Performance prediction of inverter-induction motor drives for DC-fed railways

Simulation of the steady state aspects of a converter-fed AC traction drive is demonstrated using a parallel computer. The solution of the system equations is carried out by exploiting concurrency, through multitasking software. The computer has an efficient hardware structure consisting of parallel processors with shared memory, the intelligence of each being exploited to retain operational independence from the others. Software then identifies data dependency and prioritises the tasks for execution on the separate processors. The results presented show the main system variables of an inverter-fed AC traction drive as the train accelerates through a given duty cycle and with predictable load characteristics