Representing Iterative Loops for Decision-Theoretic Planning* (Preliminary Report)
نویسندگان
چکیده
Many planning problems are most naturally solved using an interative loop of actions. For example, a natural plan to unload a truckload of boxes is to repeatedly remove a box until the truck is empty. The problem of planning with loops has been investigated in the context of classical planning [Wilkins, 1988] as well as in the context of reactive planning [Musliner, 1994] but not for decision-theoretic planning. This paper discusses representing iterative loops for decisiontheoretic planning. We present a representation of actions of the form "Repeat the following action until condition is satisfied." We address the problem of guaranteeing that such loops will terminate. Finally we discuss how to abstract iterative constructs in order to limit the number of chronicles generated in plan projection. bl0ek-or,-flo0r ¢I=OD
منابع مشابه
Bayes Network Abstraction for Decision-Theoretic Planning
This is a preliminary report on research towards the development of an abstraction-based framework for decision-theoretic planning using Bayesian networks. We discuss two problems: the representation of sets of probability distributions for Bayesian networks, and the issue of representing and abstracting actions using Bayesian networks. For the first problem, we propose the use of cc-trees to r...
متن کاملHPGP: An Abstraction-Based Framework for Decision-Theoretic Planning
This paper is a report on research towards the development of an abstraction-based framework for decision-theoretic planning. We make use of two planning approaches in the context of probabilistic planning: planning by abstraction and planning graphs. To create abstraction hierarchies our planner uses an adapted version of a hierarchical planner under uncertainty, and to search for plans, we pr...
متن کاملBackward Plan Construction for Planning with Partial Observability
We present algorithms for partially observable planning that iteratively compute belief states with an increasing distance to the goal states. The algorithms handle nondeterministic operators, but restrict to problem instances with a finite upper bound on execution length, that is plans without loops. We discuss an implementation of the algorithms which uses binary decision diagrams for represe...
متن کاملStochastic dynamic programming with factored representations
Markov decision processes (MDPs) have proven to be popular models for decision-theoretic planning, but standard dynamic programming algorithms for solving MDPs rely on explicit, state-based specifications and computations. To alleviate the combinatorial problems associated with such methods, we propose new representational and computational techniques for MDPs that exploit certain types of prob...
متن کاملLearnable Evolutionary Optimization in Autonomous Pickup & Delivery Planning
This report describes an evolutionary approach to distributed planning by transport agents. These agents, representing trucks, autonomously decide on their transport orders by interacting with cargo agents representing containers. The agents use a guided evolutionary computation method, called the learnable evolution model, to create transport plans and render optimized decisions on which cargo...
متن کامل