TA-POS-2016-047

Three algorithms, designed to search for clusters in the electromagnetic calorimeter (EMC) of the future PANDA experiment, are described and evaluated. The high rate of 2E7 interactions/s between the antiproton beam and the proton target leads to a high probability for pile-up and event mixing, complicating the search for clusters. To reduce the amount of data produced by the experiment, the detector will preprocess as much data as possible online, and select events of interest based on online reconstruction. To perform these reconstructions, input from the cluster finding algorithm, which is part of the online preprocessing, is essential. The algorithms are tested using simulated data for both their efficiency and processing time using two 5000 event data samples in the high rate
case, a lower rate case (factor 100 lower rate), and the case with completely disjoint events. The first sample is a low photon-multiplicity sample of direct two-photon production events. The efficiency, i.e. the number of successful reconstructions, is similar for all methods, around 90%. The algorithm that is most eligible for online usage runs the fastest, as expected. The second sample tests the high photon-multiplicity channel of the hc meson to 7 photons, and will also be expanded with background events. More detailed results on the performance evaluation will be presented.