Autonomous Nap-of-the-Earth Flight Path Planning for Rotorcraft
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current path planning systems for rotorcraft, especially in NoE (Nap-of-the-Earth) mode, face challenges in providing a safe and efficient flight path due to noisy terrain data, complex algorithms, and high computational requirements, which limits their practicality and ability to guarantee safety and optimal trajectories.
Innovation Solution
A computer-implemented method that calculates a shortest safe path through a rectangular grid graph, performs geometric smoothing to generate a smooth three-dimensional trajectory, and dynamic smoothing to provide a maximum possible speed profile, ensuring a safe and quasi-optimal flight path for rotorcraft, using a modified Dijkstra algorithm and sensor-based updates.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If conventional optimal NOE trajectory planning solutions use a linear combination of flight time and terrain masking as the performance index, then terrain masking is improved, but manufacturing precision deteriorates due to noisy terrain data requiring first and second partial derivatives
Solution Approach 1:
The patent segments the continuous terrain surface into discrete grid cells with predetermined length and width. This discretization allows the system to represent terrain data in a structured manner, avoiding the need to compute noisy partial derivatives of continuous terrain functions while still achieving effective terrain masking through grid-based path planning.
Solution Approach 2:
The patent uses a simplified grid-based representation instead of complex continuous terrain models. This approach trades off some modeling precision for computational efficiency and robustness against noisy sensor data, using inexpensive discrete grid cells rather than expensive high-precision continuous surface representations that require derivative calculations.
2Productivity
If complex algorithms are used for NOE path planning, then trajectory optimality is improved, but device complexity increases
Solution Approach 1:
The patent divides the flight path planning problem into discrete grid cell transitions. Instead of solving a complex continuous optimization problem, the system uses a modified Dijkstra algorithm on a segmented grid graph, which simplifies the computational approach while still finding optimal paths through the terrain.
Solution Approach 2:
The patent replaces complex continuous mathematical optimization mechanisms with a discrete graph-based algorithm. The modified Dijkstra algorithm operating on grid cells substitutes for sophisticated continuous trajectory optimization, reducing computational complexity while maintaining path optimality.
3Reliability
If high computational requirements are imposed on path planning systems, then path safety is improved, but productivity decreases due to real-time processing constraints
Solution Approach 1:
The patent segments the flight environment into a grid graph structure that enables efficient computational processing. This segmentation allows safety-critical path planning to be performed rapidly using discrete algorithms, meeting real-time processing requirements while maintaining high path safety through systematic exploration of the grid space.
Solution Approach 2:
The patent implements dynamic path planning that adapts to changing terrain and obstacle conditions in real-time. The system dynamically updates the grid-based path calculations based on current sensor data, maintaining path safety while responding to environmental changes without excessive computational burden.
Data Source
AI summary
A flight path planning approach may be deterministic and guarantee a safe, quasi-optimal path. A plurality of three-dimensional voxels may be determined as cells of a rectangular grid. The cells may have a predetermined length and width. A shortest safe path through the grid graph may be calculated from a local start to a local goal defined as points on a nominal global path. Geometric smoothing may be performed on the basis line from the local start to the local goal to generate a smooth three-dimensional trajectory that can be followed by a given rotorcraft. Dynamic smoothing may be performed on the three-dimensional trajectory to provide a maximum possible speed profile over a path defined by the dynamic smoothing. The three dimensional path information may be provided to an autopilot, which may then control the rotorcraft to fly along the defined path.