neural based fault tolerance system using fpaa

Upload: mageshkumar-gunasekar

Post on 03-Apr-2018

225 views

Category:

Documents


0 download

TRANSCRIPT

  • 7/28/2019 Neural Based Fault Tolerance System Using FPAA

    1/19

    Neural Based Fault Tolerance

    System Using FPAA

  • 7/28/2019 Neural Based Fault Tolerance System Using FPAA

    2/19

    Basic Concepts of FTS

    Process resilience

    Reliable client-server communication

    Reliable group communication Distributed commit

    Recovery

  • 7/28/2019 Neural Based Fault Tolerance System Using FPAA

    3/19

    Introduction

    What is fault tolerance?

    What are the ways in which systems can

    fail?

    What is the most basic way of achieving

    fault tolerance?

  • 7/28/2019 Neural Based Fault Tolerance System Using FPAA

    4/19

    Dependability

    Basics:A componentprovides services toclients. To provide services, the componentmay require the services from other componentsa component may depend on some other

    component. Specifically:A component Cdepends on C*if

    the correctness ofCs behavior depends on thecorrectness ofC*s behavior.

    Some properties of dependability: Availability: Readiness for usage

  • 7/28/2019 Neural Based Fault Tolerance System Using FPAA

    5/19

    Reliability: Continuity of servicedelivery Safety: Very low probability of

    catastrophes

    Maintainability: How easy can afailed system be repaired

    Note: For distributed systems,components can be either processes orchannels

  • 7/28/2019 Neural Based Fault Tolerance System Using FPAA

    6/19

    Neural Networks

    Biological approach to AI.

    Developed in 1943.

    Comprised of one or more layers of neurons Several types, well focus on feed-forward

    networks.

    https://reader009.{domain}/reader009/html5/0407/5ac8934fc7c6f/5ac893524ddd3.jpghttps://reader009.{domain}/reader009/html5/0407/5ac8934fc7c6f/5ac893524ddd3.jpg

  • 7/28/2019 Neural Based Fault Tolerance System Using FPAA

    7/19

    Back Propagation Algorithm

    Back-Propagation Algorithm:

    function BACK-PROP-LEARNING(examples, network) returns a neural

    network

    inputs: examples, a set of examples, each with input vector x and output vector

    y

    network, a multilayer network with L layers, weights Wj,i , activation

    function g

    repeat

    for each e in examples do

    for each node j in the input layer do aj xj[e]for l = 2 to M do

    inij Wj,i aj

    aig(ini)

    for each node i in the output layer do

    Dj g(inj) i Wji Di

    for l = M 1 to 1 dofor each node j in layer l do

    Dj g(inj) i Wj,i Di

    for each node i in layer l + 1 do

    Wj,iWj,i + a x aj x Diuntil some stopping criterion is satisfied

    return NEURAL-NET-HYPOTHESIS(network)

    [Russell,] Fig. 20.25 Pg. 746

  • 7/28/2019 Neural Based Fault Tolerance System Using FPAA

    8/19

    Back-Propagation Illustration

    ARTIFICIAL NEURAL NETWORKS Colin Fahey's Guide (Book CD)

  • 7/28/2019 Neural Based Fault Tolerance System Using FPAA

    9/19

    Training

    Datasets from gaming sessions of humanvs. human are best

    Must decide whether training will occurin-game, during development, or both

    Learning during play provides foradaptations against individual players

  • 7/28/2019 Neural Based Fault Tolerance System Using FPAA

    10/19

    Training Patterns of Neural Network :

  • 7/28/2019 Neural Based Fault Tolerance System Using FPAA

    11/19

    Fuzzy Logic Controller

    FUZZY SETS FOR LOAD AND SPEED

    To choose membership functions, first of all oneneeds to consider the universe of distance for allthe linguistic variables, applied to the rules

    formulation. To specify the universe of discourse, one must

    firstly determine the applicable range for acharacteristic variable in the context of the

    system designed. The range you select shouldbe carefully considered.

  • 7/28/2019 Neural Based Fault Tolerance System Using FPAA

    12/19

    Example :

    For example, if you specify a range, which is too small,

    regularly occurring data will be off the scale, that may impacton an overall system performance. Conversely, if the universefor the input is too large, a temptation will often be to havewide membership functions on the right or left to capture theextreme input values.

    Because of this situation it is usually desirable and oftennecessary to scale or normalize the universe of discourse of aninput/output variable. Normalization means applying thestandard range of [1, +1] for the universe of discourse bothfor inputs and outputs. The universe of discourse for

    percentage load, percentage speed and percentage controlvoltage is chosen as {0% to 100%). Error, change in error andcontrol voltage are taken as fuzzy variables and are assignedthe membership functions as shown in the following figure.

  • 7/28/2019 Neural Based Fault Tolerance System Using FPAA

    13/19

    NL NM NS AZ PS PM PL

    -30 -20 -10 0 10 20 30Error

    NL NM NS AZ PS PM PL

    -3 -2 -1 0 1 2 3

    Change in Error

    Control voltage in %

    NL NM NS AZ PS PM PL

    -80 -55 -30 0 30 55 80

    Fuzzy calculation:

    In the membership

    function plot

    NL Negative LargeNM Negative Medium

    NS Negative Small

    AZ Absolutely Zero

    PS Positive small

    PM Positive Medium

    PL Positive Large

    Available in Knowledge base.

  • 7/28/2019 Neural Based Fault Tolerance System Using FPAA

    14/19

    cee

    NL NM NS AZ PS PM PL

    NL NL NM NS AZ PS PM PL

    NM NL NL NL NM NS AZ PS

    NS NL NL NM NS PS PS PM

    AZ NL NL NS AZ PS PM PL

    PS NM NS AZ PS PM PL PL

    PM NS AZ PS PM PL PL PL

    PL AZ PS PM PL PL PL PL

    Rule-Base for the proposed FLC

    The Fuzzy rule base relating the variables error, change in error and control

    voltage is shown in Table below.

  • 7/28/2019 Neural Based Fault Tolerance System Using FPAA

    15/19

    FPAA and its Features

    FPAA is Field Programmable Analog

    Array.

    Its architecture is shown in the next slide.

    It is a dynamically reconfigurable device.

    It is quickly accessible.

  • 7/28/2019 Neural Based Fault Tolerance System Using FPAA

    16/19

  • 7/28/2019 Neural Based Fault Tolerance System Using FPAA

    17/19

    Special features of FPAA

  • 7/28/2019 Neural Based Fault Tolerance System Using FPAA

    18/19

    In our project

    Neural network is used as an Estimator.

    FPAA is used as a reconfigurable device.

    FLC is used as the controller.

  • 7/28/2019 Neural Based Fault Tolerance System Using FPAA

    19/19

    Block diagram of our project