The multilevel trigger system of the DIRAC experiment

TL;DR

The paper describes the design and implementation of a multilevel trigger system for the CERN DIRAC experiment, which efficiently selects pion pairs with low relative momentum, reducing data rates while maintaining high detection efficiency.

Contribution

It introduces a novel multilevel trigger system incorporating neural network algorithms and detector correlations for improved event selection in high-energy physics experiments.

Findings

Reduces event rate by a factor of 1000

Achieves detection efficiency of at least 95%

Utilizes multiple trigger processing methods including neural networks

Abstract

The multilevel trigger system of the DIRAC experiment at CERN is presented. It includes a fast first level trigger as well as various trigger processors to select events with a pair of pions having a low relative momentum typical of the physical process under study. One of these processors employs the drift chamber data, another one is based on a neural network algorithm and the others use various hit-map detector correlations. Two versions of the trigger system used at different stages of the experiment are described. The complete system reduces the event rate by a factor of 1000, with efficiency $\geq$95% of detecting the events in the relative momentum range of interest.