查看更多>>摘要：Kernel ridge regression (KRR) is a popular scheme for non-linear non-parametric learning. However, existing implementations of KRR require that all the data is stored in the main memory, which severely limits the use of KRR in contexts where data size far exceeds the memory size. Such applications are increasingly common in data mining, bioinformatics, and control. A powerful paradigm for computing on data sets that are too large for memory is the streaming model of computation , where we process one data sample at a time, discarding each sample before moving on to the next one. In this paper, we propose StreaMRAK - a streaming version of KRR. StreaMRAK improves on existing KRR schemes by dividing the problem into several levels of resolution, which allows continual refinement to the predictions. The algorithm reduces the memory requirement by continuously and efficiently integrating new samples into the training model. With a novel sub-sampling scheme, StreaMRAK reduces memory and computational complexities by creating a sketch of the original data, where the sub-sampling density is adapted to the bandwidth of the kernel and the local dimensionality of the data. We present a showcase study on two synthetic problems and the prediction of the trajectory of a double pendulum. The results show that the proposed algorithm is fast and accurate. (c) 2022 The Author(s). Published by Elsevier Inc. ( http://creativecommons.org/licenses/by/4.0/ )
查看更多>>摘要：In this article, a new fully distributed neural control is presented for periodically time varying parameterized stochastic nonlinear multi-agent systems with hybrid-order dynamics. All follower systems are not required to be nonlinear dynamics of the same-order. The unknown periodically time-varying nonlinear function is described by using neural networks and Fourier series expansion in the design. It is proved that the presented distributed adaptive method can ensure that all closed-loop signals are bounded in the sense of probability. Furthermore, the state variable of the closed-loop system can be proved to converge to an arbitrary small neighborhood of the zero in the mean square sense. Simulation results are presented to demonstrate the effectiveness of the proposed distributed algorithm.(c) 2022 Elsevier Inc. All rights reserved.
查看更多>>摘要：In this paper, the robust adaptive event-triggered fault-tolerant control (FTC) problem is addressed for a time-varying continuous-time system with perturbations and general actuator failures. In order to eliminate the adverse effects of faults and perturbations on system performance, a robust fault-tolerant controller is derived via integrating direct adaptive control technique to adjust control parameters. For the sake of reducing the communication resources and actuator actions, an event-triggered control technique is presented to adjust the control inputs based on the proposed FTC strategy. To ensure the system is free from Zeno phenomenon, a bound on the inter-event execution time is established. On the basis of the event-trigger condition, the uniformly bounded stability results of the adaptive FTC system are obtained by utilizing Lyapunov stability theory in the presence of actuator failures and perturbations. The superior of the developed control strategy is substantiated by simulation results of a rocket fairing system.(c) 2022 Elsevier Inc. All rights reserved.
查看更多>>摘要：We show that adding new chemical species into the reactions of a chemical reaction network (CRN) in such a way that the rank of the network remains unchanged preserves its capacity for multiple nondegenerate equilibria and/or periodic orbits. One consequence is that any bounded nondegenerate behaviours which can occur in a CRN can occur in a CRN with bounded stoichiometric classes. The main result adds to a family of theorems which tell us which enlargements of a CRN preserve its capacity for nontrivial dynamical behaviours. It generalises some earlier claims, and complements similar claims involving the addition of reactions into CRNs. The result gives us information on how ignoring some chemical species, as is common in biochemical modelling, might affect the allowed dynamics in differential equation models of CRNs. We demonstrate the scope and limitations of the main theorem via several examples. These illustrate how we can use the main theorem to predict multistationarity and oscillation in CRNs enlarged with additional species; but also how the enlargements can introduce new behaviours such as additional periodic orbits and new bifurcations. (c) 2022 The Authors. Published by Elsevier Inc. This is an open access article under the CC BY license ( http://creativecommons.org/licenses/by/4.0/ )
查看更多>>摘要：This article focuses on the event-triggered fault-tolerant control (ETFTC) for a family of nonlinear pure-feedback systems (PFSs) is investigated. Firstly, the differential mean-value theorem (MVT) is applicable to deal with the nonaffine function in systems and the con-trol input signals were successfully decoupled. Then, by using backstepping technique and the fuzzy logic systems (FLSs), a novel finite-time ETFTC scheme is effectively rendered. Furthermore, the newly designed fault-tolerant controller compensates for actuator faults, the trigger mechanism improves the transmission efficiency of the input signals, and the ETFTC scheme guarantees that all closed-loop signals are bounded. Comparative simulation results further validate the effectiveness of the given control method. (c) 2022 Elsevier Inc. All rights reserved.
Caraballo, L. E.Diaz-Banez, J. M.Rodriguez, F.Sanchez-Canales, V...
12页
查看更多>>摘要：Melodic similarity measurement is of key importance in music information retrieval. In this paper, we use geometric matching techniques to measure the similarity between two melodies. We represent music as sets of points or sets of horizontal line segments in the Euclidean plane and propose efficient algorithms for optimization problems inspired in two operations on melodies; linear scaling and audio compression. In the scaling problem, an incoming query melody is scaled forward until the similarity measure between the query and a reference melody is minimized. The compression problem asks for a subset of notes of a given melody such that the matching cost between the selected notes and the refer-ence melody is minimized.(c) 2022 Elsevier Inc. All rights reserved.
查看更多>>摘要：In game theory, it is an essential topic to study the emergence and maintenance of cooperative behavior in groups based on the theories of evolutionary game and complex network. Unfortunately, an in-depth analysis of cooperative behavior on maintenance and development is usually challenged by the diversity of groups in society, which is mainly caused by the single mechanism in traditional networks. More recent studies have shown that multi-layer coupled network based evolutionary game theory is promising in exploiting the transmission of cooperative behavior between individuals in the game. Meanwhile, inspired by the decisive ability of reinforcement learning in overcoming the limitation of replica, in this work, we propose to combine the game strategy of reinforcement learning with the traditional prisoner's dilemma strategy based on multiple coupled networks. The most advantage of this model is the improved capability of intelligent decision making for group behaviors. With the simulation of game evolution, the influence of individual strategy change, as well as individual ability on cooperative behavior in reinforcement learning, is also explored. Substantial validations have verified that in social dilemmas, the cooperative behavior can be maintained by adjusting the group's ability with effective guidance.(c) 2022 Elsevier Inc. All rights reserved.
Mathiyalagan, K.Nidhi, A. ShreeSu, H.Renugadevi, T....
19页
查看更多>>摘要：This work focuses on the observer design for a first order ODE-transport PDE coupled at the boundary points. A novel anti-collocated observer and an output feedback boundary control law are designed for an under-actuated coupled system using the backstepping method. The homo-directional hyperbolic type PDE is considered with in-domain coupling between the states. The stabilization of the coupled systems are discussed by Lyapunov theory and linear matrix inequality (LMI) approach is implemented to design the gains. The obtained results show that the observer value coincides with the actual ones and it has been demonstrated through numerical examples. The effectiveness of the output feed-back controller is also illustrated. (c) 2022 Elsevier Inc. All rights reserved.
查看更多>>摘要：The focus of this paper is on solving the state tracking control problem for a class of fuzzy semi-Markovian jump systems with the aid of fuzzy modeling approach. Besides, the impact of both unexpected disturbances and uncertainties in fuzzy semi-Markovian jump systems is considered. In particular, the nonlinearity in the addressed system is represented by the T-S fuzzy model, wherein the disturbances and uncertainties are estimated by precisely considering uncertainty and disturbance estimator (UDE)-based control approach. Whilst, the desired fuzzy rule-based mode-dependent control law is formulated by exploiting the parallel distributed compensation strategy and estimated output of disturbances and uncertainties. In the proposed tracking control scheme, the resilient strategy is considered to overcome the impact of gain perturbations. Further, in terms of LMIs, a set of sufficient constraints is derived in accordance with the appropriate selection of mode dependent fuzzy rule-based Lyapunov function. Under which, the asymptotic tracking and enhanced disturbance rejection performances are guaranteed for the system under consideration. Notably, the desired control gain matrices are determined by solving the derived sufficient conditions via MATLAB LMI toolbox. Conclusively, a single-link robot arm model is adopted to justify the practicability and efficacy of the theoretical findings.(c) 2022 Published by Elsevier Inc.
查看更多>>摘要：We examine the effects of ion diffusion coefficient on steady electroosmotic couple stress nanofluid flow and heat transfer in a porous microchannel bounded by two permeable beds. The study considers the nanofluid injection process through porous beds. The combined effects of ion diffusivity and injected nanofluid velocity on EDL (Electric double layer) thickness are analytically expressed. The differential transform method (DTM) is used to solve the flow equations. The DTM solution is then compared with the exact analytical solution of velocity, Zeta potential function and a square averaged residual analysis is conducted. These comparisons show excellent matching. We observe that the enhancement of ion diffusion coefficient increases nanofluid velocity at the edge of EDL but diminishes with intrinsic couple stress. The bulk nanofluid flow rate is found to follow a quadratic like relationship with the couple stress parameter. We have investigated the nanofluid temperature and nanoparticle concentration distribution in microchannel for variations of the diffusive Reynolds number, porous permeability parameter and couple stress parameter. The nanofluid temperature is found to increase with an enhancement of ion diffusion coefficient, which is related with experimental results observed by Kong et al. (Phys. Chem. Chem. Phys. 19 (2017) 7678) for NaCl electrolyte solution confined in a graphene nanochannel. Moreover, the nanofluid temperature decreases with enhancement of both porous permeability of the medium and couple stress parameter. (c) 2022 Elsevier Inc. All rights reserved.
NSTL
Elsevier