From: couchot Date: Mon, 2 Dec 2013 21:09:24 +0000 (+0100) Subject: une section de plus X-Git-Url: https://bilbo.iut-bm.univ-fcomte.fr/and/gitweb/desynchronisation-controle.git/commitdiff_plain/eff67eeac67ec967add0eb7e5e90d500ac32a491?ds=inline une section de plus --- diff --git a/IWCMC14/HLG.tex b/IWCMC14/HLG.tex index 183e00b..1008903 100644 --- a/IWCMC14/HLG.tex +++ b/IWCMC14/HLG.tex @@ -70,7 +70,6 @@ The objective is thus to find $R$, $x$, $P_s$ which minimize \item $P_{sh} > 0,\forall h \in V$ \end{enumerate} - To achieve this optimizing goal a local optimisation, the problem is translated into an equivalent one: find $R$, $x$, $P_s$ which minimize @@ -84,23 +83,54 @@ P_{si}+ \sum_{l \in L}a_{il}^{+}.c^s_l.\left( \sum_{h \in V}x_{hl} \right) \\ \end{array} $$ -The authors then apply a dual based approach with Lagrange multiplier -to solve such a problem. -They first introduce dual variables -$u_{hi}$, $v_{h}$, $\lambda_{i}$, and $w_l$ for any -$h \in V$,$ i \in N$, and $l \in L$. -They next replace the objective of reducing $\sum_{i \in N }q_i^2$ + +They thus replace the objective of reducing +$\sum_{i \in N }q_i^2$ by the objective of reducing -$$ +\begin{equation} \sum_{i \in N }q_i^2 + \sum_{h \in V, l \in L } \delta.x_{hl}^2 + \sum_{h \in V }\delta.R_{h}^2 -$$ -where $\delta$ is a \JFC{ formalisation} factor. -This allows indeed to get convex functions whose minimum value is unique. +\label{eq:obj2} +\end{equation} +where $\delta$ is a regularisation factor. +This indeed introduces quadratic fonctions on variables $x_{hl}$ and +$R_{h}$ and makes some of the functions strictly convex. + +The authors then apply a classical dual based approach with Lagrange multiplier +to solve such a problem~\cite{}. +They first introduce dual variables +$u_{hi}$, $v_{h}$, $\lambda_{i}$, and $w_l$ for any +$h \in V$,$ i \in N$, and $l \in L$. -The proposed algorithm iteratively computes the following variables +\begin{equation} +\begin{array}{l} +L(R,x,P_{s},q,u,v,\lambda,w)=\\ +\sum_{i \in N} q_i^2 + q_i. \left( +\sum_{l \in L } a_{il}w_l^{(k)}- +\lambda_iB_i +\right)\\ ++ \sum_{h \in V} +v_h.\dfrac{\ln(\sigma^2/D_h)}{\gamma P_{sh} ^{2/3}} + \lambda_h P_{sh}\\ ++ \sum_{h \in V} \sum_{l\in L} +\left( +\delta.x_{hl}^2 \right.\\ +\qquad \qquad + x_{hl}. +\sum_{i \in N} \left( +\lambda_{i}.(c^s_l.a_{il}^{+} + +c^r. a_{il}^{-} ) \right.\\ +\qquad \qquad\qquad \qquad + +\left.\left. u_{hi} a_{il} +\right) +\right)\\ + + \sum_{h \in V} +\delta R_{h}^2 +-v_h.R_{h} - \sum_{i \in N} u_{hi}\eta_{hi} +\end{array} +\end{equation} +The proposed algorithm iteratively computes the following variables +untill the variation of the dual function is less than a given threshold. \begin{enumerate} \item $ u_{hi}^{(k+1)} = u_{hi}^{(k)} - \theta^{(k)}. \left( \eta_{hi}^{(k)} - \sum_{l \in L }a_{il}x_{hl}^{(k)} \right) $ @@ -125,7 +155,7 @@ $\theta^{(k)} = \omega / t^{1/2}$ \item $q_i^{(k)} = \arg\min_{q_i>0} \left( -q^2 + q. +q_i^2 + q_i. \left( \sum_{l \in L } a_{il}w_l^{(k)}- \lambda_i^{(k)}B_i @@ -171,4 +201,4 @@ c^r. a_{il}^{-} ) \right.\\ $ \end{enumerate} where the first four elements are dual variable and the last four ones are -primal ones \ No newline at end of file +primal ones. \ No newline at end of file diff --git a/IWCMC14/argmin.tex b/IWCMC14/argmin.tex new file mode 100644 index 0000000..0f64780 --- /dev/null +++ b/IWCMC14/argmin.tex @@ -0,0 +1,87 @@ +The approach detailed previously requires to compute +the variable that minimizes four convex fonctions +on constraint domains, one +for each primal variable $q_i$, $P_{sh}$, $R_h$, and $x_{hl}$, +Among state of the art for this optimization step, there is +the L-BFGS-B algorithm~\cite{byrd1995limited}, +the truncated Newton algorithm, +the Constrained Optimization BY Linear Approximation (COBYLA) method~\cite{ANU:1770520}. +However, all these methods suffer from being iterative approaches +and need many steps of computation to obtain an approximation +of the minimal value. This approach is dramatic whilst the objective is to +reduce all the computation steps. + +A closer look to each function that has to be minimized shows that it is +differentiable and the minimal value can be computed in only one step. +The table~\ref{table:min} presents these minimal value for each primal +variable. + +\begin{table*}[t] +$$ +\begin{array}{|l|l|l|} +\hline +q_i^{(k)} & + \arg\min_{q_i>0} +\left( +q^2 + q. +\left( +\sum_{l \in L } a_{il}w_l^{(k)}- +\lambda_i^{(k)}B_i +\right) +\right) + & +\max \left(\epsilon,\dfrac{\sum_{l \in L } a_{il}w_l^{(k)}- +\lambda_i^{(k)}B_i}{2}\right) \\ +\hline +P_{sh}^{(k)}& +\arg \min_{p > 0} +\left( +v_h^{(k)}.\dfrac{\ln(\sigma^2/D_h)}{\gamma p^{2/3}} + \lambda_h^{(k)}p ++ \delta_p p^{8/3} +\right) +& +\max \left(\epsilon, +\left( +\dfrac{ +-\lambda_h^{(k)} + \sqrt{(\lambda_h^{(k)})^2 + \dfrac{64}{9}\alpha} +}{\frac{16}{3}\delta_p} +\right)^{\frac{3}{5}} +\right) \\ +\hline +R_h^{(k)} +& +\arg \min_{r \geq 0 } +\left( +\delta_r r^2 +-v_h^{(k)}.r - \sum_{i \in N} u_{hi}^{(k)} \eta_{hi} +\right) & +\max\left(0,\dfrac{v_h^{(k)}}{2\delta_r}\right) +\\ +\hline +x_{hl}^{(k)} & +\begin{array}{l} +\arg \min_{x \geq 0} +\left( +\delta_x.x^2 \right.\\ +\qquad \qquad + x. +\sum_{i \in N} \left( +\lambda_{i}^{(k)}.(c^s_l.a_{il}^{+} + +c^r. a_{il}^{-} ) \right.\\ +\qquad \qquad\qquad \qquad + +\left.\left. u_{hi}^{(k)} a_{il} +\right) +\right) +\end{array} +& +\max\left(0,\dfrac{-\sum_{i \in N} \left( +\lambda_{i}^{(k)}.(c^s_l.a_{il}^{+} + +c^r. a_{il}^{-} ) + u_{hi}^{(k)} a_{il} +\right)}{2\delta_x}\right) +\\ +\hline +\end{array} +$$ +\caption{Expression of each optimized primal variable} +\end{table*} + + \ No newline at end of file diff --git a/IWCMC14/convergence.tex b/IWCMC14/convergence.tex new file mode 100644 index 0000000..ea4d4c2 --- /dev/null +++ b/IWCMC14/convergence.tex @@ -0,0 +1,35 @@ +Let us first have a discussion on the stop criterion of the citted algorithm. +We claim that even if the variation of the dual function is less than a given +threeshold, this does not ensure that the lifetime has been maximized. +Minimizing a function on a multiple domain (as the dual function) +may indeed easilly fall into a local trap because some of introduced +variables may lead to uniformity of the output. + +\begin{figure} + to be continued + \caption{Relations between dual function threshold and $q_i$ convergence} + \label{fig:convergence:scatterplot} +\end{figure} + +Experimentations have indeed shown that even if the dual +function seems to be constant +(variations between two evaluations of this one is less than $10^{-5}$) +not all the $q_i$ have the same value. +For instance, the Figure~\ref{fig:convergence:scatterplot} presents +a scatter plot. + +The maximum amplitude rate of the sequence of $q_i$ --which is +$\frac{\max_{i \in N} q_i} {\min_{i \in N}q_i}-1$-- +is represented in $y$-coordonates + with respect to the +value of the threeshold for dual function that is represented in +$x$-coordonates. +This figure shows that a very small threshold is a necessary condition, but not +a sufficient criteria to observe convergence of $q_i$. + +In the following, we consider the system are $\epsilon$-stable if both +maximum amplitude rate and the dual function are less than a threeshold +$\epsilon$. + + + \ No newline at end of file diff --git a/IWCMC14/convexity.tex b/IWCMC14/convexity.tex new file mode 100644 index 0000000..0519ecb --- /dev/null +++ b/IWCMC14/convexity.tex @@ -0,0 +1 @@ + \ No newline at end of file diff --git a/IWCMC14/main.tex b/IWCMC14/main.tex index 0bea088..49d816d 100644 --- a/IWCMC14/main.tex +++ b/IWCMC14/main.tex @@ -75,17 +75,28 @@ -\section{Improving an Existing Scheme} +\section{Basics of Existing Scheme} \input{HLG} +\section{Improving } + + +\subsection{Convergence Detection} +\input{convergence} + + \subsection{A closer look to convexity} +\input{convexity} +\JFC{On doit parler du probleme pour Ps} \subsection{Directly retreiving the argmin} +\input{argmin} + + -\subsection{Convergence Detection} -\section{Fault-Tolerence Considerations} +\subsection{Fault-Tolerence Considerations} \