-
Notifications
You must be signed in to change notification settings - Fork 3
/
chap-cronograma.tex
200 lines (175 loc) · 14.2 KB
/
chap-cronograma.tex
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
\chapter{\label{chap:cronograma}Cronograma}
\section{\label{chap:stages}Etapas do trabalho}
O Trabalho de Conclusão é dividido em duas etapas distintas ao longo de dois
semestres do curso de Ciência da Computação. Durante o primeiro semestre,
envolve a contextualização na área de conhecimento para definir objetivos e
elaborar uma fundamentação teórica e a modelagem do que se deseja implementar.
Já a segunda etapa, durante o segundo semestre, envolve a implementação
utilizando ferramentas de software e a análise dos resultados de modo a obter
conclusões acerca dos objetivos definidos na primeira etapa~-~se foram atingidos
e por quê.
Os objetivos do trabalho foram segmentados da seguinte forma:
\begin{enumerate}
\item \textbf{Objetivos da Primeira Etapa}: fundamentação teórica e modelagem
\begin{enumerate}[label*=\arabic*.]
\item Estudo do funcionamento de elevadores e suas possibilidades de
comportamento;
\item Estudo de métodos de inteligência artificial utilizados em ou
propostos para elevadores;
\item Definição de objetivos e cenários de testes;
\item Estudo de padrões de projeto de simuladores;
\item Realização da modelagem do simulador de elevadores;
\end{enumerate}
\item \textbf{Objetivos da Segunda Etapa}: implementação, análise e conclusões
\begin{enumerate}[label*=\arabic*.]
\item Implementação do simulador;
\item Implementação de algoritmos de Inteligência Artificial;
\item Realização de simulações nos cenários pré-definidos;
\item Análise de resultados e tomada de conclusões.
\end{enumerate}
\end{enumerate}
\section{\label{section:schedule}Cronograma}
A execução do trabalho será dividida em iterações de duas semanas, respeitando
as etapas previamente definidas. Cada iteração terá um objetivo claramente
definido, conforme ilustrado pela tabela~\ref{tab:cronograma}.
\begin{table}[htb!]
\centering
\caption{Objetivos por iteração.}
\label{tab:cronograma}
\begin{tabular}{c|c|c|c|c|c|c|c|c|c|c|c|c|c|c|c|c|}
\cline{2-17}
{\bf } & \multicolumn{2}{c|}{{\bf Ago/15}} & \multicolumn{2}{c|}{{\bf Set/15}} & \multicolumn{2}{c|}{{\bf Out/15}} & \multicolumn{2}{c|}{{\bf Nov/15}} & \multicolumn{2}{c|}{{\bf Mar/16}} & \multicolumn{2}{c|}{{\bf Abr/16}} & \multicolumn{2}{c|}{{\bf Mai/16}} & \multicolumn{2}{c|}{{\bf Jun/16}} \\ \hline
\multicolumn{1}{|c|}{{\bf Objetivo}} & {\bf 1} & {\bf 2} & {\bf 3} & {\bf 4} & {\bf 5} & {\bf 6} & {\bf 7} & {\bf 8} & {\bf 9} & {\bf 10} & {\bf 11} & {\bf 12} & {\bf 13} & {\bf 14} & {\bf 15} & {\bf 16} \\ \hline
\multicolumn{1}{|c|}{1.1} & {\bf X} & {\bf X} & & & & & & & & & & & & & & \\ \hline
\multicolumn{1}{|c|}{1.2} & & {\bf X} & {\bf X} & & & & & & & & & & & & & \\ \hline
\multicolumn{1}{|c|}{1.3} & & {\bf X} & {\bf X} & & & & & & & & & & & & & \\ \hline
\multicolumn{1}{|c|}{1.4} & & & & {\bf X} & & & & & & & & & & & & \\ \hline
\multicolumn{1}{|c|}{1.5} & & & & & {\bf X} & {\bf X} & {\bf X} & {\bf X} & & & & & & & & \\ \hline
\multicolumn{1}{|c|}{2.1} & & & & & & & & & {\bf X} & {\bf X} & {\bf X} & & & & & \\ \hline
\multicolumn{1}{|c|}{2.2} & & & & & & & & & & & {\bf X} & {\bf X} & & & & \\ \hline
\multicolumn{1}{|c|}{2.3} & & & & & & & & & & & & & {\bf X} & {\bf X} & & \\ \hline
\multicolumn{1}{|c|}{2.4} & & & & & & & & & & & & & & & {\bf X} & {\bf X} \\ \hline
\end{tabular}
\end{table}
\section{\label{section:objectives}Detalhamento de objetivos}
\subsection{Primeira etapa}
O objetivo da primeira etapa foi realizar a fundamentação teórica e produzir
artefatos de modelagem de software. Os sub-objetivos da etapa são detalhados a
seguir:
\setcounter{enumi}{1}
\begin{enumerate}[label*=\arabic*.]
\item Estudo do funcionamento de elevadores e suas possibilidades de
comportamento
\begin{description}[leftmargin=!,labelwidth=\widthof{\bfseries Descrição}]
\item[Descrição] Pesquisa e leitura de artigos científicos e literatura
especializada em elevadores; compreensão das
necessidades, demandas e possibilidades da área;
elaboração da introdução deste trabalho.
\item[Iterações] 1 e 2
\item[Período] 1 a 30 de agosto de 2015
\end{description}
\item Estudo de métodos de inteligência artificial utilizados em ou
propostos para elevadores;
\begin{description}[leftmargin=!,labelwidth=\widthof{\bfseries Descrição}]
\item[Descrição] Pesquisa e leitura de artigos científicos que aplicaram
conceitos e técnicas de Inteligência Artificial a
elevadores; elaboração da descrição do problema deste
trabalho.
\item[Iterações] 2 e 3
\item[Período] 15 de agosto a 15 de setembro de 2015
\end{description}
\item Definição de objetivos e cenários de testes;
\begin{description}[leftmargin=!,labelwidth=\widthof{\bfseries Descrição}]
\item[Descrição] Definição dos objetivos das etapas deste trabalho;
pesquisa de dados praticados na área para
parametrização de cenários; definição dos cenários de
testes.
\item[Iterações] 2 e 3
\item[Período] 15 de agosto a 15 de setembro de 2015
\end{description}
\item Estudo de padrões de projeto de simuladores;
\begin{description}[leftmargin=!,labelwidth=\widthof{\bfseries Descrição}]
\item[Descrição] Pesquisa e estudo de \textit{patterns} e
\textit{anti-patterns} já utilizados e validados
para implementação de simuladores~\cite{Law,Banks,Sim}.
\item[Iterações] 4
\item[Período] 15 a 30 de setembro de 2015
\end{description}
\item Realização da modelagem do simulador de elevadores;
\begin{description}[leftmargin=!,labelwidth=\widthof{\bfseries Descrição}]
\item[Descrição] Modelagem de um simulador de elevadores baseado no
estudo realizado na iteração anterior e produção de
artefatos de engenharia de software.
\item[Iterações] 5, 6, 7 e 8
\item[Período] 01 de outubro a 30 de novembro de 2015
\end{description}
\end{enumerate}
\subsection{Segunda etapa}
O objetivo da segunda etapa é colocar em prática a fundamentação teórica e modelagem construídas na primeira etapa do trabalho. Isso envolve a implementação dos componentes de software projetados, sua execução e análise dos resultados. Os sub-objetivos da etapa são detalhados a seguir:
\begin{enumerate}[label*=\arabic*.]
\item Implementação do simulador
\begin{description}[leftmargin=!,labelwidth=\widthof{\bfseries Descrição}]
\item[Descrição] Definição de plataforma de desenvolvimento; escolha da
linguagem de programação utilizada; desenvolvimento do
simulador; desenvolvimento de testes unitários.
\item[Iterações] 9, 10 e 11
\item[Período] 01 de março a 15 de abril de 2016
\end{description}
\item Implementação de algoritmos de Inteligência Artificial
\begin{description}[leftmargin=!,labelwidth=\widthof{\bfseries Descrição}]
\item[Descrição] Implementação das abordagens multi-agentes e
aprendizado de máquina para utilização junto ao
simulador implementado nas iterações anteriores.
\item[Iterações] 11 e 12
\item[Período] 01 de abril a 30 de abril de 2016
\end{description}
\item Realização de simulações nos cenários pré-definidos
\begin{description}[leftmargin=!,labelwidth=\widthof{\bfseries Descrição}]
\item[Descrição] Simulações utilizando os algoritmos de Inteligência
Artificial e levantamento de resultados.
\item[Iterações] 13 e 14
\item[Período] 01 de maio a 30 de maio de 2016
\end{description}
\item Análise de resultados e tomada de conclusões
\begin{description}[leftmargin=!,labelwidth=\widthof{\bfseries Descrição}]
\item[Descrição] Análise dos resultados das simulações e avaliação
acerca das combinações de estratégias e cenários;
proposta de estratégias mais adequadas; avaliação se
os objetivos foram atingidos ou não.
\item[Iterações] 15 e 16
\item[Período] 01 de junho a 30 de junho de 2016
\end{description}
\end{enumerate}
\section{\label{sec:proposal:out-of-scope}Técnicas fora do escopo}
Duas técnicas princiais estão fora do escopo deste trabalho, por motivos de
limitação de tempo. No entanto, sua abordagem seria interessante.
\subsection{\label{section:multiagentes}Sistemas multi-agentes}
Um \textit{agente} é um sistema computacional que está situado em um \textit{ambiente} e é capaz de tomar \textit{ações autônomas} neste ambiente de modo a atingir seus \textit{objetivos}~\cite{Woolridge:2001:IMS:559667}. Ou seja, dadas as condições do ambiente (entradas), o agente é capaz de decidir qual ação tomar (saídas). Logo, qualquer \textit{sistema de controle} pode ser considerado um agente.
Além disto, estes agentes podem ser inteligentes. A inteligência de um agente pode ser classificada das seguintes formas:
\begin{description}[leftmargin=!,labelwidth=\widthof{\bfseries Pró-ativa}]
\item[Reativa] O agente percebe o ambiente e responde à mudanças no estado
do sistema de forma a atingir seus objetivos;
\item[Pró-ativa] Ao agente é permitida a iniciativa para tomar ações sem que
necessariamente ocorra uma mudança no ambiente
anterioremente;
\item[Social] Os agentes são capazes de interagir entre si de modo a
atingirem seus objetivos.
\end{description}
Estes conceitos podem ser aplicados a um sistema de controle de elevadores
descentralizado, removendo toda a tarefa de coordenação do EGCS e trazendo-a
para dentro dos elevadores, onde cada um é um agente inteligente. Portanto,
julga-se que uma modelagem multi-agentes pode ser interessante no contexto.
\subsection{\label{section:machinelearning}Aprendizado de máquina}
As técnicas de aprendizado de máquina são especialmente interessantes para a
solução deste problema por lidarem dinamicamente com a mudança de comportamento
do cenário~\cite{Russell:2003:AIM:773294}.
Entende-se que os padrões de tráfego de um prédio podem mudar ao longo do tempo,
e isto pode alterar drasticamente os requisitos de alocação de elevadores. Por
exemplo, uma nova empresa pode se instalar em um prédio comercial, utilizando
dois andares. Subitamente, o tráfego~\textit{interfloor}\footnote{Tráfego entre
andares, oposto a tráfego do \textit{lobby} para um andar ou de um andar para
o \textit{lobby}.} se torna significativo em algumas horas do dia, quando antes
era desprezível.
Além de lidarem bem com mudanças graduais, as tecnicas de aprendizado de máquina
tendem a obter resultados melhores com o tempo, conforme vão descobrindo padrões
existentes, mas que não foram anteriormente notados.