Knock Detection Based on Recursive Variational Mode Decomposition and Multilevel Semi-Supervised Local Fisher Discriminant Analysis

Abstract

Knock is an abnormal combustion phenomenon in gasoline engines. Strong knocks will reduce the efficiency and durability of engine, while with slight knocks engines can run on a high-efficiency state. It is necessary to detect knock and control the state of knock in order to improve the thermal efficiency of engine. This paper proposes a novel approach for detecting engine knocks in various intensities based on vibration signal of engine block using variational mode decomposition (VMD) and semi-supervised local fisher discriminant analysis (SELF). Since the quadratic penalty of recursive VMD has a great influence on decomposition results, the approach establishes the connection between the quadratic penalty and the stop condition by analyzing a large amount of data and quantifies the relationship by polynomial fitting, which reduces the complexity and subjectivity of recursive VMD. A multilevel SELF is developed for solving the problem that SELFs sometimes may not find ideal embedding space under large scale dimensionality reduction. This method adopts multi embedding spaces, with gradually decreasing dimension, to reduce the dimensionality of original data by a series of small steps. Verifications show the proposed approach can achieve high classification accuracy in knock detection and is able to identify the intensity of knock. This research contributes to the field of engine abnormality detection and can be implemented on vibration-based faults diagnosis area.INDEX TERMS Engine, knock detection, semi-supervised local fisher discriminant analysis (SELF), variational mode decomposition (VMD), vibration.

A Novel Multiple Feature-Based Engine Knock Detection System using Sparse Bayesian Extreme Learning Machine

“…The results indicated that the knock detection accuracy was 97%. AAn approach for detecting engine knocks inof various intensities based on the vibration signal of an engine block using VMD and semi-supervisedsemisupervised local fisherFisher discriminant analysis was proposed in [3], and the classification rate forof strong knocks was over 95%. As mentioned above, there is much room for improvement in the denoising performance and accuracy.…”

“…In recent years, VMD has been introduced for noise analyses of rotating machines and as [8,28,35]. Although signals are separated into a series of IMFs, IMFs depend on the values of the balancing parameter and the number of modes that are adjustable, and the results may be inaccurate when they are not set in place [3]. Therefore, an optimization method utilizing GA is proposed in this work to solve the problem of parameter optimization.…”

An intelligent fault diagnosis method for an electromechanical actuator based on sparse feature and long short-term network

“…where TP i represents the number of true positive instances, FP i represents the number of false positive instances, FN i represents the number of false negative instances and i = 1, 2, 3, 4 represents the four faults in table 1. As shown in (22), the F-measure index contains information about the precision rate and recall rate, whose value ranges from 0 (worst) to 1 (best) [51].…”

“…However, we consider the SAE-based CS method a sparse representation method to obtain a compressed matrix and perform further FDI work. Compared with the traditional DR method, PCA [16][17][18], linear discriminant analysis [19], and Fisher discriminant analysis [20][21][22] etc, this method does not require prior knowledge to obtain features and is easily operated.…”