QUBO-based track reconstruction

This article considers the reconstruction of particle tracks in high-energy detectors as a global optimisation problem. Detector hits are combined into local geometric segments and represented by binary variables within a quadratic unconstrained binary optimisation (QUBO) model. The QUBO coefficients encode geometric consistency and mutual compatibility between neighboring segments. The resulting optimisation problem is solved using simulated annealing, and the reconstruction performance is evaluated in terms of purity and efficiency. The approach achieves high reconstruction quality, while the dominant computational cost arises from the construction of the QUBO model itself. The article concludes by outlining the potential for further optimisation and the prospect of more favorable scaling using quantum optimisation methods.

Rekonstrukcija delčnih sledi z binarno optimizacijo (QUBO)

Članek obravnava rekonstrukcijo delčnih sledi v detektorjih visokih energij kot globalni optimizacijski problem. Detektorski zadetki so združeni v lokalne geometrijske segmente, ki so predstavljeni z binarnimi spremenljivkami v okviru kvadratne binarne neomejene optimizacije (QUBO). Koeficienti QUBO modela opisujejo geometrijsko skladnost in medsebojno združljivost sosednjih segmentov. Nastali optimizacijski problem se rešuje s simuliranim ohlajanjem, rezultati pa so ovrednoteni z merama čistosti in učinkovitosti rekonstrukcije. Pristop dosega visoko kakovost rekonstruiranih sledi, vendar glavna računska omejitev ostaja konstrukcija samega QUBO modela. V zaključku je poudarjen potencial nadaljnjih optimizacij ter možnost ugodnejšega skaliranja z uporabo kvantnih optimizacijskih metod.