CN113486514A - Building heat consumption prediction modeling method based on automatic calibration - Google Patents
Building heat consumption prediction modeling method based on automatic calibration Download PDFInfo
- Publication number
- CN113486514A CN113486514A CN202110760839.5A CN202110760839A CN113486514A CN 113486514 A CN113486514 A CN 113486514A CN 202110760839 A CN202110760839 A CN 202110760839A CN 113486514 A CN113486514 A CN 113486514A
- Authority
- CN
- China
- Prior art keywords
- building
- heat consumption
- data
- time
- automatic calibration
- 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
- 238000000034 method Methods 0.000 title claims abstract description 32
- 238000004088 simulation Methods 0.000 claims abstract description 28
- 238000004140 cleaning Methods 0.000 claims abstract description 8
- 230000002159 abnormal effect Effects 0.000 claims abstract description 7
- 238000003064 k means clustering Methods 0.000 claims abstract description 7
- 238000007781 pre-processing Methods 0.000 claims abstract description 6
- 238000004422 calculation algorithm Methods 0.000 claims description 13
- 238000005457 optimization Methods 0.000 claims description 9
- 239000002245 particle Substances 0.000 claims description 9
- 238000012546 transfer Methods 0.000 claims description 5
- 230000005855 radiation Effects 0.000 claims description 4
- 238000012935 Averaging Methods 0.000 claims description 2
- 238000013450 outlier detection Methods 0.000 claims description 2
- 238000005070 sampling Methods 0.000 claims description 2
- 238000013461 design Methods 0.000 description 4
- 238000005265 energy consumption Methods 0.000 description 3
- 238000010276 construction Methods 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 238000007418 data mining Methods 0.000 description 1
- 238000001514 detection method Methods 0.000 description 1
- 238000011156 evaluation Methods 0.000 description 1
- 238000000605 extraction Methods 0.000 description 1
- 238000010438 heat treatment Methods 0.000 description 1
- 238000005065 mining Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000012360 testing method Methods 0.000 description 1
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/10—Geometric CAD
- G06F30/13—Architectural design, e.g. computer-aided architectural design [CAAD] related to design of buildings, bridges, landscapes, production plants or roads
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/23—Clustering techniques
- G06F18/232—Non-hierarchical techniques
- G06F18/2321—Non-hierarchical techniques using statistics or function optimisation, e.g. modelling of probability density functions
- G06F18/23213—Non-hierarchical techniques using statistics or function optimisation, e.g. modelling of probability density functions with fixed number of clusters, e.g. K-means clustering
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/20—Design optimisation, verification or simulation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/004—Artificial life, i.e. computing arrangements simulating life
- G06N3/006—Artificial life, i.e. computing arrangements simulating life based on simulated virtual individual or collective life forms, e.g. social simulations or particle swarm optimisation [PSO]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2119/00—Details relating to the type or aim of the analysis or the optimisation
- G06F2119/08—Thermal analysis or thermal optimisation
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Geometry (AREA)
- Evolutionary Computation (AREA)
- Computer Hardware Design (AREA)
- Life Sciences & Earth Sciences (AREA)
- Artificial Intelligence (AREA)
- Health & Medical Sciences (AREA)
- Computing Systems (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Bioinformatics & Computational Biology (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Probability & Statistics with Applications (AREA)
- Biomedical Technology (AREA)
- Biophysics (AREA)
- Computational Linguistics (AREA)
- General Health & Medical Sciences (AREA)
- Molecular Biology (AREA)
- Evolutionary Biology (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Architecture (AREA)
- Civil Engineering (AREA)
- Structural Engineering (AREA)
- Computational Mathematics (AREA)
- Mathematical Analysis (AREA)
- Mathematical Optimization (AREA)
- Pure & Applied Mathematics (AREA)
- Air Conditioning Control Device (AREA)
Abstract
The invention discloses a building heat consumption prediction modeling method based on automatic calibration, which comprises the following steps: s1: acquiring the time-by-time heat consumption of a building and the original data of characteristic variables related to the heat consumption; s2: performing data preprocessing on the acquired original data, including abnormal value cleaning and missing value filling; s3: performing K-means clustering analysis on the processed time-by-time heat consumption data to obtain a typical building heat consumption model; s4: building a basic simulation model of the building by using EnergyPlus software; s5: and automatically calibrating the configuration parameters of the basic building simulation model by using the heat consumption data in each typical thermal model obtained by clustering. Therefore, an accurate building heat consumption prediction model is obtained. The invention can effectively improve the precision of the building heat consumption simulation model.
Description
Technical Field
The invention relates to the field of building heat consumption prediction modeling, in particular to a building heat consumption prediction modeling method based on automatic calibration.
Technical Field
Currently, energy consumption simulation software for buildings has been widely used in aspects of building energy saving scheme design, energy saving reconstruction, optimized operation and the like. However, a serious problem is often faced, namely that the building energy consumption simulation result is far from the actual result, which greatly influences the application range of the building energy consumption software and limits the deepening capability of the building energy consumption software in the building field. Therefore, a technical route for improving the accuracy of the building energy consumption model should be actively explored.
Whether the configuration parameters of the building heat consumption simulation model are accurately set directly determines whether the simulation model can output results which accord with the actual characteristics of the building. Therefore, improving the accuracy of the model configuration parameters is a main technical route for improving the accuracy of the building heat consumption simulation model. However, accurate configuration parameters are difficult to obtain in practice, such as heat transfer coefficients of building envelopes, and on one hand, the results caused by design data and the actual construction process have certain deviation, and it is difficult to obtain accurate values only according to the design data. On the other hand, as the building is a whole, the thermal performance data of the whole building is difficult to obtain only through a single-point test. And the configuration parameters of the simulation model are reversely calibrated according to the building operation heat consumption data, and the configuration parameters of the simulation model are accurately identified by using an intelligent optimization algorithm, so that the precision of the building heat consumption simulation model is improved, and the method is a feasible technical route.
Disclosure of Invention
Based on the technical background, the invention provides a building heat consumption prediction modeling method based on automatic calibration, which is used for improving the precision of a building time-by-time heat consumption prediction model.
The invention is realized by adopting the following technical scheme:
the invention provides a building heat consumption prediction modeling method based on automatic calibration, which comprises the following steps:
s1: acquiring the time-by-time heat consumption of a building and the original data of characteristic variables related to the heat consumption;
s2: performing data preprocessing on the acquired original data, including abnormal value cleaning and missing value filling;
s3: performing K-means clustering analysis on the processed time-by-time heat consumption data to obtain several typical heat consumption modes of the building;
s4: building a basic simulation model of the building by using EnergyPlus software;
s5: and (4) automatically calibrating the configuration parameters of the basic simulation model of the building by utilizing the heat consumption data in each typical heat mode obtained by clustering and combining a particle swarm optimization algorithm. Therefore, an accurate building heat consumption prediction model is obtained.
Further, the characteristic variables in step S1 include the hourly outdoor temperature and the hourly solar radiation intensity.
Further, all the raw data in step S1 are collected at intervals of 1 hour, and the building heat consumption and its related characteristic variable data are collected at each sampling point.
Further, the data preprocessing in step S2 includes an outlier detection and cleaning function. The Hampel filter is mainly used to detect and clean outliers. The method calculates the median of a moving window of 2k samples around each sample point and calculates the standard deviation of each sample with respect to its window median using the Median Absolute Deviation (MAD). If the sample differs from the median by more than t standard deviations, it is detected as an outlier and removed. Wherein the value of k is 12. the value of t is 3.
Further, the missing value padding in step S2 is performed by using a method of similar day data. And averaging 4 data at the time before and after the position of the missing data and at the same time before and after the position of the missing data, and filling the missing data.
Further, in step S3, the time-by-time heat consumption data of the building is subjected to typical heat pattern clustering using K-means. Firstly, normalizing the daily time-by-time heat consumption curve, then clustering the building heat consumption curves of all days by using a K-means clustering algorithm, and determining the number of clustering centers by using a Dunn index. Thereby obtaining the typical heat-using pattern characteristics of the building.
Further, in step S5, on the energy plus and GenOpt combined simulation platform, the configuration parameters of the basic simulation model are automatically calibrated by using the particle swarm optimization algorithm. The calibration parameters mainly include: building envelope heat transfer coefficient, power density of equipment and lighting, and personnel density. The calibration target is then: in each typical mode, the error between the building heat consumption output value of the basic simulation model and the actually measured value is minimum. And finally obtaining an accurate building heat consumption prediction model.
Compared with the prior art, the technical scheme of the invention has the following beneficial effects:
the invention can utilize an unsupervised data mining method to realize mining extraction of the building heat consumption typical mode, and can realize accurate calibration of the building heat consumption basic simulation model based on the typical mode. Compared with a building heat consumption simulation model which is not subjected to typical thermal mode clustering and automatic model calibration, the method effectively improves the prediction precision of the time-by-time heat consumption of the building, and the precision is improved by about 15%.
Drawings
FIG. 1 is a flow chart of the method of the present invention
Detailed Description
The invention is described in further detail below with reference to the figures and specific examples. It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.
As shown in fig. 1, a building heat consumption prediction modeling method based on automatic calibration includes the following steps:
s1: acquiring the time-by-time heat consumption of a building and the original data of characteristic variables related to the heat consumption;
in the embodiment, the heat consumption data of a certain office building in the city of noon and security is tested and collected, the heat consumption data comprises the temperature and the flow of water supply and return at a thermal inlet of the building, and the time-by-time heat consumption data of the building is calculated according to the data. In addition, a small meteorological station is used for collecting time-by-time outdoor temperature and solar radiation intensity data. The collection interval of each type of data is 1 hour, and 120 days of data in the whole heating season are collected.
S2: performing data preprocessing on the acquired original data, including abnormal value cleaning and missing value filling;
and (4) carrying out abnormal value detection and cleaning on the building heat consumption data, the outdoor temperature and the solar radiation intensity data by adopting a Hampel filter. The median of a moving window of 2k samples around each sample point is first calculated and the standard deviation of each sample with respect to its window median is calculated using the Median Absolute Deviation (MAD). If the sample differs from the median by more than t standard deviations, it is detected as an outlier and removed. Wherein the value of k is 12. the value of t is 3. Then, the missing value generated after the abnormal value is deleted is filled with data similar to the missing value.
S3: performing K-means clustering analysis on the processed time-by-time heat consumption data to obtain several typical heat consumption modes of the building;
specifically, firstly, the daily time-by-time heat consumption curve is normalized, then the K-means clustering algorithm is used for clustering the building heat consumption curves of all days, and the Dunn index is used for determining the number of clustering centers. The specific K-means algorithm is expressed as follows:
the algorithm process of clustering is as follows: (1) randomly selecting K objects from N sample data as initial clustering centers; (2) respectively calculating the cluster from each sample to each cluster center, and distributing the object to the cluster with the closest distance; (3) after all the objects are distributed, recalculating centers of the K clusters; (4) compared with the clustering center of K obtained by the previous calculation, if the clustering center changes, the process (2) is switched to, and if not, the process (5) is switched to; (5) and finishing the clustering process and outputting the result when the centroid is not changed.
Among them, the Dunn index can be calculated by the following formula:
in the formula:
d(ci,cj) Is of class ciAnd class cjIs determined by the degree of dissimilarity function of (c),
diam (C) is the diameter of class C, diam (C) maxx,y∈C{d(x,y)}；
Generally, the larger the Dunn index, the better the clustering result.
S4: building a basic simulation model of the building by using EnergyPlus software;
the basic building information including the geometric size of the building, the heat transfer coefficient of the wall, roof and window, the density of lighting, equipment and personnel and the schedule is obtained according to the building design data. And inputting the simulation model into EnergyPuls software to realize the construction of a basic simulation model of the heat consumption of the building.
S5: and (4) automatically calibrating the configuration parameters of the basic simulation model of the building by utilizing the heat consumption data in each typical heat mode obtained by clustering and combining a particle swarm optimization algorithm. Therefore, an accurate building heat consumption prediction model is obtained.
Specifically, the heat consumption data in each type of typical thermal model is averaged, the average value is used as a calibration target of the configuration parameters of the simulation model, and the particle swarm optimization algorithm is used for carrying out optimization solution on the device. The particle swarm optimization method initializes a group of random particles in a search range and then continuously iterates to find an optimal solution based on an algorithm mechanism. In each generation, the particles are updated based on two optimal values, the first being personal optimal and the other being global optimal. The update process is shown as follows.
The calibration target uses two common evaluation indexes of model calibration: MBE (Mean Bias Error) and CV (RMSE) (coeffient of Variation of the Root-Mean-Square Error).
In the formula:
yi-the actual measured value;
The whole calibration process can be realized by an EnergyPlus and GenOpt combined simulation platform. The experimental results show that the mean absolute percentage error of the model is reduced by 15% after using this method.
The present invention is not limited to the above-described embodiments. The foregoing description of the specific embodiments is intended to describe and illustrate the technical solutions of the present invention, and the above specific embodiments are merely illustrative and not restrictive. Those skilled in the art can make many changes and modifications to the invention without departing from the spirit and scope of the invention as defined in the appended claims.
Claims (8)
1. A building heat consumption prediction modeling method based on automatic calibration is characterized by comprising the following steps:
s1: acquiring the time-by-time heat consumption of a building and the original data of characteristic variables related to the heat consumption;
s2: performing data preprocessing on the acquired original data, including abnormal value cleaning and missing value filling;
s3: performing K-means clustering analysis on the processed time-by-time heat consumption data to obtain several typical heat consumption modes of the building;
s4: building a basic simulation model of the building by using EnergyPlus software;
s5: utilizing the heat consumption data in each typical heat mode obtained by clustering and combining a particle swarm optimization algorithm to automatically calibrate the configuration parameters of the basic simulation model of the building; therefore, an accurate building heat consumption prediction model is obtained.
2. The building heat consumption prediction modeling method based on automatic calibration as claimed in claim 1, characterized in that the characteristic variables in step S1 include time-by-time outdoor temperature and time-by-time solar radiation intensity.
3. The method for predictive modeling of building heat consumption based on automatic calibration as claimed in claim 1, wherein all raw data in step S1 are collected at 1 hour intervals, and building heat consumption and its related characteristic variable data are collected at each sampling point.
4. The method for predictive modeling of building heat consumption based on automatic calibration as claimed in claim 1, wherein the data preprocessing in step S2 includes outlier detection and cleaning functions; detecting and cleaning abnormal values by adopting a Hampel filter, calculating the median of a moving window consisting of each sample point and 2k samples around the sample point by adopting the method, and calculating the standard deviation of each sample about the median of the window by using the Median Absolute Deviation (MAD); if the sample differs from the median by more than t standard deviations, the sample is detected as an outlier and removed.
5. The building heat consumption prediction modeling method based on automatic calibration as claimed in claim 1, characterized in that the missing value filling in step S2 is performed by using a similar day data method, and the missing data is filled by averaging 4 data at the time before and after the position of the missing data and the same time before and after the same day.
6. The building heat consumption prediction modeling method based on automatic calibration according to claim 1, characterized in that in step S3, the building time-by-time heat consumption data is subjected to typical heat consumption mode clustering by using K-means; firstly, normalizing the daily time-by-time heat consumption curve, then clustering the building heat consumption curves of all days by using a K-means clustering algorithm, and determining the number of clustering centers by using a Dunn index; thereby obtaining the typical heat-using pattern characteristics of the building.
7. The method according to claim 1, wherein the basic simulation model in step S4 is constructed from basic building research information including building geometry, heat transfer coefficient of walls, roofs, and windows, lighting, equipment and personnel density, and schedule.
8. The building heat consumption prediction modeling method based on automatic calibration as claimed in claim 1, wherein in step S5, the configuration parameters in the basic simulation model are automatically calibrated using a particle swarm optimization algorithm, and the calibration parameters mainly comprise: building envelope heat transfer coefficient, power density of equipment and lighting, and personnel density; the calibration target is then: in each type of typical mode, the error between the building heat consumption output value of the basic simulation model and the actual measured value is minimum; and finally obtaining an accurate building heat consumption prediction model.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110760839.5A CN113486514A (en) | 2021-07-06 | 2021-07-06 | Building heat consumption prediction modeling method based on automatic calibration |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110760839.5A CN113486514A (en) | 2021-07-06 | 2021-07-06 | Building heat consumption prediction modeling method based on automatic calibration |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN113486514A true CN113486514A (en) | 2021-10-08 |
Family
ID=77941148
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110760839.5A Pending CN113486514A (en) | 2021-07-06 | 2021-07-06 | Building heat consumption prediction modeling method based on automatic calibration |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113486514A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117970859A (en) * | 2024-04-02 | 2024-05-03 | 山西顺达胜业通信工程有限公司 | Intelligent control method for equipment under low-power-consumption wide area Internet of things |
Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109636677A (en) * | 2019-01-17 | 2019-04-16 | 天津大学 | Building thermal technique performance estimating method based on model calibration |
-
2021
- 2021-07-06 CN CN202110760839.5A patent/CN113486514A/en active Pending
Patent Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109636677A (en) * | 2019-01-17 | 2019-04-16 | 天津大学 | Building thermal technique performance estimating method based on model calibration |
Non-Patent Citations (2)
| Title |
|---|
| 李万程: "居住建筑热负荷模型分步校准和案例研究", 《中国优秀硕士学位论文全文数据库 工程科技Ⅱ辑》 * |
| 马文辉等: "基于 K-means 聚类的热计量用户负荷特性及行为节能分析", 《建筑节能》 * |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117970859A (en) * | 2024-04-02 | 2024-05-03 | 山西顺达胜业通信工程有限公司 | Intelligent control method for equipment under low-power-consumption wide area Internet of things |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN106059492B (en) | Photovoltaic module shade fault type judges method based on power prediction | |
| CN104616079B (en) | Power grid daily-use power predicating method based on temperature change | |
| Sangogboye et al. | Performance comparison of occupancy count estimation and prediction with common versus dedicated sensors for building model predictive control | |
| Shi et al. | Prediction of indoor temperature and relative humidity based on cloud database by using an improved BP neural network in Chongqing | |
| CN109948281B (en) | Direct dispersion separation modeling method based on weather type effective identification and combined prediction | |
| CN112288164B (en) | Wind power combined prediction method considering spatial correlation and correcting numerical weather forecast | |
| Nonnenmacher et al. | Verification of the SUNY direct normal irradiance model with ground measurements | |
| CN111639823A (en) | Building cold and heat load prediction method constructed based on feature set | |
| Hassan et al. | High-resolution estimates of diffuse fraction based on dynamic definitions of sky conditions | |
| CN112183826A (en) | Building energy consumption prediction method based on deep cascade generation countermeasure network and related product | |
| CN112819312A (en) | Method and system for evaluating drought socioeconomic exposure degree under climate change scene | |
| Wang et al. | Short time air temperature prediction using pattern approximate matching | |
| CN115204444A (en) | Photovoltaic power prediction method based on improved cluster analysis and fusion integration algorithm | |
| CN113486514A (en) | Building heat consumption prediction modeling method based on automatic calibration | |
| Balti et al. | Big data based architecture for drought forecasting using LSTM, ARIMA, and Prophet: Case study of the Jiangsu Province, China | |
| Chen et al. | ATLAS: software for analysing the relationship between urban microclimate and urban morphology in a tropical city | |
| Mantri et al. | Weather prediction and classification using neural networks and k-nearest neighbors | |
| Li et al. | Assessing urban micro-climates with vertical and horizontal building morphological cutting deep transfer learning neural networks | |
| CN111814399B (en) | Model parameter optimization extraction method and measurement data prediction method of solar photovoltaic cell system | |
| CN117370813A (en) | Atmospheric pollution deep learning prediction method based on K line pattern matching algorithm | |
| CN110796292A (en) | Photovoltaic power short-term prediction method considering haze influence | |
| CN116796291A (en) | LSTM-MEA-SVR-based air quality forecasting system | |
| Li et al. | Transfer learning strategy for data-driven correction of wind measurement around buildings using cup anemometers | |
| CN110188964A (en) | A kind of photovoltaic power generation output forecasting method based on correlation | |
| CN116776263A (en) | Data-driven-based method and system for detecting abnormal high energy consumption of block-scale building |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20211008 |
|
| RJ01 | Rejection of invention patent application after publication |