Information-Driven Planning and Control by Tuntematon

Author:Tuntematon [Tuntematon]
Language: eng
Format: epub
Tags: Perception; autonomy; active sensing; sensor networks; unmanned vehicles; wireless; coverage; detection; localization; tracking; classification; environmental adaptation; path planning; performance modelling; performance prediction; information theory; collaborative; coordinated; information-driven; feedback; control; Optimal control; information theory; probability theory; graph theory;Autonomous vehicles; sensor measurements; targets; environment.;Coverage; tracking; detection; classification;Packing algorithms; multi-objective optimization; metaheuristic optimization;Trajectory optimization; path planning; feedback control;
Publisher: MIT Press

where 𝒵 is the range of z(tk). Then, the FIM can be obtained by substituting the above marginal likelihood in (15.37) as follows:

To take into account variable noise statistics, the joint measurement PDF in (15.39) must be carefully derived from the measurement model (15.26) and the environmental variability model (10.1). Then, the joint measurement PDF can be formulated as a function of Λ or parameterized by the estimate , both of which are functions of position, ξ ∈𝒲, and time t ∈ [T0, Tf], as shown in chapter 10. For simplicity, the joint PDF’s dependency on Λ is left implicit in the remainder of this section.

When planning a sequence of measurements, the range of all possible values grows exponentially and, therefore, computing the marginal likelihood (15.39) can become intractable. Another alternative is to follow the expectation maximization (EM) approach, which computes the expected log-likelihood function and then maximizes it with respect to the next parameter estimate. In sensor planning, the expected log-likelihood is obtained by taking the expectation of the log-likelihood in (15.32) with respect to the conditional distribution of the unknown measurement, z(tk), given the prior observations and latest parameter estimate, or

Download

Copyright Disclaimer:
This site does not store any files on its server. We only index and link to content provided by other sites. Please contact the content providers to delete copyright contents if any and email us, we'll remove relevant links or contents immediately.