arXiv subjects

Massimo Cencini

Publications and source records attributed to Massimo Cencini.

At least 37 records · Page 2Linked to original sources

arXiv2022

Taming Lagrangian Chaos with Multi-Objective Reinforcement Learning

We consider the problem of two active particles in 2D complex flows with the multi-objective goals of minimizing both the dispersion rate and the energy consumption of the pair. We approach the problem by means of Multi Objective Reinforcement Learning (MORL), combining scalarization techniques together with a Q-learning algorithm, for Lagrangian drifters that have variable swimming velocity. We show that MORL is able to find a set of trade-off solutions forming an optimal Pareto frontier. As a benchmark, we show that a set of heuristic strategies are dominated by the MORL solutions. We consider the situation in which the agents cannot update their control variables continuously, but only after a discrete (decision) time, $τ$. We show that there is a range of decision times, in between the Lyapunov time and the continuous updating limit, where Reinforcement Learning finds strategies that significantly improve over heuristics. In particular, we discuss how large decision times require enhanced knowledge of the flow, whereas for smaller $τ$ all a priori heuristic strategies become Pareto optimal.

physics.flu-dyn↗

arXiv2022

Non-equilibrium Ensembles for the three-dimensional Navier-Stokes Equations

At the molecular level fluid motions are, by first principles, described by time reversible laws. On the other hand, the coarse grained macroscopic evolution is suitably described by the Navier-Stokes equations, which are inherently irreversible, due to the dissipation term. Here, a reversible version of three-dimensional Navier-Stokes is studied, by introducing a fluctuating viscosity constructed in such a way that enstrophy is conserved, along the lines of the paradigm of microcanonical versus canonical treatment in equilibrium statistical mechanics. Through systematic simulations we attack two important questions: (a) What are the conditions that must be satisfied in order to have a statistical equivalence between the two non-equilibrium ensembles? (b) What is the empirical distribution of the fluctuating viscosity observed by changing the Reynolds number and the number of modes used in the discretization of the evolution equation? The latter point is important also to establish regularity conditions for the reversible equations. We find that the probability to observe negative values of the fluctuating viscosity becomes very quickly extremely small when increasing the effective Reynolds number of the flow in the fully resolved hydro dynamical regime, at difference from what was observed previously.

physics.flu-dyn↗

arXiv2022

Reinforcement learning for pursuit and evasion of microswimmers at low Reynolds number

We consider a model of two competing microswimming agents engaged in a pursue-evasion task within a low-Reynolds-number environment. Agents can only perform simple maneuvers and sense hydrodynamic disturbances, which provide ambiguous (partial) information about the opponent's position and motion. We frame the problem as a zero-sum game: The pursuer has to capture the evader in the shortest time, while the evader aims at deferring capture as long as possible. We show that the agents, trained via adversarial reinforcement learning, are able to overcome partial observability by discovering increasingly complex sequences of moves and countermoves that outperform known heuristic strategies and exploit the hydrodynamic environment.

physics.flu-dyn↗

arXiv2021

Instability of a dusty Kolmogorov flow

Suspended particles can significantly alter the fluid properties and, in particular, can modify the transition from laminar to turbulent flow. We investigate the effect of heavy particle suspensions on the linear stability of the Kolmogorov flow by means of a multiple scale expansion of the Eulerian model originally proposed by Saffman (1962). We find that, while at small Stokes numbers particles always destabilize the flow (as already predicted by Saffman in the limit of very thin particles), at sufficiently large Stokes numbers the effect is non-monotonic in the particle mass fraction and particles can both stabilize and destabilize the flow. Numerical analysis is used to validate the analytical predictions. We find that in a region of the parameter space the multiple-scale expansion overestimates the stability of the flow and that this is a consequence of the breakdown of the scale separation assumptions.

physics.flu-dyn↗

arXiv2021

Optimal collision avoidance in swarms of active Brownian particles

The effectiveness of collective navigation of biological or artificial agents requires to accommodate for contrasting requirements, such as staying in a group while avoiding close encounters and at the same time limiting the energy expenditure for manoeuvring. Here, we address this problem by considering a system of active Brownian particles in a finite two-dimensional domain and ask what is the control that realizes the optimal tradeoff between collision avoidance and control expenditure. We couch this problem in the language of optimal stochastic control theory and by means of a mean-field game approach we derive an analytic mean-field solution, characterized by a second-order phase transition in the alignment order parameter. We find that a mean-field version of a classical model for collective motion based on alignment interactions (Vicsek model) performs remarkably close to the optimal control. Our results substantiate the view that observed group behaviors may be explained as the result of optimizing multiple objectives and offer a theoretical ground for biomimetic algorithms used for artificial agents.

cond-mat.stat-mech↗

arXiv2021

Gyrotactic swimmers in turbulence: shape effects and role of the large-scale flow

We study the dynamics and the statistics of dilute suspensions of gyrotactic swimmers, a model for many aquatic motile microorganisms. By means of extensive numerical simulations of the Navier-Stokes equations at different Reynolds numbers, we investigate preferential sampling and small scale clustering as a function of the swimming (stability and speed) and shape parameters, considering in particular the limits of spherical and rod-like particles. While spherical swimmers preferentially sample local downwelling flow, for elongated swimmers we observe a transition from downwelling to upwelling regions at sufficiently high swimming speed. The spatial distribution of both spherical and elongated swimmers is found to be fractal at small scales in a wide range of swimming parameters. The direct comparison between the different shapes shows that spherical swimmers are more clusterized at small stability and speed numbers, while for large values of the parameters elongated cells concentrate more. The relevance of our results for phytoplankton swimming in the ocean is briefly discussed.

physics.flu-dyn↗

arXiv2020

Effective models and predictability of chaotic multiscale systems via machine learning

We scrutinize the use of machine learning, based on reservoir computing, to build data-driven effective models of multiscale chaotic systems. We show that, for a wide scale separation, machine learning generates effective models akin to those obtained using multiscale asymptotic techniques and, remarkably, remains effective in predictability also when the scale separation is reduced. We also show that predictability can be improved by hybridizing the reservoir with an imperfect model.

nlin.AO↗

arXiv2019

Collective olfactory search in a turbulent environment

Finding the distant source of an odor dispersed by a turbulent flow is a vital task for many organisms, either for foraging or for mating purposes. At the level of individual search, animals like moths have developed effective strategies to solve this very difficult navigation problem based on the noisy detection of odor concentration and wind velocity alone. When many individuals concurrently perform the same olfactory search task, without any centralized control, sharing information about the decisions made by the members of the group can potentially increase the performance. But how much of this information is actually valuable and exploitable for the collective task ? Here we show that, in a model of a swarm of agents inspired by moth behavior, there is an optimal way to blend the private information about odor and wind detections with the publicly available information about other agents' heading direction. At optimality, the time required for the first agent to reach the source is essentially the shortest flight time from the departure point to the target. Conversely, agents who discard public information are several fold slower and groups that do not put enough weight on private information perform even worse. Our results then suggest an efficient multi-agent olfactory search algorithm that could prove useful in robotics, for instance in the identification of sources of harmful volatile compounds.

physics.bio-ph↗

arXiv2019

Gyrotactic phytoplankton in laminar and turbulent flows: a dynamical systems approach

Gyrotactic algae are bottom heavy, motile cells whose swimming direction is determined by a balance between a buoyancy torque directing them upwards and fluid velocity gradients. Gyrotaxis has, in recent years, become a paradigmatic model for phytoplankton motility in flows. The essential attractiveness of this peculiar form of motility is the availability of a mechanistic description which, despite its simplicity, revealed predictive, rich in phenomenology, easily complemented to include the effects of shape, feed-back on the fluid and stochasticity (e.g. in cell orientation). In this review we consider recent theoretical, numerical and experimental results to discuss how, depending on flow properties, gyrotaxis can produce inhomogeneous phytoplankton distributions on a wide range of scales, from millimeters to kilometers, in both laminar and turbulent flows. In particular, we focus on the phenomenon of gyrotactic trapping in nonlinear shear flows and in fractal clustering in turbulent flows. We shall demonstrate the usefulness of ideas and tools borrowed from dynamical systems theory in explaining and interpreting these phenomena.

nlin.CD↗

arXiv2018

The Role of Data in Model Building and Prediction: A Survey Through Examples

The goal of Science is to understand phenomena and systems in order to predict their development and gain control over them. In the scientific process of knowledge elaboration, a crucial role is played by models which, in the language of quantitative sciences, mean abstract mathematical or algorithmical representations. This short review discusses a few key examples from Physics, taken from dynamical systems theory, biophysics, and statistical mechanics, representing three paradigmatic procedures to build models and predictions from available data. In the case of dynamical systems we show how predictions can be obtained in a virtually model-free framework using the methods of analogues, and we briefly discuss other approaches based on machine learning methods. In cases where the complexity of systems is challenging, like in biophysics, we stress the necessity to include part of the empirical knowledge in the models to gain the minimal amount of realism. Finally, we consider many body systems where many (temporal or spatial) scales are at play and show how to derive from data a dimensional reduction in terms of a Langevin dynamics for their slow components.

cond-mat.stat-mech↗

arXiv2018

Reaction fronts in persistent random walks with demographic stochasticity

Standard Reaction-Diffusion (RD) systems are characterized by infinite velocities and no persistence in the movement of individuals, two conditions that are violated when considering living organisms. Here we consider a discrete particle model in which individuals move following a persistent random walk with finite speed and grow with logistic dynamics. We show that when the number of individuals is very large, the individual-based model is well described by the continuous Reactive Cattaneo Equation (RCE), but for smaller values of the carrying capacity important finite-population effects arise. The effects of fluctuations on the propagation speed are investigated both considering the RCE with a cutoff in the reaction term and by means of numerical simulations of the individual-based model. Finally, a more general Lévy walk process for the transport of individuals is examined and an expression for the front speed of the resulting traveling wave is proposed.

physics.bio-ph↗

arXiv2018

Equivalence of nonequilibrium ensembles in turbulence models

Understanding under what conditions it is possible to construct equivalent ensembles is key to advancing our ability to connect microscopic and macroscopic properties of non-equilibrium statistical mechanics. In the case of fluid dynamical systems, one issue is to test whether different models for viscosity lead to the same macroscopic properties of the fluid systems in different regimes. Such models include, besides the standard choice of constant viscosity, cases where the time symmetry of the evolution equations is exactly preserved, as it must be in the corresponding microscopic systems, when available. Here a time-reversible dynamics is obtained by imposing the conservation of global observables. We test the equivalence of reversible and irreversible ensembles for the case of a multiscale shell model of turbulence. We verify that the equivalence is obeyed for the mean values of macroscopic observables, up to an error that vanishes as the system becomes more and more chaotic.

physics.flu-dyn↗

arXiv2018

Time irreversibility in reversible shell models of turbulence

Turbulent flows governed by the Navier-Stokes equations (NSE) generate an out-of-equilibrium time irreversible energy cascade from large to small scales. In the NSE, the energy transfer is due to the nonlinear terms that are formally symmetric under time reversal. As for the dissipative term: first it explicitly breaks time reversibility; second it produces a small-scale sink for the energy transfer that remains effective even in the limit of vanishing viscosity. As a result, it is not clear how to disentangle the time irreversibility originating from the non-equilibrium energy cascade from the explicit time-reversal symmetry breaking due to the viscous term. To this aim, in this paper we investigate the properties of the energy transfer in turbulent Shell models by using a reversible viscous mechanism, avoiding any explicit breaking of the $t \rightarrow -t$ symmetry. We probe time-irreversibility by studying the statistics of Lagrangian power, which is found to be asymmetric under time reversal also in the time-reversible model. This suggests that the turbulent dynamics converges to a strange attractor where time-reversibility is spontaneously broken and whose properties are robust for what concerns purely inertial degrees of freedoms, as verified by the anomalous scaling behavior of the velocity structure functions.

physics.flu-dyn↗

arXiv2017

Stochastic spatial models in ecology: a statistical physics approach

Ecosystems display a complex spatial organization. Ecologists have long tried to characterize them by looking at how different measures of biodiversity change across spatial scales. Ecological neutral theory has provided simple predictions accounting for general empirical patterns in communities of competing species. However, while neutral theory in well-mixed ecosystems is mathematically well understood, spatial models still present several open problems, limiting the quantitative understanding of spatial biodiversity. In this review, we discuss the state of the art in spatial neutral theory. We emphasize the connection between spatial ecological models and the physics of non-equilibrium phase transitions and how concepts developed in statistical physics translate in population dynamics, and vice versa. We focus on non-trivial scaling laws arising at the critical dimension D = 2 of spatial neutral models, and their relevance for biological populations inhabiting two-dimensional environments. We conclude by discussing models incorporating non-neutral effects in the form of spatial and temporal disorder, and analyze how their predictions deviate from those of purely neutral theories.

q-bio.PE↗

arXiv2017

Energetic funnel facilitates facilitated diffusion

Transcription factors are able to associate to their binding sites on DNA faster than the physical limit posed by diffusion. Such high association rates can be achieved by alternating between three-dimensional diffusion and one-dimensional sliding along the DNA chain, a mechanism dubbed Facilitated Diffusion. By studying a collection of transcription factor binding sites of Escherichia coli from the RegulonDB database and of Bacillus subtilis from DBTBS, we reveal a funnel in the binding energy landscape around the target sequences. We show that such a funnel is linked to the presence of gradients of AT in the base composition in the DNA region around the binding sites. An extensive computational study of the stochastic sliding process along the energetic landscapes obtained from the database shows that the funnel can significantly enhance the probability of transcription factors to find their target sequences when sliding in their proximity. We demonstrate that this enhancement leads to a speed-up of the association process.

q-bio.BM↗

arXiv2017

Time irreversibility and multifractality of power along single particle trajectories in turbulence

The irreversible turbulent energy cascade epitomizes strongly non-equilibrium systems. At the level of single fluid particles, time irreversibility is revealed by the asymmetry of the rate of kinetic energy change, the Lagrangian power, whose moments display a power-law dependence on the Reynolds number, as recently shown by Xu et al. [H. Xu et al, Proc. Natl. Acad. Sci. U.S.A. 111, 7558 (2014)]. Here Lagrangian power statistics are rationalized within the multifractal model of turbulence, whose predictions are shown to agree with numerical and empirical data. Multifractal predictions are also tested, for very large Reynolds numbers, in dynamical models of the turbulent cascade, obtaining remarkably good agreement for statistical quantities insensitive to the asymmetry and, remarkably, deviations for those probing the asymmetry. These findings raise fundamental questions concerning time irreversibility in the infinite-Reynolds-number limit of the Navier-Stokes equations.

physics.flu-dyn↗

arXiv2016

Clustering and Turbophoresis in a Shear Flow without Walls

We investigate the spatial distribution of inertial particles suspended in the bulk of a turbulent inhomogeneous flow. By means of direct numerical simulations of particle trajectories transported by the turbulent Kolmogorov flow, we study large and small scale mechanisms inducing inhomogeneities in the distribution of heavy particles. We discuss turbophoresis both for large and weak inertia, providing heuristic arguments for the functional form of the particle density profile. In particular, we argue and numerically confirm that the turbophoretic effect is maximal for particles of intermediate inertia. Our results indicate that small-scale fractal clustering and turbophoresis peak in different ranges in the particles' Stokes number and the separation of the two peaks increases with the flow's Reynolds number.

physics.flu-dyn↗

arXiv2015

On the evolution of particle-puffs in turbulence

We study the evolution of turbulent puffs by means of high-resolution numerical simulations. Puffs are bunches of passive particles released from an initially spherical distribution at regular time intervals of the order of the Kolmogorov time. The instantaneous shapes of particle puffs, in particular their asphericity and prolateness, are characterized by measuring the gyration tensor. Analysis has been performed by following, up to one large scale eddy-turn-over time, more than $10^4$ different puffs, each made of 2000 particle tracers, emitted from different places in a homogeneous and isotropic turbulent fluid with Taylor-scale Reynolds number $Re_λ\sim 300$. We also analyze the probability of hitting a given target placed downstream with respect to the local wind at the time of emission, presenting data for three different cases: (i) without any reconstruction of the shape, i.e. considering the bunch of point tracers, and approximating the particle-puff as a (ii) sphere or as an (iii) ellipsoid. The results show a strong dependence on the fluctuations of the instantaneous wind at the moment of the emission and appear to be robust with respect to the approximations (i)-(iii).

physics.flu-dyn↗