update slides

slides/main.md

 ---
 title: 'Under Control: A Control Theory Introduction for Computer Scientists'
 short_title: 'Under Control: A Control Theory Introduction for Computer Scientists'
-subtitle: 'First try'
-date: 21/04/2023
+subtitle: 'Session \@ LIG'
+date: 13/06/2023
 authors:
   - firstname: "Quentin"
     lastname: "Guilloteau"
@@ -63,7 +63,7 @@ header-includes:
     
   - $\rightsquigarrow$ unpredictable
 
-  - but want to guarantee a behavior, a Quality-of-Serive
+  - but want to guarantee a behavior, a Quality-of-Service
   
 - Usual approaches:
 
@@ -115,15 +115,15 @@ Use the knowledge about the current state of the system to respond
   
 - main tool: MAPE-K loop
 
-- $\neq$ implementations of AC (rules, AI, control)
-
 - Separation of concerns
 
+- $\neq$ implementations of AC (rules, AI, control)
+
 
 :::
 ::::::::::::::
 
-# Examples
+# Some Examples
 
 - Regulate the heat of a processor based on its frequency
 
@@ -131,17 +131,23 @@ Use the knowledge about the current state of the system to respond
   
   - Actuator: DVFS
   
+  - Reference: Desired CPU temperature
+  
 - Regulate the waiting time of users based on the number of available servers
 
   - Sensor: Waiting time of a request
   
   - Actuator: Number of servers to add/remove
   
+  - Reference: Mean waiting time for the requests
+  
 - Regulate the FPS of an online video rendering based on the quality
 
   - Sensor: FPS
   
   - Actuator: depth of computation
+  
+  - Reference: 60 FPS
 
 
 
@@ -192,7 +198,7 @@ Map the error to the next input to reach desired system state with guarantees
 
 - Feedforward (proactive reaction to disturbances)
 
-- Adaptive (Change behavior at runtime)
+- Adaptive (change behavior at runtime)
 
 - Model Predictive
 
@@ -208,9 +214,10 @@ Map the error to the next input to reach desired system state with guarantees
 
 ![Control Theory Methodology](figs/methodo.pdf){width=90% height=50%}
 
-# The (famous) PID Controller
 
-\begin{block}{First, \textbf{a Model ...} (i.e. how does the system behave (Open-Loop))}
+# The (famous) PID Controller (discretized)
+
+\begin{block}{First, \textbf{a Model ...} (i.e., how does the system behave (Open-Loop))}
 
 \begin{center}
 \scalebox{0.85}{
@@ -219,7 +226,7 @@ Map the error to the next input to reach desired system state with guarantees
 \end{center}
 \end{block}
 
-\begin{block}{... then \textbf{a PID Controller} (i.e. the Closed-Loop behavior)}
+\begin{block}{... then \textbf{a PID Controller} (i.e., the Closed-Loop behavior)}
 \begin{center}
 \scalebox{0.8}{
 $\displaystyle Output = \textbf{K}_p \times Error + \textbf{K}_i \times \sum_k Error_k + \textbf{K}_d \times \left(Error_k - Error_{k-1}\right)$
@@ -266,7 +273,7 @@ The controller gains define this behavior!
 
 1. Play with a dummy system
 
-2. Implement a naive "Bang-Bang" controller 
+2. Implement a naive Threshold-based controller 
 
 3. First introduction with Control Theory: P Controller
 
@@ -281,6 +288,6 @@ The controller gains define this behavior!
 
 
 \begin{center}
-\url{https://tinyurl.com/CtrlComputing}
+\url{https://tinyurl.com/Control4Computing}
 \end{center}