Artificial Intelligence for Optimisation and Demand Side Response of Built Environment Management
Lead Research Organisation:
Brunel University London
Department Name: Mechanical and Aerospace Engineering
Abstract
The project will be looking at the improvement of aspects of Mitie's existing Building Analytics architecture
and at furthering its Demand Side Response capability. Both of these areas are seen as a key part of Mitie's
energy strategy to deliver not only energy solutions, but also a fully-connected workplace.
One of the two aspects that the project will look to enhance is Building Analytics core (BA). A big problem
with BA is initial data acquisition and translation to onboard new client's data into the BA rules engine. The
goal is to replace the manual element with mathematical techniques and self-learning algorithms.
As an approach to this machine learning problem, Deep Learning techniques (Neural Networks) are going to
be implemented in order to build an accurate model which responds well to the data with a minimun error,
and to make the corresponding decisions in order to enhance the system's capabilities, as well as to ensure a
statistical monitoring afterwards, to continuously improve the system's predictions.
As prerequisites for implementing this method, I will be provided with significant historical data to train the
algorithm and the company has a recurring need for predicting a behaviour in order to cut costs and create
value.
and at furthering its Demand Side Response capability. Both of these areas are seen as a key part of Mitie's
energy strategy to deliver not only energy solutions, but also a fully-connected workplace.
One of the two aspects that the project will look to enhance is Building Analytics core (BA). A big problem
with BA is initial data acquisition and translation to onboard new client's data into the BA rules engine. The
goal is to replace the manual element with mathematical techniques and self-learning algorithms.
As an approach to this machine learning problem, Deep Learning techniques (Neural Networks) are going to
be implemented in order to build an accurate model which responds well to the data with a minimun error,
and to make the corresponding decisions in order to enhance the system's capabilities, as well as to ensure a
statistical monitoring afterwards, to continuously improve the system's predictions.
As prerequisites for implementing this method, I will be provided with significant historical data to train the
algorithm and the company has a recurring need for predicting a behaviour in order to cut costs and create
value.
People |
ORCID iD |
JOSE JIMENEZ (Student) |
Publications
Mesa Jiménez J
(2020)
Modelling energy demand response using long short-term memory neural networks
in Energy Efficiency
Mesa-Jiménez J
(2020)
Machine learning for BMS analysis and optimisation
in Engineering Research Express
Studentship Projects
Project Reference | Relationship | Related To | Start | End | Student Name |
---|---|---|---|---|---|
EP/N509437/1 | 30/09/2016 | 29/09/2021 | |||
2295162 | Studentship | EP/N509437/1 | 30/09/2017 | 29/09/2020 | JOSE JIMENEZ |
EP/R512990/1 | 30/09/2018 | 29/09/2023 | |||
2295162 | Studentship | EP/R512990/1 | 30/09/2017 | 29/09/2020 | JOSE JIMENEZ |