Nemanja Ilić

Consensus based overlapping decentralized fault detection and isolation

In this paper a new distributed fault detection and isolation (FDI) methodology is proposed in the form of a multi-agent network representing a combination of a consensus based FDI observer for residual generation and a consensus based decision making strategy for change detection, applicable in real time. The proposed observer is based on overlapping system decomposition and a combination between the local optimal stochastic FDI observers and a dynamic consensus strategy. It is shown how the proposed algorithm can generate residuals which provide, under general conditions concerning local models and the network topology, high efficiency, scalability and robustness. The proposed decision making strategy provides solutions for two particular cases: a) local detection for non-overlapping parts of the identified subsystems; b) a consensus based strategy for FDI in the overlapping parts. One selected example illustrates the applicability of the proposed methodology in practice.

Adaptive Consensus-Based Distributed Target Tracking in Sensor Networks With Limited Sensing Range

In this brief, consensus-based algorithms for distributed target tracking are discussed, and a new algorithm with decentralized adaptation is proposed for sensor networks with limited sensing range. The given comparative analysis shows that tuning of the gains in the consensus scheme is of crucial importance for getting simple, yet efficient tracking algorithms, requiring the exchange of only local state estimates between the nodes. The proposed algorithm, which gives higher importance to the nodes currently receiving measurements, is based on the exchange of additional binary information and represents a robust and efficient tool for practice. Stability of the algorithm is theoretically discussed on a given scenario. Selected examples illustrate its performance in terms of the estimation error and the disagreement between the nodes.

Consensus based distributed change detection using Generalized Likelihood Ratio Methodology

In this paper a novel distributed recursive algorithm based on the Generalized Likelihood Ratio methodology is proposed for real time change detection using sensor networks. The algorithm is based on a combination of recursively generated local statistics and a global consensus strategy, and does not require any fusion center, so that the state of any node can be tested w.r.t. a given common threshold. Two different problems are discussed: detection of an unknown change in the mean and in the variance of an observed random process. Performance of the algorithm for change detection in the mean is analyzed in the sense of a measure of the error with respect to the corresponding centralized algorithm. The analysis encompasses constant and randomly time varying matrices describing communications in the network. Simulation results illustrate characteristic properties of the algorithms.

Consensus based distributed change detection using Generalized Likelihood Ratio methodology

In this paper a novel distributed recursive algorithm based on the Generalized Likelihood Ratio methodology is proposed for real time change detection using sensor networks. The algorithm is based on a combination of recursively generated local statistics and a global consensus strategy, and does not require any fusion center, so that the state of any node can be tested w.r.t. a given common threshold. Two different problems are discussed: detection of an unknown change in the mean and in the variance of an observed random process. Performance of the algorithm for change detection in the mean is analyzed in the sense of a measure of the error with respect to the corresponding centralized algorithm. The analysis encompasses constant and randomly time varying matrices describing communications in the network. Simulation results illustrate characteristic properties of the algorithms.

Consensus based overlapping decentralized observer for fault detection and isolation

In this paper a new fault detection and isolation (FDI) observer is proposed in the form of a multi-agent network based on a combination of local optimal stochastic FDI observers and a dynamic consensus strategy. It is shown that the proposed algorithm can generate residuals which provide, under general conditions concerning local models and the network topology, high efficiency, scalability and robustness of the whole estimation scheme. It is demonstrated that the consensus gains can be obtained by minimizing a suitably defined criterion function. A number of examples illustrate the applicability of the proposed methodology in practice.

Distributed Stochastic Approximation: Weak Convergence and Network Design

This paper studies distributed stochastic approximation algorithms based on broadcast gossip on communication networks represented by digraphs. Weak convergence of these algorithms is proved, and an associated ordinary differential equation (ODE) is formulated connecting convergence points with local objective functions and network properties. Using these results, a methodology is proposed for network design, aimed at achieving the desired asymptotic behavior at consensus. Convergence rate of the algorithm is also analyzed and further improved using an attached stochastic differential equation. Simulation results illustrate the theoretical concepts.

Distributed consensus based IPDAF for tracking in vision networks

In this paper consensus based algorithms for distributed target tracking in large scale camera networks are discussed and a new adaptive algorithm is proposed. Camera networks are typically characterized by sparse communication and coverage topologies, as well as by the presence of multiple targets and clutter. The proposed algorithm (IPDA-ACF) is a result of the introduction of the probabilities of target perceivability and target existence in the basic distributed consensus based tracking algorithm (ACF). The distributed adaptation scheme for information fusion allows obtaining robustness in the cases of high level clutter and occulted targets, together with high level agreement between the nodes. A comparison with analogous methods derived from the Kalman Consensus Filter (KCF) and the Information Consensus Filter (ICF) shows that the proposed method achieves better performance, along with reduced communication requirements.

Distributed Change Detection Based on a Consensus Algorithm

In this paper a novel consensus based distributed recursive algorithm is proposed for real time change detection using sensor networks. Convergence of the algorithm to the optimal centralized solution defined by a weighted sum of the results of local signal processing is proved in the cases of constant and time varying forgetting factors of the underlying recursions, assuming correlated data and different local values of the parameter changes. Simulation results illustrate characteristic properties of the algorithms.

Adaptive sensor networks for consensus based distributed estimation

In this paper consensus based algorithms for distributed estimation in sensor networks are discussed and a new algorithm with decentralized adaptation is proposed for solving the problem where the state of a monitored process is observed only by a relatively small percentage of the sensors at each iteration of the algorithm. The given analysis shows that adaptation of the gains in the consensus scheme is of crucial importance for getting simple yet efficient estimation algorithms. It is also shown that the exchange of an additional binary information between the nodes on whether or not a node has received the observation, along with the information on state estimates, is sufficient to obtain a robust and efficient tool for practice. Selected examples illustrate performance of the proposed algorithm in terms of the mean square estimation error and the disagreement between the nodes.