Intelligent System Developing Of Extracting Characteristic Parameters Of Random Load And Predicting Fatigue Life

The existent study on the random load during analyses of fatigue life prediction usually considers amplitude and mean of the load, in this way the random load can't be described exactly, accordingly the fatigue life can't be predicted accurately.To solve this problem, using signal analyses technology, mathematical statistics method and fractal theory, the author abstracted parameters of the random load from these four aspects, amplitude, time domain, frequency domain, and shape of the random load.In this way, the random load can be described more exactly.Based on these parameters, the author set up the similitude-fatigue life prediction method. If the two random loads are similar, use the relationship of the two equivalent stress amplitudes of the two similar load to predict the unknown life; If the two random loads are dissimilar, use the generally life prediction method, in detail. The P-S-N curve and Miner principle are used in life prediction with generally method.The author also developed the fatigue life prediction system which was based on the parameters of random load for realizing the methd through software. It comprised two primary parts, extracting characteristic parameters and predicting fatigue life. Using technology of VB and MATLAB, the system took the advantages of them, so the system is provided with fast operational speed and friendly interface. Using the interface technology of VB and Microsoft Access software, the system was provided with convenience operation and perfect function.