چكيده به لاتين
The vital and undeniable role of the rail transportation system in the economic and social growth and development of countries on the one hand and the high costs related to the repair of rail lines on the other hand have caused the existence of a comprehensive management system for the repair and maintenance of rail lines to a It becomes necessary to maintain the level of service and safety of the line at an optimal level and to reduce maintenance costs to the minimum possible amount. Therefore, the engineers and technicians active in the railway industry are always looking for a solution to reduce the wear and tear of parts and achieve the appropriate time measurement for the timely maintenance of equipment and tools used in the rail transportation system.
The use of new sciences in this industry, especially the science of machine learning, along with the intelligentization of the various systems used, makes it possible for us to detect defects in specific parts such as rails. Now, in this research, which includes two parts of simulation and coding, the method of using machine learning science has been discussed. The way the work is done is that for the simulation part of a passenger car in the dynamic simulator software, with the desired conditions for a certain profile of wheels and rails and standard irregularities of railway lines at several different and specific speeds. and the result of this simulation is receiving the output signal of head axis acceleration, which is used as the input of another part of the research, that is, the input of machine learning. After receiving the signals at any speed and any failure class according to the type of fault, the machine learning algorithm will have the ability to detect any type of line class based on the FRA standard with the speed being known. This intelligent system is designed in several levels of deterrence, warning and replacement or repair warning to predict possible risks. In this research, several machine learning algorithms have been used to compare the performance and accurately predict the classification of failures, and the accuracy of each has been investigated, and finally, as a result, it can be mentioned that this intelligent system saves a lot. It brings in the time and cost associated with the maintenance of railway lines for engineers and inspectors
