Study On Diesel Engine Fault Diagnosis Technology Based On Multisensor Information Fusion For Valve Train

As the most common power mechanical, Diesel Engine due to its high heatefficiency,high power per quality, is widely used for ships, automobiles， powergenerations,mechanical equipments,etc．But because of its complex structure and severe working environment, diesel enginebreakdown easily, especially for the partial parts, such as valve train. It will affect theoverall performance and could cause an accident. So diesel engine fault diagnosis is the themain direction of engine management.As this background paper proceed a study on dieselengine fault diagnosis technology based on multisensor information fusion.Firstly analyse the rationale and symptom for failure of valve train, based on the valvestructure and working principle.Second designed a multi-sensor information fusion methodof diagnosis taking the vibration parameters as main data source.Final Set up theexperiment rig and state monitoring system.In the experiment simulated two faults: diesel engine valve leakage and valve gapabnormal. Establish the operating parameters database for normal and fault conditions. Thediagnostic system fusion the instantaneous speed cylinder head vibration signal and TDCsignal, to resample vibration signal.Get vibration transformation form the time domain toangle domain. Solved influence to the results by loss or increased pulse in instantaneousspeed measurement process. And analysed the affect form angular resolution interpolationalgorithm to resample data.then analysed the fault segment data with Hilbert-HuangTransform.Components off the cylinder head vibration signal at different excitationsource.using the Hilbert Spectrum Entropy and Mean-Square deviation quantifycharacteristic parameters. Comparative and analyse of diesel engines conventionalparameters such as: oil pressure, air pressure signal and the relationship betweenthem.Finally design of the RBF network for information integration.Experimental results demonstrate the information fusion process and the design areeffectiveness, and have a huge space for development.