Generated Images and updated some slides.

report_slides/typst/main.typ

@@ -20,7 +20,7 @@
   box-task-title: none
 )
 
-#show figure.caption: set text(size: 15pt)
+#show figure.caption: set text(size: 14pt)
 
 #slide[
   = Overview
@@ -33,7 +33,7 @@
 ]
 
 #slide[
-  = What is Iterative Learning Control ?
+  = Iterative Learning Control in a nutshell
 
   Iterative Learning Control @bristowSurveyIterativeLearning2006 (ILC) generally concerns the control of a repeated task. It does so by:
 
@@ -58,12 +58,6 @@
 
 ]
 
-#slide[
-  = Contributions of the paper
-  - Treatment of the case in which *\#inputs<\#outputs*
-  - 
-]
-
 #slide[
   == System Set up
   #task[
@@ -100,8 +94,19 @@
   = Simulations
   #side-by-side()[  
     == Carts
+
+    #image("images/MSD.png")
+
+    Mass-spring-damper system actuated just on the last cart
+
   ][
     == Basketball in the wind
+
+  #side-by-side()[
+    #image("images/freethrow_wind.svg")
+  ][
+    Basket hoops in 
+  ]
   ]
 
 ]
@@ -136,6 +141,36 @@
   ]
 ]
 
+#slide()[
+  ===  Basketball - *$d tilde(T) = 0.2s$*
+  #side-by-side()[
+    #figure(
+      image("images/basket_8iters_0.2dT.svg", fit: "contain"),
+      caption: "8 Iterations",
+    )
+  ][
+    #figure(
+      image("images/basket_100iters_0.2dT.svg", fit: "contain"),
+      caption: "100 Iterations",
+    )
+  ]
+]
+
+#slide()[
+  ===  Basketball - *$d tilde(T)$* = 0.5s
+  #side-by-side()[
+    #figure(
+      image("images/basket_8iters_0.5dT.svg", fit: "contain"),
+      caption: "8 Iterations",
+    )
+  ][
+    #figure(
+      image("images/basket_100iters_0.5dT.svg", fit: "contain"),
+      caption: "100 Iterations",
+    )
+  ]
+]
+
 
 #slide[
   #set text(size: 0.7em)

scripts/basketball.jl

@@ -42,7 +42,8 @@ for i in 1:Niter
     # Set up and solve ODE
     p = (A, B, C, αrray[:, :, i], T, dT, utotconstr, dist)
     prob = ODEProblem(LinearDynamicsDisturbed!, u0, tspan, p)
-    sol = solve(prob, ROCK2(), abstol=1e-9)
+    sol = solve(prob, ROCK2(), abstol=1e-9)     
+    # Heun, ROCK2, Ralston all work-> if you put a standard solver like Tsit5 it gives incorrect results. I think this is due to the stiffness of the input and the adaptive timestep of some solvers miss the step function.
     solarray[i] = sol
 
     # Extract and store results
@@ -54,7 +55,7 @@ for i in 1:Niter
     # Update α for next iteration
     if i < Niter
         yⁱ[:, :, i] = xⁱ[:, :, i]
-        αrray[:, :, i+1] = αrray[:, :, i] + reshape(L * (ywant[:] - yⁱ[:, :, i][:]), No, o)
+        αrray[:, :, i+1] = αrray[:, :, i] + reshape(L * (ywant[:] - yⁱ[:, :, i][:]), No, o)     
     end
 end
 
@@ -66,72 +67,77 @@ v_vals = [solarray[end](t)[N+1:2N] for t in t_vals]  # Velocity data
 u_vals = hcat([utotconstr(A, B, C, αrray[:, :, end], t, T,dT) for t in t_vals]...)
 
 # Creating the figure
-f = Figure(resolution=(1200, 1000))
-
-plotcolors = ColorSchemes.matter
-
-# 1. Position plot
-ax1 = Axis(f[1, 1],
-    title="Positions",
-    xlabel="Time",
-    ylabel=L"q_{i}",
-    ylabelrotation=0,  # Rotates y-label 90 degrees
-    ylabelsize=18,
-    xticks=t_vals[1]:0.5:t_vals[end]
-)
-for i in 1:N
-    lines!(ax1, t_vals, getindex.(x_vals, i), label=L"q_{%$i}")  # Positions
-    scatter!(ax1, T, ywant[i, :], color=:black, marker=:x, markersize=8, label=L"$\overline{y}$")  # Desired positions 
-end
-axislegend(ax1, unique=true, orientation=:horizontal, position = :lt)
-
-# 2. Velocity plot
-ax2 = Axis(f[2, 1],
-    title="Velocities",
-    xlabel="Time",
-    ylabel=L"\dot{q}_{i}",
-    ylabelrotation=0,  # Rotates y-label 90 degrees
-    ylabelsize=18,
-    xticks=t_vals[1]:0.5:t_vals[end]
-)
-for i in 1:N
-    lines!(ax2, t_vals, getindex.(v_vals, i), label=L"\dot{q}_{%$i}")  # Velocities
-end
-axislegend(ax2, unique=true, orientation=:horizontal, position = :lt)
-#
-base_vals = zeros(o, l, No, length(t_vals))
-# 3. Base functions plot using pi_paper
-ax3 = Axis(f[3, 1],
-    title="Base Functions",
-    xlabel="Time",
-    ylabel=L"$\pi_{i}$",
-    ylabelrotation=0,  # Rotates y-label 90 degrees
-    ylabelsize=18,
-    xticks=t_vals[1]:0.5:t_vals[end]
-)
-for i in 1:o
-    base_vals[i, :, :, :] .= reshape(hcat([pi_constr(A, B, C, t, T, dT, i) for t in t_vals]...),(l,No,length(t_vals)))
-    for k in 1:l
-        for j in 1:No# Using pi_paper to compute the base functions
-            #color_index = ((i - 1) * 2N + j) / (o * 2N)
-            lines!(ax3, t_vals, base_vals[i, k, j, :])#, color=plotcolors[color_index])
+    f = Figure(resolution=(1200, 1000))
+
+    plotcolors = ColorSchemes.matter
+
+    # 1. Position plot
+    ax1 = Axis(f[1, 1],
+        title="Positions",
+        xlabel="Time",
+        ylabel=L"q_{i}",
+        ylabelrotation=0,  # Rotates y-label 90 degrees
+        ylabelsize=18,
+        xticks=t_vals[1]:0.5:t_vals[end]
+    )
+    for i in 1:N
+        lines!(ax1, t_vals, getindex.(x_vals, i), label=L"q_{%$i}")  # Positions
+        scatter!(ax1, T, ywant[i, :], color=:black, marker=:x, markersize=8, label=L"$\overline{y}$")  # Desired positions 
+    end
+    axislegend(ax1, unique=true, orientation=:horizontal, position = :lt)
+
+    # 2. Velocity plot
+    ax2 = Axis(f[2, 1],
+        title="Velocities",
+        xlabel="Time",
+        ylabel=L"\dot{q}_{i}",
+        ylabelrotation=0,  # Rotates y-label 90 degrees
+        ylabelsize=18,
+        xticks=t_vals[1]:0.5:t_vals[end]
+    )
+    for i in 1:N
+        lines!(ax2, t_vals, getindex.(v_vals, i), label=L"\dot{q}_{%$i}")  # Velocities
+    end
+    axislegend(ax2, unique=true, orientation=:horizontal, position = :lt)
+    #
+    base_vals = zeros(o, l, No, length(t_vals))
+    # 3. Base functions plot using pi_paper
+    ax3 = Axis(f[3, 1],
+        title="Base Functions",
+        xlabel="Time",
+        ylabel=L"$\pi_{i}$",
+        ylabelrotation=0,  # Rotates y-label 90 degrees
+        ylabelsize=18,
+        xticks=t_vals[1]:0.5:t_vals[end]
+    )
+    for i in 1:o
+        base_vals[i, :, :, :] .= reshape(hcat([pi_constr(A, B, C, t, T, dT, i) for t in t_vals]...),(l,No,length(t_vals)))
+        for k in 1:l
+            for j in 1:No# Using pi_paper to compute the base functions
+                color_index = ((i-1)*l*No + (k-1)*No + j) / (o*l*No)
+                lines!(ax3, t_vals, base_vals[i, k, j, :], color=plotcolors[color_index])
+            end
         end
     end
-end
-#
-# 4. Control Action plot (input uₙ)
-ax4 = Axis(f[4, 1],
-    title="Control Action",
-    xlabel="Time",
-    ylabel=L"$u(t)$",   
-    ylabelrotation=0,  # Rotates y-label 90 degrees
-    ylabelsize=18,
-    xticks=t_vals[1]:0.5:t_vals[end]
-)
-for i in 1:l
-    lines!(ax4, t_vals, u_vals[i, :],label=L"u_{%$i}")
-end
-axislegend(ax4, unique=true, orientation=:horizontal, position = :lt)
+    #
+    # 4. Control Action plot (input uₙ)
+    ax4 = Axis(f[4, 1],
+        title="Control Action",
+        xlabel="Time",
+        ylabel=L"$u(t)$",   
+        ylabelrotation=0,  # Rotates y-label 90 degrees
+        ylabelsize=18,
+        xticks=t_vals[1]:0.5:t_vals[end]
+    )
+    for i in 1:l
+        lines!(ax4, t_vals, u_vals[i, :],label=L"u_{%$i}")
+    end
+    axislegend(ax4, unique=true, orientation=:horizontal, position = :lt)
+
+    # Display the figure
+    display(f)
 
-# Display the figure
-display(f)
+## Save figure 
+imgpath = "./report_slides/typst/images/"       
+figname = "basket_$(Niter)iters_$(dT)dT.svg"
+save(imgpath * figname, f)