Distributed and Adaptive Target Tracking with a Sensor Network
Abstract
Ensuring the robustness and resilience of safety-critical systems from civil aviation to military surveillance technologies requires improvements to target tracking capabilities. Implementing target tracking as a distributed function can improve the quality and availability of information for end users. Any errors in the model of a target’s dynamics or a sensor network’s measurement process will result in estimates with degraded accuracy or even filter divergence. This dissertation solves a distributed estimation problem for estimating the state of a dynamical system and the parameters defining a model of that system. The novelty of this work lies in the ability of a sensor network to maintain consensus on state and parameter estimates through local communications between sensor platforms. The system for the target dynamics and sensor network’s measurement process was defined as a parameterized distributed state-space model (PDSSM). Two solutions were presented for this distributed state and parameter estimation problem: the Adaptive Centralized Kalman Filter (ACKF) and the Adaptive Distributed Kalman Filter (ADKF). The algorithms were derived in terms of state, parameter, and consensus estimation. These solutions were designed to highlight the difference between utilizing a Kalman Filter and a Distributed Kalman Filter for the state estimation at each sensor platform. An analysis was provided for the computational complexity, communication cost, optimality, stability, and simulation-based performance. The algorithms provided very similar results under nominal conditions, but the ADKF had a much high communication cost. Furthermore, the ADKF is likely to present a significant challenge to realize a communication network with a sufficient message size and broadcast rate for target tracking applications. Then, the ACKF was implemented in a civil aviation simulation in order to demonstrate the capability for improving the safety of an airspace. The safety of an airspace was quantified by calculating the time until a future conflict will exist. The ACKF can improve the availability and quality of estimated aircraft state data, thus reducing the uncertainty an aircraft’s safety assessment. The availability of state data refers to the degree to which aircraft and air traffic control stations can calculate state estimates of each aircraft for traffic management. Currently, the ACKF is the recommended algorithm over the ADKF based on the significantly smaller communication cost for calculating similar results. In order to make the ADKF a more robust solution than the ACKF, the parameter estimation routine can be altered at the expense of a higher communication cost.
Degree
Ph.D.
Advisors
DeLaurentis, Purdue University.
Subject Area
Communication|Electrical engineering|Mathematics
Off-Campus Purdue Users:
To access this dissertation, please log in to our
proxy server.