Estimation and Prediction of the Condition of the Vehicle Engine Based on the Correlation Dimension

This paper applies the fractal dimension as a characteristic to describe the engine抯 operating condition and its developmental trend. A correlation dimension is one of the quantities that are usually used to characterize a strange attractor. With the operation of the phase space reconstruction, respective correlation dimensions of a series of vibration signals obtained under different conditions are calculated to find the intrinsic relationship between the indicator and the operating condition. The experiment result shows that the correlation dimension is sensitive to the condition evolution and convenient for the identification of abnormal operational states. In advanced prognostic algorithm based on the BP neural network is then applied on the correlation dimensions to predict the short-term running conditions in order to avoid severe faults and realize in-time maintenance. Experimental results are presented to illustrate the proposed methodology.

基于振动信号的车用发动机运行状态预测

考虑到车用发动机结构复杂、振源较多,振动信号为非平稳信号的特点,文中利用混沌与分形理论对多组不同运行状态的振动序列进行研究,探讨关联维数与运行状态之间的内在联系,并利用混沌与神经网络相结合的方法对主要状态参量进行单变量及多变量预测。实验结果表明,关联维数能敏感反应发动机的磨损状态,而多变量的预测效果比单变量效果理想。