Release Time:2019-03-13  Hits:

Indexed by: Conference Paper

Date of Publication: 2018-01-01

Journal: IFAC PAPERSONLINE

Included Journals: CPCI-S

Volume: 51

Issue: 31

Page Number: 125-130

Key Words: SI Engine; Statistic Control; Knock; Beta Distribution; Bayes' rule; Likelihood

Abstract: Knock phenomenon as a stochastic process requires feedback control for its relation to engine efficiency, noise and cylinder damage. In this paper, a knock probability estimation method using Bayes' updating rule and beta distribution is proposed based on the independent and identically distributed (iid) characteristic analysis of the knock event sequence. A stochastic control algorithm using the estimation method and likelihood ratio test is also proposed. The proposed control algorithm is validated on a production spark ignition engine and shows the ability to maintain knock probability close to the target. (C) 2018, IFAC (International Federation of Automatic Control) Hosting by Elsevier Ltd. All rights reserved.