Drone Inspection Path Planning With Obstacle-Aware Eulerian Graphs

The patent segments the path planning process into multiple distinct modules: a first graph generation module that creates a comprehensive graph including all nodes and edges, a second graph generation module that selects specific inspection object edges, and a third graph generation module that generates the final Eulerian path. This segmentation allows each module to handle a specific aspect of obstacle consideration, improving reliability while managing complexity through modular design.

The patent transforms the path planning problem from a traditional 2D plane into a graph theory problem with multiple dimensions. By representing the flight area as a graph with nodes and edges, and incorporating obstacle information as constraints on edge selection, the system adds a dimensional layer of graph structure that systematically handles obstacle avoidance while maintaining computational tractability.

The patent applies preliminary action by pre-generating the first comprehensive graph that includes all possible nodes and edges in the flight area before actual path planning occurs. This pre-established graph structure serves as a foundation that can be quickly processed by subsequent modules, reducing the computational time required during actual operation while still enabling complete flight path optimization from departure to return.

The patent extracts only the necessary inspection object edges from the comprehensive first graph to create the second graph. By taking out only the relevant edges that correspond to inspection objects, the system reduces the computational burden on the third graph generation module while ensuring that the entire flight path (including departure, inspection, and return) is still optimized. This extraction approach balances productivity improvement with reduced computational time.