pso(particle swam optimization)

Upload: sandeep036

Post on 30-May-2018

236 views

Category:

Documents


0 download

TRANSCRIPT

  • 8/14/2019 PSO(Particle Swam Optimization)

    1/34

    Particle Swarm Optimization (PSO)Particle Swarm Optimization (PSO)

    PSO is a robust stochastic optimization technique based on

    the movement and intelligence of swarms. PSO applies the concept of social interaction to problem

    solving.

    It was developed in 1995 by James Kennedy (social-

    psychologist) and Russell Eberhart (electrical engineer).

    It uses a number of agents (particles) that constitute a

    swarm moving around in the search space looking for the

    best solution.

    Each particle is treated as a point in a N-dimensional spacewhich adjusts its flying according to its own flying

    experience as well as the flying experience of other

    particles.

  • 8/14/2019 PSO(Particle Swam Optimization)

    2/34

    Each particle keeps track of its coordinates in the solutionspace which are associated with the best solution (fitness)

    that has achieved so far by that particle. This value is called

    personal best , pbest.

    Another best value that is tracked by the PSO is the best

    value obtained so far by any particle in the neighborhood of

    that particle. This value is called gbest.

    The basic concept of PSO lies in accelerating each particle

    toward its pbest and the gbest locations, with a random

    weighted accelaration at each time step as shown in Fig.1

    Particle Swarm Optimization (PSO)Particle Swarm Optimization (PSO)

  • 8/14/2019 PSO(Particle Swam Optimization)

    3/34

    Fig.1 Concept of modification of a searching point by PSO

    sk : current searching point.

    sk+1 : modified

    searching point.

    vk: current velocity.

    vk+1 : modified

    sk

    vk

    vpbest

    vgbest

    sk+1

    vk+1

    sk

    vk

    vpbest

    vgbest

    sk+1

    vk+1

    Particle Swarm Optimization (PSO)Particle Swarm Optimization (PSO)

    x

    y

  • 8/14/2019 PSO(Particle Swam Optimization)

    4/34

    Particle Swarm Optimization (PSO)Particle Swarm Optimization (PSO) Each particle tries to modify its position using the following

    information: the current positions, the current velocities, the distance between the current position and pbest, the distance between the current position and the gbest.The modification of the particles position can be mathematically

    modeled according the following equation :

    Vik+1

    =w

    Vik

    +c1 rand1() x (pbesti-sik

    ) + c2 rand2() x (gbest-sik

    ) .. (1)where, vi

    k : velocity of agent i at iteration k,

    w: weighting function,

    cj : weighting factor,

    rand : uniformly distributed random number

    between 0 and 1, sik : current position of agent i at

    iteration k, pbesti : pbest of agent i,

  • 8/14/2019 PSO(Particle Swam Optimization)

    5/34

    Thefollowing weighting function is usually utilized in (1)

    w= wMax-[(wMax-wMin) x iter]/maxIter (2)

    where wMax= initial weight,

    wMin = final weight,

    maxIter = maximum iteration number,

    iter = current iteration number.

    sik+1

    = sik

    + Vik+1

    (3)

    Particle Swarm Optimization (PSO)Particle Swarm Optimization (PSO)

  • 8/14/2019 PSO(Particle Swam Optimization)

    6/34

    Comments on the Inertial weight factor:Comments on the Inertial weight factor:

    A large inertia weight (A large inertia weight (ww) facilitates a global search while) facilitates a global search whilea small inertia weight facilitates a local search.a small inertia weight facilitates a local search.

    By linearly decreasing the inertia weight from a relativelyBy linearly decreasing the inertia weight from a relativelylarge value to a small value through the course of thelarge value to a small value through the course of the

    PSO run gives the best PSO performance comparedPSO run gives the best PSO performance comparedwith fixed inertia weight settings.with fixed inertia weight settings.

    Larger w ----------- greater global search abilityLarger w ----------- greater global search ability

    Smaller w ------------ greater local search ability.Smaller w ------------ greater local search ability.

    Particle Swarm OptimizationParticle Swarm Optimization

    (PSO)(PSO)

  • 8/14/2019 PSO(Particle Swam Optimization)

    7/34

    Particle Swarm Optimization (PSO)Particle Swarm Optimization (PSO)

    Flow chart depicting the General PSO Algorithm:

    Start

    Initialize particles with random position

    and velocity vectors.

    For each particles position (p)

    evaluate fitness

    If fitness(p) better than

    fitness(pbest) then pbest= pLoop

    untilall

    particles

    exhaust

    Set best of pBests as gBest

    Update particles velocity (eq. 1) and

    position (eq. 3)

    Loopun

    tilmaxiter

    Stop: giving gBest, optimal solution.

  • 8/14/2019 PSO(Particle Swam Optimization)

    8/34

    Comparison with other evolutionaryComparison with other evolutionary

    computation techniques.computation techniques. Unlike in genetic algorithms, evolutionary programming andevolutionary strategies, in PSO, there is no selection operation.

    All particles in PSO are kept as members of the population through

    the course of the run

    PSO is the only algorithm that does not implement the survival of

    the fittest.

    No crossover operation in PSO.

    eq 1(b) resembles mutation in EP. In EP balance between the global and local search can be adjusted

    through the strategy parameter while in PSO the balance is

    achieved through the inertial weight factor (w) of eq. 1(a)

  • 8/14/2019 PSO(Particle Swam Optimization)

    9/34

    Variants of PSOVariants of PSO

    Discrete PSO can handle discrete binaryvariables

    MINLP PSO can handle both discrete binary and

    continuous variables.

    Hybrid PSO. Utilizes basic mechanism of PSO

    and the natural selection mechanism, which is usually

    utilized by EC methods such as GAs.

  • 8/14/2019 PSO(Particle Swam Optimization)

    10/34

    Application ofPSOALGORITHMto Optimize a

    Meander-line Polarizer for LICP conversion

    Intialization parameters used for PSO:

    wMax=0.41

    wMin=0.4

    (Note:The inertial weight ,w is linearly decreased from wMaxto wMinaccording the Eq. (2), w is chosen virtually constant in this case for better

    local search near the Suns Optimized parameters.)

    c1=c2=1.49

    maxIter=2000

    The above parameters are used in conjuction with eqs.

    (1) & (2)

    Swarm size/Population size used for solution search : 25

  • 8/14/2019 PSO(Particle Swam Optimization)

    11/34

    Application ofPSOALGORITHMto Optimize a

    Meander-line Polarizer for LICP conversion

    Frequency band of interest: 3.5 to 6.5 (GHz)

    (evaluated at 12 frequency points)

    Desired VSWR

  • 8/14/2019 PSO(Particle Swam Optimization)

    12/34

    Mean best & Best fitness over50 runs

  • 8/14/2019 PSO(Particle Swam Optimization)

    13/34

    VSWR

  • 8/14/2019 PSO(Particle Swam Optimization)

    14/34

    Axial Ratio (dB)

  • 8/14/2019 PSO(Particle Swam Optimization)

    15/34

    4 Layers for CP

    Spacer

    (inches)

    loi

    dielectric(inches)

    l i

    Line Width

    (inches)

    w1 w2

    Height

    (inches)

    h

    Period

    (inches)

    b

    Pitch

    (inches)

    a

    Layer

    ----8.4705018E-

    03

    1.5984001E-

    02

    2.2501351E-

    020.25205650.72833820.34493604

    0.54281283.4676325E-

    03

    9.4296653E-

    03

    5.3999661E-

    030.47074200.87043980.37984693

    0.42118453.4676325E-

    03

    9.4296653E-

    03

    5.3999661E-

    030.47074200.87043980.37984692

    0.54281288.4705018E-

    03

    1.5984001E-

    02

    2.2501351E-

    020.25205650.72833820.34493601

    Dielectric

    Sheet

    Metal

    Layer

    Spacer

    55.2i = 15.1i0 =Dielectric constants:

    Optimized dimensions for 4-layer

    Meander Line Polarizer

  • 8/14/2019 PSO(Particle Swam Optimization)

    16/34

    Application ofPSOALGORITHMto Optimize a

    Meander-line Polarizer for LICP conversion

    Frequency bands of interest:

    Band1: 3.7 to 4.2 (GHz)

    Band2: 5.9 to 6.4 (GHz)

    (evaluated at 2 frequency points: 3.95 (GHz), 6.15 (GHz))

    Desired VSWR

  • 8/14/2019 PSO(Particle Swam Optimization)

    17/34

    Mean best & Best fitness over50 runs

  • 8/14/2019 PSO(Particle Swam Optimization)

    18/34

    VSWR

  • 8/14/2019 PSO(Particle Swam Optimization)

    19/34

    Axial Ratio (dB)

  • 8/14/2019 PSO(Particle Swam Optimization)

    20/34

    4 Layers for CP

    Spacer

    (inches)

    loi

    dielectric(inches)

    l i

    Line Width

    (inches)

    w1 w2

    Height

    (inches)

    h

    Period

    (inches)

    b

    Pitch

    (inches)

    a

    Layer

    ----4.1147252E-

    03

    2.7704202E-

    02

    4.0456183E-

    020.24801430.80816690.25289134

    0.45338233.9593712E-

    03

    4.4981677E-

    02

    5.0358579E-

    020.34337220.95296580.54257243

    0.44821953.9593712E-

    03

    4.4981677E-

    02

    5.0358579E-

    020.34337220.95296580.54257242

    0.45338234.1147252E-

    03

    2.7704202E-

    02

    4.0456183E-

    020.24801430.80816690.25289131

    Dielectric

    Sheet

    Metal

    Layer

    Spacer

    55.2i = 15.1i0 =Dielectric constants:

    Optimized dimensions for 4-layer

    Meander Line Polarizer

  • 8/14/2019 PSO(Particle Swam Optimization)

    21/34

    Application ofPSOALGORITHMto Optimize a

    Meander-line Polarizer forLP rotation

    Intialization parameters used for PSO:

    wMax=0.41

    wMin=0.4

    (Note:The inertial weight ,w is linearly decreased from wMaxto wMinaccording the Eq. (2), w is chosen virtually constant in this case for better

    local search near the Suns Optimized parameters.)

    c1=c2=1.3

    maxIter=1000

    The above parameters are used in conjuction with eqs.

    (1) & (2)

    Swarm size/Population size used for solution search : 25

  • 8/14/2019 PSO(Particle Swam Optimization)

    22/34

    Application ofPSOALGORITHMto Optimize a

    Meander-line Polarizer forLP rotation

    Frequency band of interest: 3.5 to 6.5 (GHz)

    (evaluated at 12 frequency points)

    Desired VSWR

  • 8/14/2019 PSO(Particle Swam Optimization)

    23/34

    Mean best & Best fitness over15 runs

  • 8/14/2019 PSO(Particle Swam Optimization)

    24/34

    VSWR

  • 8/14/2019 PSO(Particle Swam Optimization)

    25/34

    Axial Ratio (dB)

  • 8/14/2019 PSO(Particle Swam Optimization)

    26/34

    Phase Difference

  • 8/14/2019 PSO(Particle Swam Optimization)

    27/34

    8 Layers for LP

    Spacer

    (inches)

    loi

    dielectric(inches)

    l i

    Line Width

    (inches)

    w1 w2

    Height

    (inches)

    h

    Period

    (inches)

    b

    Pitch

    (inches)

    a

    Layer

    ----2.1623570E-

    02

    3.0709708E-

    02

    2.8606838E-

    020.21480451.0206000.36318784, 8

    0.45507974.0660784E-

    02

    3.4942929E-

    02

    4.1542474E-

    020.43749990.82257650.38481703, 7

    0.39594684.0660784E-

    02

    3.4942929E-

    02

    4.1542474E-

    020.43749990.82257650.38481702, 6

    0.45507972.1623570E-

    02

    3.0709708E-

    02

    2.8606838E-

    020.21480451.0206000.36318781, 5

    Dielectric

    Sheet

    Metal

    Layer

    Spacer

    55.2i = 15.1i0 =Dielectric constants:

    Optimized dimensions for 8-layer

    Meander Line Polarizer

  • 8/14/2019 PSO(Particle Swam Optimization)

    28/34

    Application ofPSOALGORITHMto Optimize a

    Meander-line Polarizer forLP rotation

    Frequency bands of interest:

    Band1: 3.7 to 4.2 (GHz)

    Band2: 5.9 to 6.4 (GHz)

    (evaluated at 2 frequency points: 3.95 (GHz), 6.15 (GHz))

    Desired VSWR

  • 8/14/2019 PSO(Particle Swam Optimization)

    29/34

    Mean best & Best fitness over15 runs

  • 8/14/2019 PSO(Particle Swam Optimization)

    30/34

    VSWR

  • 8/14/2019 PSO(Particle Swam Optimization)

    31/34

    Axial Ratio (dB)

  • 8/14/2019 PSO(Particle Swam Optimization)

    32/34

  • 8/14/2019 PSO(Particle Swam Optimization)

    33/34

    8 Layers for LP

    Spacer

    (inches)

    loi

    dielectric(inches)

    l i

    Line Width

    (inches)

    w1 w2

    Height

    (inches)

    h

    Period

    (inches)

    b

    Pitch

    (inches)

    a

    Layer

    ----2.2299249E-

    02

    3.2010745E-

    02

    3.3202391E-

    020.27475691.0555960.31508694, 8

    0.35447253.1127717E-

    020.1126298

    6.0811251E-

    020.35656081.1645760.40858013, 7

    0.31593893.1127717E-

    020.1126298

    6.0811251E-

    020.35656081.1645760.40858012, 6

    0.35447252.2299249E-

    02

    3.2010745E-

    02

    3.3202391E-

    020.27475691.0555960.31508691, 5

    Dielectric

    Sheet

    Metal

    Layer

    Spacer

    55.2i = 15.1i0 =Dielectric constants:

    Optimized dimensions for 8-layer

    Meander Line Polarizer

  • 8/14/2019 PSO(Particle Swam Optimization)

    34/34

    END