ALPSO

Augmented Lagrangian Particle Swarm Optimizer (ALPSO) is a PSO method that uses the augmented Lagrangian approach to handle constraints.

Options

ALPSO Default Options

Name	Type	Default value	Description
SwarmSize	int	40	Number of Particles (Depends on Problem dimensions)
maxOuterIter	int	200	Maximum Number of Outer Loop Iterations (Major Iterations)
maxInnerIter	int	6	Maximum Number of Inner Loop Iterations (Minor Iterations)
minInnerIter	int	6	Minimum Number of Inner Loop Iterations (Dynamic Inner Iterations)
dynInnerIter	int	0	Dynamic Number of Inner Iterations Flag
stopCriteria	int	1	Stopping Criteria Flag (0 - maxIters, 1 - convergence)
stopIters	int	5	Consecutive Number of Iterations for which the Stopping Criteria must be Satisfied
etol	float	0.001	Absolute Tolerance for Equality constraints
itol	float	0.001	Absolute Tolerance for Inequality constraints
rtol	float	0.01	Relative Tolerance for Lagrange Multipliers
atol	float	0.01	Absolute Tolerance for Lagrange Function
dtol	float	0.1	Relative Tolerance in Distance of All Particles to Terminate (GCPSO)
printOuterIters	int	0	Number of Iterations Before Print Outer Loop Information
printInnerIters	int	0	Number of Iterations Before Print Inner Loop Information
rinit	float	1.0	Initial Penalty Factor
xinit	int	0	Initial Position Flag (0 - no position, 1 - position given)
vinit	float	1.0	Initial Velocity of Particles in Normalized [-1, 1] Design Space
vmax	float	2.0	Maximum Velocity of Particles in Normalized [-1, 1] Design Space
c1	float	2.0	Cognitive Parameter
c2	float	1.0	Social Parameter
w1	float	0.99	Initial Inertia Weight
w2	float	0.55	Final Inertia Weight
ns	int	15	Number of Consecutive Successes in Finding New Best Position of Best Particle Before Search Radius will be Increased (GCPSO)
nf	int	5	Number of Consecutive Failures in Finding New Best Position of Best Particle Before Search Radius will be Increased (GCPSO)
dt	float	1.0	Time step
vcrazy	float	0.0001	Craziness Velocity (Added to Particle Velocity After Updating the Penalty Factors and Langangian Multipliers)
fileout	int	1	Flag to Turn On Output to filename
filename	str	ALPSO.out	We could probably remove fileout flag if filename or fileinstance is given
seed	int	0	Random Number Seed (0 - Auto-Seed based on time clock)
HoodSize	int	40	Number of Neighbours of Each Particle
HoodModel	str	gbest	Neighbourhood Model (dl/slring - Double/Single Link Ring, wheel - Wheel, Spatial - based on spatial distance, sfrac - Spatial Fraction)
HoodSelf	int	1	Selfless Neighbourhood Model (0 - Include Particle i in NH i, 1 - Don’t Include Particle i)
Scaling	int	1	Design Variables Scaling Flag (0 - no scaling, 1 - scaling between [-1, 1])
parallelType	str	None	Type of parallelization None: No parallel function evaluations EXT: Use parallel function evaluations

API

class pyoptsparse.pyALPSO.pyALPSO.ALPSO(*args, **kwargs)

ALPSO Optimizer Class - Inherited from Optimizer Abstract Class

Keyword arguments:*

• pll_type -> STR: ALPSO Parallel Implementation (None, SPM- Static, DPM- Dynamic, POA-Parallel Analysis), Default = None

This is the base optimizer class that all optimizers inherit from. We define common methods here to avoid code duplication.

Parameters:

namestr

Optimizer name

categorystr

Typically local or global

defaultOptionsdictionary

A dictionary containing the default options

informsdict

Dictionary of the inform codes

__call__(optProb, storeHistory=None, hotStart=None, **kwargs)

This is the main routine used to solve the optimization problem.

Parameters:

optProbOptimization or Solution class instance

This is the complete description of the optimization problem to be solved by the optimizer

storeHistorystr

File name of the history file into which the history of this optimization will be stored

hotStartstr

File name of the history file to “replay” for the optimization. The optimization problem used to generate the history file specified in ‘hotStart’ must be IDENTICAL to the currently supplied ‘optProb’. By identical we mean, EVERY SINGLE PARAMETER MUST BE IDENTICAL. As soon as he requested evaluation point from ALPSO does not match the history and function evaluations revert back to normal evaluations.

Notes

The kwargs are there such that the sens= argument can be supplied (but ignored here in alpso)