Experimental testing of a random neural network smart controller using a single zone test chamber

Abbas Javed, Hadi Larijani, Ali Ahmadini, Rohinton Emmanuel, Desmond Gibson, Caspar Clark

Research output: Contribution to journalArticle

Abstract

Monitoring and analysis of energy use and indoor environmental conditions is an urgent need in large buildings to respond to changing conditions in an efficient manner. Correct estimation of occupancy will further improve energy performance. In this work, a smart controller for maintaining a comfortable environment using multiple random neural networks (RNNs) has been developed. The implementation of RNN-based controller is demonstrated to be more efficient on hardware and requires less memory compared to both artificial neural networks and model predictive controllers. This controller estimates the number of room occupants by using the information from wireless sensor nodes placed in the Heating, Ventilation and Air Conditioning (HVAC) duct and the room. For an occupied room, the controller can switch between thermal comfort mode (based on predicted mean vote set points) and user defined mode (i.e. occupant defined set points for heating/cooling/ventilation). Furthermore, the hybrid particle swarm optimisation with sequential quadratic programming training algorithms are used (for the first time to the best of the authors' knowledge) for training the RNN and results show that this algorithm outperforms the widely used gradient descent algorithm for RNN. The results show that occupancy estimation by smart controller is 83.08% accurate.
Original languageEnglish
Pages (from-to)350-358
Number of pages8
JournalIET Networks
Volume4
Issue number6
DOIs
Publication statusPublished - Nov 2015

Fingerprint

Random Networks
Neural Networks
Neural networks
Controller
Controllers
Testing
Ventilation
Point Sets
Heating
Air conditioning ducts
Hybrid Optimization
Descent Algorithm
Thermal comfort
Training Algorithm
Wireless Sensors
Gradient Algorithm
Gradient Descent
Quadratic programming
Vote
Energy

Keywords

  • energy usage monitoring
  • occupancy estimation
  • comfortable environment
  • thermal comfort mode
  • artificial neural networks
  • energy usage analysis
  • random neural network smart controller
  • sequential quadratic programming training algorithms
  • heating-ventilation-air conditioning
  • indoor environment
  • building energy control system
  • gradient descent algorithm
  • environmental conditions
  • experimental testing
  • model predictive controllers
  • single zone test chamber
  • RNN-based controller
  • HVAC duct
  • wireless sensor nodes
  • hybrid particle swarm optimisation
  • predicted mean vote-based set points

Cite this

Javed, Abbas ; Larijani, Hadi ; Ahmadini, Ali ; Emmanuel, Rohinton ; Gibson, Desmond ; Clark, Caspar . / Experimental testing of a random neural network smart controller using a single zone test chamber. In: IET Networks. 2015 ; Vol. 4, No. 6. pp. 350-358.
@article{a5859554026d40ef9b0d17be59af1a4e,
title = "Experimental testing of a random neural network smart controller using a single zone test chamber",
abstract = "Monitoring and analysis of energy use and indoor environmental conditions is an urgent need in large buildings to respond to changing conditions in an efficient manner. Correct estimation of occupancy will further improve energy performance. In this work, a smart controller for maintaining a comfortable environment using multiple random neural networks (RNNs) has been developed. The implementation of RNN-based controller is demonstrated to be more efficient on hardware and requires less memory compared to both artificial neural networks and model predictive controllers. This controller estimates the number of room occupants by using the information from wireless sensor nodes placed in the Heating, Ventilation and Air Conditioning (HVAC) duct and the room. For an occupied room, the controller can switch between thermal comfort mode (based on predicted mean vote set points) and user defined mode (i.e. occupant defined set points for heating/cooling/ventilation). Furthermore, the hybrid particle swarm optimisation with sequential quadratic programming training algorithms are used (for the first time to the best of the authors' knowledge) for training the RNN and results show that this algorithm outperforms the widely used gradient descent algorithm for RNN. The results show that occupancy estimation by smart controller is 83.08{\%} accurate.",
keywords = "energy usage monitoring, occupancy estimation, comfortable environment, thermal comfort mode, artificial neural networks, energy usage analysis, random neural network smart controller, sequential quadratic programming training algorithms, heating-ventilation-air conditioning, indoor environment, building energy control system, gradient descent algorithm, environmental conditions, experimental testing, model predictive controllers, single zone test chamber, RNN-based controller, HVAC duct, wireless sensor nodes, hybrid particle swarm optimisation, predicted mean vote-based set points",
author = "Abbas Javed and Hadi Larijani and Ali Ahmadini and Rohinton Emmanuel and Desmond Gibson and Caspar Clark",
year = "2015",
month = "11",
doi = "10.1049/iet-net.2015.0020",
language = "English",
volume = "4",
pages = "350--358",
journal = "IET Networks",
issn = "2047-4954",
publisher = "Institution of Engineering and Technology",
number = "6",

}

Experimental testing of a random neural network smart controller using a single zone test chamber. / Javed, Abbas ; Larijani, Hadi ; Ahmadini, Ali ; Emmanuel, Rohinton ; Gibson, Desmond; Clark, Caspar .

In: IET Networks, Vol. 4, No. 6, 11.2015, p. 350-358.

Research output: Contribution to journalArticle

TY - JOUR

T1 - Experimental testing of a random neural network smart controller using a single zone test chamber

AU - Javed, Abbas

AU - Larijani, Hadi

AU - Ahmadini, Ali

AU - Emmanuel, Rohinton

AU - Gibson, Desmond

AU - Clark, Caspar

PY - 2015/11

Y1 - 2015/11

N2 - Monitoring and analysis of energy use and indoor environmental conditions is an urgent need in large buildings to respond to changing conditions in an efficient manner. Correct estimation of occupancy will further improve energy performance. In this work, a smart controller for maintaining a comfortable environment using multiple random neural networks (RNNs) has been developed. The implementation of RNN-based controller is demonstrated to be more efficient on hardware and requires less memory compared to both artificial neural networks and model predictive controllers. This controller estimates the number of room occupants by using the information from wireless sensor nodes placed in the Heating, Ventilation and Air Conditioning (HVAC) duct and the room. For an occupied room, the controller can switch between thermal comfort mode (based on predicted mean vote set points) and user defined mode (i.e. occupant defined set points for heating/cooling/ventilation). Furthermore, the hybrid particle swarm optimisation with sequential quadratic programming training algorithms are used (for the first time to the best of the authors' knowledge) for training the RNN and results show that this algorithm outperforms the widely used gradient descent algorithm for RNN. The results show that occupancy estimation by smart controller is 83.08% accurate.

AB - Monitoring and analysis of energy use and indoor environmental conditions is an urgent need in large buildings to respond to changing conditions in an efficient manner. Correct estimation of occupancy will further improve energy performance. In this work, a smart controller for maintaining a comfortable environment using multiple random neural networks (RNNs) has been developed. The implementation of RNN-based controller is demonstrated to be more efficient on hardware and requires less memory compared to both artificial neural networks and model predictive controllers. This controller estimates the number of room occupants by using the information from wireless sensor nodes placed in the Heating, Ventilation and Air Conditioning (HVAC) duct and the room. For an occupied room, the controller can switch between thermal comfort mode (based on predicted mean vote set points) and user defined mode (i.e. occupant defined set points for heating/cooling/ventilation). Furthermore, the hybrid particle swarm optimisation with sequential quadratic programming training algorithms are used (for the first time to the best of the authors' knowledge) for training the RNN and results show that this algorithm outperforms the widely used gradient descent algorithm for RNN. The results show that occupancy estimation by smart controller is 83.08% accurate.

KW - energy usage monitoring

KW - occupancy estimation

KW - comfortable environment

KW - thermal comfort mode

KW - artificial neural networks

KW - energy usage analysis

KW - random neural network smart controller

KW - sequential quadratic programming training algorithms

KW - heating-ventilation-air conditioning

KW - indoor environment

KW - building energy control system

KW - gradient descent algorithm

KW - environmental conditions

KW - experimental testing

KW - model predictive controllers

KW - single zone test chamber

KW - RNN-based controller

KW - HVAC duct

KW - wireless sensor nodes

KW - hybrid particle swarm optimisation

KW - predicted mean vote-based set points

U2 - 10.1049/iet-net.2015.0020

DO - 10.1049/iet-net.2015.0020

M3 - Article

VL - 4

SP - 350

EP - 358

JO - IET Networks

JF - IET Networks

SN - 2047-4954

IS - 6

ER -