Fig. 7 Three-dimensional figure of vibration (190 hours)
Fig. 8 Three-dimensional figure of vibration (2300 hours)
Fig. 9 Three-dimensional figure of vibration (6000 hours)
If the regularity can be understood, generation of an abnormality in the changing normal pattern can be understood. From the above-mentioned viewpoint, in the following chapter, the event is examined by applying the statistical analysis.
4. Data analysis
In order to grasp a normal condition and an abnormal condition of an engine from the pattern of the vibration data to be changed with the service time of the engine, the relationship between the running time of the engine and the vibration must be examined. Thus, the possibility of explaining the relationship using the multiple regression analysis was considered. The multiple regression analysis is a technique to explain the dependent variables by explanatory variables of multivariates, and in this study, the multiple regression equation was obtained with the value of the vibration data obtained through the frequency analysis as the explanatory variable, and with the running time of the engine as the dependent variable. In preparing the above-mentioned three-dimensional figures, the required number of the variables is 24000 in total, because one cycle is composed of 120 sections and one section is composed of 200 variables.
