The paper analyses the possibility of studying dynamic loads in transmissions and load-bearing systems of wheeled and tracked machines by bringing a non-stationary random process, which describes the system, to a stationary one by using a finite-difference method of correlation analysis of a process of additive form. This method is particularly interesting for the study of the widespread methodological approach to the study of load modes of car elements, smoothness of running, in which the amplitude-frequency information about the mode is obtained by processing the records of the car movement on a measured section with a constant speed. For comparison, we present a graph of the correlation function of the same random process constructed by the method of preliminary smoothing of the initial process with simultaneous approximation by a cubic parabola and subsequent centering of the initial process relative to the parabola. It was found that faster decay of the correlation function indicates larger errors and noise caused by them, inherent in the algorithm of the method with the smoothing operator. The possibility of using the proposed methodology has been analyzed and recommendations for its use have been proposed. .

错误:搜索内容不能为空,请输入英文关键词
错误:关键词超出字数限制,请精简
高级检索

On Reduction of a Non-stationary Random Process to a Stationary One in the Study of Dynamic Loads

  • Yevhen Kalinin,
  • Volodymyr Boldovskyi,
  • Volodymyr Demianyshyn,
  • Kateryna Boldovska,
  • Oleksandr Hasiuk

摘要

The paper analyses the possibility of studying dynamic loads in transmissions and load-bearing systems of wheeled and tracked machines by bringing a non-stationary random process, which describes the system, to a stationary one by using a finite-difference method of correlation analysis of a process of additive form. This method is particularly interesting for the study of the widespread methodological approach to the study of load modes of car elements, smoothness of running, in which the amplitude-frequency information about the mode is obtained by processing the records of the car movement on a measured section with a constant speed. For comparison, we present a graph of the correlation function of the same random process constructed by the method of preliminary smoothing of the initial process with simultaneous approximation by a cubic parabola and subsequent centering of the initial process relative to the parabola. It was found that faster decay of the correlation function indicates larger errors and noise caused by them, inherent in the algorithm of the method with the smoothing operator. The possibility of using the proposed methodology has been analyzed and recommendations for its use have been proposed. .