Thumbnail Image
Publication or External Link
Park, Shinkyu
Martins, Nuno C.
In this dissertation, we consider three problems: in the first we investigate distributed state estimation of linear time-invariant (LTI) plants; in the second we study optimal remote state estimation of Markov processes; while in the third we examine stability of evolutionary game dynamics in large populations. Problem 1: Consider that an autonomous LTI plant is given and that each member of a network of LTI observers accesses a portion of the output of the plant. The dissemination of information within the network is dictated by a pre-specified directed graph in which each vertex represents an observer. This work proposes a distributed estimation scheme that is a natural generalization of consensus in which each observer computes its own state estimate using only the portion of the output vector accessible to it and the state estimates of other observers that are available to it, according to the graph. Unlike straightforward high-order solutions in which each observer broadcasts its measurements throughout the network, the average size of the state of each observer in the proposed scheme does not exceed the order of the plant plus one. We determine necessary and sufficient conditions for the existence of a parameter choice for which the proposed scheme attains asymptotic omniscience of the state of the plant at all observers. The conditions reduce to certain detectability requirements that imply that if omniscience is not possible under the proposed scheme then it is not viable under any other scheme -- including higher order LTI, nonlinear, and time-varying ones -- subject to the same graph. We apply the proposed scheme to distributed tracking of a group of water buffaloes. Problem 2: Consider a two-block remote estimation framework in which a sensing unit accesses the full state of a Markov process and decides whether to transmit information about the state to a remotely located estimator given that each transmission incurs a communication cost. The estimator finds the best state estimate of the process using the information received from the sensing unit. The main purpose of this work is to design transmission policies and estimation rules that dictate decision making of the sensing unit and estimator, respectively, and that are optimal for a cost functional which combines the expectation of squared estimation error and communication costs. Our main results establish the existence of transmission policies and estimation rules that are jointly optimal, and propose an iterative procedure to find ones. Our convergence analysis shows that the sequence of sub-optimal solutions generated by the proposed procedure has a convergent subsequence, and the limit of any convergent subsequence is a person-by-person optimal solution. We apply the proposed scheme to remote estimation of location of a water buffalo. Problem 3: We investigate an energy conservation and dissipation (passivity) aspect of evolutionary dynamics in evolutionary game theory. We define a notion of passivity for evolutionary dynamics, and describe conditions under which dynamics exhibit passivity. For dynamics that are defined on a finite-dimensional state space, we show that the conditions can be characterized in connection with state-space realizations of the dynamics. In addition, we establish stability of passive dynamics in terms of dissipation of stored energy defined by passivity, and present stability results in population games. We provide implications of stability for various passive dynamics both analytically and by means of numerical simulations.