Multi-Hypothesis Tracking of charged particles through drift tube arrays

We report on the application of techniques from the target tracking literature to the experimental study of the distribution of charged particles. In the experiment we have multiple planes of position-sensitive detectors able to detect locations of particle hits in each sensitive plane. The problem we address here is determining how to combine a series of particular hits into a particle trajectory. This simple task is complicated by the specifics of the measurements, missed hits and background. In the target tracking literature, the challenge of associating hits in different images is called the association problem, and current best practice is Multi-Hypothesis Tracking or MHT. Recent work has investigated the possibility of using Markov chain Monte Carlo (MCMC) methods for addressing the multi-hypothesis challenge. In our study we report on results from both approaches and on the trade off between computational budget and accuracy. We extend the usual MHT formalism to select tracks with certain qualities and set quality flags as required by the experiment. In our presentation we describe in more detail our approach and preliminary results of its application to the experimental data.