data management Methods 0.000 claims description 4
- 238000013500 data storage Methods 0.000 claims description 4
- 230000006855 networking Effects 0.000 claims description 4
- 241000254032 Acrididae Species 0.000 claims description 3
- 238000012423 maintenance Methods 0.000 claims description 3
- 230000009467 reduction Effects 0.000 claims description 3
- 238000012552 review Methods 0.000 claims description 3
- ORILYTVJVMAKLC-UHFFFAOYSA-N Adamantane Natural products C1C(C2)CC3CC1CC2C3 ORILYTVJVMAKLC-UHFFFAOYSA-N 0.000 claims description 2
- 238000004458 analytical method Methods 0.000 claims description 2
- 238000004883 computer application Methods 0.000 claims description 2
- 230000010354 integration Effects 0.000 claims description 2
- 230000008439 repair process Effects 0.000 claims description 2
- 238000009423 ventilation Methods 0.000 claims description 2
- 238000012800 visualization Methods 0.000 claims description 2
- 239000008186 active pharmaceutical agent Substances 0.000 description 1
- 238000013475 authorization Methods 0.000 description 1
- 238000013079 data visualisation Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000001537 neural effect Effects 0.000 description 1
- 230000008569 process Effects 0.000 description 1
- 238000011084 recovery 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
- G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
- G06Q10/063—Operations research, analysis or management
- G06Q10/0637—Strategic management or analysis, e.g. setting a goal or target of an organisation; Planning actions based on goals; Analysis or evaluation of effectiveness of goals
- G06Q10/06375—Prediction of business process outcome or impact based on a proposed change
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
- G06F16/26—Visual data mining; Browsing structured data
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
-
- 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
- G06N3/08—Learning methods
- G06N3/082—Learning methods modifying the architecture, e.g. adding, deleting or silencing nodes or connections
-
- 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
- G06Q50/10—Services
- G06Q50/26—Government or public services
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T17/00—Three dimensional [3D] modelling, e.g. data description of 3D objects
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Business, Economics & Management (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Human Resources & Organizations (AREA)
- Data Mining & Analysis (AREA)
- General Engineering & Computer Science (AREA)
- Strategic Management (AREA)
- Artificial Intelligence (AREA)
- General Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- Tourism & Hospitality (AREA)
- Software Systems (AREA)
- Evolutionary Computation (AREA)
- Economics (AREA)
- Educational Administration (AREA)
- Health & Medical Sciences (AREA)
- Mathematical Physics (AREA)
- Molecular Biology (AREA)
- General Business, Economics & Management (AREA)
- Computing Systems (AREA)
- Marketing (AREA)
- Computational Linguistics (AREA)
- Biophysics (AREA)
- Biomedical Technology (AREA)
- Databases & Information Systems (AREA)
- Entrepreneurship & Innovation (AREA)
- Development Economics (AREA)
- Operations Research (AREA)
- Computer Graphics (AREA)
- Geometry (AREA)
- Evolutionary Biology (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Bioinformatics & Computational Biology (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Primary Health Care (AREA)
- Quality & Reliability (AREA)
- Game Theory and Decision Science (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention relates to an intelligent prediction method, in particular to an intelligent prediction method for carbon emission in a building operation stage of a neural network optimization algorithm. An in-building IoT platform module is built for collecting and processing various data related to building carbon emissions and categorizing according to the following five broad categories: (1) building energy consumption, (2) material consumption, (3) meteorological data, (4) building activity conditions, and (5) renewable energy sources; recording and monitoring various data; and calculating the related data of the energy consumption in the 5-class data to obtain the corresponding carbon emission. The system can predict and calculate daily carbon emission of the building based on a neural network optimization algorithm, so that strategy suggestions for reducing the carbon emission are provided for building operators.
Description
Technical Field
The invention relates to an intelligent prediction method, in particular to an intelligent prediction method for carbon emission in a building operation stage of a neural network optimization algorithm.
Background
With the increased global concern for carbon emissions, the building industry is also challenged to reduce carbon emissions, with carbon emissions up to 60% of the total amount in building operation over the entire life cycle of the building, and with different building types (e.g., public buildings such as offices, exhibitions, malls, education, etc. and various residential buildings) having different sources of carbon emissions in operation. The carbon emission during the operation of the building is calculated using a single model, and the calculated result may bring about a certain deviation, which affects the carbon reduction strategy adopted by the building operator.
Meanwhile, the carbon emission amount in the building operation stage is not constant, and some external factors such as climate, building people flow, energy price fluctuation in operation and the like can influence the final carbon emission amount, so that a dynamic and intelligent carbon emission prediction system in the building operation stage needs to be designed.
Disclosure of Invention
The invention mainly solves the defects existing in the prior art, and provides the intelligent prediction method for the carbon emission in the building operation stage of the neural network optimization algorithm, which can predict and calculate the daily carbon emission of the building based on the neural network optimization algorithm, thereby providing the strategy proposal for reducing the carbon emission for a building operator.
The technical problems of the invention are mainly solved by the following technical proposal:
the intelligent prediction method for the carbon emission in the building operation stage of the neural network optimization algorithm is implemented by a carbon emission intelligent prediction system according to the following steps:
step (1): an in-building IoT platform module is built for collecting and processing various data related to building carbon emissions and categorizing according to the following five broad categories: (1) building energy consumption, (2) material consumption, (3) meteorological data, (4) building activity conditions, and (5) renewable energy sources; recording and monitoring various data; calculating the related data of the energy consumption in the 5 types of data to obtain the corresponding carbon emission:
obtaining various energy consumption of building operation within a period of days, and calculating the energy consumption as carbon emission;
obtaining various material losses of building operation within a period of days, and calculating the material carbon emission as carbon emission according to a material carbon emission factor;
acquiring meteorological data of building operation within a period of days;
acquiring the activity condition of the building during running within a period of days;
acquiring energy generated by building auxiliary renewable energy in a period of days, and calculating the energy as carbon offset;
step (2): according to the requirements of building operators, formulating the total carbon emission of the annual building operation and the average daily carbon emission of the building operation;
step (3): establishing a carbon emission calculation module, screening platform data, and calculating carbon emission;
step (4): building a neural network module, processing input data, and predicting daily carbon emission;
step (5): an optimization algorithm module is built, parameter weights in the neural network are adjusted by using the optimization algorithm, prediction errors are optimized, and an optimization result is fed back to the neural network module;
step (6): generating a building operation carbon emission report and a strategy suggestion based on the predicted carbon emission result;
step (7): and updating each weight of the neural network according to the carbon emission report and the strategy suggestion, and performing directional model optimization.
Preferably, the method comprises the following steps:
step 001: building an Internet of things (IoT) platform module, and collecting data of an operation stage through sensors, monitoring instruments and an IoT platform in a building, wherein the data comprise five data of (1) building energy consumption, (2) material consumption, (3) meteorological data, (4) building activity conditions and (5) renewable energy sources;
step 002: establishing a carbon emission calculation module, screening data directly related to the carbon emission amount of building operation in 5 major types of data, and taking other types of data as input characteristic data in a weight matrix;
step 003: building a neural network model, and calculating the actual daily carbon emission C of building operation according to the stored data in the system act ；
And (3) constructing a computing system:
step S001, collecting and classifying data from the building IoT platform module, and importing valid data into the carbon emission calculation module;
step S002, calculating the carbon emission amount in the construction operation stage, and leading the calculated carbon emission amount and effective data into a neural network module, wherein the neural network module comprises an input layer, a hidden layer and an output layer;
step S003, various carbon emission amounts are processed in the input layer;
input layer x= [ X ] 1 ，x 2 ，...x i ]Wherein x is i Representing the carbon emission of the corresponding large class in the construction operation;
step S004, for a given input carbon emission X, forward direction is performed through a neural networkPropagating, calculating output value of each layer, and calculating predicted daily carbon emission C of output layer pre ：
H 1 ＝ReLU(X×w 1 +b 1 )
H 2 ＝ReLU(H 1 ×W 2 +b 2 )
C pre ＝ReLU(H 2 ×w 3 +b 3 )
Wherein W is 1 ，W 2 ，W 3 Is a weight matrix input to the hidden layer of the neural network, b 1 ，b 2 ，b 3 Is a bias vector; weight matrixWhere i is the input characteristic data class number, j is the neuron number, each weight w i,j Representing the connection strength between the i-th input characteristic data and the j-th hidden layer neuron;
step S005, outputting a predicted carbon emission value, and simultaneously outputting weight and bias variables in the neural network module to an optimization algorithm module;
step 004, an optimization algorithm module is built, the parameter weight in the neural network is adjusted by using an optimization algorithm, the prediction error is optimized, and the loss is calculated by using a mean square error function;
step S006, according to the loss functionLoss optimization is performed, where n is the number of days of recorded data, C act,i Is the actual daily carbon emission on day i, C pre，i Is the predicted daily carbon emission on day i;
step S007, in calculating the predicted value C for each day pre In this case, the weights and bias values in the neural network module 003 at the next iteration are updated by the back propagation algorithm, where α is the learning rate as follows:
calculating the gradient of the output layer: delta out ＝(C act -C pre )×ReLU′(C act )；
Calculating a second hidden layer gradient: delta hid2 ＝W z δ out ×ReLU′(H 2 )；
Calculating a first hidden layer gradient: delta hid1＝ w 1 δ hid2 ×ReLU′(H 1 )；
Updating the weights of the second hidden layer to the output layer: Δw 3 ＝-α×δ out ×H 2 ；
Updating weights of the first hidden layer to the second hidden layer: Δw 2 ＝-α×δ hid2 ×H 1 ；
Updating the weights of the input layer to the first hidden layer: Δw 1 ＝-α×δ hid1 ×X；
Updating the bias vector of the second hidden layer to the output layer: Δb 3 ＝-α×δ out ；
Updating the bias vector of the first hidden layer to the second hidden layer: Δb 2 ＝-α×δ hid2 ；
Updating the bias vector of the input layer to the first hidden layer: Δb 1 ＝-α×δ hid1 ；
Step S008, according to the special requirement of the back end of the operator on carbon emission control, directivity adjustment is input to the neural network module;
step 005: based on the algorithm, more accurate predicted daily carbon emission is obtained, and a building operation carbon emission report and a strategy suggestion are generated:
the system may visualize the carbon emission report through the building IoT platform for viewing by building operators and users in the in-building control panel;
the system can upload the carbon emission report to the cloud after networking, so that a building operator and a user can remotely review at the mobile terminal;
the system can set energy consumption thresholds when various systems in the building run according to the predicted carbon emission, and after the energy consumption exceeds the thresholds, the system can send early warning messages to building operators to adjust the running conditions of the systems in the building in time;
step 006: according to the carbon emission report and the strategy proposal, the building operator can directionally adjust and update each parameter of the neural network, and further optimize the prediction system:
according to the operation strategy suggestion generated by the system, a building operator can input the operation strategy into various energy consumption systems in advance, and monitor the operation working conditions of the various energy consumption systems in real time by utilizing the system, so that the carbon emission of the building operation stage is reduced;
according to the carbon emission report generated by the system, annual carbon emission in a unit building area is calculated, building operation conditions are evaluated, and data support is provided for subsequent low-carbon building and other authentications;
the data generated by the system carbon emission report can be exported to form a CSV format file, and most of the existing data analysis software can be imported for further research by scientific research institutions;
and the data generated by the system carbon emission report can be displayed in the Rhino software through a Grasshopper platform to form a three-dimensional carbon emission distribution model and a two-dimensional carbon emission distribution model, and a DWG format file can be derived and formed.
The in-building IoT platform module has a visualization function that can project carbon emission data to a presentation interface.
The in-building IoT platform module has a user feedback system that allows in-building users to provide feedback regarding the indoor environment. And the operation of the equipment is adjusted according to the feedback of the user, so that the comfort requirement is met, and the carbon emission target is achieved.
The in-building IoT platform module has a device efficiency reporting function that can analyze the energy efficiency of various building devices and provide reports to operators and advice for device upgrades or replacement, improving energy efficiency and reducing carbon emissions.
The carbon emission calculation module has an energy source analysis function, and can select building energy sources including solar energy, wind energy, electric energy and the like.
The carbon emission calculation module has a device priority setting function, can select different building devices to set priorities, and when carbon emission needs to be reduced, the system can be adjusted according to the priorities of the devices. Different building equipment refers to air conditioning, lighting, elevators, etc.
The neural network module has a network structure selection function, can be adjusted according to the data in the IoT platform module, and selects to use a shallow network or a deep network.
The neural network module has an activation function selection function, can be adjusted according to the data classification in the IoT platform module, and selects activation functions such as ReLU, sigmoid and the like to use.
The neural network module has a back propagation algorithm function, and can update weight parameters in the iterative neural network according to data such as actual daily carbon emission of building operation.
The optimization algorithm module has an algorithm selection function, and can select and use an optimization algorithm such as SGD, momentum, adam according to the optimization accuracy requirement.
The optimization algorithm module has a learning rate adjusting function and can select a fixed learning rate or a dynamic learning rate according to the optimization depth requirement.
The intelligent carbon emission prediction system has the authority of being integrated with other systems, and is convenient to access to a regional carbon emission monitoring system.
An in-building IoT platform module may monitor and record building energy consumption over a period of time, including energy consumption from air conditioning, lighting, ventilation, electrical outlets, elevators, others, and the like.
An in-building IoT platform module may monitor and record material consumption in building operations over a period of time, including material consumption resulting from in-building maintenance repairs, in-building use component replacement, in-building secondary finishing, and the like.
An in-building IoT platform module may monitor and record meteorological data over a period of time, including air temperature, relative humidity, solar illuminance, wind speed, and the like.
An in-building IoT platform module may monitor and record building activity conditions over a period of time, including instantaneous personnel density, instantaneous personnel occupancy, instantaneous indoor personnel activity intensity, and the like.
The carbon emission calculation module can dynamically count and calculate the classification and the item of the carbon emission and the renewable energy carbon reduction in the construction operation stage.
And the carbon emission calculation module can increase the purchasing function of the carbon sink and the external carbon sink of the building green land according to the later-period requirement of a building operator. And (5) carrying out dynamic statistics and calculation on classification items of building carbon offset.
The neural network module can calculate various carbon emission amounts in the construction operation stage based on a neural network calculation method and forecast an internal carbon emission value in a future time period.
The neural network module can adjust the depth layer number of the neural network according to the conditions of building area, energy system scale in the building, building complexity and the like so as to meet the calculation efficiency of the system.
And the optimization algorithm module can be used for feeding back various data and actual carbon emission based on the optimization algorithm in the construction operation stage, so that the accuracy of the predicted value of the internal carbon emission in the future time period is optimized.
The intelligent carbon emission prediction system comprises a database of carbon emission related data, and can independently store the carbon emission related data information for a long time.
The intelligent carbon emission prediction system can be written based on Python, matlab, C ++ and other computer open source languages, and is convenient for updating and maintaining in later period.
The intelligent carbon emission prediction system can build an algorithm model by adopting an open source framework such as TensorFlow, pyTorch, matlab and the like, and is convenient for the creation and training of the model.
The intelligent carbon emission prediction system can divide areas of a large building complex with multiple purposes and properties according to conditions such as building functions, room purposes and the like, classify and generalize rooms with similar functions, monitor and predict carbon emission, and further improve prediction accuracy.
The intelligent carbon emission prediction system can perform inquiry, early warning, recording and downloading of carbon emission data in real time and produce a building carbon emission report.
The intelligent carbon emission prediction system can provide computer operation software by combining with computer application, and comprises a data collection and preprocessing module, a neural network model training module, an optimization algorithm module, a real-time monitoring and feedback module, a data storage and management module, a user interface module, an API (application program interface) and integration module and a security and authority management module
The intelligent carbon emission prediction system has data collection and preprocessing module capable of collecting data from IoT devices and other data sources, screening, normalizing and converting the data, and providing input data in proper format for the neural network model.
The data collection and preprocessing module of the carbon emission intelligent prediction system can add more data sources such as weather forecast, energy price and the like according to requirements so as to provide more accurate supporting data.
A neural network model training module trains a neural network model using historical data of building operations to calculate and predict carbon emissions.
The carbon emission intelligent prediction system is provided with an optimization algorithm module, can optimize based on the calculation result of the neural network, and provides a self-optimized building operation stage carbon emission prediction calculation method.
The intelligent carbon emission prediction system is provided with an actual monitoring and feedback module, can display current carbon emission data, prediction results and optimization suggestions in real time, has an alarm function, and automatically sends a notification when the carbon emission exceeds a threshold value.
The intelligent carbon emission prediction system is an actual monitoring and feedback module, and can provide optimization suggestions for building operation, such as adjusting the operation time of an air conditioning system, changing a lighting strategy and the like.
The intelligent carbon emission prediction system comprises a data storage and management module, wherein the data storage and management module can store historical data, model parameters and optimization results and support data query and report generation. Data backup and recovery functions can be added, and data safety is ensured.
The carbon emission intelligent prediction system has a user interface module with a friendly operation interface, so that a user can easily view prediction results, optimization suggestions and other relevant information. And can add multi-language support, custom dashboards and data visualization tools.
And the API and the integrated module of the carbon emission intelligent prediction system can provide APIs for the system, so that software and other building management systems, energy management systems and the like are integrated for use.
The carbon emission intelligent prediction system has the advantages that the safety and authority management module can be added with other advanced safety functions such as multistage authentication, data encryption and the like, ensures the safety of data, and provides the functions of user authentication and authorization.
Therefore, according to the intelligent prediction method for the carbon emission in the building operation stage based on the neural network optimization algorithm, according to the operation strategy suggestion generated by the system, a building operator can input the operation strategy into various energy consumption systems in advance, and the operation conditions of the various energy consumption systems are monitored in real time by utilizing the system, so that the carbon emission in the building operation stage is reduced; according to the carbon emission report generated by the system, annual carbon emission in unit building area is calculated, and building operation conditions are evaluated; data support is provided for subsequent authentication of low-carbon buildings and the like; the data generated by the system carbon emission report can be exported to form a CSV format file, and most of the existing data analysis software can be imported for further research by scientific research institutions; and the data generated by the system carbon emission report can be displayed in the Rhino software through a Grasshopper platform to form a three-dimensional carbon emission distribution model and a two-dimensional carbon emission distribution model, and a DWG format file can be derived and formed.
Drawings
FIG. 1 is a flow chart of the present invention;
FIG. 2 is a flow chart of the internal calculation of the prediction operation software according to the present invention.
Detailed Description
The technical scheme of the invention is further specifically described below through examples and with reference to the accompanying drawings.
Example 1: as shown in the figure, the intelligent prediction method for the carbon emission in the building operation stage of the neural network optimization algorithm is implemented by a carbon emission intelligent prediction system according to the following steps:
step 001: building an Internet of things (IoT) platform module, and collecting data of an operation stage through sensors, monitoring instruments and an IoT platform in a building, wherein the data comprise five data of (1) building energy consumption, (2) material consumption, (3) meteorological data, (4) building activity conditions and (5) renewable energy sources;
step 002: establishing a carbon emission calculation module, screening data directly related to the carbon emission amount of building operation in 5 major types of data, and taking other types of data as input characteristic data in a weight matrix;
step 003: building a neural network model, and calculating the actual daily carbon emission C of building operation according to the stored data in the system act ；
And (3) constructing a computing system:
step S001, collecting and classifying data from the building IoT platform module, and importing valid data into the carbon emission calculation module;
step S002, calculating the carbon emission amount in the construction operation stage, and leading the calculated carbon emission amount and effective data into a neural network module, wherein the neural network module comprises an input layer, a hidden layer and an output layer;
step S003, various carbon emission amounts are processed in the input layer;
input layer x= [ X ] 1 ，x 2 ，...x i ]Wherein x is i Representing the carbon emission of the corresponding large class in the construction operation;
step S004, for a given input carbon emission X, forward propagation is performed through a neural network, output values of all layers are calculated, and the predicted daily carbon emission C of the output layer can be calculated pre ：
H 1 ＝ReLU(X×W 1 +b 1 )
H 2 ＝ReLU(H 1 ×W 2 +b 2 )
C pre ＝ReLU(H 2 ×w 3 +b 3 )
Wherein W is 1 ，W 2 ，W 3 Is a weight matrix input to the hidden layer of the neural network, b 1 ，b 2 ，b 3 Is a bias vector; weight matrixWhere i is the input characteristic data class number, j is the neuron number, each weight w i,j Representing the connection strength between the i-th input characteristic data and the j-th hidden layer neuron;
step S005, outputting a predicted carbon emission value, and simultaneously outputting weight and bias variables in the neural network module to an optimization algorithm module;
step 004, an optimization algorithm module is built, the parameter weight in the neural network is adjusted by using an optimization algorithm, the prediction error is optimized, and the loss is calculated by using a mean square error function;
step S006, according to the loss functionLoss optimization is performed, where n is the number of days of recorded data, C act，i Is the actual daily carbon emission on day i, C pre，i Is the predicted daily carbon emission on day i;
step S007, in calculating the predicted value C for each day pre In this case, the weights and bias values in the neural network module 003 at the next iteration are updated by the back propagation algorithm, where α is the learning rate as follows:
calculating the gradient of the output layer: delta out ＝(C act -C pre )×ReLU′(C act )；
Calculating a second hidden layer gradient: delta hid2 ＝W z δ out ×ReLU′(H 2 )；
Calculating a first hidden layer gradient: delta hid1 ＝w 1 δ hid2 ×ReLU′(H 1 )；
Updating the weights of the second hidden layer to the output layer: Δw 3 ＝-α×δ out ×H 2 ；
Updating weights of the first hidden layer to the second hidden layer: Δw 2 ＝-α×δ hid2 ×H 1 ；
Updating the weights of the input layer to the first hidden layer: Δw 1 ＝-α×δ hid1 ×X；
Updating the bias vector of the second hidden layer to the output layer: Δb 3 ＝-α×δ out ；
Updating the bias vector of the first hidden layer to the second hidden layer: Δb 2 ＝-α×δ hid2 ；
Updating inputLayer-to-first hidden layer bias vector: Δb 1 ＝-α×δ hid1 ；
Step S008, according to the special requirement of the back end of the operator on carbon emission control, directivity adjustment is input to the neural network module;
step 005: based on the algorithm, more accurate predicted daily carbon emission is obtained, and a building operation carbon emission report and a strategy suggestion are generated:
the system may visualize the carbon emission report through the building IoT platform for viewing by building operators and users in the in-building control panel;
the system can upload the carbon emission report to the cloud after networking, so that a building operator and a user can remotely review at the mobile terminal;
the system can set energy consumption thresholds when various systems in the building run according to the predicted carbon emission, and after the energy consumption exceeds the thresholds, the system can send early warning messages to building operators to adjust the running conditions of the systems in the building in time;
step 006: according to the carbon emission report and the strategy proposal, the building operator can directionally adjust and update each parameter of the neural network, and further optimize the prediction system:
according to the technical scheme, all working flows are carried out and completed by the computer software, and the functions of monitoring, calculating, predicting and optimizing the carbon emission in the building operation stage can be realized by the computer software only by setting building total index data and networking the building internet of things (IoT) platform in the early stage by a building operator, so that the operator does not need to collect and process the data such as energy consumption, and the labor cost and the time cost are saved.
The above is only a specific embodiment of the present invention, but the scope of the present invention is not limited thereto, and it should be understood by those skilled in the art that the present invention includes but is not limited to the accompanying drawings and the description of the above specific embodiment. Any modifications which do not depart from the functional and structural principles of the present invention are intended to be included within the scope of the appended claims.
Claims (11)
1. The intelligent prediction method for the carbon emission in the building operation stage of the neural network optimization algorithm is characterized by comprising the following steps of:
step (1): an in-building IoT platform module is built for collecting and processing various data related to building carbon emissions and categorizing according to the following five broad categories: (1) building energy consumption, (2) material consumption, (3) meteorological data, (4) building activity conditions, and (5) renewable energy sources; recording and monitoring various data; calculating the related data of the energy consumption in the 5 types of data to obtain the corresponding carbon emission:
obtaining various energy consumption of building operation within a period of days, and calculating the energy consumption as carbon emission;
obtaining various material losses of building operation within a period of days, and calculating the material carbon emission as carbon emission according to a material carbon emission factor;
acquiring meteorological data of building operation within a period of days;
acquiring the activity condition of the building during running within a period of days;
acquiring energy generated by building auxiliary renewable energy in a period of days, and calculating the energy as carbon offset;
step (2): according to the requirements of building operators, formulating the total carbon emission of the annual building operation and the average daily carbon emission of the building operation;
step (3): establishing a carbon emission calculation module, screening platform data, and calculating carbon emission;
step (4): building a neural network module, processing input data, and predicting daily carbon emission;
step (5): an optimization algorithm module is built, parameter weights in the neural network are adjusted by using the optimization algorithm, prediction errors are optimized, and an optimization result is fed back to the neural network module;
step (6): generating a building operation carbon emission report and a strategy suggestion based on the predicted carbon emission result;
step (7): and updating each weight of the neural network according to the carbon emission report and the strategy suggestion, and performing directional model optimization.
2. The intelligent prediction method for carbon emission in the construction operation stage of the neural network optimization algorithm according to claim 1, which is characterized by comprising the following steps:
step 001: building an Internet of things (IoT) platform module, and collecting data of an operation stage through sensors, monitoring instruments and an IoT platform in a building, wherein the data comprise five data of (1) building energy consumption, (2) material consumption, (3) meteorological data, (4) building activity conditions and (5) renewable energy sources;
step 002: establishing a carbon emission calculation module, screening data directly related to the carbon emission amount of building operation in 5 major types of data, and taking other types of data as input characteristic data in a weight matrix;
step 003: building a neural network model, and calculating the actual daily carbon emission C of building operation according to the stored data in the system act ；
And (3) constructing a computing system:
step S001, collecting and classifying data from the building IoT platform module, and importing valid data into the carbon emission calculation module;
step S002, calculating the carbon emission amount in the construction operation stage, and leading the calculated carbon emission amount and effective data into a neural network module, wherein the neural network module comprises an input layer, a hidden layer and an output layer;
step S003, various carbon emission amounts are processed in the input layer;
input layer x= [ X ] 1 ，x 2 ，...x i ]Wherein x is i Representing the carbon emission of the corresponding large class in the construction operation;
step S004, for a given input carbon emission X, forward propagation is performed through a neural network, output values of all layers are calculated, and the predicted daily carbon emission C of the output layer can be calculated pre ：
H 1 ＝ReLU(X×W 1 +b 1 )
H 2 ＝ReLU(H 1 ×W 2 +b 2 )
C pre ＝ReLU(H 2 ×W 3 +b 3 )
Wherein W is 1 ，W 2 ，W 3 Is a weight matrix input to the hidden layer of the neural network, b 1 ，b 2 ，b 3 Is a bias vector; weight matrixWhere i is the input characteristic data class number, j is the neuron number, each weight w i,j Representing the connection strength between the i-th input characteristic data and the j-th hidden layer neuron;
step S005, outputting a predicted carbon emission value, and simultaneously outputting weight and bias variables in the neural network module to an optimization algorithm module;
step 004, an optimization algorithm module is built, the parameter weight in the neural network is adjusted by using an optimization algorithm, the prediction error is optimized, and the loss is calculated by using a mean square error function;
step S006, according to the loss functionLoss optimization is performed, where n is the number of days of recorded data, C act，i Is the actual daily carbon emission on day i, C pre，i Is the predicted daily carbon emission on day i;
step S007, in calculating the predicted value C for each day pre In this case, the weights and bias values in the neural network module 003 at the next iteration are updated by the back propagation algorithm, where α is the learning rate as follows:
calculating the gradient of the output layer: delta out ＝(C act -C pre )×ReLU′(C act )；
Calculating a second hidden layer gradient: delta hid2 ＝W z δ out ×ReLU′(H 2 )；
Calculating a first hidden layer gradient: delta hid1 ＝W 1 δ hid2 ×ReLU′(H 1 )；
Updating the weights of the second hidden layer to the output layer: ΔW (delta W) 3 ＝-α×δ out ×H 2 ；
Updating a first hidden layer to a second hidden layerWeight of the reservoir: ΔW (delta W) 2 ＝-α×δ hid2 ×H 1 ；
Updating the weights of the input layer to the first hidden layer: ΔW (delta W) 1 ＝-α×δ hid1 ×X；
Updating the bias vector of the second hidden layer to the output layer: Δb 3 ＝-α×δ out ；
Updating the bias vector of the first hidden layer to the second hidden layer: Δb 2 ＝-α×δ hid2 ；
Updating the bias vector of the input layer to the first hidden layer: Δb 1 ＝-α×δ hid1 ；
Step S008, according to the special requirement of the back end of the operator on carbon emission control, directivity adjustment is input to the neural network module;
step 005: based on the algorithm, more accurate predicted daily carbon emission is obtained, and a building operation carbon emission report and a strategy suggestion are generated:
the system may visualize the carbon emission report through the building IoT platform for viewing by building operators and users in the in-building control panel;
the system can upload the carbon emission report to the cloud after networking, so that a building operator and a user can remotely review at the mobile terminal;
the system can set energy consumption thresholds when various systems in the building run according to the predicted carbon emission, and after the energy consumption exceeds the thresholds, the system can send early warning messages to building operators to adjust the running conditions of the systems in the building in time;
step 006: according to the carbon emission report and the strategy proposal, the building operator can directionally adjust and update each parameter of the neural network, and further optimize the prediction system:
according to the operation strategy suggestion generated by the system, a building operator can input the operation strategy into various energy consumption systems in advance, and monitor the operation working conditions of the various energy consumption systems in real time by utilizing the system, so that the carbon emission of the building operation stage is reduced;
according to the carbon emission report generated by the system, annual carbon emission in a unit building area is calculated, building operation conditions are evaluated, and data support is provided for subsequent low-carbon building authentication;
the data generated by the system carbon emission report can be exported to form a CSV format file, and most of the existing data analysis software can be imported for further research by scientific research institutions;
and the data generated by the system carbon emission report can be displayed in the Rhino software through a Grasshopper platform to form a three-dimensional carbon emission distribution model and a two-dimensional carbon emission distribution model, and a DWG format file can be derived and formed.
3. The intelligent prediction method for carbon emission in the construction operation stage of the neural network optimization algorithm according to claim 2, wherein: the in-building IoT platform module has a visualization function, and can project carbon emission data to a display interface;
the in-building IoT platform module has a user feedback system that allows in-building users to provide feedback about the indoor environment; and according to the operation of the user feedback adjustment equipment, the comfort requirement is met, and the carbon emission target is achieved;
the in-building IoT platform module has a device efficiency reporting function that can analyze the energy efficiency of various building devices and provide reports to operators and advice for device upgrades or replacement, improving energy efficiency and reducing carbon emissions.
4. The intelligent prediction method for carbon emission in the construction operation stage of the neural network optimization algorithm according to claim 2, wherein:
the carbon emission calculation module has an energy source analysis function and can select building energy sources;
the carbon emission calculation module has a device priority setting function, can select different building devices to set priorities, and when carbon emission needs to be reduced, the system can be adjusted according to the priorities of the devices.
5. The intelligent prediction method for carbon emission in the construction operation stage of the neural network optimization algorithm according to claim 2, wherein:
the neural network module has a network structure selection function, can be adjusted according to the data in the IoT platform module, and selects to use a shallow network or a deep network;
the neural network module has an activation function selection function, can be adjusted according to the data classification in the IoT platform module, and selects and uses a ReLU and Sigmoid activation function;
the neural network module has a back propagation algorithm function, and can update weight parameters in the iterative neural network according to the relevant data of the actual daily carbon emission of the building operation.
6. The intelligent prediction method for carbon emission in the construction operation stage of the neural network optimization algorithm according to claim 2, wherein:
the optimization algorithm module has an algorithm selection function and can select SGD, momentum, adam optimization algorithm according to the optimization accuracy requirement;
the optimization algorithm module has a learning rate adjusting function and can select a fixed learning rate or a dynamic learning rate according to the optimization depth requirement.
7. The intelligent prediction method for carbon emission in the construction operation stage of the neural network optimization algorithm according to claim 2, wherein: the system has the authority of being integrated with other systems, and is convenient to access the regional carbon emission monitoring system:
an in-building IoT platform module that can monitor and record building energy consumption over a period of time, including air conditioning, lighting, ventilation, electrical outlets, elevators, other generated energy consumption;
an in-building IoT platform module that can monitor and record material consumption in building operations over a period of time, including in-building maintenance repairs, in-building use component replacement, in-building secondary finishing;
an in-building IoT platform module that can monitor and record meteorological data including air temperature, relative humidity, solar illuminance, wind speed over a period of time;
an in-building IoT platform module may monitor and record building activity conditions over a period of time, including instantaneous personnel density, instantaneous personnel occupancy, instantaneous indoor personnel activity intensity.
8. The intelligent prediction method for carbon emission in the construction operation phase of the neural network optimization algorithm according to claim 7, wherein:
the carbon emission calculation module can dynamically count and calculate the classification and the item of the carbon emission and the renewable energy carbon reduction in the construction operation stage;
and the carbon emission calculation module can increase the purchasing functions of building green carbon sink and external carbon sink according to the later-period requirements of building operators, and dynamically count and calculate the classification and the sub-term of the building carbon offset.
9. The intelligent prediction method for carbon emission in the construction operation phase of the neural network optimization algorithm according to claim 7, wherein:
the neural network module can calculate various carbon emission amounts in the construction operation stage based on a neural network calculation method and forecast an internal carbon emission value in a future time period;
the neural network module can adjust the depth layer number of the neural network according to the building area, the scale of the energy system in the building and the related conditions of the building complexity so as to meet the calculation efficiency of the system.
10. The intelligent prediction method for carbon emission in the construction operation phase of the neural network optimization algorithm according to claim 7, wherein:
and the optimization algorithm module can be used for feeding back various data and actual carbon emission based on the optimization algorithm in the construction operation stage, so that the accuracy of the predicted value of the internal carbon emission in the future time period is optimized.
11. The intelligent prediction method for carbon emission in the construction operation phase of the neural network optimization algorithm according to claim 7, wherein:
the intelligent carbon emission prediction system comprises a database of carbon emission related data, and can independently store the carbon emission related data information for a long time;
the carbon emission intelligent prediction system can be written based on Python, matlab, C ++ computer open source language, so that later updating and maintenance are facilitated;
the intelligent carbon emission prediction system can adopt a TensorFlow, pyTorch, matlab open source framework to build an algorithm model, so that the model is convenient to build and train;
the intelligent carbon emission prediction system can divide areas of a large building complex with multiple purposes and properties according to building functions and room use conditions, classify and generalize rooms with similar functions, monitor and predict carbon emission, and further improve prediction accuracy;
the intelligent carbon emission prediction system can perform inquiry, early warning, recording and downloading of carbon emission data in real time and produce a building carbon emission report;
the intelligent carbon emission prediction system can provide computer operation software by combining with computer application, and comprises a data collection and preprocessing module, a neural network model training module, an optimization algorithm module, a real-time monitoring and feedback module, a data storage and management module, a user interface module, an API (application program interface) and integration module and a security and authority management module;
the data collection and preprocessing module of the operation software of the intelligent carbon emission prediction system can collect data from the IoT devices and other data sources, screen, normalize and convert the data, and provide input data with proper format for the neural network model;
the data collection and preprocessing module of the operation software of the intelligent carbon emission prediction system can increase more data sources such as weather forecast and energy price according to requirements so as to provide more accurate support data;
the neural network model training module of the operation software of the intelligent carbon emission prediction system trains a neural network model by using historical data of building operation so as to calculate and predict the carbon emission.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311306117.8A CN117495119A (en) | 2023-10-10 | 2023-10-10 | Intelligent prediction method for carbon emission in building operation stage of neural network optimization algorithm |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311306117.8A CN117495119A (en) | 2023-10-10 | 2023-10-10 | Intelligent prediction method for carbon emission in building operation stage of neural network optimization algorithm |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN117495119A true CN117495119A (en) | 2024-02-02 |
Family
ID=89673413
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202311306117.8A Pending CN117495119A (en) | 2023-10-10 | 2023-10-10 | Intelligent prediction method for carbon emission in building operation stage of neural network optimization algorithm |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117495119A (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN118313563A (en) * | 2024-06-04 | 2024-07-09 | 清华大学深圳国际研究生院 | Building full life cycle carbon emission calculation method, calculation interaction method and system |
| CN118410950A (en) * | 2024-07-01 | 2024-07-30 | 三峡集团浙江能源投资有限公司 | Green license withholding energy consumption method and system based on blockchain |
| CN118425425A (en) * | 2024-04-29 | 2024-08-02 | 江苏省星霖工程咨询有限公司 | Intelligent carbon emission detection system and method based on big data |
-
2023
- 2023-10-10 CN CN202311306117.8A patent/CN117495119A/en active Pending
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN118425425A (en) * | 2024-04-29 | 2024-08-02 | 江苏省星霖工程咨询有限公司 | Intelligent carbon emission detection system and method based on big data |
| CN118425425B (en) * | 2024-04-29 | 2024-10-11 | 江苏省星霖工程咨询有限公司 | Intelligent carbon emission detection system and method based on big data |
| CN118313563A (en) * | 2024-06-04 | 2024-07-09 | 清华大学深圳国际研究生院 | Building full life cycle carbon emission calculation method, calculation interaction method and system |
| CN118313563B (en) * | 2024-06-04 | 2024-09-13 | 清华大学深圳国际研究生院 | Building full life cycle carbon emission calculation method, calculation interaction method and system |
| CN118410950A (en) * | 2024-07-01 | 2024-07-30 | 三峡集团浙江能源投资有限公司 | Green license withholding energy consumption method and system based on blockchain |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN117495119A (en) | Intelligent prediction method for carbon emission in building operation stage of neural network optimization algorithm | |
| Yan et al. | Optimization and prediction in the early design stage of office buildings using genetic and XGBoost algorithms | |
| CN116579506B (en) | Building energy consumption data intelligent management method and system based on big data | |
| Qiang et al. | Building automation systems for energy and comfort management in green buildings: A critical review and future directions | |
| Hosamo et al. | Digital Twin of HVAC system (HVACDT) for multiobjective optimization of energy consumption and thermal comfort based on BIM framework with ANN-MOGA | |
| CN115796393B (en) | Energy management optimization method, system and storage medium based on multi-energy interaction | |
| CN107392368A (en) | Meteorological forecast-based office building dynamic heat load combined prediction method | |
| CN112415924A (en) | Energy-saving optimization method and system for air conditioning system | |
| CN109858700A (en) | BP neural network heating system energy consumption prediction technique based on similar screening sample | |
| CN111598225B (en) | Air conditioner cold load prediction method based on self-adaptive deep confidence network | |
| CN112524751B (en) | Dynamic air conditioning system energy consumption prediction model construction and prediction method and device | |
| CN114322199A (en) | Ventilation system autonomous optimization operation regulation and control platform and method based on digital twins | |
| CN116205425A (en) | Low-carbon park cold-hot electric load prediction method based on typical database | |
| CN117553404A (en) | Method and system for improving energy efficiency of large water-cooling central air conditioning system | |
| CN115882455A (en) | Distributed photovoltaic power generation prediction method, system and terminal | |
| Gao et al. | Hybrid forecasting model of building cooling load based on combined neural network | |
| Das et al. | A study on the application of artificial intelligence techniques for predicting the heating and cooling loads of buildings | |
| Kang et al. | Integrated passive design method optimized for carbon emissions, economics, and thermal comfort of zero-carbon buildings | |
| CN118396780A (en) | Building environment and energy coupling intelligent regulation and control system and method based on digital twin | |
| Zhu | Research on adaptive combined wind speed prediction for each season based on improved gray relational analysis | |
| Moreno et al. | Exploiting IoT-based sensed data in smart buildings to model its energy consumption | |
| CN111144611A (en) | Spatial load prediction method based on clustering and nonlinear autoregression | |
| CN117634678A (en) | Low-carbon park carbon emission prediction method based on actual operation scene | |
| CN118517769B (en) | Operation management method and system of high-efficiency energy-saving air conditioner room | |
| KR102720660B1 (en) | Design method of smart farm cooling package optimal design support system |
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 |