Showing papers on "Admissible heuristic published in 2002"
TL;DR: This work describes a new technique for designing more accurate admissible heuristic evaluation functions, based on pattern databases, that can be improved on the Fifteen Puzzle by a factor of over 2000, and to find optimal solutions to 50 random instances of the Twenty-Four Puzzle.
217 citations
28 Jul 2002
TL;DR: A plnning algorithm is described that integrates two approaches to solving Markov decision processes with large state spaces in a novel way that exploits symbolic model-checking techniques and demonstrates their usefulness for decision-theoretic planning.
Abstract: We describe a plnning algorithm that integrates two approaches to solving Markov decision processes with large state spaces. State abstraction is used to avoid evaluating states individually. Forward search from a start state, guided by an admissible heuristic, is used to avoid evaluating all states. We combine these two approaches in a novel way that exploits symbolic model-checking techniques and demonstrates their usefulness for decision-theoretic planning.
104 citations
15 Dec 2002
TL;DR: The novel block depth first search (BDFS) algorithm using an admissible heuristic is deployed to assign cells to switches (i.e., to group base stations into location areas), so as to minimize the paging, updating, and physical infrastructure costs.
Abstract: This paper tries to solve the problem of cell to switch assignment (CSA), which plays an important role in designing an optimal location area for a mobile cellular network. The problem is conventionally formulated as a combinatorial optimization problem, which has been shown to be NP hard. So to solve this problem in real time, efficient heuristics and search strategies are necessary. In this work, the novel block depth first search (BDFS) algorithm using an admissible heuristic is deployed to assign cells to switches (i.e., to group base stations into location areas), so as to minimize the paging, updating, and physical infrastructure costs. The algorithm is flexible enough to handle both memory and time constraints while providing a satisfactory solution, which is very much useful for the mobile service providers in reconfiguring location areas online. Additionally, if the time constraint is relaxed, BDFS guarantees to produce the optimal solution, which may help a designer to properly plan a mobile cellular network in the pre-deployment stage.
13 citations
12 Nov 2002
TL;DR: The pomset-based model presented in this paper takes into account the precedence constraints in order to obtain a better estimation for the second heuristic function, so that the performance of the algorithm could be improved.
Abstract: This paper presents a model based on pomsets (partially ordered multisets) for estimating the minimum number of setups in the workcells in Assembly Sequence Planning. This problem is focused through the minimization of the makespan (total assembly time) in a multirobot system. The planning model considers, apart from the durations and resources needed for the assembly tasks, the delays due to the setups in the workcells. An A* algorithm is used to meet the optimal solution. It uses the And/Or graph for the product to assemble, that corresponds to a compressed representation of all feasible assembly plans. Two basic admissible heuristic functions can be defined from relaxed models of the problem, considering the precedence constraints and the use of resources separately. The pomset-based model presented in this paper takes into account the precedence constraints in order to obtain a better estimation for the second heuristic function, so that the performance of the algorithm could be improved.
2 citations
01 Jan 2002
TL;DR: A real time admissible heuristic learning algorithm that allows the tour configuration to change as the tour is built to transform the traveling salesman problem into a state space model is described.
Abstract: This paper describes the application of a real time admissible heuristic learning algorithm that allows the tour configuration to change as the tour is built. We present a state space transformation process to transform the traveling salesman problem into a state space model. A state space transformation process that defines state, state transition operator and state transition cost will be given. The heuristic evaluation function of the algorithm considers both local and global estimated distance information. The heuristic estimation of a state is computed using minimal spanning tree. The heuristic learning mechanism of this approach allows the heuristic estimates of visited states to be updated, and hence modify the tour configuration along the search process.
1 citations