An M/PH/1 Queue with Catastrophes

Liu, Youxin; Liu, Liwei

doi:10.21203/rs.3.rs-2634820/v1

Cited by 3 publications

(3 citation statements)

References 25 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…A modern CNN architecture can contain hundreds of layers and tens of millions trainable parameters. Neural networks and deep learning has been utilized before in the context of heavy-ion collisions for various different applications, such as impact parameter estimation, identifying quenched jets or determination of the QCD matter phase transition [17][18][19][20][21]. In Ref.…”

Section: Introductionmentioning

confidence: 99%

Deep learning for flow observables in ultrarelativistic heavy-ion collisions

Hirvonen¹,

Eskola²,

Niemi³

2023

Preprint

View full text Add to dashboard Cite

We train a deep convolutional neural network to predict hydrodynamic results for flow coefficients, average transverse momenta and charged particle multiplicities in ultrarelativistic heavy-ion collisions from the initial energy density profiles. We show that the neural network can be trained accurately enough so that it can reliably predict the hydrodynamic results for the flow coefficients and, remarkably, also their correlations like normalized symmetric cumulants, mixed harmonic cumulants and flow-transverse-momentum correlations. At the same time the required computational time decreases by several orders of magnitude. To demonstrate the advantage of the significantly reduced computation time, we generate 10M initial energy density profiles from which we predict the flow observables using the neural network, which is trained using 5k, and validated using 90k events per collision energy. We then show that increasing the number of collision events from 90k to 10M can have significant effects on certain statistics-expensive flow correlations, which should be taken into account when using these correlators as constraints in the determination of the QCD matter properties.

show abstract

Section: Introductionmentioning

confidence: 99%

Deep learning for flow observables in ultrarelativistic heavy-ion collisions

Hirvonen¹,

Eskola²,

Niemi³

2023

Preprint

View full text Add to dashboard Cite

show abstract

“…[16] Additionally, modulating geometric parameters could be a more straightforward approach for experiment. [39] In short, these modulations make the Hamiltonian non-reciprocal and cause NHSE, which manifests the directed aggregation of temperature. Figure 1(a) illustrates the modulation of the nonreciprocal coupling coefficient within a single cell.…”

mentioning

confidence: 99%

Two-Dimensional Thermal Regulation Based on Non-Hermitian Skin Effect

Huang 黄,

Liu 刘,

Cao 曹

et al. 2023

Chinese Phys. Lett.

View full text Add to dashboard Cite

The non-Hermitian skin effect has been applied in multiple fields. However, there are relatively few models in the field of thermal diffusion that utilize the non-Hermitian skin effect for achieving thermal regulation. Here, we propose two non-Hermitian Su-Schrieffer-Heeger (SSH) models for thermal regulation: one capable of achieving edge states, and the other capable of achieving corner states within the thermal field. By analyzing the energy band structures and the generalized Brillouin zone, we predict the appearance of the non-Hermitian skin effect in these two models. Furthermore, we analyze the time-dependent evolution results and assess the robustness of the models. The results indicate that the localized thermal effects of the models align with our predictions. In summary, this work presents two models based on the non-Hermitian skin effect for regulating the thermal field, injecting vitality into the design of non-Hermitian thermal diffusion systems.

show abstract

“…[17][18][19] Recently, diffusion system, [20][21][22][23][24][25][26] with its dissipation nature, has recently been embraced as a novel platform to reveal topological physics, which has garnered significant interest within the realms of artificial metamaterials and condensed matter physics. [27][28][29][30][31][32][33] In recent years, topological physics has been employed in plasma physics to elucidate certain unique phenomena. [34][35][36][37] However, most of these investigations see the plasma as a fluid wherein electromagnetic forces are pivotal.…”

mentioning

confidence: 99%

Topological Plasma Transport from a Diffusion View

Liu 刘,

Huang 黄

2023

Chinese Phys. Lett.

View full text Add to dashboard Cite

Recent studies have identified plasma as a topological material. Yet, these researches often depict plasma as a fluid governed by electromagnetic fields, i.e., a classical wave system. Indeed, plasma transport can be characterized by a unique diffusion process distinguished by its collective behaviors. In this work, we adopt a simplified diffusion-migration method to elucidate the topological plasma transport. Drawing parallels to the thermal conduction-convection system, we introduce a double ring model to investigate the plasma density behaviors in the anti-parity-time reversal (APT) unbroken and broken phases. Subsequently, by augmenting the number of rings, we have established a coupled ring chain structure. This structure serves as a medium for realizing the APT symmetric one-dimensional (1D) reciprocal model, representing the simplest tight-binding model with a trivial topology. To develop a model featuring topological properties, we should modify the APT symmetric 1D reciprocal model from the following two aspects: hopping amplitude and onsite potential. From the hopping amplitude, we incorporate the non-reciprocity to facilitate the non-Hermitian skin effect, an intrinsic non-Hermitian topology. Meanwhile, from the onsite potential, the quasiperiodic modulation has been adopted onto the APT symmetric 1D reciprocal model. This APT symmetric 1D Aubry-André-Harper model is of topological nature. Additionally, we suggest the potential applications for these diffusive plasma topological states. This study establishes a diffusion-based approach to realizing topological states in plasma, potentially inspiring further advancements in plasma physics.

show abstract

An M/PH/1 Queue with Catastrophes

Cited by 3 publications

References 25 publications

Deep learning for flow observables in ultrarelativistic heavy-ion collisions

Deep learning for flow observables in ultrarelativistic heavy-ion collisions

Two-Dimensional Thermal Regulation Based on Non-Hermitian Skin Effect

Topological Plasma Transport from a Diffusion View

Contact Info

Product

Resources

About