|Table of Contents|

 Assumed internal sensor-based neural network left-inversefor tension identification of two-motor speed regulation(PDF)

[ISSN:1007-449X/CN:23-1 408/TM]

Issue:
2018年01
Page:
23-28
Research Field:
Publishing date:

Info

Title:
 Assumed internal sensor-based neural network left-inversefor tension identification of two-motor speed regulation
Author(s):
 systemLIU Guo-hai CHEN Jie ZHAO Wen-xiang YUAN Jun XU Liang
 ( School of Electrical and Information Engineering,Jiangsu University,Zhenjiang 212013,China)
Keywords:
 inherent sensor neural network left inverse two-motor speed-regulation system tension i-dentification
PACS:
TM 343
DOI:
10.15938/j.emc.2018.01.004
Abstract:
 In order to solve the problem of high cost,high installation requirements,material and environ-ment constraints of high precision tension sensor of two-motor drive system,
a tension identification meth-od based on assumed internal sensor neural network left inverse was proposed. To identify the tension oftwo-motor speed regulation system,
the tension subsystem of two-motor drive system was established andthe existence of its left-inverse was proved based on the named inherent sensor. Considering that the com-plexity of the mathematic model and time-variant system parameter in the left-inverse system,a novel i-dentification strategy based on neural network left inverse ( NNLI) was proposed,in which the back prop-agation neural network ( BPNN) was used to approximate the left-inverse system of tension. Then,it isconnected in series with the original system to realize the estimation of tension. The simulation and exper-imental results of two-motor drive system are given,verifying that the proposed strategy can identify theactual tension quickly and accurately.

References:

Memo

Memo:
-
Last Update: 2018-06-28