Simón Bolívar University

About: Simón Bolívar University is a education organization based out in Caracas, Venezuela. It is known for research contribution in the topics: Population & Crystallization. The organization has 5912 authors who have published 8294 publications receiving 126152 citations.

Planning as Heuristic Search: New Results

Abstract: In the recent AIPS98 Planning Competition, the hsp planner, based on a forward state search and a domain-independent heuristic, showed that heuristic search planners can be competitive with state of the art Graphplan and Satisfiability planners. hsp solved more problems than the other planners but it often took more time or produced longer plans. The main bottleneck in hsp is the computation of the heuristic for every new state. This computation may take up to 85% of the processing time. In this paper, we present a solution to this problem that uses a simple change in the direction of the search. The new planner, that we call hspr, is based on the same ideas and heuristic as hsp , but searches backward from the goal rather than forward from the initial state. This allows hspr to compute the heuristic estimates only once. As a result, hspr can produce better plans, often in less time. For example, hspr solves each of the 30 logistics problems from Kautz and Selman in less than 3 seconds. This is two orders of magnitude faster than blackbox. At the same time, in almost all cases, the plans are substantially smaller. hspr is also more robust than hsp as it visits a larger number of states, makes deterministic decisions, and relies on a single adjustable parameter than can be fixed for most domains. hspr, however, is not better than hsp accross all domains and in particular, in the blocks world, hspr fails on some large instances that hsp can solve. We discuss also the relation between hspr and Graphplan, and argue that Graphplan can also be understood as a heuristic search planner with a precise heuristic function and search algorithm.

Successive self-nucleation/annealing (SSA): A novel technique to study molecular segregation during crystallization

Abstract: A new procedure to fractionate ethylene/α-olefin copolymers using DSC is presented. This procedure allows melt/melt and melt/solid segregation to occur during thermal cycles that promote self-nucleation, crystallization and annealing processes (Successive Self-Nucleation/ Annealing, SSA). The SSA has been compared with the Step-Crystallization (SC) method proposed earlier in the literature to qualitatively characterize chain branching distribution in a faster and easier way than Temperature Rising Elution Fractionation (TREF). In general, SSA produces better fractionation than SC and the DSC derived chain branching distribution by SSA can be qualitatively comparable to that obtained by TREF. The SSA technique could have important applications for the characterization of polymers that crystallize over a broad temperature range.

Decoupling gravitational sources in general relativity: from perfect to anisotropic fluids

Abstract: We show a simple, systematic and direct approach to decoupling gravitational sources in general relativity. As a direct application, a robust and simple way to generate anisotropic solutions for self-gravitating systems from perfect fluid solutions is presented.

Ant colony system algorithm for the planning of primary distribution circuits

Abstract: The planning problem of electrical power distribution networks, stated as a mixed nonlinear integer optimization problem, is solved using the ant colony system algorithm (ACS). The behavior of real ants has inspired the development of the ACS algorithm, an improved version of the ant system (AS) algorithm, which reproduces the technique used by ants to construct their food recollection routes from their nest, and where a set of artificial ants cooperate to find the best solution through the interchange of the information contained in the pheromone deposits of the different trajectories. This metaheuristic approach has proven to be very robust when applied to global optimization problems of a combinatorial nature, such as the traveling salesman and the quadratic assignment problem, and is favorably compared to other solution approaches such as genetic algorithms (GAs) and simulated annealing techniques. In this work, the ACS methodology is coupled with a conventional distribution system load-flow algorithm and adapted to solve the primary distribution system planning problem. The application of the proposed methodology to two real cases is presented: a 34.5-kV system with 23 nodes from the oil industry and a more complex 10-kV electrical distribution system with 201 nodes that feeds an urban area. The performance of the proposed approach outstands positively when compared to GAs, obtaining improved results with significant reductions in the solution time. The technique is shown as a flexible and powerful tool for the distribution system planning engineers.

A robust and fast action selection mechanism for planning

Abstract: The ability to plan and react in dynamic environments is central to intelligent behavior yet few algorithms have managed to combine fast planning with a robust execution. In this paper we develop one such algorithm by looking at planning as real time search. For that we develop a variation of Korf's Learning Real Time A* algorithm together with a suitable heuristic function. The resulting algorithm interleaves lookahead with execution and never builds a plan. It is an action selection mechanism that decides at each time point what to do next. Yet it solves hard planning problems faster than any domain independent planning algorithm known to us, including the powerful SAT planner recently introduced by Kautz and Selman. It also works in the presence of perturbations and noise, and can be given a fixed time window to operate. We illustrate each of these features by running the algorithm on a number of benchmark problems.