Combined Trajectory Generation and Path Planning for Mobile Robots Using Lattices with Hybrid Dimensionality

摘要: Safe navigation for mobile robots in unstructured and dynamic environments is still a challenging research topic. Most approaches use separate algorithms global path planning local obstacle avoidance. However, this generally results globally sub-optimal strategies. In paper, we present an algorithm which combines these two tasks single integrated approach. For purpose, introduce novel search space, namely, × lattice with hybrid dimensionality. We describe procedure generating high-quality motion primitives robot four-wheel steering to define the lattice. Our computes solution problem consisting of trajectory (i.e., time component) imminent future, dynamically feasible near kinematically remaining goal. Finally, provide some our action prove its high quality real-time capability.