UAV Flight Control Using Adaptive Cost Functions
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing UAV systems require constant human control to achieve flight objectives and ensure sufficient battery power, limiting their autonomous operation, and manual symbolic differentiation is cumbersome and error-prone under time constraints.
Innovation Solution
A flight control subsystem on the UAV accesses cost functions to autonomously control movement, adjusting them based on user input or conflicts, and employs computer-implemented symbolic differentiation using multithreaded processing to enhance processing speed.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If manual symbolic differentiation is used for real-time control, then implementation is simple, but processing speed is insufficient and error-prone
Solution Approach 1:
The patent replaces manual symbolic differentiation with computer-implemented automatic differentiation. The system uses a computation graph to automatically compute gradients through forward and backward passes, eliminating the need for manual derivation while providing exact analytical gradients. This substitution of mechanical/manual processes with automated computational systems resolves the contradiction between processing speed and accuracy.
Solution Approach 2:
The patent introduces a computation graph as an intermediary structure that represents the cost function and its derivatives. This graph serves as a mediator between the control objectives and the gradient computation, enabling automatic differentiation through structured forward and backward passes. The computation graph resolves the contradiction by providing a systematic framework that automates gradient calculation while maintaining mathematical precision.
2Extent of automation
If constant human control is maintained, then flight objectives can be achieved, but autonomous operation is limited and human intervention is required
Solution Approach 1:
The patent enables the UAV to perform self-service through autonomous cost function optimization. The system automatically defines cost functions representing flight objectives, computes gradients through automatic differentiation, and adjusts control inputs without human intervention. This self-service capability resolves the contradiction by enabling autonomous operation while maintaining reliability through mathematically rigorous optimization.
Solution Approach 2:
The patent implements continuous feedback through the optimization loop where cost functions evaluate flight performance, gradients guide control adjustments, and the system iteratively improves toward objectives. This feedback mechanism resolves the contradiction by enabling autonomous decision-making based on real-time state assessment while maintaining reliability through systematic error correction.
3Adaptability or versatility
If multiple cost functions are optimized simultaneously, then multifaceted goals are achieved, but computational complexity increases
Solution Approach 1:
The patent merges multiple cost functions into a unified optimization framework where gradients from different objectives are combined through the computation graph. The system defines cost functions for various flight objectives (positioning, imaging, battery management) and automatically computes their combined gradients through structured forward and backward passes. This merging approach resolves the contradiction by handling multifaceted goals through systematic gradient aggregation rather than separate computations.
Solution Approach 2:
The patent creates a universal automatic differentiation system that handles diverse cost functions through a common computation graph framework. The same forward-backward pass mechanism efficiently computes gradients for positioning, imaging quality, battery consumption, and other objectives. This multi-functional approach resolves the contradiction by providing a versatile yet computationally efficient framework for optimizing multiple conflicting objectives simultaneously.
Data Source
AI summary
A computer of an unmanned aerial vehicle (UAV) accesses, from a memory unit, a problem definition comprising cost functions associated with travel of the UAV. The computer causes movement of the UAV based on the cost functions. The computer adjusts one or more of the cost functions during a flight of the UAV. The computer causes further movement of the UAV based on the adjusted one or more of the cost functions.