PWL Approximation for Dense Mapping and Associated Dijkstra Processes for the Concurrent Synthesis of Multiple Full Cost-to-Go Functions
Karime Pereida, José Guivant · 2013
This work proposes and illustrates the synthesis of multiple cost-to-go functions based on PWL approximation of a dense environment. The goal is to create a data structure based on PWL quadtree that accurately describes the working space allowing optimisation algorithms to explore greater areas of the working space in an ecient manner. To illustrate the proposed approach, a Dijsktra algorithm will be used to compute cost-to-go functions in both, the working space and the data structure and the results will be presented. The methodology includes adaptations to the algorithms used in order to achieve higher eciency of the computational cost.