1,720,999 research outputs found
Asymptotic properties of statistical estimators using multivariate Chi-squared measurements
No full textThis paper studies the problem of estimating a parameter vector from measurements having a multivariate chi-squared distribution. Maximum likelihood estimation in this setting is unfeasible because the multivariate chi-squared distribution has no closed form expression. The typical approach to go around this consists in considering a sub-optimal solution by replacing the chi-squared distribution with a normal one. We investigate the theoretical properties of this approximation as the number of measurements approach infinity. More precisely, we show that this approximation is strongly consistency, asymptotically normal and asymptotically efficient. We consider a source localization problem as a case study.Fil: Marelli, Damian Edgardo. Consejo Nacional de Investigaciones Científicas y Técnicas. Centro Científico Tecnológico Conicet - Rosario. Centro Internacional Franco Argentino de Ciencias de la Información y de Sistemas. Universidad Nacional de Rosario. Centro Internacional Franco Argentino de Ciencias de la Información y de Sistemas; Argentina. Guandong University of Technology; ChinaFil: Fu, Minyue. Universidad de Newcastle; Australia. Guandong University of Technology; Chin
Convergence and Accuracy Analysis for a Distributed Static State Estimator based on Gaussian Belief Propagation
No full textThis paper focuses on the distributed static estimation problem and a Belief Propagation (BP) based estimation algorithm is proposed. We provide a complete analysis for convergence and accuracy of it. More precisely, we offer conditions under which the proposed distributed estimator is guaranteed to converge and we give concrete characterizations of its accuracy. Our results not only give a new algorithm with good performance but also provide a useful analysis framework to learn the properties of a distributed algorithm. It yields better theoretical understanding of the static distributed state estimator and may generate more applications in the future.Fil: Marelli, Damian Edgardo. Consejo Nacional de Investigaciones Científicas y Técnicas. Centro Científico Tecnológico Conicet - Rosario. Centro Internacional Franco Argentino de Ciencias de la Información y de Sistemas. Universidad Nacional de Rosario. Centro Internacional Franco Argentino de Ciencias de la Información y de Sistemas; ArgentinaFil: Sui, Tianju. Dalian University of Technology; ChinaFil: Fu, Minyue. Universidad de Newcastle; AustraliaFil: Sun, Ximing. Dalian University of Technology; Chin
Distributed Kalman estimation with decoupled local filters
No full textWe study a distributed Kalman filtering problem in which a number of nodes cooperate without central coordination to estimate a common state based on local measurements and data received from neighbors. This is typically done by running a local filter at each node using information obtained through some procedure for fusing data across the network. A common problem with existing methods is that the outcome of local filters at each time step depends on the data fused at the previous step. We propose an alternative approach to eliminate this error propagation. The proposed local filters are guaranteed to be stable under some mild conditions on certain global structural data, and their fusion yields the centralized Kalman estimate. The main feature of the new approach is that fusion errors introduced at a given time step do not carry over to subsequent steps. This offers advantages in many situations including when a global estimate is only needed at a rate slower than that of measurements or when there are network interruptions. If the global structural data can be fused correctly asymptotically, the stability of local filters is equivalent to that of the centralized Kalman filter. Otherwise, we provide conditions to guarantee stability and bound the resulting estimation error. Numerical experiments are given to show the advantage of our method over other existing alternatives.Fil: Marelli, Damian Edgardo. Consejo Nacional de Investigaciones Científicas y Técnicas; Argentina. Guangdong University Of Technology; ChinaFil: Sui, Tianju. Dalian University Of Technology; ChinaFil: Fu, Minyue. Universidad de Newcastle; Australi
Multiple-Vehicle Localization Using Maximum Likelihood Kalman Filtering and Ultra-Wideband Signals
No full textIn this article we study the problem of localizing a fleet of vehicles in an indoor environment using ultra-wideband (UWB) signals. This is typically done by placing a number of UWB anchors with respect to which vehicles measure their distances. The localization performance is usually poor in the vertical axis, due to the fact that anchors are often placed at similar heights. To improve accuracy, we study the use of inter-vehicle distance measurements. These measurements introduce a technical challenge, as this requires the joint estimation of positions of all vehicles, and currently available methods become numerically complex. To go around this, we use a recently proposed technique called maximum likelihood Kalman filtering (MLKF). We present experiments using real data, showing how the addition of inter-vehicle measurements improves the localization accuracy by about 60%. Experiments also show that the MLKF achieves a localization error similar to the best among available methods, while requiring only about 20% of computational time.Fil: Wang, Wenxu. Guandong University Of Technology; ChinaFil: Marelli, Damian Edgardo. Consejo Nacional de Investigaciones Científicas y Técnicas. Centro Científico Tecnológico Conicet - Rosario. Centro Internacional Franco Argentino de Ciencias de la Información y de Sistemas. Universidad Nacional de Rosario. Centro Internacional Franco Argentino de Ciencias de la Información y de Sistemas; ArgentinaFil: Fu, Minyue. Universidad de Newcastle; Australi
Statistical Approach to Detection of Attacks for Stochastic Cyber-Physical Systems
No full textWe study the problem of detecting an attack on a stochastic cyber-physical system. We aim to treat the problem in its most general form. We start by introducing the notion of asymptotically detectable attacks, as those attacks introducing changes to the system's output statistics which persist asymptotically. We then provide a necessary and sufficient condition for asymptotic detectability. This condition preserves generality as it holds under no restrictive assumption on the system and attacking scheme. To show the importance of this condition, we apply it to detect certain attacking schemes which are undetectable using simple statistics. Our necessary and sufficient condition naturally leads to an algorithm which gives a confidence level for attack detection. We present simulation results to illustrate the performance of this algorithm.Fil: Marelli, Damian Edgardo. Consejo Nacional de Investigaciones Científicas y Técnicas. Centro Científico Tecnológico Conicet - Rosario. Centro Internacional Franco Argentino de Ciencias de la Información y de Sistemas. Universidad Nacional de Rosario. Centro Internacional Franco Argentino de Ciencias de la Información y de Sistemas; ArgentinaFil: Sui, Tianju. Dalian University Of Technology; ChinaFil: Fu, Minyue. Universidad de Newcastle; AustraliaFil: Lu, Renquan. Guandong University Of Technology; Chin
Multi-sensor state estimation over lossy channels using coded measurements
No full textThis paper focuses on a networked state estimation problem for a spatially large linear system with a distributed array of sensors, each of which offers partial state measurements, and the transmission is lossy. We propose a measurement coding scheme with two goals. Firstly, it permits adjusting the communication requirements by controlling the dimension of the vector transmitted by each sensor to the central estimator. Secondly, for a given communication requirement, the scheme is optimal, within the family of linear causal coders, in the sense that the weakest channel condition is required to guarantee the stability of the estimator. For this coding scheme, we derive the minimum mean-square error (MMSE) state estimator, and state a necessary and sufficient condition with a trivial gap, for its stability. We also derive a sufficient but easily verifiable stability condition, and quantify the advantage offered by the proposed coding scheme. Finally, simulations results are presented to confirm our claims.Fil: Sui, Tianju. Dalian University of Technology; República de ChinaFil: Marelli, Damian Edgardo. Consejo Nacional de Investigaciones Científicas y Técnicas. Centro Científico Tecnológico Conicet - Rosario. Centro Internacional Franco Argentino de Ciencias de la Información y de Sistemas. Universidad Nacional de Rosario. Centro Internacional Franco Argentino de Ciencias de la Información y de Sistemas; Argentina. Guangdong University of Technology; República de ChinaFil: Sun, Ximing. Dalian University of Technology; República de ChinaFil: Fu, Minyue. Universidad de Newcastle; Australi
Distributed Target Tracking Using Maximum Likelihood Kalman Filter with Non-Linear Measurements
No full textWe propose a distributed method for tracking a target with linear dynamics and non-linear measurements acquired by a number of sensors. The proposed method is based on a Bayesian tracking technique called maximum likelihood Kalman filter (MLKF), which is known to be asymptotically optimal, in the mean squared sense, as the number of sensors becomes large. This method requires, at each time step, the solution of a maximum likelihood (ML) estimation problem as well as the Hessian matrix of the likelihood function at the optimal. In order to obtain a distributed method, we compute the ML estimate using a recently proposed fully distributed optimization method, which yields the required Hessian matrix as a byproduct of the optimization procedure. We call the algorithm so obtained the distributed MLKF (DMLKF). Numerical simulation results show that DMLKF largely outperforms other available distributed tracking methods, in terms of tracking accuracy, and that it asymptotically approximates the optimal Bayesian tracking solution, as the number of sensors and inter-node information fusion iterations increase.Fil: Huang, Zenghong. Guangdong University of Technology; ChinaFil: Marelli, Damian Edgardo. Guangdong University of Technology; China. Consejo Nacional de Investigaciones Científicas y Técnicas. Centro Científico Tecnológico Conicet - Rosario. Centro Internacional Franco Argentino de Ciencias de la Información y de Sistemas. Universidad Nacional de Rosario. Centro Internacional Franco Argentino de Ciencias de la Información y de Sistemas; ArgentinaFil: Xu, Yong. Guangdong University of Technology; ChinaFil: Fu, Minyue. Universidad de Newcastle; Australi
Stability of networked nonlinear systems: Generalization of small-gain theorem and distributed testing
No full textThis paper studies the stability problem for networked control systems. A general result is presented to determine either global uniform boundedness (GUB), global asymptotic stability (GAS) or input-to-state stability (ISS), for interconnected nonlinear systems. This result checks stability in terms of a scalar called the network gain, hence we call the result the network gain theorem. The result generalizes the previously known matrix small-gain theorem and cyclic small-gain theorem for ISS. As in these results, our theorem does not assess the stability of a given networked system, but of a whole family of networked systems satisfying certain common assumption. An advantage of our stability condition is that it is not only sufficient, but also necessary, in the sense that, if not met, there exists an unstable networked system within that family. To complement our theoretical result, we propose a fully distributed algorithm to compute the network gain. We present simulation results to illustrate the proposed algorithm.Fil: Marelli, Damian Edgardo. Consejo Nacional de Investigaciones Científicas y Técnicas. Centro Científico Tecnológico Conicet - Rosario. Centro Internacional Franco Argentino de Ciencias de la Información y de Sistemas. Universidad Nacional de Rosario. Centro Internacional Franco Argentino de Ciencias de la Información y de Sistemas; ArgentinaFil: Fu, Minyue. Southern University Of Science And Technology; Chin
Quaternion-based fault-tolerant control design for spacecraft attitude stabilization: an anti-saturation method
No full textThis paper investigates the problem of spacecraft attitude stabilization using an anti-saturation strategy. Taking into account the actuator faults or failures, input saturation, modeling uncertainties and external disturbances, we propose a novel adaptive neural network fault-tolerant scheme, in which a terminal sliding mode is embedded in a fault-tolerant controller (FTC) that is implemented based on radial basis function neural networks (RBFNNs). The proposed approach not only shows the ro- bustness and adaptivity with respect to unknown mass properties and external disturbances but also is capable of accommodating actuator faults or failures. Moreover, as the designed adaptive parameters are scalars, it only requires light computational load and can avoid redesign process of the controller during spacecraft operation. Finally, the feasibility of the proposed method is illustrated via a numerical example
A Data-driven Approach to Solve a Production Constrained Build-order Optimization Problem
Full text linkProduction constrained build-order optimization problems challenge artificial intelligence research in computer game applications due to an uncertain set of constraints. Traditional approaches provide subjective values in the constraint formulation therefore resulting in unexpected performance of the optimal build-order in a game. In this article, we propose a data-driven approach to solve a build-order optimization problem in StarCraft. We formulate the constraint by learning the parameter values from game replay data, which complements more precise problem formulation. To solve the optimization, we use the improved genetic algorithm by learning initial solutions from the data. We show the performance of the data-driven methods in a StarCraft simulation platform
- …
