An accelerated recurrent network training algorithm using IIR filter model and recursive least squares method

1997 ◽  
Vol 44 (11) ◽  
pp. 1082-1086 ◽  
Author(s):  
T.W.S. Chow ◽  
Siu-Yeung Cho
Keyword(s):  
Least Squares ◽  
Least Squares Method ◽  
Recurrent Network ◽  
Recursive Least Squares ◽  
Training Algorithm ◽  
Filter Model ◽  
Iir Filter ◽  
Network Training ◽  
Recursive Least Squares Method

Online parameter estimation and adaptive control with limited information in structures

2021 ◽  
pp. 107754632110191
Author(s):  
Fereidoun Amini ◽  
Elham Aghabarari
Keyword(s):  
Parameter Estimation ◽  
Adaptive Control ◽  
Least Squares ◽  
Control Algorithm ◽  
Least Squares Method ◽  
Recursive Least Squares ◽  
Limited Information ◽  
Online Identification ◽  
Recursive Least Squares Method ◽  
Adaptive Control Algorithm

An online parameter estimation is important along with the adaptive control, that is, a time-dependent plant. This study uses both online identification and the simple adaptive control algorithm with velocity feedback. The recursive least squares method was used to identify the stiffness and damping parameters of the structure’s stories. Identification was carried out online without initial estimation and only by measuring the structural responses. The limited information regarding sensor measurements, parameter convergence, and the effects of the covariance matrix is examined. The integration of the applied online identification, the appropriate reference model selection in simple adaptive control, and adopting the proportional integral filter was used to limit the structural control response error. Some numerical examples are simulated to verify the ability of the proposed approach. Despite the limited information, the results show that the simultaneous use of online identification with the recursive least squares method and simple adaptive control algorithm improved the overall structural performance.

MEMS Gyro Random Drift Model Parameter Identification Based on Two-Stage Recursive Least Squares Method

2012 ◽  
Vol 220-223 ◽  
pp. 1044-1047 ◽  
Author(s):  
Zhao Hua Liu ◽  
Jia Bin Chen ◽  
Yu Liang Mao ◽  
Chun Lei Song
Keyword(s):  
Least Squares ◽  
Least Squares Method ◽  
Moving Average ◽  
Recursive Least Squares ◽  
Model Parameters ◽  
Random Drift ◽  
Two Stage ◽  
Autoregressive Moving Average Model ◽  
Drift Model ◽  
Recursive Least Squares Method

Autoregressive moving average model (ARMA) was usually used for gyro random drift modeling. Because gyro random drift was a non-stationary, weak non-linear and time-variant random signal, model parameters were random and time-variant, too. For improving precision of gyro and reducing effects of random drift, this paper adopted two-stage recursive least squares method for ARMA parameter estimation. This method overcame the shortcomings of the conventional recursive extended least squares (RELS) algorithm. At the same time, the forgetting factor was introduced to adapt the model parameters change. The simulation experimental results showed that this method is effective.

scholarly journals Simple Orbit Determination Using GPS Based on a Least-Squares Algorithm Employing Sequential Givens Rotations

2007 ◽  
Vol 2007 ◽  
pp. 1-8 ◽  
Author(s):  
Rodolpho Vilhena de Moraes ◽  
Aurea Aparecida da Silva ◽  
Helio Koiti Kuga
Keyword(s):  
Least Squares ◽  
Orbit Determination ◽  
Least Squares Method ◽  
Artificial Satellite ◽  
Low Cost ◽  
Recursive Least Squares ◽  
Gps Receiver ◽  
Gravity Field Model ◽  
Givens Rotations ◽  
Recursive Least Squares Method

A low-cost computer procedure to determine the orbit of an artificial satellite by using short arc data from an onboard GPS receiver is proposed. Pseudoranges are used as measurements to estimate the orbit via recursive least squares method. The algorithm applies orthogonal Givens rotations for solving recursive and sequential orbit determination problems. To assess the procedure, it was applied to the TOPEX/POSEIDON satellite for data batches of one orbital period (approximately two hours), and force modelling, due to the full JGM-2 gravity field model, was considered. When compared with the reference Precision Orbit Ephemeris (POE) of JPL/NASA, the results have indicated that precision better than 9 m is easily obtained, even when short batches of data are used.

Sign in / Sign up

Export Citation Format

Share Document