A Fault Detection Method for Injectors of Diesel Engine Common Rail System Based on Measurable Data Features

Abstract

Disclosed in the present invention is a method for detecting faults of fuel injectors in common rail system of diesel engines based on measurable data features. The method first measures the fuel injectors from four aspects: time domain, frequency domain, pressure-rate, and empirical characteristics. Parameters are used for feature extraction, and a complete fault feature set is constructed; then principal component analysis (PCA) is used to reduce the redundancy between features, and effective low-dimensional orthogonal features are extracted; then, support vector machine (SVM) is used to analyze the fault mode Classification to identify the type of failure. The invention solves the problems that the existing marine high-power diesel engine electronically controlled high-pressure common rail system fuel injector fault symptoms are weak and difficult to distinguish, the fault types are diverse, and the fault samples are sparse, which makes it difficult to realize real-time online fault detection, and realizes fast and accurate diagnosis of fuel injectors function of the fault.

Description

technical field [0001] The invention belongs to the fault diagnosis technology of marine high-power high-pressure common-rail diesel engines, and relates to a method for detecting faults of diesel engine common-rail system fuel injectors based on measurable data characteristics, and specifically relates to a certain type of marine high-power high-pressure common-rail diesel engine electronically controlled fuel injection Diagnosis technology for small faults such as control plunger and needle valve coupler. Background technique [0002] As a high-efficiency, low-power consumption, and light-emission fuel injection technology, high-pressure common rail technology has become an important development direction for marine diesel engines. The electronically controlled fuel injector is one of the most complex and important components in the high-pressure common rail fuel injection system. According to the instructions of the controller, it controls the switch of the solenoid valve...

AI Technical Summary

Problems solved by technology

[0004] Existing detection of marine diesel engines mainly relies on the experience of staff and on-site instruments for detection. The detection and alarm system equipped is only a simple parameter over-limit alarm and does not have a fault diagnosis function. At present, researchers have also used some machine learning Methods The fault diagnosis of marine diesel engines is studied, mainly using fault tree, Bayesian network, statistical learning, neural network and other methods to diagnose the faults of diesel engine pressure signals, acoustic signals, vibration signals, etc. However, with the type of fault increase, the expert knowledge required by Bayesian network and other methods has increased significantly, and the complexity of modeling has limited its application in actual engineering. Difficult to meet the training requirements for neural networks, etc., and the fault diagnosis effect is not ideal

Images