Flight Path Generation Using Visibility Polygons for Obstacle Avoidance
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing flight path generation algorithms do not effectively account for obstacle geometry, leading to inefficient and time-consuming computations, as they often result in inadmissible positions and long computing times due to the lack of anticipation of collisions with environmental obstacles.
Innovation Solution
A method and device that determine lateral extreme apices on either side of the direct flight path, assign ratings based on distance to these apices, and iteratively store and reconstitute the optimum flight path, using an obstacle expansion function to adapt the computation algorithm to the environmental situation, thereby reducing unnecessary computations and focusing on viable solutions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If the algorithm generates new candidate positions in each iteration and checks a posteriori that these positions are admissible with respect to environmental constraints, then the method is robust with regard to different possible configurations, but the computing time becomes long and unacceptable for onboard functions
Solution Approach 1:
The patent applies preliminary action by determining obstacle apices and constructing visibility polygons before generating candidate positions. The visibility polygon pre-identifies all positions that are guaranteed to be admissible (non-colliding) with respect to obstacles, allowing the algorithm to restrict candidate position generation to this pre-computed safe region. This eliminates the need for a posteriori collision checks and prevents generation of inadmissible positions, thereby resolving the contradiction between robustness and computing time.
2Device complexity
If the algorithm goes straight to the joining point until it meets an obstacle, then the path generation is simple, but it results in useless computations of positions that are inadmissible
Solution Approach 1:
The patent applies local quality by making the candidate position generation locally adaptive to the obstacle environment. Instead of uniformly generating candidates along the direct path to the joining point, the algorithm constructs a visibility polygon that locally adapts to obstacle geometry. This visibility polygon defines a local safe region where candidate positions are guaranteed to be admissible, concentrating computational effort only in relevant safe areas rather than exploring useless inadmissible positions.
3Ease of manufacture
If the algorithm does not take account of the location and geometry of obstacles to anticipate collisions, then the algorithm is simpler to implement, but it produces long computing times due to lack of anticipation
Solution Approach 1:
The patent determines obstacle apices (extreme points of obstacle polygons) as preliminary action before path generation. These apices are used to construct the visibility polygon, which pre-anticipates collision-prone areas. By incorporating obstacle geometry information in advance through the visibility polygon construction, the algorithm anticipates which regions are safe for candidate position generation, eliminating the need for complex real-time collision anticipation while maintaining implementation simplicity.
Data Source
AI summary
A device includes a database including data related to obstacles, a first determination module for determining at least one obstacle intercepted by a direct flight path connecting an initial point to the target point, a second determination module for determining at least one lateral extreme apex of at least one intercepted obstacle, an evaluation module for assigning a rating to each section corresponding to a segment between a lateral extreme apex and the initial point, a storage module for storing in a memory the section corresponding to the optimum extreme apex having the best rating, the optimum flight path corresponding to all of the sections stored in the memory, the preceding modules being used iteratively, the device further including a transmission module for transmitting the optimum flight path to a user device.