Research On Diesel Engine Fault Warning System Based On Fuzzy Neural Network

As the core of mechanical power,diesel engines are of great significance for monitoring their condition and providing fault prediction.However,the composition of the body is complex,and the working mechanism involves many fields,making it difficult to obtain effective and accurate fault characteristics when troubleshooting diesel engines.Aiming at the difficulty of extracting diesel engine fault characteristics,the lack of clear relationship between fault location and fault characteristics,and the difficulty of fault location,this paper takes a diesel engine as the research object and carries out the research and design of diesel engine typical fault early warning system.The contents are as follows:(1)Aiming at the problem that the rotational speed of the diesel engine is not constant and the vibration signal of the body cannot be accurately collected,a fixed speed signal acquisition system is designed.The sampling process is triggered by the limiting condition and the sampling signal is processed by the conditioning circuit to obtain stable stability.Vibration signal.(2)A method based on empirical mode decomposition(EMD)and time series model(AR)is proposed.The original vibration signal is extracted by EMD-AR method.(3)Combining fuzzy algorithm with neural network,establishing integrated fuzzy neural network(FNN)system and improving Takagi-Sugeno model for diesel engine fault diagnosis,effectively solving the ambiguity and difficulty of positioning in the diagnosis process.(4)The diesel engine typical fault early warning system is designed.The above research method is successfully embedded into the early warning system as a sub-module to realize the fault diagnosis function and verify the effectiveness of the method.