A platform for research: civil engineering, architecture and urbanism
A platform for research: civil engineering, architecture and urbanism
Predicting indoor air temperature and thermal comfort in occupational settings using weather forecasts, indoor sensors, and artificial neural networks
https://orcid.org/0000-0001-5588-2985
https://orcid.org/0000-0003-3864-1703
Abstract We present an approach to generate location-specific forecasts of indoor temperature (T i) and thermal comfort and issue indoor heat warnings for occupational settings. Indoor forecasts are generated using standard outdoor weather forecasting products and an artificial neural network (ANN) trained on-site using local indoor measurements from a low-cost sensor system measuring T i and indoor physiologically equivalent temperature (PETi). The outcomes are hourly indoor T i and PETi forecasts. Different ANN-based forecast products using different predictors were concurrently tested at 121 workplaces in agricultural, industrial, storage, and office buildings using data for an entire annual cycle. A forecast was considered skillful when the T i and PETi forecast was <2 K from actual measurements. The best-performing model used the predictors time of year, week, and day; solar position; and outdoor weather forecast variables to train and run an ANN to predict T i or PETi. It had an annual average mean absolute forecast error of 0.87 K for T i and 0.99 K for PETi over the next 24 h, with Pearson correlation coefficients of 0.98 and 0.97, respectively. Overall, 91% of T i forecasts and 88% of PETi forecasts were skillful. Indoor forecasts showed larger errors in the summer than in the winter. We conclude that combining indoor data with weather forecasts using ANNs could be implemented widely to provide location-specific indoor weather forecasts to improve and localize heat and health warning systems.
Graphical abstract Display Omitted
Highlights Neural networks predict indoor temperatures, thermal comfort and heat warnings. Neural networks trained using indoor sensor data and outdoor weather forecasts. Workflow concurrently tested at 121 typical workplaces over full annual cycle. Best model predicts indoor temperature with <1 K error, 92% of all forecasts <2 K. Approach enables skilful and location-specific indoor heat and health warnings.
Predicting indoor air temperature and thermal comfort in occupational settings using weather forecasts, indoor sensors, and artificial neural networks
https://orcid.org/0000-0001-5588-2985
https://orcid.org/0000-0003-3864-1703
Abstract We present an approach to generate location-specific forecasts of indoor temperature (T i) and thermal comfort and issue indoor heat warnings for occupational settings. Indoor forecasts are generated using standard outdoor weather forecasting products and an artificial neural network (ANN) trained on-site using local indoor measurements from a low-cost sensor system measuring T i and indoor physiologically equivalent temperature (PETi). The outcomes are hourly indoor T i and PETi forecasts. Different ANN-based forecast products using different predictors were concurrently tested at 121 workplaces in agricultural, industrial, storage, and office buildings using data for an entire annual cycle. A forecast was considered skillful when the T i and PETi forecast was <2 K from actual measurements. The best-performing model used the predictors time of year, week, and day; solar position; and outdoor weather forecast variables to train and run an ANN to predict T i or PETi. It had an annual average mean absolute forecast error of 0.87 K for T i and 0.99 K for PETi over the next 24 h, with Pearson correlation coefficients of 0.98 and 0.97, respectively. Overall, 91% of T i forecasts and 88% of PETi forecasts were skillful. Indoor forecasts showed larger errors in the summer than in the winter. We conclude that combining indoor data with weather forecasts using ANNs could be implemented widely to provide location-specific indoor weather forecasts to improve and localize heat and health warning systems.
Graphical abstract Display Omitted
Highlights Neural networks predict indoor temperatures, thermal comfort and heat warnings. Neural networks trained using indoor sensor data and outdoor weather forecasts. Workflow concurrently tested at 121 typical workplaces over full annual cycle. Best model predicts indoor temperature with <1 K error, 92% of all forecasts <2 K. Approach enables skilful and location-specific indoor heat and health warnings.
Predicting indoor air temperature and thermal comfort in occupational settings using weather forecasts, indoor sensors, and artificial neural networks
https://orcid.org/0000-0001-5588-2985
https://orcid.org/0000-0003-3864-1703
Abstract We present an approach to generate location-specific forecasts of indoor temperature (T i) and thermal comfort and issue indoor heat warnings for occupational settings. Indoor forecasts are generated using standard outdoor weather forecasting products and an artificial neural network (ANN) trained on-site using local indoor measurements from a low-cost sensor system measuring T i and indoor physiologically equivalent temperature (PETi). The outcomes are hourly indoor T i and PETi forecasts. Different ANN-based forecast products using different predictors were concurrently tested at 121 workplaces in agricultural, industrial, storage, and office buildings using data for an entire annual cycle. A forecast was considered skillful when the T i and PETi forecast was <2 K from actual measurements. The best-performing model used the predictors time of year, week, and day; solar position; and outdoor weather forecast variables to train and run an ANN to predict T i or PETi. It had an annual average mean absolute forecast error of 0.87 K for T i and 0.99 K for PETi over the next 24 h, with Pearson correlation coefficients of 0.98 and 0.97, respectively. Overall, 91% of T i forecasts and 88% of PETi forecasts were skillful. Indoor forecasts showed larger errors in the summer than in the winter. We conclude that combining indoor data with weather forecasts using ANNs could be implemented widely to provide location-specific indoor weather forecasts to improve and localize heat and health warning systems.
Graphical abstract Display Omitted
Highlights Neural networks predict indoor temperatures, thermal comfort and heat warnings. Neural networks trained using indoor sensor data and outdoor weather forecasts. Workflow concurrently tested at 121 typical workplaces over full annual cycle. Best model predicts indoor temperature with <1 K error, 92% of all forecasts <2 K. Approach enables skilful and location-specific indoor heat and health warnings.
Predicting indoor air temperature and thermal comfort in occupational settings using weather forecasts, indoor sensors, and artificial neural networks
Sulzer, Markus (author) / Christen, Andreas (author) / Matzarakis, Andreas (author)
Building and Environment ; 234
2023-02-04
Article (Journal)
Electronic Resource
English
Indoor Thermal Comfort Predicting Correlations in Subtropical Plateau Monsoon Climate
| British Library Conference Proceedings | 2012
Thermal diary: Connecting temperature history to indoor comfort
| Online Contents | 2008
Thermal diary: Connecting temperature history to indoor comfort
| Online Contents | 2008
Thermal diary: Connecting temperature history to indoor comfort
| British Library Online Contents | 2008