DEVELOPMENT OF NOVEL METHODS FOR THE PREDICTION AND REAL-TIME DETECTION OF ADVERSE PILOT-VEHICLE INTERACTIONS



Fasiello Bernardi, Simone
(2024) DEVELOPMENT OF NOVEL METHODS FOR THE PREDICTION AND REAL-TIME DETECTION OF ADVERSE PILOT-VEHICLE INTERACTIONS. PhD thesis, University of Liverpool.

[thumbnail of 201314714_Jun2024.pdf] Text
201314714_Jun2024.pdf - Author Accepted Manuscript

Download (56MB)

Abstract

Nowadays, the complexity of high-speed civil transport and highly augmented rotorcraft has led to an increase in the chances of encountering adverse pilot-vehicle system (PVS) interactions, such as the so-called aircraft/rotorcraft-pilot couplings (A/RPCs), whose unpredictability is a serious problem for mission safety. To better understand the causes that can lead to PVS interactions for advanced helicopter configurations, modelling and simulation techniques for the prediction and real-time detection of adverse A/RPC events have been used, explored, and enhanced through the study reported in this thesis. First of all, the phase-aggression criterion (PAC) has been developed as a parameter to indicate the presence of A/RPC events in real-time, this can be linked to a head-up display for alerting the pilot that they are in, or about to enter, an A/RPC. Moreover, atmospheric turbulence and swashplate actuator rate limits have been included in the simulations to stimulate the onset of A/RPC events and to assess its effect on pilot workload. Lastly, to enable objective and quantitative comparisons of human behaviour during adverse PVS interactions, the cybernetic approach with its pilot modelling techniques has been used for identification of manual control behaviour and adaptation to variations in aircraft dynamics. Hence, the results of this research contribute towards the facilitation of unmasking and objectively detecting in real-time unstable phenomena due to unwanted interactions between aircraft/rotorcraft and pilot. One key research question to be answered through the efforts carried out within this work is “How can one better detect in real-time embedded tendencies to A/RPCs in modern aircraft?” To answer this question, initially, an assessment of the efficacy of the PAC has been undertaken both as a means of alerting the pilot to conditions likely to lead to the onset of a pilot-induced oscillation (PIO), and as a means to assist the pilot in alleviating the PIO condition itself. Results from flight simulation trials to explore how best to achieve this have been reported. Objective and subjective measures, designed in past research projects to reduce the A/RPC susceptibility in rotorcraft, were used to develop and implement a novel real-time detection tool useful for piloted simulated (and potentially real) flights, along with a warning system able to provide useful cues to the pilot. Pilot-in-the-loop experiments, in which different test pilots participated, were run inside the University of Liverpool’s and Delft University of Technology’s flight simulation facilities. The purpose of these piloted simulation test campaigns was not only to develop, test and validate a real-time detection system to alleviate and/or reduce the presence of A/RPCs, but also to perform an objective estimation of human control behaviour and adaptation to changes in the controlled aircraft dynamics. The research reported in this thesis contributes to enlarging the state-of-the-art and the know-how in relation to adverse pilot-vehicle interactions, such as A/RPC events, with the aim of preventing their incipience, or detecting and alleviating them, prior to catastrophic occurrences, through the use of novel tools which can be implemented in the aircraft/rotorcraft. Overall, the results show a good correlation between objective and subjective evaluations, and that it is possible to detect PIOs in real-time and provide a timely indication of the PIO and its severity to the pilot. Furthermore, a significant effect of turbulence on A/RPCs incipience and pilot workload is observed. Finally, the results from the cybernetic approach show consistent adaptation of manual control behaviour to variations in both short-period parameters and a worsening of task performance with decreased short-period natural frequency and decreased damping ratio settings.

Item Type: Thesis (PhD)
Depositing User: Symplectic Admin
Date Deposited: 09 Sep 2024 14:55
Last Modified: 08 Feb 2025 03:04
DOI: 10.17638/03182342
Supervisors:
URI: https://livrepository.liverpool.ac.uk/id/eprint/3182342