ENPI / e-JIKEI Network Promotion Institute

Journal of Mechanical and Electrical Intelligent System (JMEIS, J. Mech. Elect. Intel. Syst.)

An international open-access peer-reviewed journal

ISSN 2433-8273

Home

Home > Archives > Vol.5, No.3

Vol.5, No.3

TABLE OF CONTENTS

Articles

Extended Leslie-Singh Architecture of 1^st order Delta-Sigma AD Modulator with Multi-bit DAC

Lengkhang Nengvang, Shogo Katayama, Jianglin Wei, Lei Sha, Anna Kuwana, Hiroshi Tanimoto, Tatsuji Matsuura, Kazufumi Naganuma, Kiyoshi Sasai, Junichi Saito, Katsuaki Morishita, and Haruo Kobayashi

Journal of Mechanical and Electrical Intelligent System, Vol.5, No.3, pp.1-8, 2022.

Abstract: This paper presents extended Leslie-Singh architecture of the 1^st order ΔΣAD modulator using multi-bit internal ADC and DAC. The original Leslie-Singh architecture ΔΣ AD modulator uses a multi-bit ADC and a single-bit DAC inside the modulator. Here we consider an m-bit ADC and an n-bit DAC with 𝑚 ≥ 𝑛 ≥ 1. SQNDR of the modulator for various (𝑚, 𝑛) is investigated by simulations and it is found that as m increases by 1, SQNDR improves by 6dB, while as n increases by 1, SQNDR improves by 3dB for 𝑚 ≫ 𝑛 but it saturates for 𝑚 ≈ 𝑛. We have clarified that as the DAC resolution increases by 1 bit, the SQNDR improves by 3dB since the input range for the modulator stable operation is extended.

Full Paper (PDF)

Variational Bayes estimation of unknown parameter distribution in 1-DoF vibratory system which subjected to random excitation

Soichiro Takata and Hirofumi Inoue

Journal of Mechanical and Electrical Intelligent System, Vol.5, No.3, pp.9-18, 2022.

Abstract: This paper presents an expansion method for parameter distribution estimation based on variational Bayes inference in a linear single-degree-of-freedom system using Gaussian random vibration responses. The likelihood function of the proposed method is defined by the analytical solution of the Fokker–Planck equation derived in a previous study. The unknown parameters are estimated using the variational Bayes formula. Furthermore, numerical identifications are conducted using random responses from the results of the 4^th-order Runge–Kutta method. The estimated performance of the proposed method was verified in terms of the dependence on the sample size. Benchmark tests were conducted to compare the accuracy of the variational Bayes and maximum likelihood estimations. The variational Bayes estimation exhibited higher accuracy than the maximum likelihood estimation for small sample sizes. Furthermore, a high-accuracy implementation trial was conducted with a focus on the dependence of the calculation sequence on the expected value of the variance.

Full Paper (PDF)