A novel diagnostic system for gasoline-engine leakage detection

作者：Chen P C^*

来源：Proceedings of the Institution of Mechanical Engineers - Part D: Journal of Automobile Engineering , 2011, 225(D5): 673-685.

DOI：10.1177/0954407010395687

• leakage detection
• crankcase ventilation
• gasoline engine

The current paper presents a leakage diagnosis system for elucidating the leakage source and identifying the leakage degree of gasoline engines. The proposed diagnostic system consists of a crankcase-ventilation detecting system, a pressure-regulator detecting system, a leakage-source detecting system, and a decision system. Being based on neural networks, the three detecting systems are trained using the steepest-descent method and a back-propagation algorithm. Although the engine non-linearity and noise may spoil the leakage identification, the decision system can be employed to enhance the identification of leakage source and leakage degree. Genetic algorithms are used to train the initial weights and biases of the neural networks such that the training performance is raised. Experimental results indicate that the proposed diagnostic system for elucidating the leaking source and leakage degree is feasible and effective.

• 出版日期2011