Building Energy Consumption Raw Data Forecasting Using Data Cleaning and Deep Recurrent Neural Networks
With the rising focus on building energy big data analysis, there lacks a framework for raw data preprocessing to answer the question of how to handle the missing data in the raw data set. This study presents a methodology and framework for building energy consumption raw data forecasting. A case bu...
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2019-09-01
|
| Series: | Buildings |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2075-5309/9/9/204 |
| _version_ | 1818263037500456960 |
|---|---|
| author | Junjing Yang Kok Keng Tan Mat Santamouris Siew Eang Lee |
| author_facet | Junjing Yang Kok Keng Tan Mat Santamouris Siew Eang Lee |
| author_sort | Junjing Yang |
| collection | DOAJ |
| description | With the rising focus on building energy big data analysis, there lacks a framework for raw data preprocessing to answer the question of how to handle the missing data in the raw data set. This study presents a methodology and framework for building energy consumption raw data forecasting. A case building is used to forecast the energy consumption by using deep recurrent neural networks. Four different methodologies to impute missing data in the raw data set are compared and implemented. The question of sensitivity of gap size and available data percentage on the imputation accuracy was tested. The cleaned data were then used for building energy forecasting. While the existing studies explored only the use of small recurrent networks of 2 layers and less, the question of whether a deep network of more than 2 layers would be performing better for building energy consumption forecasting should be explored. In addition, the problem of overfitting has been cited as a significant problem in using deep networks. In this study, the deep recurrent neural network is then used to explore the use of deeper networks and their regularization in the context of an energy load forecasting task. The results show a mean absolute error of 2.1 can be achieved through the 2*32 gated neural network model. In applying regularization methods to overcome model overfitting, the study found that weights regularization did indeed delay the onset of overfitting. |
| first_indexed | 2024-12-12T19:12:39Z |
| format | Article |
| id | doaj.art-e711077f76a64d08b4e7679ad006c121 |
| institution | Directory Open Access Journal |
| issn | 2075-5309 |
| language | English |
| last_indexed | 2024-12-12T19:12:39Z |
| publishDate | 2019-09-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Buildings |
| spelling | doaj.art-e711077f76a64d08b4e7679ad006c1212022-12-22T00:14:49ZengMDPI AGBuildings2075-53092019-09-019920410.3390/buildings9090204buildings9090204Building Energy Consumption Raw Data Forecasting Using Data Cleaning and Deep Recurrent Neural NetworksJunjing Yang0Kok Keng Tan1Mat Santamouris2Siew Eang Lee3Department of Building, National University of Singapore, Singapore 117566, SingaporeDepartment of Building, National University of Singapore, Singapore 117566, SingaporeDepartment of Building, National University of Singapore, Singapore 117566, SingaporeDepartment of Building, National University of Singapore, Singapore 117566, SingaporeWith the rising focus on building energy big data analysis, there lacks a framework for raw data preprocessing to answer the question of how to handle the missing data in the raw data set. This study presents a methodology and framework for building energy consumption raw data forecasting. A case building is used to forecast the energy consumption by using deep recurrent neural networks. Four different methodologies to impute missing data in the raw data set are compared and implemented. The question of sensitivity of gap size and available data percentage on the imputation accuracy was tested. The cleaned data were then used for building energy forecasting. While the existing studies explored only the use of small recurrent networks of 2 layers and less, the question of whether a deep network of more than 2 layers would be performing better for building energy consumption forecasting should be explored. In addition, the problem of overfitting has been cited as a significant problem in using deep networks. In this study, the deep recurrent neural network is then used to explore the use of deeper networks and their regularization in the context of an energy load forecasting task. The results show a mean absolute error of 2.1 can be achieved through the 2*32 gated neural network model. In applying regularization methods to overcome model overfitting, the study found that weights regularization did indeed delay the onset of overfitting.https://www.mdpi.com/2075-5309/9/9/204energy forecastingdeep recurrent neural networksdata imputation |
| spellingShingle | Junjing Yang Kok Keng Tan Mat Santamouris Siew Eang Lee Building Energy Consumption Raw Data Forecasting Using Data Cleaning and Deep Recurrent Neural Networks Buildings energy forecasting deep recurrent neural networks data imputation |
| title | Building Energy Consumption Raw Data Forecasting Using Data Cleaning and Deep Recurrent Neural Networks |
| title_full | Building Energy Consumption Raw Data Forecasting Using Data Cleaning and Deep Recurrent Neural Networks |
| title_fullStr | Building Energy Consumption Raw Data Forecasting Using Data Cleaning and Deep Recurrent Neural Networks |
| title_full_unstemmed | Building Energy Consumption Raw Data Forecasting Using Data Cleaning and Deep Recurrent Neural Networks |
| title_short | Building Energy Consumption Raw Data Forecasting Using Data Cleaning and Deep Recurrent Neural Networks |
| title_sort | building energy consumption raw data forecasting using data cleaning and deep recurrent neural networks |
| topic | energy forecasting deep recurrent neural networks data imputation |
| url | https://www.mdpi.com/2075-5309/9/9/204 |
| work_keys_str_mv | AT junjingyang buildingenergyconsumptionrawdataforecastingusingdatacleaninganddeeprecurrentneuralnetworks AT kokkengtan buildingenergyconsumptionrawdataforecastingusingdatacleaninganddeeprecurrentneuralnetworks AT matsantamouris buildingenergyconsumptionrawdataforecastingusingdatacleaninganddeeprecurrentneuralnetworks AT sieweanglee buildingenergyconsumptionrawdataforecastingusingdatacleaninganddeeprecurrentneuralnetworks |