Track reconstruction in the TRD and MuCh detectors

The successful application of the TRD tracking algorithms [1] lead to the idea to generalize and optimize those algorithms in a flexible way in order to make them applicable for other, similar detectors. In this contribution we present how this idea was realized and applied for the TRD layout optimization and for track finding in the MuCh detector. Since both tracking detector setups (TRD and MuCh) are analogous in many details, the same approach can be used for tracking in these detectors. Therefore, a generalized algorithm of track finding in the CBM tracking detectors has been developed. It is based on the Kalman Filter and track following methods where the next track point is searched in an area surrounding a predicted point. The main idea is to apply a general algorithm of track finding to different detectors and only adopt it to a specific one by changing the tracking routine parameters. This helps to decrease code duplication, thus significantly easing the software support. In our implementation, tracking is accomplished in an iterative way. In each iteration, tracking parameters have to be specified, and hits belonging to tracks found in the current iteration are deleted from the hit array. In order to use the tracking routine, one has to specify the number of tracking iterations, the maximum number of allowed missing hits in a detector station, start and end station for the tracking, the sigma coefficient determining the searching region, and other parameters of the tracking. The searching region can be determined in two different ways: (1) using the covariance matrix of the predicted track parameters and position errors of the hits; (2) calculating the maximal deviation between the predicted position and a hit on the basis of a look-up table obtained from a simulation with large statistics. For the TRD, two different approaches have been used: a standalone TRD track finder (using only TRD information), and an algorithm based on the information from tracks found in preceding detectors (STS-based or MuChbased). For the MuCh detector, which consists of a sequence of several absorber and detector layers, vertex tracks reconstructed in the STS have been used as seeds for track reconstruction. The software was embedded into the CbmRoot framework and tested on central Au+Au collisions at 25 AGeV beam energy from UrQMD. Using the TRD track finder, a detector layout study has been performed in order to optimize the detector setup while keeping high reconstruction efficiency. The aim of this study is to minimize the costs of the detector. The