In the previous chapter, the deterministic awareness-based decision-making strategies for search and classification are developed assuming perfect sensing under both centralized and decentralized system architectures. This chapter focuses on the development of real-time decision-making criteria given limited sensory resources under probabilistic frameworks. The uncertainties in sensor perception is taken into account during MAV decision-making. Bayesian-based and information-theoretic search versus classification decision-making strategies are developed that result in guaranteed detection and classification of all the unknown objects in the domain. The basic problem setup is first introduced in Section 4.1. In Section 4.2, the probabilistic counterparts of the task metrics are provided. Based on the problem formulation and task metrics, a Bayesian-based decision-making strategy is developed in Section 4.3. Both a numerical example and Monte-Carlo simulation experiments are presented in Section 4.4 to study the performance of the proposed decision-making strategy.
Bayesian-Based Decision-Making Strategy
Search and Classification Using Multiple Autonomous Vehicles. Decision-Making and Sensor Management ; 79-88
2012
10 Seiten
Article/Chapter (Book)
English
Bayesian Decision Making for Planetary Micro-Rovers
AIAA | 2012
|2012-2440 Bayesian Decision Making for Planetary Micro-Rovers
British Library Conference Proceedings | 2012
|Bayesian network-based security decision-making method for the full-ocean-depth “Wukong” AUV
SAGE Publications | 2023
|Collective Decision-Making and Change Detection with Bayesian Robots in Dynamic Environments
BASE | 2023
|