Engine Anticipation Algorithm for Rotary-Wing Aircraft
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional engine power anticipation algorithms in rotary-wing aircraft, such as collective pitch based and predictive anticipators, face limitations in accurately anticipating power requirements due to dependence on multiple factors and certification issues with neural networks, respectively.
Innovation Solution
A flight control system with a module that determines expected power required data using a model following control law and utilizes this data to control engine speed, providing a deterministic engine power anticipation algorithm that merges flight control system model following algorithms with power required computations, allowing for timely engine speed adjustments.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If collective pitch based anticipators are used, then engine power anticipation can be implemented, but performance is imperfect because power required depends on multiple factors
Solution Approach 1:
The patent merges flight control system model following algorithms with power required computations into a unified deterministic algorithm. This integration allows the system to consider multiple factors (air speed, gross weight, maneuver, etc.) simultaneously through the model following framework, resolving the limitation of collective pitch based anticipators that only monitor control displacement without accounting for other power-affecting variables.
Solution Approach 2:
The patent replaces mechanical or electronic feedback implementations with a deterministic computational algorithm based on model following control laws. This substitution enables more accurate and flexible power anticipation by using mathematical models to predict power requirements rather than relying solely on physical feedback from collective pitch changes.
2Reliability
If predictive anticipators with neural networks are used, then power required changes can be predicted, but certification issues arise because the neural network is not deterministic
Solution Approach 1:
The patent inverts the approach of predictive anticipators by using deterministic model following algorithms instead of non-deterministic neural networks. Rather than attempting to train a neural network to predict power requirements, the system uses analytically determined model following control laws that inherently provide deterministic predictions, thereby achieving both accuracy and certifiability.
Solution Approach 2:
The patent changes the fundamental parameter of the prediction algorithm from non-deterministic neural network weights to deterministic model following control parameters. This parameter change transforms the system from uncertifiable to certifiable while maintaining the ability to predict power required changes accurately across various flight conditions.
3Reliability
If engine power anticipation is improved, then rotor speed control performance improves, but the system complexity increases
Solution Approach 1:
The patent makes the model following control algorithm multi-functional by using it for both flight control and power required computation. This universal approach allows the same deterministic framework to serve dual purposes, improving rotor speed control performance without proportionally increasing system complexity, as the algorithm already exists for flight control purposes.
Data Source
AI summary
A flight control system and method which determines an expected power required data in response to a flight control command of the at least one model following control law and utilizes the expected power required data to perform at least one action to control an engine speed.