UDC 681.2.089
COMPARATIVE ANALYSIS OF TWO METHODS OF NONLINEAR OPTIMIZATION IN THE PROBLEM OF CALIBRATING THREE-AXIAL MEMS ACCELEROMETER
I. S. Kholopov, Ph.D. (Tech.), Dean of the Faculty of Radio Engineering and Telecommunications, RSREU, Ryazan, Russia;
orcid.org/0000-0001-5220-0811, e-mail: This email address is being protected from spambots. You need JavaScript enabled to view it.
I. A. Lyutkov, master student, Department of Radio Engineering Systems, RSREU, Ryazan, Russia;
orcid.org/0000-0003-0272-0696, e-mail: This email address is being protected from spambots. You need JavaScript enabled to view it.
The aim of the work is a comparative analysis of Gauss – Newton and Levenberg – Marquardt optimization algorithms in the calibration problem of triaxial micro electromechanical accelerometer. A mathematical description of these algorithms is given; their advantages and disadvantages in solving the problem of optimizing the parameters of a simplified mathematical model of triaxial micro electromechanical accelerometer by the criterion of minimum error squared are noted. It is shown that with identical parameters of the model, Levenberg – Marquardt algorithm requires an average of 30 % more iterations to complete than Gauss – Newton algorithm, however, when applied, the error in estimating the elements of the vector of calibration parameters decreases by 20-30 % .
Key words: MEMS accelerometer calibration, Gauss – Newton method, Levenbergh – Marquardt method.