TY - EJOU
AU - Xiao, Rongge
AU - Jin, Wenbo
TI - STUDY ON WAX DEPOSITION RATE OPTIMIZATION ALGORITHM BASED ON LEVENBERG-MARQUARDT ALGORITHM AND GLOBAL OPTIMIZATION
T2 - Frontiers in Heat and Mass Transfer
PY - 2019
VL - 12
IS - 1
SN - 2151-8629
AB - In order to accurately obtain the wax deposition rate model, according to the kinetic principle of wax deposition, several factors affecting the wax
deposition rate were selected, and by a optimization software of First Optimization(1stOpt), The parameters of two typical wax deposition
rate models are solved respectively based on optimization algorithm combined by Levenberg-Marquardt (L-M) algorithm and global optimization
and the calculated data were compared. The results show that: compared with the model parameters obtained by least squares method, the model
parameters obtained by this optimization algorithm can describe the variation of wax deposition rate more accurately. The maximum error is reduced
from 30% to 10%, and the average error is reduced from 10.3% to 2.42%; Alike, the mathematical model obtained by this optimization algorithm is
also better than that solved by L-M algorithm alone. The maximum error is reduced from 13.62% to 11%, and the average error is reduced
from 6.46% to 4.77%. To a certain extent, this optimization algorithm avoids the premature phenomenon caused by using Levenberg-Marquardt
alone. In addition, the use of the optimization algorithm does not require suitable initial values, prior knowledge and programming, easy to use, and
has important use value.
KW - *Oil pipeline; Wax deposition rate; mathematical model; optimization algorithm; error analysis*
DO - 10.5098/hmt.12.28