Effective testing of computerized equipment for EMC immunity to fast electrical transients

One of the major problems in testing computerized equipment for immunity to electrical transients is their time-variant susceptibility. Since complete immunity testing of an operational cycle with hundreds of susceptible time windows is prohibitive this paper aims at providing methods for affordable estimation of the mean and maximum malfunction probability in an operational cycle. For the latter reinforcement learning algorithms are used which automatically restrict testing to most susceptible time windows. This allows affordable testing also for safety-related equipment and supports effective redesign