Realization of dynamic test and measurement system for railway electrification systems

Today, the demand for railway transportation is increasing. Studies in this field in Turkey have accelerated after 2007. Currently, high-speed trains operate on Ankara-Konya and Ankara-Eskişehir lines, reaching a speed of 250 km / hour and the travel time is significantly shortened. While rail infrastructure work in Turkey remains , this rate the appropriateness of made roads and train sets are of great importance in terms of road and passenger safety. The tests performed to ensure road and passenger safety can be grouped under 3 main headings. The first of these is the controls on the electrification line. The energy required for the movement of the train is provided from the power line. In order for the rail system vehicle to be continuously energized, there must be continuous contact between the power line and the pantograph. Even a short-term non-contact between the pantograph and the catenary adversely affects the rail system vehicle and the electronic systems inside. For this reason, pantograph and catenary interaction should be tested dynamically and statically in certain periods. In Turkey, tests and measurements of railways are made by companies of foreign origin by paying certain fees per km or systems purchased from those companies are purchased. This situation causes our country resources to go out. With this work, a system equivalent to the test systems made abroad has been developed with our own hardware and software configuration. Labview software was preferred in the study in terms of both its compatibility with National Instrument products and its efficient operation. While the software was being developed, each module and the sensors it will work with were tested separately and then integrated into the detailed software one by one. Four load cell graphs to measure four load cells simultaneously and display the graphs on the graphical interface, four accelerometric sensor graphs to measure four accelerometer sensors simultaneously and display the graphs on the screen, a detachment graph, a graph measuring total force, There is a screen showing GPS data and one map showing the real time location of the train integrated with Google maps program. The data coming from the sensors and the GPS device are synchronized and saved to the computer in real time in csv file format. In the software, the records are only taken when the train is moving, and no record is taken when the train stops, which prevents extra space occupation. The system is now above the capacity of trains that measure on a national basis, and Google maps support takes the system one step further. In this study, the measurement results are recorded in csv format that can be accessed by everyone. The next step after saving the data is to interpret them by an expert personnel and even to determine if there is a malfunction. For this purpose, deep learning algorithms were used and the training results were tested successfully up to 98,54%.