Search

In this work, we consider the dynamics of repairable systems characterized by three distinct states: one signifying normal operational states, another representing degraded conditions and a third denoting failed conditions. These systems are characterized by their ability to be repaired when failures and/or degradation occur. Typically described by transport equations, these systems exhibit a coupled nature, interlinked through integro-differential equations and integral boundary conditions that dictate the transitions among all the states. In this paper, we address two less-explored facets: 1) the well-posedness and the asymptotic behavior of such systems with maximum repair time being finite; and 2) the bilinear controllability of the system via repair actions. In particular, we focus on the case where only one degraded and one failed states exist. We first discuss part 1) for given time-independent repair rates and then design the space-time dependent repair strategies that can manipulate system dynamics to achieve the desired level over a finite horizon. Our objective is to enhance the system availability -- the probability of being operational when needed over a fixed period of time. We present rigorous analysis and develop control strategies that leverage the bilinear structure of the system model.

Reparable systems are systems that are characterized by their ability to undergo maintenance actions when failures occur. These systems are often described by transport equations, all coupled through an integro-differential equation. In this paper, we address the understudied aspect of the controllability of reparable systems. In particular, we focus on a two-state reparable system and our goal is to design a control strategy that enhances the system availability -- the probability of being operational when needed. We establish bilinear controllability, demonstrating that appropriate control actions can manipulate system dynamics to achieve desired availability levels. We provide theoretical foundations and develop control strategies that leverage the bilinear structure of the equations.

Coherent optics has profoundly impacted diverse applications ranging from communications, LiDAR to quantum computations. However, building coherent systems in integrated photonics previously came at great expense in hardware integration and energy efficiency: the lack of a power-efficient way to generate highly coherent light necessitates bulky lasers and amplifiers, while frequency and phase recovery schemes require huge digital signal processing resources. In this work, we demonstrate a high-coherence parallelization strategy that facilitates advanced integrated coherent systems at a minimum price. Using a self-injection locked microcomb to injection lock a distributed feedback laser array, we boost the microcomb power by a record high gain of up to 60 dB on chip with no degradation in coherence. This strategy enables tens of highly coherent channels with an intrinsic linewidth down to the 10 Hz level and power of more than 20 dBm. The overall electrical to optical wall-plug efficiency reaches 19%, comparable with that of the state-of-the-art semiconductor lasers. Driven by this parallel source, we demonstrate a silicon photonic communication link with an unprecedented data rate beyond 60 Tbit/s. Importantly, the high coherence we achieve reduces the coherent-related DSP consumption by 99.999% compared with the traditional III-V laser pump scheme. This work paves a way to realizing scalable, high-performance coherent integrated photonic systems, potentially benefiting numerous applications.

This work develops an efficient and accurate optimization algorithm to study the optimal mixing problem driven by boundary control of unsteady Stokes flows, based on the theoretical foundation laid by Hu and Wu in a series of work. The scalar being mixed is purely advected by the flow and the control is a force exerted tangentially on the domain boundary through the Navier slip conditions. The control design has potential applications in many industrial processes such as rotating wall driven mixing, micromixers with acoustic waves, and artificial cilia mixing. The numerical algorithms have high complexity, high accuracy demand, and high computing expense, due to the multiscale nature of the mixing problem and the optimization requirements. A crucial problem is the computation of the G$\^a$teaux derivative of the cost functional. To this end, a hybrid approach based on variational formula and finite difference is built with high accuracy and efficiency to treat various types of control input functions. We have experimented with various optimization schemes including the steepest descent algorithm, the conjugate gradient method and two line search options (backtracking and exact line search). We are able to identify and implement the best combinations. The numerical simulations show that the mixing efficacy is limited when only one single type of control is applied, but can be enhanced when more diverse control types and more time segmentation are utilized. The mix-norm in the optimal mixings decays exponentially. The numerical study in this work demonstrates that boundary control alone could be an effective strategy for mixing in incompressible flows., Comment: 36 pages

A new approach, iteration projection method, is proposed to solve the saddle point problem obtained after the full discretization of the unsteady Navier-Stokes equations. The proposed method iterates projections in each time step with a proper convection form. We prove that the projection iterations converge with a certain parameter regime. Optimal iteration convergence can be achieved with the modulation of parameter values. This new method has several significant improvements over the Uzawa method and the projection method both theoretically and practically. First, when the iterative projections are fully convergent in each time step, the numerical velocity is weakly divergence free (pointwise divergence free in divergence free finite element methods), and the stability and error estimate are rigorously proven. With proper parameters, this method converges much faster than the Uzawa method. Second, numerical simulations show that with rather relaxed stopping criteria which require only a few iterations each time step, the numerical solution preserves stability and accuracy for high Reynolds numbers, where the convectional projection method would fail. Furthermore, this method retains the efficiency of the traditional projection method by decoupling the velocity and pressure fields, which splits the saddle point system into small elliptic problems. Three dimensional simulations with Taylor-Hood P2/P1 finite elements are presented to demonstrate the performance and efficiency of this method. More importantly, this method is a generic approach and thus there are many potential improvements and extensions of the iterative projection method, including utilization of various convection forms, association with stabilization techniques for high Reynolds numbers, applications to other saddle point problems.

This work develops scientific computing techniques to further the exploration of using boundary control alone to optimize mixing in Stokes flows. The theoretical foundation including mathematical model and the optimality conditions for solving the optimal control has been established by Hu and Wu in a series of work. The scalar being mixed is purely advected by the flow and the control is exerted tangentially on the domain boundary through the Navier slip conditions. The control design is motivated by the physical observations that the moving or rotating walls accelerate mixing. A gradient descent-based optimization algorithm is designed. A critical problem is the computation of the Gateaux derivative or the gradient of the cost functional. Two methods are proposed: one is based on the Variational Formula (VF) and one utilizes Algorithmic Differentiation (AD). The convergence of the algorithm is studied and various designs of boundary control using cosine and sine functions with time segmentation are computed. The algorithm has a first order convergence rate and the VF method is more efficient by taking only one third of the time as the AD method when the dimension of control basis is large. The numerical implementations show that the boundary control produces similar mixing results as internal mixings in the existing literature. The mixing effect becomes better when more diverse basis control functions and more time segmentation are utilized. It is shown that the mixing decay rate in time follows power rules, approximately. The numerical study in this work suggests that boundary control alone could be an effective strategy for mixing in incompressible fluid flows., Comment: This article was significantly upgraded, and the new version is available in arXiv:2306.10690

This paper is concerned with a bilinear control problem for enhancing convection-cooling via an incompressible velocity field. Both optimal open-loop control and closed-loop feedback control designs are addressed. First and second order optimality conditions for characterizing the optimal solution are discussed. In particular, the method of instantaneous control is applied to establish the feedback laws. Moreover, the construction of feedback laws is also investigated by directly utilizing the optimality system with appropriate numerical discretization schemes. Computationally, it is much easier to implement the closed-loop feedback control than the optimal open-loop control, as the latter requires to solve the state equations forward in time, coupled with the adjoint equations backward in time together with a nonlinear optimality condition. Rigorous analysis and numerical experiments are presented to demonstrate our ideas and validate the efficacy of the control designs., Comment: 27 pages, 7 figures, 3 tables

We study temperature and velocity output tracking problem for a two-dimensional room model with the fluid dynamics governed by the linearized translated Boussinesq equations. Additionally, the room model includes finite-dimensional models for actuation and sensing dynamics, thus the complete model dynamics are governed by an ODE-PDE-ODE system. As the main result, we design a low-dimensional internal model based controller for robust output racking of the room model. Efficiency of the controller is demonstrated through a numerical example of velocity and temperature tracking., Comment: 26 pages, 9 figures, submitted

This paper is concerned with the problem of enhancing convection-cooling via active control of the incompressible velocity field, described by a stationary diffusion-convection model. This essentially leads to a bilinear optimal control problem. A rigorous proof of the existence of an optimal control is presented and the first order optimality conditions are derived for solving the control using a variational inequality. Moreover, the second order sufficient conditions are established to characterize the local minimizer. Finally, numerical experiments are conducted utilizing finite elements methods together with nonlinear iterative schemes, to demonstrate and validate the effectiveness of our control design., Comment: 24 pages, 17 figures

We consider robust output regulation of a partial differential equation model describing temperature evolution in a room. More precisely, we examine a two-dimensional room model with the velocity field and temperature evolution governed by the incompressible steady state Navier-Stokes and advection-diffusion equations, respectively, which coupled together form a simplification of the Boussinesq equations. We assume that the control and observation operators of our system are distributed, whereas the disturbance acts on a part of the boundary of the system. We solve the robust output regulation problem using a finite-dimensional low-order controller, which is constructed using model reduction on a finite element approximation of the model. Through numerical simulations, we compare performance of the reduced-order controller to that of the controller without model reduction as well as to performance of a low-gain robust controller., Comment: 12 pages, 5 figures. Accepted for publication in the Proceedings of the 24th International Symposium on Mathematical Theory of Networks and Systems, 23-27 August, 2021

Finding or monitoring subgraph instances that are isomorphic to a given pattern graph in a data graph is a fundamental query operation in many graph analytic applications, such as network motif mining and fraud detection. The state-of-the-art distributed methods are inefficient in communication. They have to shuffle partial matching results during the distributed multiway join. The partial matching results may be much larger than the data graph itself. To overcome the drawback, we develop the Batch-BENU framework (B-BENU) for distributed subgraph enumeration. B-BENU executes a group of local search tasks in parallel. Each task enumerates subgraphs around a vertex in the data graph, guided by a backtracking-based execution plan. B-BENU does not shuffle any partial matching result. Instead, it stores the data graph in a distributed database. Each task queries adjacency sets of the data graph on demand. To support dynamic data graphs, we propose the concept of incremental pattern graphs and turn continuous subgraph enumeration into enumerating incremental pattern graphs at each time step. We develop the Streaming-BENU framework (S-BENU) to enumerate their matches efficiently. We implement B-BENU and S-BENU with the local database cache and the task splitting techniques. The extensive experiments show that B-BENU and S-BENU can scale to big data graphs and complex pattern graphs. They outperform the state-of-the-art methods by up to one and two orders of magnitude, respectively., Comment: Modify some terms;Fix typos; Unify line styles in Fig. 13 and Fig. 14

Multiple-choice reading comprehension (MCRC) is the task of selecting the correct answer from multiple options given a question and an article. Existing MCRC models typically either read each option independently or compute a fixed-length representation for each option before comparing them. However, humans typically compare the options at multiple-granularity level before reading the article in detail to make reasoning more efficient. Mimicking humans, we propose an option comparison network (OCN) for MCRC which compares options at word-level to better identify their correlations to help reasoning. Specially, each option is encoded into a vector sequence using a skimmer to retain fine-grained information as much as possible. An attention mechanism is leveraged to compare these sequences vector-by-vector to identify more subtle correlations between options, which is potentially valuable for reasoning. Experimental results on the human English exam MCRC dataset RACE show that our model outperforms existing methods significantly. Moreover, it is also the first model that surpasses Amazon Mechanical Turker performance on the whole dataset., Comment: 6 pages, 2 tables

We propose a hybridizable discontinuous Galerkin (HDG) method to approximate the solution of a tangential Dirichlet boundary control problem for the Stokes equations with an $ L^2 $ penalty on the boundary control. The contribution of this paper is twofold. First, we obtain well-posedness and regularity results for the tangential Dirichlet control problem on a convex polygonal domain. The analysis contains new features not found in similar Dirichlet control problems for the Poisson equation; an interesting result is that the optimal control has higher local regularity on the individual edges of the domain compared to the global regularity on the entire boundary. Second, under certain assumptions on the domain and the target state, we prove a priori error estimates for the control for the HDG method. In the 2D case, our theoretical convergence rate for the control is superlinear and optimal with respect to the global regularity on the entire boundary. We present numerical experiments to demonstrate the performance of the HDG method.

In the first part of this work, we analyzed a Dirichlet boundary control problem for an elliptic convection diffusion PDE and proposed a new hybridizable discontinuous Galerkin (HDG) method to approximate the solution. For the case of a 2D polygonal domain, we also proved an optimal superlinear convergence rate for the control under certain assumptions on the domain and on the target state. In this work, we revisit the convergence analysis without these assumptions; in this case, the solution can have low regularity and we use a different analysis approach. We again prove an optimal convergence rate for the control, and present numerical results to illustrate the convergence theory.

We propose a new hybridizable discontinuous Galerkin (HDG) method to approximate the solution of a Dirichlet boundary control problem governed by an elliptic convection diffusion PDE. Even without a convection term, Dirichlet boundary control problems are well-known to be very challenging theoretically and numerically. Although there are many works in the literature on Dirichlet boundary control problems for the Poisson equation, the authors are not aware of any existing theoretical or numerical analysis works for convection diffusion Dirichlet control problems. We make two contributions. First, we obtain well-posedness and regularity results for the Dirichlet control problem. Second, under certain assumptions on the domain and the target state, we obtain optimal a priori error estimates in 2D for the control for the new HDG method. As far as the authors are aware, there are no existing comparable results in the literature. We present numerical experiments to demonstrate the performance of the HDG method., Comment: arXiv admin note: text overlap with arXiv:1712.02931, arXiv:1801.00082

We propose a hybridizable discontinuous Galerkin (HDG) method to approximate the solution of a distributed optimal control problem governed by an elliptic convection diffusion PDE. We derive optimal a priori error estimates for the state, adjoint state, their fluxes, and the optimal control. We present 2D and 3D numerical experiments to illustrate our theoretical results., Comment: arXiv admin note: substantial text overlap with arXiv:1712.10106, arXiv:1712.01403, arXiv:1712.02931

We begin an investigation of hybridizable discontinuous Galerkin (HDG) methods for approximating the solution of Dirichlet boundary control problems governed by elliptic PDEs. These problems can involve atypical variational formulations, and often have solutions with low regularity on polyhedral domains. These issues can provide challenges for numerical methods and the associated numerical analysis. We propose an HDG method for a Dirichlet boundary control problem for the Poisson equation, and obtain optimal a priori error estimates for the control. Specifically, under certain assumptions, for a 2D convex polygonal domain we show the control converges at a superlinear rate. We present 2D and 3D numerical experiments to demonstrate our theoretical results.

We consider a distributed optimal control problem governed by an elliptic convection diffusion PDE, and propose a hybridizable discontinuous Galerkin (HDG) method to approximate the solution. We use polynomials of degree $k+1$ and $k \ge 0$ to approximate the state, dual state, and their fluxes, respectively. Moreover, we use polynomials of degree $k$ to approximate the numerical traces of the state and dual state on the faces, which are the only globally coupled unknowns. We prove optimal a priori error estimates for all variables when $ k > 0 $. Furthermore, from the point of view of the number of degrees of freedom of the globally coupled unknowns, this method achieves superconvergence for the state, dual state, and control when $k\geq 1$. We illustrate our convergence results with numerical experiments.

We consider an approximating control design for optimal mixing of a non-dissipative scalar field $\theta$ in unsteady Stokes flows. The objective of our approach is to achieve optimal mixing at a given final time $T>0$, via the active control of the flow velocity $v$ through boundary inputs. Due to the zero diffusivity of the scalar field $\theta$, establishing the well-posedness of its G\^{a}teaux derivative requires $\sup_{t\in[0,T]}\|\nabla \theta\|_{L^2}<\infty$, which in turn demands the flow velocity field to satisfy the condition $ \int^{T}_{0}\|\nabla v\|_{L^{\infty}(\Omega)}\, dt<\infty$. This condition results in the need to penalize the time derivative of the boundary control in the cost functional. As a result, the optimality system becomes difficult to solve \cite{hu2017boundary}. Our current approximating approach will provide a more transparent optimality system, with the set of admissible controls being $L^{2}$ in both time and space. This is achieved by first introducing a small diffusivity to the scalar equation and then establishing a rigorous analysis of convergence of the approximating control problem to the original one as the diffusivity approaches to zero. Uniqueness of the optimal solution is obtained for the two dimensional case.

Recent researches have shown that machine learning based malware detection algorithms are very vulnerable under the attacks of adversarial examples. These works mainly focused on the detection algorithms which use features with fixed dimension, while some researchers have begun to use recurrent neural networks (RNN) to detect malware based on sequential API features. This paper proposes a novel algorithm to generate sequential adversarial examples, which are used to attack a RNN based malware detection system. It is usually hard for malicious attackers to know the exact structures and weights of the victim RNN. A substitute RNN is trained to approximate the victim RNN. Then we propose a generative RNN to output sequential adversarial examples from the original sequential malware inputs. Experimental results showed that RNN based malware detection algorithms fail to detect most of the generated malicious adversarial examples, which means the proposed model is able to effectively bypass the detection algorithms.

We discuss the problem of optimal impulse control representing the preventive maintenance of a simple reparable system. The system model is governed by coupled transport and integro-differential equations in a nonreflexive Banach space. The objective of this paper is to construct nonnegative impulse control inputs at given system running times that minimize the probability of the system in failure mode. To guarantee the nonnegativity of the controlled system, we consider the control inputs to depend on the system state. This essentially leads to a bilinear control problem. We first present a rigorous proof of existence of an optimal controller and then apply the variational inequality to derive the first-order necessary conditions of optimality.

Machine learning has been used to detect new malware in recent years, while malware authors have strong motivation to attack such algorithms. Malware authors usually have no access to the detailed structures and parameters of the machine learning models used by malware detection systems, and therefore they can only perform black-box attacks. This paper proposes a generative adversarial network (GAN) based algorithm named MalGAN to generate adversarial malware examples, which are able to bypass black-box machine learning based detection models. MalGAN uses a substitute detector to fit the black-box malware detection system. A generative network is trained to minimize the generated adversarial examples' malicious probabilities predicted by the substitute detector. The superiority of MalGAN over traditional gradient based adversarial example generation algorithms is that MalGAN is able to decrease the detection rate to nearly zero and make the retraining based defensive method against adversarial examples hard to work.

A coupled-wave model is developed for photonic-crystal quantum cascade lasers. The analytical model provides an efficient analysis of full three-dimensional large-area device structure, and the validity is confirmed via simulations and previous experimental results., Comment: 21 pages and 8 figures

Wildfires emit significant amounts of pollutants that degrade air quality. Plumes from three wildfires in the western U.S. were measured from aircraft during the Studies of Emissions and Atmospheric Composition, Clouds and Climate Coupling by Regional Surveys (SEAC4RS) and the Biomass Burning Observation Project (BBOP), both in summer 2013. This study reports an extensive set of emission factors (EFs) for over 80 gases and 5 components of submicron particulate matter (PM1) from these temperate wildfires. These include rarely, or never before, measured oxygenated volatile organic compounds and multifunctional organic nitrates. The observed EFs are compared with previous measurements of temperate wildfires, boreal forest fires, and temperate prescribed fires. The wildfires emitted high amounts of PM1 (with organic aerosol (OA) dominating the mass) with an average EF that is more than 2 times the EFs for prescribed fires. The measured EFs were used to estimate the annual wildfire emissions of carbon monoxide, nitrogen oxides, total non methane organic compounds, and PM1 from 11 western U.S. states. The estimated gas emissions are generally comparable with the 2011 National Emissions Inventory (NEI). However, our PM1 emission estimate (1530 +/- 570 Gg/yr) is over 3 times that of the NEI PM2.5 estimate and is also higher than the PM2.5 emitted from all other sources in these states in the NEI. This study indicates that the source of OA from biomass burning in the western states is significantly underestimated. In addition, our results indicate that prescribed burning may be an effective method to reduce fine particle emissions.

Oxidation flow reactors (OFRs) using low-pressure Hg lamp emission at 185 and 254 nm produce OH radicals efficiently and are widely used in atmospheric chemistry and other fields. However, knowledge of detailed OFR chemistry is limited, allowing speculation in the literature about whether some non-OH reactants, including several not relevant for tropospheric chemistry, may play an important role in these OFRs. These non-OH reactants are UV radiation, O(1D), O(3P), and O3. In this study, we investigate the relative importance of other reactants to OH for the fate of reactant species in OFR under a wide range of conditions via box modeling. The relative importance of non-OH species is less sensitive to UV light intensity than to relative humidity (RH) and external OH reactivity (OHRext), as both non-OH reactants and OH scale roughly proportional to UV intensity. We show that for field studies in forested regions and also the urban area of Los Angeles, reactants of atmospheric interest are predominantly consumed by OH. We find that O(1D), O(3P), and O3 have relative contributions to VOC consumption that are similar or lower than in the troposphere. The impact of O atoms can be neglected under most conditions in both OFR and troposphere. Under pathological OFR conditions of low RH and/or high OHRext, the importance of non-OH reactants is enhanced because OH is suppressed. Some biogenics can have substantial destructions by O3, and photolysis at non-tropospheric wavelengths (185 and 254 nm) may also play a significant role in the degradation of some aromatics under pathological conditions. Working under low O2 with the OFR185 mode allows OH to completely dominate over O3 reactions even for the biogenic species most reactive with O3. Non-tropospheric VOC photolysis may have been a problem in some laboratory and source studies, but can be avoided or lessened in future studies by diluting source emissions and working at lower precursor concentrations in lab studies, and by humidification. SOA photolysis is shown to be insignificant for most functional groups, except for nitrates and especially aromatics, which may be photolyzed at high UV flux settings. Our work further establishes the OFR's usefulness as a tool to study atmospheric chemistry and enables better experiment design and interpretation, as well as improved future reactor design.

One of the most wideranging reforms in public education in the last decade has been the reorganization of large comprehensive high schools into small schools with roughly 100 students per grade. We use assignment lotteries embedded in New York City's high school match to estimate the effects of attendance at a new small high school on student achievement. More than 150 unselective small high schools created between 2002 and 2008 have enhanced autonomy, but operate withindistrict with traditional public school teachers, principals, and collectivelybargained work rules. Lottery estimates show positive score gains in Mathematics, English, Science, and History, more credit accumulation, and higher graduation rates. Small school attendance causes a substantial increase in college enrollment, with a marked shift to CUNY institutions. Students are also less likely to require remediation in reading and writing when at college. Detailed school surveys indicate that students at small schools are more engaged and closely monitored, despite fewer course offerings and activities. Teachers report greater feedback, increased safety, and improved collaboration. The results show that school size is an important factor in education production and highlight the potential for within-district reform strategies to substantially improve student achievement.