상하개폐식 철도 승강장 안전문의 제어 알고리즘 개발

Abstract

PSD (Platform Screen Door) effectively prevented accident at railway station. No accident has been reported since PSDs were installed in all the Seoul metro stations. Nevertheless, PSD demonstrated an issue when trains are manually driven and where different types of trains operate. For existing types of PSD, trains must stop exactly at the designated position. In the manual driving mode, train operators often cause minor errors in the stopping position and the screen door is not opened for safety when the stopping position is beyond ±350mm from the designated position. The same problem occurs at the train stations where different types and sizes of trains are operated.

On the other hand, RPSD (Rope type Platform Safety Door) allows the stopping position error up to ±1,950mm. This allows more flexible operation of trains even in the manual driving mode and enables serving for various types of trains. This study presents an algorithm to control the vertical operation for RPSD. Literature review showed that three methods have been widely used to control the PSD door including ATO (Automatic Train Operation), RF(Radio Frequency), and door detection types. The ATO method requires train signal system. For the RF type, it is essential to install vehicle and track infrastructures and also upgrade of the RF device in vehicle. Finally, the door detection type installs door detectors at fixed locations and thus has a limitation in applying to the stations where different types of trains with various sizes are operated and the stopping location of train doors vary.

In chapter three, an algorithm is presented to control the RPSD door to mitigate the above mentioned limitations of the existing methods. The algorithm opens the screen doors before trains completely stop at the designated position and closes the doors after trains depart. The algorithm relies on a sensor to detect the exact location of trains. For evaluation of the algorithm we used VISSIM simulator and its add-on module including Com-interface and Visual Basic Application Programming Interface (API). The control algorithm demonstrated outstanding performances under four testing scenarios.

In chapter four, another control algorithm is proposed that allows passengers to manually open and close the RPSD screen doors. This algorithms utilize the obstacle sensors installed at the front and back of RPSD. This algorithm was evaluated in the simulation environments. The test results indicate that the manual control of the screen doors was only allowed when certain amount of external force is detected from the train side and it was not allowed when external force is detected from the terminal side. The fifth chapter evaluates the screen door control algorithm after installing it in the Door Control Unit (DCU). The test was conducted at Moonyang station and the RPSD was operated total 127,680 times of screen door opening and closing. A control error occurred sixteen times and eighteen times for the downward direction and the upward direction, respectively. Therefore, the algorithm’s availability was found to be at the level of 99.99%. This test result represents that the control algorithm includes the robust reliability. The test also revealed that the control errors were caused by the sensor errors. Thus, it is expected that improving the sensing technology will enhance the algorithm’s availability.