parameters passed to the genetic hybrid algorithm gha
play

Parameters passed to the genetic hybrid algorithm (GHA) parameter - PDF document

Parameters passed to the genetic hybrid algorithm (GHA) parameter = value : accepted values description -------------------------------------------------------------------------------------------------


  1. Parameters passed to the genetic hybrid algorithm (GHA) parameter = value : accepted values description ------------------------------------------------------------------------------------------------- --------------------------------------------- PROJECT_NAME = TRIM_LOSS GET_HESSIAN = FALSE : BOOLEAN calculate Hessian matrix PRINT_HESSIAN = FALSE : BOOLEAN print_level = 1 : integer (0-3) for file output: 0=none 1=function values 2=1+weights+actions 3=debug_print superGA_ITERATIONS = 0 : integer > 0 number iterations in superstructure superMC_ITERATIONS = 3 : integer > 0 superGA_SIBS = 4 : integer > 0 number of siblings in superstructure GA_sibs = 16 : integer > 0 number of siblings GA_iterations = 20 : integer > 0 number of times the GHA is applied to a population GA_runs = 3 : integer > 0 Number of times a new population will be created CALL_FIX = TRUE : BOOLEAN always call fixvalue() dbl_SpaceLimit = 0.75 : real > 0 USE_CONVERGENCE_CRIT = FALSE : BOOLEAN dbl_ConvergenceMean = 0.0001 : real > 0 Convergence criteria Mean dbl_ConvergenceBest = 0.0005 : real > 0 Convergence criteria Best PenaltyFunction = LINEAR : LINEAR, QUADRATIC, CONCAVE, EXPONENTIAL, CONSTANT Penalty function to be used REDUCTION = NONE : NONE, NEUTRAL, FAST, SLOW allow reduction of search space ksi_interval = 5 : integer > 0 Interval for reduction of search space and mut. window LINEMETHOD = 2 : integer (1,2,3) 1:dfp_min+lnsrch, 2:dfp_min+cube_step, 3:newton t_incr = 0 : integer > 0 interval for quadratic/cubic interpolations t_final = 20 : integer > 0 t_slope = 0 : integer > 0 f_incr = 3 : integer > 0 interval for strong mutation in fixvalue()

  2. ADD_HISTORY = FALSE : BOOLEAN freeze history as part of current population all times RCOND_LIMIT = 1000.0 : real GE 0 Limit for reciprocal condition number Min_Penalty = 0.1 : real GE 0 minimum penalty c = 0.3 : real [0.005,0.5] penalty coefficient Running_Mean = 3 : integer > 0 beta = 2.0 : real non-uniformity parameter, (default=2) (Michalewicz [1994]) window = 0.5 : real initial mutation window echo_interval = 1 : integer > 0 number indicating how often iteration result should be echoed to resultfile small_zone = 0 : integer (0/1) number indicating whether a smaller zone will be used in the initialization lower_zone = 0.0 : double upper_zone = 0.0 : double rnd_startpoints = 3000 : integer > 0 rnd_accelerator = 1 : integer (0/1) delta_f_min = 0.0 : double digit indicating the limit value when individuals are considered equal crossover = 4 : integer 0:None, 1:Arithmetic, 2:Permuted, 3:Onepoint, 4:Twopoint, 5:Threepoint, 6:Random mutation = 2 : integer (1) 1:Nonuniform, 2:Gradient diversity_check = 2 : integer (0=NONE/1=STOCHASTIC/2=FIXED) diversity check for stochastic immigration grad_filter = 2.0 : double digit gradient filter (Max Steplength) RANDOMIZE_w = TRUE : BOOLEAN RANDOM_DENSITY = 0 : integer 0:Uniform 1:Normal MC_iterations = 3 : integer if SIM_FLIP=TRUE then > 0 number of Monte Carlo Iterations CHECK_KSI = FALSE : BOOLEAN CHECK_VARIABLES = TRUE : BOOLEAN UPDATEw_in = FALSE : BOOLEAN MUTATE_SIGN = FALSE : BOOLEAN s_incr = 10 : integer > 0 GET_RCOND = TRUE : BOOLEAN

  3. RCOND_ITERATIONS = 300 : integer GE 0 STABLE_GRAD = TRUE : BOOLEAN

Download Presentation
Download Policy: The content available on the website is offered to you 'AS IS' for your personal information and use only. It cannot be commercialized, licensed, or distributed on other websites without prior consent from the author. To download a presentation, simply click this link. If you encounter any difficulties during the download process, it's possible that the publisher has removed the file from their server.

Recommend


More recommend