The Application Of Stochastic Resonance On Vibration Analysis Of Civil Aero-engine Rotor

Weak signal detection is usually being interfered or covered by the noise, which makes signal detection more difficult. Therefore, the "noise" is often considered as disgusting things. Stochastic resonance is a new method to detect weak signal. Compared with the traditional methods which simply remove the noise, Stochastic resonance is a phenomenon that enhances the output of nonlinear system which is triggered by the effect of both weak signal and noise under certain nonlinear condition.The civil aviation engine is a typical representative of rotating machinery. The rotor system which is the core components of civil aviation engine will trigger vibration due to design, manufacture, assembly and other factors. In order to reduce the interference of noise effectively in the extraction of vibration signal of civil aero-engine rotor and improve the accuracy of the rotor vibration signal extraction, stochastic resonance can be used to deal with the vibration signal of a rotor.The vibration signal of civil aviation engine rotor is a kind of the Aperiodic signal, not periodic signal which stochastic resonance often deal with. Introducing ASR which comes from cross-correlation measurement method and measure of the information theory(average mutual information). The application scope of stochastic resonance expands from weak periodic signal to aperiodic signal. Considering the noise in the engine rotor vibration signal is unknown and time invariant in real engineering application, the way of adjusting the structure parameters of the bistable system was adopted to achieve stochastic resonance.Simulation algorithm of bistable stochastic resonance was studied, and MATLAB was used in simulation experiments. Achieving stochastic resonance processing of rotor vibration signal in one or a few engines, and successfully obtaining the stochastic resonance output waveform of high or low pressure rotor. Bistable stochastic resonance can effectively extract the aero-engine rotor vibration signals and has outstanding advantages of extracting vibration signal of rotor from heavy noise, and it can monitor the short-term failure phenomenon when the vibration data is collected, in addition, it is an effective method to reflect the real vibration of the rotor. Considering one layer of stochastic resonance can't extract the best output waveform of the engine rotor vibration signal in some cases, the engine rotor vibration signal was carried on layers of stochastic resonance processing, and come to the conclusion that choosing the layer number of stochastic resonance according to need.Client/server development mode was adopted, and Microsoft SQL Sever 2008 was used to set up a database for the stochastic resonance analysis system of engine rotor. The server which SQL Sever 2008 works on the local server. Microsoft Visual Studio 2010 was used to develop the application of engine rotor stochastic resonance analysis system. The connection between client and server can be considered as the connection between the application and the database, and this system can call the database mainly through ADO.NET. C++ client applications connect the Server SQL Server 2008 database through local server to access the required engine rotor vibration data. The stochastic resonance analysis system of engine rotor R1.0 can realize the multilayer stochastic resonance processing on one or a few engines. This application is easy to operate,and the interface is simple, and there is a good prospect to popularise and apply it in industry.