Filtering a Random Loading Process in the Problem of Assessing the Durability of Rolling Stock Parts

The current problem of choosing a method for filtering a random loading process, which in some cases may be nonstationary, is considered. This issue is closely related to the task of assessing durability, since its solution determines how fully the components of the oscillations (including high-frequency ones) will be represented in the output load vector. The problem is considered using the example of loading processes obtained during operational measurements of a real object at several speeds of movement used most frequently. Various filtering methods are considered: using low-pass median filters, using the level method, and using mode decomposition. Since the processing involves subsequent assessment of the durability, conditional durability assessments were carried out in order to show the similarities and differences of the filtering methods when calculating the filtered processes. Recommendations are given for preprocessing of processes.