Determination of Electrical Parameters of a Traction Current Circuit Using Measured Data in the Traction Power Supply System of an 825-V Subway

Abstract

Providing selectivity of the relay protection of a dc network has always been associated with difficulties in distinguishing between operating and emergency modes, which is especially noticeable in the traction power supply system for an 825-V subway due to the high traffic intensity and the use of electric rolling stock (ERS) with asynchronous traction motors. The practice of calculating protection settings shows that the maximum current protection settings (MCPs), both for an electromechanical relay and for microprocessor terminals, as a rule, are set off from the calculated minimum short circuit currents, which can be lower than the startup currents of the ERS. This practice of calculations and modern conditions of operation of a subway power supply system can lead to false activations, which, in turn, can lead to a rapid loss of the switching capacity of high-speed switches, as well as lead to a delay in train movement. An analysis of the current situation leads to the conclusion that simple types of protection, such as MCPs, cannot provide selectivity of outages. The situation is different with multiparameter protection of the type of protection by current increment (PCI) and current buildup rate (PCBR). The essence of the operation of the PCI and PCBR is based on the analysis of changes in the current graph, which allows determination that is faster, and sometimes more accurate, than MCP, whether or not the change in current is connected with the occurrence of a short circuit. However, taking into account the choice of protection settings for changes in current and its buildup rate requires the calculation of transients. The accuracy of such calculations directly depends on the accuracy of determining the electrical parameters of the traction circuit. In modern conditions of development of measuring and computing equipment, one of the most simple and effective methods of calculation is the method of minimizing the distance between a set function or model and the measured data. It is worth noting that the use of this method directly in the algorithmic part of microprocessor-based protection terminals together with self-learning functions will allow one to automatically select the desired values of the PCI and PCBR settings corresponding to the operating conditions of the traction power supply system.