Mixture Density Networks-Based Knock Simulator

被引:29
作者
Shen, Xun [1 ,2 ]
Ouyang, Tinghui [3 ]
Khajorntraidet, Chanyut [4 ]
Li, Yuanchao [5 ]
Li, Shuai [6 ]
Zhuang, Jiancang [1 ,7 ]
机构
[1] SOKENDAI, Dept Stat Sci, Grad Univ Adv Studies, Mishima, Shizuoka 4118540, Japan
[2] Inst Stat Math, Tachigawa 1908562, Japan
[3] Natl Inst Adv Ind Sci & Technol, Koto Ku, Koto City, Tokyo 1350064, Japan
[4] King Mongkuts Univ Technol North Bangkok, Fac Engn & Technol, Bangkok 10800, Thailand
[5] Honda Res & Dev Co Ltd, Honda Innovat Lab, Minato Ku, Tokyo 1076238, Japan
[6] Univ Tennessee, Dept Civil & Environm Engn, Knoxville, TN 37996 USA
[7] Inst Stat Math, Tokyo 1908562, Japan
关键词
Engines; Combustion; Bars; Timing; Probability distribution; Mechatronics; IEEE transactions; Accept-reject method; engine knock; mixture density networks (MDNs); statistical analysis; statistical simulator;
D O I
10.1109/TMECH.2021.3059775
中图分类号
TP [自动化技术、计算机技术];
学科分类号
0812 ;
摘要
The engine knock simulator is useful for the evaluation of the feedback knock controllers and also the calibration of the feedforward control input without experiments in spark-ignition engines. This article proposes a mixture density network (MDN)-based statistical simulator of the engine knock for spark-ignition engines. The simulator can output the simulated knock intensity by the operating condition, which has a consistent probability distribution with the real engine. The statistical analysis is conducted based on the experimental data. According to the analysis results, several important properties about the knock intensity have been revealed. The logarithm of knock intensity is independent and identically distributed under an identical operating condition, whose probability distribution can be approximated by Gaussian mixture model (GMM). The parameter vector of the GMM is a function of the engine's operation condition. Based on these statistical properties of engine knock, we formulate the problem of establishing a statistical simulator, which includes two subproblems. The first one is how to approximate the function from the operating condition to the parameters of the GMM. The second one is how to output the simulated random data of logarithm of knock intensity that obeys a given distribution. The MDN and the accept-reject algorithm are applied to solve the two subproblem, respectively. Finally, we conducted experimental data-based validations to verify the proposed method.
引用
收藏
页码:159 / 168
页数:10
相关论文
共 30 条
[1]   Development of High-Reliability EV and HEV IM Propulsion Drive With Ultra-Low Latency HIL Environment [J].
Adzic, Evgenije M. ;
Adzic, Milan S. ;
Katic, Vladimir A. ;
Marcetic, Darko P. ;
Celanovic, Nikola L. .
IEEE TRANSACTIONS ON INDUSTRIAL INFORMATICS, 2013, 9 (02) :630-639
[2]   A new knock event definition for knock detection and control optimization [J].
Bares, P. ;
Selmanaj, D. ;
Guardiola, C. ;
Onder, C. .
APPLIED THERMAL ENGINEERING, 2018, 131 :80-88
[3]   KERNEL DENSITY ESTIMATION VIA DIFFUSION [J].
Botev, Z. I. ;
Grotowski, J. F. ;
Kroese, D. P. .
ANNALS OF STATISTICS, 2010, 38 (05) :2916-2957
[4]   Knock detection in automobile engines [J].
Boubal, O .
IEEE INSTRUMENTATION & MEASUREMENT MAGAZINE, 2000, 3 (03) :24-28
[5]   Overview of power management in hybrid electric vehicles [J].
Chau, KT ;
Wong, YS .
ENERGY CONVERSION AND MANAGEMENT, 2002, 43 (15) :1953-1968
[6]   MAXIMUM LIKELIHOOD FROM INCOMPLETE DATA VIA EM ALGORITHM [J].
DEMPSTER, AP ;
LAIRD, NM ;
RUBIN, DB .
JOURNAL OF THE ROYAL STATISTICAL SOCIETY SERIES B-METHODOLOGICAL, 1977, 39 (01) :1-38
[7]  
Devore J., 2009, PROBABILITY STAT ENG
[8]   Simulation of knock probability in an internal combustion engine [J].
Di, Huanyu ;
Shen, Tielong .
PHYSICAL REVIEW E, 2018, 98 (01)
[9]  
Eriksson Lars., 2014, MODELING CONTROL ENG
[10]   An On-Board Calibration Scheme for Map-Based Combustion Phase Control of Spark-Ignition Engines [J].
Gao, Jinwu ;
Zhang, Yahui ;
Shen, Tielong .
IEEE-ASME TRANSACTIONS ON MECHATRONICS, 2017, 22 (04) :1485-1496