Gaussian Mixture Model Clustering-Based Knock Threshold Learning in Automotive Engines

In this article, a Gaussian mixture model (GMM) clustering-based method is proposed to learn the optimal threshold of the knock intensity metric, which minimizes the probability of the judgment error of the knock event. First, statistical analysis of knock intensity on an experimental database measured from a gasoline engine test bench is conducted and the results show that GMM has better fitting performance than the single Gaussian model. Then, the learning problem which consists of two subproblems is formulated based on the assumption that the probability distribution model of knock intensity is a two-component GMM. The first subproblem is a clustering problem in which parameters of the two-component GMM are optimized to maximize the likelihood of obtaining the measurements. The second subproblem is also an optimization problem in which the probability of the judgment error is formulated as a function of the threshold. The solution to the problem is the required threshold decision. The convergence analysis of the proposed method is implemented. Furthermore, the proposed method is also validated in the experimental database. The proposed method is expected to be applied in the real-world application to save time, labor effort on the knock threshold calibration.

[1]  Giulio Panzani,et al.  Engine Knock Margin Estimation Using In-Cylinder Pressure Measurements , 2017, IEEE/ASME Transactions on Mechatronics.

[2]  Jesse Frey,et al.  Characterization of knock intensity distributions: Part 2: parametric models , 2013 .

[3]  Michael I. Jordan,et al.  On Convergence Properties of the EM Algorithm for Gaussian Mixtures , 1996, Neural Computation.

[4]  Geoffrey J. McLachlan,et al.  A Block EM Algorithm for Multivariate Skew Normal and Skew $t$ -Mixture Models , 2016, IEEE Transactions on Neural Networks and Learning Systems.

[5]  Jinwu Gao,et al.  Probabilistic Guaranteed Gradient Learning-Based Spark Advance Self-Optimizing Control for Spark-Ignited Engines , 2018, IEEE Transactions on Neural Networks and Learning Systems.

[6]  Jesse Frey,et al.  Optimizing knock thresholds for improved knock control , 2014 .

[7]  Fengrong Bi,et al.  Knock Detection Based on Recursive Variational Mode Decomposition and Multilevel Semi-Supervised Local Fisher Discriminant Analysis , 2019, IEEE Access.

[8]  Eric R. Ziegel,et al.  Probability and Statistics for Engineering and the Sciences , 2004, Technometrics.

[9]  Tao Xu,et al.  The engine knock analysis – An overview , 2012 .

[10]  Georg Brasseur,et al.  Analysis of Autoregressive Coefficients of Knock Sensor Signals for Misfire Detection in Internal Combustion Engines , 2019, 2019 IEEE International Instrumentation and Measurement Technology Conference (I2MTC).

[11]  O. Boubai.  Knock detection in automobile engines , 2000 .

[12]  E. Galloni,et al.  Dynamic knock detection and quantification in a spark ignition engine by means of a pressure based method , 2012 .

[13]  Jesse Frey,et al.  Characterization of knock intensity distributions: Part 1: statistical independence and scalar measures , 2014 .

[14]  Giulio Panzani,et al.  Adaptive and Unconventional Strategies for Engine Knock Control , 2019, IEEE Transactions on Control Systems Technology.

[15]  Hui Xie,et al.  Trade-off on fuel economy, knock, and combustion stability for a stratified flame-ignited gasoline engine , 2018, Applied Energy.

[16]  Jesse Frey,et al.  A Bayesian Knock Event Controller , 2020, IEEE Transactions on Control Systems Technology.

[17]  Alan L. Yuille,et al.  Statistical Physics, Mixtures of Distributions, and the EM Algorithm , 1994, Neural Computation.

[18]  Tielong Shen,et al.  Real-time statistical learning-based stochastic knock limit control for spark-ignition engines , 2017 .

[19]  Yorgos Goletsis,et al.  Hierarchical Similarity Transformations Between Gaussian Mixtures , 2013, IEEE Transactions on Neural Networks and Learning Systems.

[20]  D. Rubin,et al.  Maximum likelihood from incomplete data via the EM - algorithm plus discussions on the paper , 1977 .

[21]  Enrico Corti,et al.  Knock Indexes Thresholds Setting Methodology , 2007 .

[22]  P. Bares,et al.  A new knock event definition for knock detection and control optimization , 2018 .

[23]  Mitsuji Sampei,et al.  Knocking detection in gasoline engines based on probability density functions: A mixed Gaussian distribution approach , 2015, 2015 54th IEEE Conference on Decision and Control (CDC).

[24]  Yahui Zhang,et al.  Adaptive Internal Model Based Control of the RGF Using Online Map Learning and Statistical Feedback Law , 2020, IEEE/ASME Transactions on Mechatronics.

[25]  Jiancang Zhuang,et al.  Forecasting the magnitude of the largest expected earthquake , 2019, Nature Communications.

[26]  A. Amell,et al.  Engine operation just above and below the knocking threshold, using a blend of biogas and natural gas , 2018, Energy.

[27]  Asok Ray,et al.  Bayesian Nonparametric Regression Modeling of Panel Data for Sequential Classification , 2018, IEEE Transactions on Neural Networks and Learning Systems.

[28]  Jinwu Gao,et al.  An On-Board Calibration Scheme for Map-Based Combustion Phase Control of Spark-Ignition Engines , 2017, IEEE/ASME Transactions on Mechatronics.

[29]  Guoming Zhu,et al.  Stochastic limit control and its application to spark limit control using ionization feedback , 2005, Proceedings of the 2005, American Control Conference, 2005..

[30]  Jon Rigelsford,et al.  Automotive Control Systems: For Engine, Driveline and Vehicle , 2004 .

[31]  Yuhu Wu,et al.  Design, implementation and experimental verification of a compensator-based triple-step model reference controller for an automotive electronic throttle , 2020 .

[32]  Lars Eriksson,et al.  Modeling and Control of Engines and Drivelines , 2014 .

[33]  James C. Peyton Jones,et al.  Likelihood-Based Control of Engine Knock , 2013, IEEE Transactions on Control Systems Technology.

[34]  Jinwu Gao,et al.  Challenges and solutions in automotive powertrain systems , 2018 .