Call for Papers - International Journal of Intelligent Systems and Applicatio...
Presentation_PhD_DOCEIS_RicardoSantos_.pptx
1. A New Approach to Provide Sustainable Solutions
for Residential Sector
Ricardo Santos 1 Antonio Abreu 3
João Matias 2
08-10 May, 2019 | Caparica, Lisbon – Portugal
1 University of Aveiro, Portugal 2Dept. of Economics, Management, Industrial
Engineering and Tourism (DEGEIT)
University of Aveiro, Portugal
C-MasT – University of Beira Interior
Govcopp - University of Aveiro
3 ISEL- Instituto Superior de Engenharia de Lisboa,
Instituto Politécnico de Lisboa
CTS Uninova, Faculdade de Ciências e Tecnologia,
Universidade Nova de Lisboa, Portugal
ajfa@dem.isel.ipl.pt
Technological Innovation for Industrial and
Service Systems
2. • Building sector
Represents about
30-40 % final energy
consumed
• 27% of final energy
(building sector)
concerns to residential
group
Potential energy savings on residential sector
INE/DGEG, 2010
Fonte: EIA,2017
1.Introduction|Motivation
10th Doctoral Conference on Computing, Electrical and
Industrial Systems
Technological Innovation for Industrial
and Service Systems 08-10 May, 2019 | Caparica, Lisbon – Portugal
Ricardo Santos | João Matias |Antonio Abreu
Energy Efficiency in buildings by using evolutionary algorithms 2
• High “weight” of
electrical energy
on total consumer
energy expenses
Fonte: EIA,2017
• High
dependency on
electrical
energy
producion
through fóssil
fuels
INE/DGEG, 2010
3. 3
1.Introduction|Motivation
Research Question
Is it possible to develop a holistic model (Economic, Social and Environmental dimensions) to
support decision making, based on evolutionary algorithms (EA), that allows the decision
agent to obtain sustainable solutions?
H1: If by using Multi-objective Evolutionary Algorithms (EA), combined with Multicriteria analysis, it’s
possible to achieve several sustainable solutions to the consumer, by selecting household appliances
from the market.
Hypothesis
10th Doctoral Conference on Computing, Electrical and
Industrial Systems
Technological Innovation for Industrial
and Service Systems 08-10 May, 2019 | Caparica, Lisbon – Portugal
Ricardo Santos | João Matias |Antonio Abreu
Energy Efficiency in buildings by using evolutionary algorithms
4. 4
2.Relationship to industrial and service systems
Some reasons to integrate resilience…
To be resilient to that changes, the system should include the
capability to adapt and preserve at the same time its original
functions.
• due to the technology innovation, economic crisis (consumer
surplus), and climate change.
This occurs particularly with energy services….
• increased rate of disruptive events regarding energy area (e.g.
climate change, economic crisis, technological evolution, etc.)
challenging the way that systems are designed.
10th Doctoral Conference on Computing, Electrical and
Industrial Systems
Technological Innovation for Industrial
and Service Systems 08-10 May, 2019 | Caparica, Lisbon – Portugal
Ricardo Santos | João Matias |Antonio Abreu
Energy Efficiency in buildings by using evolutionary algorithms
5. 5
3.Research method
Proposed
Approach
10th Doctoral Conference on Computing, Electrical and
Industrial Systems
Technological Innovation for Industrial
and Service Systems 08-10 May, 2019 | Caparica, Lisbon – Portugal
Ricardo Santos | João Matias |Antonio Abreu
Energy Efficiency in buildings by using evolutionary algorithms
6. 6
3.Research method
Attributes used to define problem dimensions, regarding each energy service considered
Dimension Criteria used Quantity
Air
Conditioner
types of air conditioner considered:
zone to be heated/ cooled by the air
conditioner
minimum capacity required
wall (mono
split)
wall (multisplit)
Portable
living room
9905,6 BTU
Washing
Machine
capacity, based on the number of
household’s occupants [18]
7 kg
Dishwasher
machine
load capacity. 12 cutlery
Oven
useful volume, available for cooking
according to the nr. of occupants [18]
47 cm x 68 cm
Dryer machine
type of dryer machines
load capacity from [18]
by exhaust ;
7 kg
Lighting
technology Halogen
CFL
Fluorescent
Led
Refrigerator
capacity of the refrigerators [18]
type of refrigerator according to the
number of occupants [18]
150 liters
refrigerator
Combined
type.
10th Doctoral Conference on Computing, Electrical and
Industrial Systems
Technological Innovation for Industrial
and Service Systems 08-10 May, 2019 | Caparica, Lisbon – Portugal
Ricardo Santos | João Matias |Antonio Abreu
Energy Efficiency in buildings by using evolutionary algorithms
7. 7
3.Research method
Attributes, used to define problem dimensions, regarding each energy service considered
10th Doctoral Conference on Computing, Electrical and
Industrial Systems
Technological Innovation for Industrial
and Service Systems 08-10 May, 2019 | Caparica, Lisbon – Portugal
Ricardo Santos | João Matias |Antonio Abreu
Energy Efficiency in buildings by using evolutionary algorithms
8. 8
3.Research method
Problem formulation
10th Doctoral Conference on Computing, Electrical and
Industrial Systems
Technological Innovation for Industrial
and Service Systems 08-10 May, 2019 | Caparica, Lisbon – Portugal
Ricardo Santos | João Matias |Antonio Abreu
Energy Efficiency in buildings by using evolutionary algorithms
( ) ( ) ( )
( )
jt jt jt
g g g
ij ij ij
x v x
Based on MAVT, there is a value 𝑣𝑖𝑗
(𝑔𝑗𝑡)
𝑥𝑖𝑗
(𝑔𝑗𝑡)
, associated to the attribute 𝑥𝑖𝑗
(𝑔𝑗𝑡)
, such as:
(g ) (g )
(1)
.
(2)
(g ) (g )
better. .
( ) ( )
( )
( ) ( )
jt jt
ij
jt jt
ij ij
ij ij worst ij
ij ij
ij worst ij
v x v x
v x
v x v x
Scaling function, to preform 𝑣𝑖𝑗
(1)
𝑥𝑖𝑗
(𝑔𝑗𝑡)
𝑣𝑖𝑗
(2)
𝑥𝑖𝑗
(𝑔𝑗𝑡)
transformation:
Evaluation table, belonging to each energy service j.
9. 9
3.Research method
Problem formulation
10th Doctoral Conference on Computing, Electrical and
Industrial Systems
Technological Innovation for Industrial
and Service Systems 08-10 May, 2019 | Caparica, Lisbon – Portugal
Ricardo Santos | João Matias |Antonio Abreu
Energy Efficiency in buildings by using evolutionary algorithms
Example of evaluation table (Lighting energy service))
𝑥𝑗
(𝑔𝑗𝑡)
𝑣𝑖𝑗 𝑥𝑗
𝑔𝑗𝑡
10. 10
3.Research method
Problem dimensions and case study
Washing
machine Dishwasher
Refrigerator Oven
Air Conditioner
Lighting Dryer machine
x12 x13 x14 x15 x16 x17
x11
x22 x23 x24 x25 x26 x27
x21
x32 x33 x34 x35 x36 x37
x31
x102 x103 x104 x105 x106 x107
x101
Option
i
Solutions
Option
Xij
J Dimension
j
i
Consumer’s decisions space
10th Doctoral Conference on Computing, Electrical and
Industrial Systems
Technological Innovation for Industrial
and Service Systems 08-10 May, 2019 | Caparica, Lisbon – Portugal
Ricardo Santos | João Matias |Antonio Abreu
Energy Efficiency in buildings by using evolutionary algorithms
11. 11
3.Research method
Problem formulation
subject to
10th Doctoral Conference on Computing, Electrical and
Industrial Systems
Technological Innovation for Industrial
and Service Systems 08-10 May, 2019 | Caparica, Lisbon – Portugal
Ricardo Santos | João Matias |Antonio Abreu
Energy Efficiency in buildings by using evolutionary algorithms
max , / ,
/ ( ) ( ), ( )
m
T
m A B
V x c m A B
subject to x X c V x V x V x
with
( ) (B )
1 1 1
( ), ( ) . ( ) . ( ) ( ) ( )
A B
j j j
jt jt
n n
n
A
A B A A B B A j j B j j
j t t
V V x V x V x V x v x v x
dim dim
1 . .
1 1
: ( ) jt
n n
A
j j disp j disp
j j
r I x availablebudget x
Economic – Budget:
Environment – Noise
: . .
jt
B
j j j j j
r Noise Max Noise x Max Noise
Objective functions:
( )
1 1
:max ( ) ( )
A
j jt
jt
n
n
A
A j j
j t
Economic Well being V x v x
(B )
1 1
:max ( ) ( )
B
j jt
jt
n
n
B j j
j t
Environment Well being V x v x
12. 12
3.Research method
• NSGAs adopted parameters:
NSGAII’s individual framework (codification)
Optimization technique used: Non Sorting Genetic Algorithm (NSGA)
• Real codification
Initial population: 100 individuals
Tournament method
Single crossover point method
Crossover rate: 0,45
Mutation operator: Normal Random
Mutation rate: 0,015
10th Doctoral Conference on Computing, Electrical and
Industrial Systems
Technological Innovation for Industrial
and Service Systems 08-10 May, 2019 | Caparica, Lisbon – Portugal
Ricardo Santos | João Matias |Antonio Abreu
Energy Efficiency in buildings by using evolutionary algorithms
13. 13
4.Simulation & Results
• 4 Occupants
• 7 Energy services to acquire (Lighting,
Air Conditioner, Washing Machine,
Dryer Machine, Oven, Refrigerator.
Dishwasher Machine)
• Available budget: 2700 €
Case study:
10th Doctoral Conference on Computing, Electrical and
Industrial Systems
Technological Innovation for Industrial
and Service Systems 08-10 May, 2019 | Caparica, Lisbon – Portugal
Ricardo Santos | João Matias |Antonio Abreu
Energy Efficiency in buildings by using evolutionary algorithms
17 m
16
m
• Air conditioner and lighting are
dependent on house dimensions
14. 14
4.Simulation & Results
Stopping criteria
• Nr. Of iterations/generations: 100
• Convergence
Software implementation: Matlab code, importing data from excel
10th Doctoral Conference on Computing, Electrical and
Industrial Systems
Technological Innovation for Industrial
and Service Systems 08-10 May, 2019 | Caparica, Lisbon – Portugal
Ricardo Santos | João Matias |Antonio Abreu
Energy Efficiency in buildings by using evolutionary algorithms
Optimization method
Electrical appliances data-base
Ms Excel Matlab code
15. 15
4.Simulation & Results
Pareto Frontier
10th Doctoral Conference on Computing, Electrical and
Industrial Systems
Technological Innovation for Industrial
and Service Systems 08-10 May, 2019 | Caparica, Lisbon – Portugal
Ricardo Santos | João Matias |Antonio Abreu
Energy Efficiency in buildings by using evolutionary algorithms
for 90th and 100th generations different parameters (90th Generation)
16. 16
Example of an efficient solution
4.Simulation & Results
10th Doctoral Conference on Computing, Electrical and
Industrial Systems
Technological Innovation for Industrial
and Service Systems 08-10 May, 2019 | Caparica, Lisbon – Portugal
Ricardo Santos | João Matias |Antonio Abreu
Energy Efficiency in buildings by using evolutionary algorithms
if the consumer, opts for the solutions set, provided by this approach, he can save
up to 2112,3 € per year, further contributing to a 1458,9 kg of CO2 and 740 litres
of water, both savings/year, given the 10 years life cycle considered
17. 17
• An approach was presented to provide sustainable household appliances from the
market, to the decision-agent (consumer), by considering two objectives,
regarding sustainability; environment and economic well-being.
• The main objective, was to maximize the consumer well-being (environment and
economics) regarding the lifecycle of each appliance, during its usage phase.
• The social wellbeing was also promoted, by suit the obtained solutions to the
consumer needs.
• The relative importance, given by the DA (consumer), was also considered, in
order to weight the DA decision through both dimensions to be maximized;
• NSGA-II can also find the Pareto frontier of the solutions, providing therefore
several alternative solutions to the consumer.
• The achievements presented on this work, allows to proceed in a way of getting
a more completed approach that maximizes both dimensions of sustainability
(Economical, Environmental and Social) with the social dimension, being
therefore integrated into the model developed here, in order to better express
the consumer preferences
5.Conclusions
10th Doctoral Conference on Computing, Electrical and
Industrial Systems
Technological Innovation for Industrial
and Service Systems 08-10 May, 2019 | Caparica, Lisbon – Portugal
Ricardo Santos | João Matias |Antonio Abreu
Energy Efficiency in buildings by using evolutionary algorithms
18. 18
Ricardo Santos 1 Antonio Abreu 3
João Matias 2
02-04 May, 2018 | Caparica, Lisbon – Portugal
1 University of Aveiro, Portugal 2Dept. of Economics, Management, Industrial
Engineering and Tourism (DEGEIT)
University of Aveiro, Portugal
C-MasT – University of Beira Interior
Govcopp - University of Aveiro
3 ISEL- Instituto Superior de Engenharia de Lisboa,
Instituto Politécnico de Lisboa
CTS Uninova, Faculdade de Ciências e Tecnologia,
Universidade Nova de Lisboa, Portugal
ajfa@dem.isel.ipl.pt
Thank you
A New Approach to Provide Sustainable Solutions for Residential
Sector
Editor's Notes
As it mentioned before, the industry and service sectors are going through profound transformation towards digitalization and integration of new levels of “smartness”, originating therefore the 4th industrial revolution.
This transformation, is led by terms such as Industry 4.0, Smart Manufacturing and Economy 4.0, giving therefore an interdisciplinary character, expressed by an increasing digitalization and interconnection of systems, products, services and business models. The link between the physical and the cyber worlds, as well as the integration of the “exponential technologies”, are key features of this innovation trend.
The approach presented here, was developed to support a Decision-Agent (DA), who wants to acquire a set of electrical appliances (energy services) from the market.
On Fig.1, it is presented an approach to provide an optimal set of appliances, regarding each energy service, needed by the DA (e.g. Consumer
The available solutions were pre-selected, according to a set of criteria (Table 1), and based on number of occupants /consumers, to improve the use of the efficient appliances. In this work, it was considered four (e.g. family) occupants.
Additionally, the consumer wants to explore the consequences of his choices in future, regarding those made on present, and by predicting an eventual situation of indirect rebound effect, between light and air conditioner appliances.
This can be estimated by using scenario simulations during the air conditioner life cycle (previously selected), and by making changes around different light appliances (Table 2) according to a set of scenarios (Table 3).
The consumption profile was performed, by making a set of assumptions based on the hours, which was then extrapolated to a weekly and year base. However, the decision-agent (consumer) can also define its usage profile according to its needs, or by using the profile, considered in the case study presented here, as a default.
This attributes, are usually measured on different measurement scales. Therefore, in order to transform the criteria to follow the same scale and units, it was used an expression to stablish the relationship between the new and the previous value of 𝑥 𝑖𝑗 ( 𝑔 𝑗𝑡 ) , respective
Based on the value attributes previously achieved, it was used the additive model to aggregate them, referred to each option i, regarding energy service j, which was further improved, by applying optimization techniques, by using NSGAII algorithm.
The approach presented here, was developed to support a Decision-Agent (DA), who wants to acquire a set of electrical appliances (energy services) from the market.
On Fig.1, it is presented an approach to provide an optimal set of appliances, regarding each energy service, needed by the DA (e.g. Consumer
Based on the diagram of Fig.1, the decision variables can be defined as:
Where the objective is to maximize the savings for the consumer:
According to [6], it was considered the following objective function ( 𝑉 𝑅 ):
The codification used, was binary, given the complexity with the use of real formulation to tackle this problem, mainly in terms of the additional number of constraints.
The initial population was 100 individuals, and the parents were selected by using Roulette method, with a single crossover point method and crossover rate of 0,45.
The selected mutation operator, was Normal Random with a mutation rate of 0,01.
The model will be applied by using a case study, and considering a consumer (DA), who wishes to acquire 7 types of appliances. His building has four occupants (him included) and the relative importance’s ( 𝜔 𝐴 and 𝜔 𝐵 ) are respectively 0,7 and 0,3. All the remaining assumptions
The remained NSGA-II parameters (initial population, crossover and mutation rate) were defined after several computational experiments.
The parameter of max generations, was tested at first, where it was selected a maximum generation number of 90 to show that if 90 iterations were enough to find the Pareto frontier. Other parameters were also tested, such as the population size (100 individuals),the tournament size (10), the crossover rate (0.9), and the mutation rate (0.3). The obtained results, regarding 90th and 100th generations, are presented on Fig. 6 a), where it can be seen that both cases, have similar Pareto frontier. In this sense it was selected the max number iterations/generations of 90. Then, it was preformed several combinations of crossover and mutation rates of NSGA-II
On Table 4, it is presented an example of a feasible solution obtained from GAs, as well as its savings in terms of CO2 emissions, for a budget of 2600 Euros.
CO2 savings were calculated by using a carbon footprint indicator (emission factor), obtained from [20].
According to Table 4, the consumer can save up to € 215.6 (€ 1,447.1 € - 1231.5), corresponding to a reduction of 1452,9 kg of CO2 and during a period of 10 years.
In this work, it was presented an approach to provide sustainable household appliances from the market, to the decision-agent (consumer), by considering two objectives,