This paper presents the diagnostic methods of self-propelled gun diesel engine. Telemetry technique is used to realize wireless transfer It makes performance parameters testing possible during self-propelled gun advanced. Hardware composition of characteristic signal extracting is analyzed and introduced, the theory and application of transient speed and cylinder vibration signals are studied systematically for fault diagnosis. Angle domain average analysis method is proposed to isolate useful signal from the compound cylinder vibration signal enhanced the SNR (Signal-to-Noise). Finally the BP neural network is applied for pattern recognition and fault diagnosis of diesel engine according to the characteristic feature extracted from cylinder vibration signals and transient rotation speed waveform. It has obtained a better effect in practical application which makes fault diagnosis theory of self-propelled gun diesel engine abundant and perfect.
