Artificial Neural Networks Application in Estimating the Impact Parameter in Heavy Ion Collisions Using the Microchannel Plate Detector Data.

Evaluation of the impact parameter in a single event of relativistic heavy ion collision is crucial for correct and efficient data processing and analysis. In this work we have studied the possibility of estimating the impact parameter in heavy ion collisions by using artificial neural networks applied to the charged particle data from fast microchannel plate (MCP) detectors. Charged particles' multiplicity, their spatial distribution and time-of-flight data were used as event features to be analyzed by the artificial neural network algorithms. We investigated two different configurations of microchannel plate detector layout, that have different data and computational requirements. We have shown that the developed artificial neural networks technique is capable of providing sufficiently good and fast results on the impact parameter determination in a single heavy ion collision event for both configurations of MCP detectors layout. In our first exercises, the proposed algorithm has successfully identified more than 90 of Au Au collision events with the impact parameter less than 5 fm or less than 1 fm, which suggests its use as a fast trigger. [ABSTRACT FROM AUTHOR]