Envelope Protection Systems for Piloted and Unmanned Rotorcraft

Open Access
Author:
Sahani, Nilesh A.
Graduate Program:
Aerospace Engineering
Degree:
Doctor of Philosophy
Document Type:
Dissertation
Date of Defense:
October 05, 2005
Committee Members:
  • Joseph Francis Horn, Committee Chair
  • Edward C Smith, Committee Member
  • Farhan S Gandhi, Committee Member
  • Asok Ray, Committee Member
Keywords:
  • Envelope protection
  • Limit avoidance
  • Rotorcraft
  • Pilot cueing
  • Adaptive control
  • Integrator windup
Abstract:
Performance and agility of rotorcraft can be improved using envelope protection systems (or carefree maneuvering systems), which allow the aircraft to use the full flight envelope without risk of exceeding structural or controllability limits. Implementation of such a system can be divided into two necessary parts: “Limit Prediction” which detects the impending violation of the limit parameter, and “Limit Avoidance” where a preventive action is taken in the form of pilot cueing or autonomous limiting. Depending upon the underlying flight control system, implementation of the envelope limiting system was categorized into two different structures: “Inceptor Constraint Architecture” and “Command Limiting Architecture”. <P> The Inceptor Constraint Architecture is applicable to existing rotorcraft with conventional flight control system where control input proportionally affects control surfaces. The relationship between control input and limit parameter is complex which requires advanced algorithms for predicting impending limit violations. This research focuses on limits that exceed in transient response. A new algorithm was developed for predicting transient response using non-linear functions of measured aircraft states. The functions were generated off-line using simulation data from a non-real-time simulation model to demonstrate the procedure for extracting them from flight test data. Constraints on pilot control stick were generated using control input sensitivity of transient peak. The stick constraints were conveyed to the pilot using a softstop cue on an active control stick. The system was evaluated in real-time piloted simulation for longitudinal hub moment limit avoidance using two different aggressive maneuvers. Results showed that the prediction algorithm was effective, as the system reduced frequency and severity of hub moment limit violations without significantly affecting the achievable agility of the aircraft. Pilot comments suggested a borderline objectionable movement of softstop cue. This was corrected by distribution of constraints between softstop cue and autonomous protection according to the frequency content. <P> Modern rotorcraft flight control systems are designed to accurately track certain aircraft states like roll and pitch attitudes which are either specified as command inputs in unmanned rotorcraft or mapped to control stick in piloted aircrafts. In the Command Limiting Architecture applicable to these systems, performance constraints were generated on the command input corresponding to the envelope limit. To simulate this flight control system, an adaptive model inversion controller was applied to a non-linear, blade element simulation model of a helicopter. The controller generated fully-coupled lateral, longitudinal, vertical and yaw axis control inputs using a single design point linear model. An adaptive neural network compensated for inversion errors. Ability of the controller to accurately track aircraft states was exploited to implement a longitudinal hub moment limiting system. A torque protection system was implemented with coupled constraints in longitudinal and vertical axis. Real-time piloted simulation results showed that aircraft was able to achieve maximum rate of climb, maximum forward acceleration or maximum forward velocity without violating the torque limit. <P> For unmanned rotorcraft (UAV) applications, an Inner Loop – Outer Loop controller was implemented to achieve trajectory tracking. In this controller structure, inner loop constraints generated by the envelope protection system place saturation limits in closed loop feedback path. This causes integrator windup in the presence of integral action in outer loop controller. An outer loop constraint method was implemented to generate constraints on outer loop command corresponding to inner loop constraints. For outer loop controller with feed-forward gain, the method was evaluated for the coupled longitudinal and vertical axis torque limit using non-linear simulation model. For controllers without feed-forward gain, it was evaluated using linear model simulations. Results showed that the outer loop constraint method resulted in smaller transient peaks and small settling times compared to saturation limits on inner loop command.