부산대학교 도서관

A COMET Experiment Phase-I is designed to search for a muon to an electron conversion with a 100 times better sensitivity than the current upper limit on its branching ratio in order to investigate the new physics beyond the standard model. To achieve such an unprecedented sensitivity, a high intensity muon beam is mandatory. In consequence, an extremely high hit rate is predicted and it can cause a high trigger rate more than tens of kHz dominated by the background events. Then the high trigger rate would lead to the severe dead-time in the data acquisition. Therefore it is essential for a success of the experiment to reduce the trigger rate down to a few kHz within a processing time shorter than 5 μs while maintaining the signal efficiency high enough. We are developing an event classification algorithm utilising a CDC hit information and a dedicated fast online trigger system to accomplish this aim. The event classification based on a boosted decision tree shows 3.8 kHz of trigger rate, which almost satisfies the rate requirement, with a good compatibility to the trigger electronics. The total processing time to generate the trigger decision is estimated to be ~3 μs which also fulfils the timing requirement.