The Dubins path problem had enormous applications in path planning for autonomous vehicles. In this paper, we consider a generalization of the Dubins path planning problem, which is to find a shortest Dubins path that starts from a given initial position and heading, and ends on a given target circle with the heading in the tangential direction. This problem has direct applications in Dubins neighborhood traveling salesman problem, obstacle avoidance Dubins path planning problem etc. We characterize the length of the four CSC paths as a function of angular position on the target circle, and derive the conditions which to find the shortest Dubins path to the target circle. The Dubins path problem had enormous applications in path planning for autonomous vehicles. In this paper, we consider a generalization of the Dubins path planning problem, which is to find a shortest Dubins path that starts from a given initial position and heading, and ends on a given target circle with the heading in the tangential direction. This problem has direct applications in Dubins neighborhood traveling salesman problem, obstacle avoidance Dubins path planning problem etc. We characterize the length of the four CSC paths as a function of angular position on the target circle, and derive the conditions which to find the shortest Dubins path to the target circle.
We employ a genetic algorithm approach to solving the persistent visitation problem for UAVs. The objective is to minimize the maximum weighted revisit time over all the sites in a cyclicly repeating walk. In general, the optimal length of the walk is not known, so this method (like the exact methods) assume some fixed length. Exact methods for solving the problem have recently been put forth, however, in the absence of additional heuristics, the exact method scales poorly for problems with more than 10 sites or so. By using a genetic algorithm, performance and computation time can be traded off depending on the application. The main contributions are a novel chromosome encoding scheme and genetic operators for cyclic walks which may visit sites more than once. Examples show that the performance is comparable to exact methods with better scalability.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.