Particle Identification Algorithms Research Articles

Deployment of Water-based Liquid Scintillator in the Accelerator Neutrino Neutron Interaction Experiment

The Accelerator Neutrino Neutron Interaction Experiment (ANNIE) is a 26-ton water Cherenkov neutrino detector installed on the Booster Neutrino Beam (BNB) at Fermilab. Its main physics goals are to perform a measurement of the neutron yield from neutrino-nucleus interactions, as well as a measurement of the charged-current cross section of muon neutrinos. An equally important focus is the research and development of new detector technologies and target media. Specifically, water-based liquid scintillator (WbLS) is of interest as a novel detector medium, as it allows for the simultaneous detection of Cherenkov light and scintillation. This paper presents the deployment of a 366 L WbLS vessel in ANNIE in March 2023 and the subsequent detection of both Cherenkov light and scintillation from the WbLS. This proof-of-concept allows for the future development of reconstruction and particle identification algorithms in ANNIE, as well as dedicated analyses within the WbLS volume, such as the search for neutral-current events and the hadronic scintillation component.

Muon and Pion Identification at BESIII Based on Machine Learning Algorithm

Muon and Pion Identification at BESIII Based on Machine Learning Algorithm

Bioaerosol Identification by Wide Particle Size Range Single Particle Mass Spectrometry

The properties of bioaerosols are complex and diverse, and have a direct impact on the environment, climate, and human health. The effective identification of bioaerosols in the atmosphere is very significant with regard to accurately obtaining the atmospheric chemical characteristics of bioaerosols and making biological early warnings and predictions. To improve the detection of large particle bioaerosol and non-bioaerosol interference in the process of bioaerosol recognition this study detected a variety of bioaerosols and abiotic aerosols based on a single particle aerosol mass spectrometer (SPAMS). Furthermore, the bioaerosol particle identification and classification algorithm based on Zawadowicz the ratio of phosphate to organic nitrogen is optimized to distinguish bioaerosols from abiotic aerosols. The influence of ionized laser energy on classification methods is thoroughly explored here. The results show that 15 kinds of pure fungal aerosols were detected by SPAMS based on a wide size range sampling system, and that fungal aerosols with a particle size of up to 10 μm can be detected. Through the mass spectra peak ratio method of PO3−/PO2− and CNO−/CN−, when discriminating abiotic aerosols such as disruptive biomass combustion particles, automobile exhaust, and dust from pure bacterial aerosols, the discrimination degree is up to 97.7%. The optimized ratio detection method of phosphate to organic nitrogen has strong specificity, which can serve as the discriminant basis for identifying bioaerosols in SPAMS analytical processes.

Deep learning particle identification in LHCb RICH

The use of Ring Imaging Cherenkov detectors (RICH) offers a powerful technique for identifying the particle species in particle physics. These detectors produce 2D images formed by rings of individual photons superimposed on a background of photon rings from other particles.The RICH particle identification (PID) is essential to the LHCb experiment at CERN. While the current PID algorithm has performed well during LHC data-taking periods between 2010 to 2018, its complexity poses a challenge for LHCb computing infrastructure upgrades towards multi-core architectures. The high particle multiplicity environment of future LHC runs strongly motivates shifting towards high-throughput computing for the online event reconstruction.In this contribution, we introduce a convolutional neural network (CNN) approach to particle identification in LHCb RICH. The CNN takes binary input images from the two RICH detectors to classify particle species. The input images are polar-transformed sub-sections of the RICH photon-detection planes. The model is hyperparameter-optimised and trained on classification accuracy with simulated collision data for the upcoming LHC operation starting in 2022. The PID performance of the CNN is comparable to the conventional algorithm, and its simplicity renders it suitable for fast online reconstruction through parallel processing.We show that under conditions of reduced combinatorial background, as expected from the introduction of timing resolution to the RICH detectors in future upgrades, the network achieves a particle identification performance close to 100 %, with simultaneous misclassification of the most prevalent particle species approaching 0 %.

Robust Neural Particle Identification Models

The volume of data processed by the Large Hadron Collider experiments demands sophisticated selection rules typically based on machine learning algorithms. One of the shortcomings of these approaches is their profound sensitivity to the biases in training samples. In the case of particle identification (PID), this might lead to degradation of the efficiency for some decays not present in the training dataset due to differences in input kinematic distributions. In this talk, we propose a method based on the Common Specific Decomposition that takes into account individual decays and possible misshapes in the training data by disentangling common and decay specific components of the input feature set. We show that the proposed approach reduces the rate of efficiency degradation for the PID algorithms for the decays reconstructed in the LHCb detector.

Application of the Quantum Kernel Algorithm on the Particle Identification at the BESIII Experiment

Particle identification is one of most fundamental tools in various particle physics experiments. For the BESIII experiment on the BEPCII, the realization of numerous physical goals heavily relies on advanced particle identification algorithms. In recent years, the emerging of quantum machine learning could potentially arm particle physics experiments with a powerful new toolbox. In this work, targeting at the muon/pion discrimination problem at BESIII, we have developed a quantum SVM classifier for the Noisy Intermediate-Scale Quantum (NISQ) device. By studying and optimizing various encoding circuits, the quantum SVM trained with the BESIII MC data shows comparable discrimination power than other traditional machine learning models. This has demonstrated the potential of using quantum machine learning techniques to form a new approach for particle identification in particle physics experiments. In this talk, we present the application of the quantum SVM for particle identification at BESIII, and demonstrate how to construct and optimize the quantum kernel. Furthermore, we present results obtained from the noisy simulator as well as the small-scale hardware to show the potential advantage of the quantum SVM algorithm.

Volumetric particle tracking velocimetry with improved algorithms using a two-view shadowgraph system

Determining the time-resolved three-dimensional (3D) three-component (3C) velocity is essential for complex turbulent flow measurements. The current study is an extension of a recently developed temporal-spatial three-dimensional particle tracking velocimetry (TS 3D-PTV) technique established for two-view imaging systems. Two improvements have been embedded in TS 3D-PTV algorithm to improve the accuracy at high particle image densities (up to 0.03 ppp). One is using the neighboring particle information to correct the predicted positions and select the temporal particles with higher probability; the other is to iteratively optimize the 2D particle positions during the tracking process using the temporal and image information. The synthetic particle tests indicate that the correctness can be increased by 4.7%–5.8%, to reach a value about 92% with the improved algorithm around 0.03 ppp. The comparative results also indicate that using an advanced particle identification algorithm can improve the correctness over 20%. Two experiments, including a buoyancy jet in water and a transient droplet splashing process, have been conducted with a two-view shadowgraph imaging system. Different tracking algorithms have been conducted to determine the 3D trajectories of seeding particles or secondary droplets comparatively. The new algorithm has shown the best performance with much longer and more reliable trajectories, which indicates the tracking interruption caused by particle overlapping is reduced. The newly developed algorithms have further improved the performance under high seeding density conditions, which makes the two-view shadowgraph 3D PTV system adaptable to more experimental conditions.

Cloud and Precipitation Particle Identification Using Cloud Radar and Lidar Measurements: Retrieval Technique and Validation

AbstractThis paper describes a technique for identifying hydrometeor particle types using airborne HIAPER Cloud Radar (HCR) and High Spectral Resolution Lidar (HSRL) observations. HCR operates at a frequency of 94 GHz (3 mm wavelength), while HSRL is an eye‐safe lidar system operating at a wavelength of 532 nm. Both instruments are deployed on the NSF‐NCAR HIAPER aircraft. HCR is designed to fly in an underwing pod and HSRL is situated in the cabin. The HCR and HSRL data used in this study were collected during the Southern Ocean Clouds, Radiation, Aerosol Transport Experimental Study (SOCRATES). Comprehensive observations of the vertical distributions of liquid and mixed‐phase clouds were obtained using in situ probes on the aircraft and remote sensing instruments. Hydrometeor particle types were retrieved from HCR, HSRL, and temperature fields with a newly developed fuzzy logic particle identification (PID) algorithm. The PID results were validated with in situ measurements collected onboard the HIAPER aircraft by a 2D‐Stereo (2D‐S) cloud probe. Particle phases derived from the PID results compare well with those obtained from the 2D‐S observations and agree in over 70% of cases. Size distributions are also consistent between the two methods of observation. Knowledge of the particle type distribution gained from the PID results can be used to constrain microphysical parameterizations and improve the representation of cloud radiative effects in weather and climate models.

Calorimetric classification of track-like signatures in liquid argon TPCs using MicroBooNE data

The MicroBooNE liquid argon time projection chamber located at Fermilab is a neutrino experiment dedicated to the study of short-baseline oscillations, the measurements of neutrino cross sections in liquid argon, and to the research and development of this novel detector technology. Accurate and precise measurements of calorimetry are essential to the event reconstruction and are achieved by leveraging the TPC to measure deposited energy per unit length along the particle trajectory, with mm resolution. We describe the non-uniform calorimetric reconstruction performance in the detector, showing dependence on the angle of the particle trajectory. Such non-uniform reconstruction directly affects the performance of the particle identification algorithms which infer particle type from calorimetric measurements. This work presents a new particle identification method which accounts for and effectively addresses such non-uniformity. The newly developed method shows improved performance compared to previous algorithms, illustrated by a 93.7% proton selection efficiency and a 10% muon mis-identification rate, with a fairly loose selection of tracks performed on beam data. The performance is further demonstrated by identifying exclusive final states in νμCC interactions. While developed using MicroBooNE data and simulation, this method is easily applicable to future LArTPC experiments, such as SBND, ICARUS, and DUNE.

A spatial-temporal algorithm for three-dimensional particle tracking velocimetry using two-view systems

In the present study, a novel temporal three-dimensional particle tracking velocimetry (3D PTV) algorithm for flow measurements with only two views is developed and validated with synthetic particles. The spatial information in image and object spaces, as well as the temporal predictions, are strongly coupled to improve the particle tracking accuracy. A well-designed cost function, simultaneously penalizing the reconstruction and tracking processes, is minimized to determine the most reliable traces. The algorithm shows a correctness over 98% up to 0.0273 ppp (particles per pixel) when using ideal synthetic particle positions, which is superior to several artificial intelligence methods. Moreover, an improved particle identification algorithm is proposed to handle overlapped particles and reduce the error introduced into the 3D PTV scheme. The algorithm adopts a particle position shifting process to tackle the correct particle numbers iteratively, which shows better performance than some other methods. A comparative study indicates that particle identification accuracy has a significant effect on the subsequent 3D reconstruction and tracking processes. The 3D PTV and particle identification algorithms show good consistencies under two types of flow conditions: a homogeneous isotropic turbulent flow and a vortex ring flow. Comparing with multiple-view setups, two-view systems are more compact and cost-effective, especially in conditions requiring high-speed cameras. With the newly established algorithms, a two-view system is now able to handle higher particle-seeding densities and thus can resolve higher spatial resolutions, which is significant for applications in turbulent flow and particle motion measurements.

Particle identification with the Cherenkov detector in the VES experiment

The Cherenkov detector is a key element of the particle identification system in the VES fixed target experiment at IHEP (Protvino). It consists of a 1.6 m R-22 gas radiator and a system of 28 mirrors with photomultiplier tubes (PMTs) as photon detectors. This paper is giving an overview of a description of the detector design, its model used in a simulation, particle identification algorithm, and discussing procedures of obtaining calibration constants. The particle identification performance is demonstrated with the Monte-Carlo and experimental data analyses.

FPGA accelerated computing for particle identification in high-energy physics experiments

The key challenges in the recent nuclear and High-Energy Physics (HEP) experiments are high reaction rate and data rate. The data rate, the product of the event size and the reaction rate, ranges from 107 up to more than 1011 Bytes/s. This huge amount of data cannot be entirely recorded and processed during offline analysis. Interestingly, physicists have interest only in a very small proportion of all the event data that occur only once within one million interactions. Hence it provides the possibility to utilize an efficient online Data Acquisition (DAQ) and Trigger system to reject uninteresting events while identify and retain interesting ones in real-time. This work assesses the machine learning assisted particle identification algorithms on Field-Programmable-Gate-Array (FPGA) based computing platform for HEP experiments. To achieve the highest classification accuracy in particle identification the combination of particle recognition algorithms and random decision forests algorithms are laid on an accelerated computing platform utilizing the FPGA based computing platform to classify the particles of interest. The simulation results of this implementation achieve many benefits through dynamic and intelligent design management, such as efficient resource utilization and high performance-cost ratio.

Selection and processing of calibration samples to measure the particle identification performance of the LHCb experiment in Run 2

Since 2015, with the restart of the LHC for its second run of data taking, the LHCb experiment has been empowered with a dedicated computing model to select and analyse calibration samples to measure the performance of the particle identification (PID) detectors and algorithms. The novel technique was developed within the framework of the innovative trigger model of the LHCb experiment, which relies on online event reconstruction for most of the datasets, reserving offline reconstruction to special physics cases. The strategy to select and process the calibration samples, which includes a dedicated data-processing scheme combining online and offline reconstruction, is discussed. The use of the calibration samples to measure the detector PID performance, and the efficiency of PID requirements across a large range of decay channels, is described. Applications of the calibration samples in data-quality monitoring and validation procedures are also detailed.

Calibration of the Belle II aerogel ring imaging detector

Abstract For efficient separation of hadrons in the forward end-cap of the Belle II spectrometer , an aerogel proximity focusing Ring Imaging Detector is installed in the high magnetic field between the central drift chamber and electromagnetic calorimeter . Cherenkov photons , emitted in the double layer aerogel radiator are expanded through the empty space and detected on the photon detector consisting of Hybrid Avalanche Photo diodes . The readout electronics working in a threshold mode records hit patterns registered during beam collisions. A particle identification algorithm based on the two dimensional extended maximum likelihood technique is used to assign probabilities for different particle hypotheses of tracks traversing the aerogel RICH detector. For efficient discrimination, the Aerogel RICH detector has to be calibrated. We present the key calibration steps used to optimize the detector performance.

Machine Learning based Global Particle Identification Algorithms at the LHCb Experiment

One of the most important aspects of data processing at flavor physics experiments is the particle identification (PID) algorithm. In LHCb, several different sub-detector systems provide PID information: the Ring Imaging Cherenkov detectors, the hadronic and electromagnetic calorimeters, and the muon chambers. The charged PID based on the sub-detectors response is considered as a machine learning problem solved in different modes: one-vs-rest, one-vs-one and multi-classification, which affect the models training and prediction. To improve charged particle identification for pions, kaons, protons, muons and electrons, neural network and gradient boosting models have been tested. This paper presents these models and their performance evaluated on Run 2 data and simulation samples. A discussion of the performances is also presented.

Trackdem: Automated particle tracking to obtain population counts and size distributions from videos in r

Abstract The possibilities for image analysis in scientific research are substantial: the costs of digital cameras and data storage are sharply decreasing, and automated image analyses greatly increase the scale, reproducibility and robustness of biological studies. However, automated image analysis in ecological and evolutionary studies is still in its infancy. There is a clear need for easy to use and accessible tools. Here, we provide a general purpose method to obtain estimates of population densities, individual body sizes and behavioural metrics from video material of moving organisms. The methods are supplied as a new r‐package trackdem, which provides a flexible, easy to install and use, generally applicable and accurate way to analyse ecological video data. The package can detect and track moving particles, count individuals and estimate individual sizes using background detection, particle identification and particle tracking algorithms. Machine learning is implemented to reduce the influence of noise in lower quality videos or to distinguish a single species in multi‐species systems. We show that trackdem provides accurate population counts and body size distributions. Using a series of simulations, we show that our estimates are robust against high levels of noise in videos. When applied to live populations of Daphnia magna, our methods obtained accurate and unbiased estimates of population counts, individual sizes and size distributions, as verified by manual counting and measuring. The package trackdem is also directly usable for movement analysis, for instance in behavioural ecology, as illustrated by the tracking of insects, fish, cars and humans. Within 24 hr, we obtained 192 accurate population counts and body sizes of 22,154 individuals. Such results underscore that automated analysis can improve robustness and reproducibility, and greatly increase the scope of studies in ecology and evolution.

Particle identification algorithms for the PANDA Endcap Disc DIRC

The Endcap Disc DIRC has been developed to provide an excellent particle identification for the future PANDA experiment by separating pions and kaons up to a momentum of 4 GeV/c with a separation power of 3 standard deviations in the polar angle region from 5̂ to 22̂. This goal will be achieved using dedicated particle identification algorithms based on likelihood methods and will be applied in an offline analysis and online event filtering. This paper evaluates the resulting PID performance using Monte-Carlo simulations to study basic single track PID as well as the analysis of complex physics channels. The online reconstruction algorithm has been tested with a Virtex4 FGPA card and optimized regarding the resulting constraints.

Particle identification technique using grid ionization chamber at China Spallation Neutron Source

Back-n white neutron source at China Spallation Neutron Source facility requires a detector with charged-particle identification ability for measuring nuclear cross sections. Therefore, a grid ionization chamber (GIC) with an adjustable distance between the cathode and grid was designed and built. Moreover, a corresponding particle identification algorithm is proposed in order to use the signals of the anode (QA) and cathode (QC), or the signal (QA) of the anode and the time difference (Δ t) between the cathode and anode signals, based on the interactive behaviors of charged particles and the GIC's working principle. First, the GIC performance was measured using a 241Am α source for preliminary performance testing, which provided an energy resolution of 5.25%. Its good particle identification performance was demonstrated by identifying 3H and 4He from neutron reactions with 6Li using a 2D QC-QA plot and Δ t-QA plot. Furthermore, 3H and 4He can be completely distinguished from the updated 2D QC-QA plot and Δ t-QA plot, in which 3H with a small incident angle is selected. Finally, it was concluded that our GIC with the proposed particle identification algorithm could identify charged particles effectively for the future measurement of nuclear cross sections.

Addendum: Measurement of charged particle yields from PMMA irradiated by a 220 MeV/u 12C beam

In this paper we report the re-analysis of the data published in Piersanti et al (2014 Phys. Med. Biol. 59 1857) documenting the charged secondary particles production induced by the interaction of a 220 MeV/u 12C ion beam impinging on a polymethyl methacrylate (PMMA) target, measured in 2012 at the GSI facility in Darmstadt (Germany). This re-analysis takes into account the inhomogeneous light response of the LYSO crystal in the experimental setup measured in a subsequent experiment (2014) performed in the Heidelberg Ion-Beam Therapy Center. A better description of the detector and re-calculation of the geometrical efficiencies have been implemented as well, based on an improved approach that accounts also for the energy dependence of the emission spectrum.The new analysis has little effect on the total secondary charged flux, but has an impact on the production yield and emission velocity distributions of the different particle species (protons, deuterons and tritons) at different angles with respect to the beam direction ( and ). All these observables indeed depend on the particle identification algorithms and hence on the LYSO detector energy response.The results of the data re-analysis presented here are intended to supersede and replace the results published in Piersanti et al (2014 Phys. Med. Biol. 59 1857).

The CHIPS R&D Programme: Reconstruction

The CHIPS R&D programme aims to demonstrate significant cost reduction in the building of large neutrino detectors, whilst contributing to the global knowledge on the CP-violating phase in neutrino oscillations. A series of event reconstruction algorithms have been developed considering the charge and time information from all of the photomultiplier tubes to reconstruct charged particle tracks. A particle identification algorithm was developed using two artificial neural networks to select charged-current νe interactions from νµ backgrounds. It has been demonstrated that using a 6% photocathode coverage of 3” photomultiplier tubes can provide the required performance from the particle identification to achieve the goals of the experiment.