A method of inspecting the rattle generated in a vehicle cabin has been developed. The waveformdata of overall in-cabin noise are analyzed using Wigner distribution, a kind of time-frequencyanalysis, and the rattle component contained in the data is condensed and separated from thebackground shake noise. Then the rattle component is classified into three levels, strong, middle andundetected, using the neural network. Fuzzy inference is also used to select regular waveform data.Experimental results show that the correct classification ratio of the method is more than 90%, whichequals the ability of skilled inspection operators.
展开▼