Herman J. Damveld

Identification of the Feedforward Component in Manual Control With Predictable Target Signals

In the manual control of a dynamic system, the human controller (HC) often follows a visible and predictable reference path. Compared with a purely feedback control strategy, performance can be improved by making use of this knowledge of the reference. The operator could effectively introduce feedforward control in conjunction with a feedback path to compensate for errors, as hypothesized in literature. However, feedforward behavior has never been identified from experimental data, nor have the hypothesized models been validated. This paper investigates human control behavior in pursuit tracking of a predictable reference signal while being perturbed by a quasi-random multisine disturbance signal. An experiment was done in which the relative strength of the target and disturbance signals were systematically varied. The anticipated changes in control behavior were studied by means of an ARX model analysis and by fitting three parametric HC models: two different feedback models and a combined feedforward and feedback model. The ARX analysis shows that the experiment participants employed control action on both the error and the target signal. The control action on the target was similar to the inverse of the system dynamics. Model fits show that this behavior can be modeled best by the combined feedforward and feedback model.

A review of visual driver models for system identification purposes

The aim of this study was to find a realistic control-theoretic visual driver model for curve driving that does not only show simular performance as actual drivers but also applies the same inputs and uses the same information. The model structure must enable system identification and parameter estimation of the model parameters. A large number of existing and adapted models have been evaluated and simulated, and when possible, frequency response functions have been identified using two system identification methods. A significant part of the paper is devoted to review these models. The evaluation shows that two-point models comply best with all system identification requirements while still governing realistic driving behavior. It is recommended to investigate further the positioning and perception part of the two-point models using eye-tracking in driving experiments with real human drivers.

Measuring the Contribution of the Neuromuscular System during a Pitch Control Task

Current pilot models oversimplify the neuromuscular system as a second-order low-pass filter, merely focusing on its role in limiting the stick position bandwidth. However, the neuromuscular system also functions as a fast feedback control system due to reflexive activity and inherent muscle visco-elasticity, allowing pilots to respond intuitively to control column forces, much faster than visual or vestibular cues would allow. Models that neglect this property of the neuromuscular system erroneously attribute its activity to the visual or vestibular system when their parameters are estimated by system identification methods. This paper’s aim is to show a proof-of-concept for a novel method to supplement the currently used pilot models with measurements for the neuromuscular system. These measurements will form the basis for more detailed neuromuscular models, allowing a better description of the contribution of the visual, vestibular, and neuromuscular feedback to the pilot’s control output. In this paper the novel method’s modeling outcome (lumped arm inertia, viscosity and stiffness) will be compared to the conventional neuromuscular estimation (indirect estimation of natural frequency and relative damping). A limited study was performed to provide data for parameter fits of a linearized pilot model for 1 DOF pitch control tasks. In the motion-based SIMONA Research Simulator (SRS) a pilot was instructed to perform a pursuit pitch-tracking task, in face of turbulence on the aircraft, and control force perturbations on the control column. In a novel perturbation method, these three forcing functions (perturbing the visual tracking signal, disturbing the aircraft’s elevator deflection and the control-force cues) were designed to contain power at three different frequency sets, allowing simultaneous identification of three corresponding frequency response functions. Additionally, a separate experiment was done to demonstrate the adaptability of the neuromusculoskeletal system, showing that a pilot can become approximately ten times more stiff or compliant than during relaxed conditions. The parameters in the visual, vestibular and neuromuscular system models are estimated by a combination of model-based system identification techniques in the frequency domain. The novel method provides estimated values for the pilot’s lumped arm dynamics (inertia, viscosity and stiffness) while executing a pitch control task. For this task, the corresponding relative damping and natural frequency are in the same order of magnitude as those estimated in the conventional method.

Pilot equalization in manual control of aircraft dynamics

In continuous manual control tasks, pilots adapt their control strategy to the dynamics of the controlled element to yield adequate performance of the combined pilot-vehicle system. For a controlled element representing the linearized pitch dynamics of a small jet aircraft, the pilot models described in literature were found to lack the required freedom in the pilot equalization term to accurately model the adopted pilot compensation. An additional lead term in the pilot equalization transfer function was found to significantly increase the accuracy in modeling manual control behavior of aircraft pitch dynamics.

Modeling Wide-Frequency-Range Pilot Equalization for Control of Aircraft Pitch Dynamics

In continuous manual control tasks, human controllers adapt their control strategy to the dynamics of the controlled element. This compensation for the controlled-element dynamics is performed around the pilot–vehicle system crossover frequency, in order to obtain satisfactory performance of the combined pilot–vehicle system, but is also seen to extend to frequencies well above crossover. For a controlled element representing the linearized pitch dynamics of a small conventional jet aircraft, an extension to the models for pilot equalization described in the literature was found to be needed for the modeling of the adopted pilot equalization dynamics over a wide frequency range. Measured pilot describing functions revealed that pilots use a combination of low-frequency lag and high-frequency lead equalization to compensate for the characteristics of these typical aircraft pitch dynamics around the short-period mode. An additional high-frequency lead term in the pilot equalization transfer function was found to allow for the modeling of these adopted equalization dynamics over a wide frequency range, thereby also yielding a significant increase in the percentage of measured control inputs that is explained by the pilot model. Furthermore, for this controlled element the extended model for the equalization dynamics was found to be important for the interpretation of the changes in pilot control behavior that occur due to the presence of physical motion feedback.

Investigation into Crossover Regression in Compensatory Manual Tracking Tasks

This paper investigates the crossover-regression phenomenon in compensatory manual-control tasks. The adjustment, between-subject variation, and accuracy of linear human-operator models are analyzed in detail. A theoretical investigation into closed-loop error minimization will be presented. Our main hypothesis was that crossover regression is caused by an operators inability to sufficiently decrease the time delays required to limit forcing-function resonance. To test the hypothesis and explore the use of linear-operator models in regressed conditions, an experiment very similar to McRuers landmark 1965 experiment was conducted. A comparison between regressive and nonregressive conditions revealed that crossover regression is indeed a strategy to reduce forcing-function resonance. The bandwidth of the forcing-function signal at which participants regressed their crossover frequency was found to vary considerably between participants. In regressed conditions, the between-subject variability in frequency-domain performance increased. Additionally, the operator control behavior became increasingly nonlinear, resulting in larger uncertainties and a higher between-subject variability in the linear-model parameter estimates.

Effects of Controlled Element Dynamics on Human Feedforward Behavior in Ramp-Tracking Tasks

In real-life manual control tasks, human controllers are often required to follow a visible and predictable reference signal, enabling them to use feedforward control actions in conjunction with feedback actions that compensate for errors. Little is known about human control behavior in these situations. This paper investigates how humans adapt their feedforward control dynamics to the controlled element dynamics in a combined ramp-tracking and disturbance-rejection task. A human-in-the-loop experiment is performed with a pursuit display and vehicle-like controlled elements, ranging from a single integrator through second-order systems with a break frequency at either 3, 2, or 1 rad/s, to a double integrator. Because the potential benefits of feedforward control increase with steeper ramp segments in the target signal, three steepness levels are tested to investigate their possible effect on feedforward control with the various controlled elements. Analyses with four novel models of the operator, fitted to time-domain data, reveal feedforward control for all tested controlled elements and both (nonzero) tested levels of ramp steepness. For the range of controlled element dynamics investigated, it is found that humans adapt to these dynamics in their feedforward response, with a close to perfect inversion of the controlled element dynamics. No significant effects of ramp steepness on the feedforward model parameters are found.

Envelope Determination of Damaged Aircraft

Many studies have been undertaken to identify the aerodynamic parameters of damaged aircraft and to build controllers to cope with the new parameters and conditions. This paper reflects the results of a different approach. Instead of building a suitable controller the objective has been to define a new safe operating envelope based on the measured change in the aircraft control parameters. Secondary objectives were firstly, that the system should be fast, and stable. In this paper the lateral mode is investigated because this mode is considered the most challenging. Failures evaluated are engine loss, loss of control effectiveness and asymmetric mass distribution. As basis for the parameter identification the two-step method was applied. In the first step the best estimate for the aircraft state is determined. In the second step the equation error method is used to find the stability derivatives. For this second step five different methods were compared: Recursive Least Squares (RLS), Forgetting Algorithm (FA), Kalman Filtering (KF), Extended Kalman Filtering(EKF) and Batch Algorithm (BA). The tests show that most methods were capable of predicting an accurate roll performance but there were significant differences in the stability of the solution.

Identification of the Feedback Component of the Neuromuscular System in a Pitch Control Task

This goal of this study is to understand which parts of the the neuromuscular system contribute during a pitch control task. A novel method developed at the Delft University of Technology allows us to determine the contribution of the neuromuscular feedback system by identifying the admittance, which is the frequency response function of the yielded displacement due to an external force perturbation which applied to control inceptor. In an experiment in a full-motion flight simulator, the neuromuscular admittance was identified during a longitudinal pitch tracking task with a side stick, for two different side stick configurations, an approach configuration with a relatively low stick stiffness, and a cruise configuration with a high stiffness. Besides the admittance, also the muscle activity of eleven muscles was measured. To validate whether the external force perturbation changed the control behavior of the pilot, the visual and vestibular response functions were identified as well. From the measured results it could be concluded that the variations of the control inceptor settings had a significant effect on the neuromuscular feedback system (admittance), although the overall lumped neuromuscular system did not change significantly. A very interesting finding were the very high levels of co-contraction measured during the pitch tracking tasks. And lastly it could be concluded that the required external force perturbation did not affect the control behavior.

An Empirical Human Controller Model for Preview Tracking Tasks

Real-life tracking tasks often show preview information to the human controller about the future track to follow. The effect of preview on manual control behavior is still relatively unknown. This paper proposes a generic operator model for preview tracking, empirically derived from experimental measurements. Conditions included pursuit tracking, i.e., without preview information, and tracking with 1 s of preview. Controlled element dynamics varied between gain, single integrator, and double integrator. The model is derived in the frequency domain, after application of a black-box system identification method based on Fourier coefficients. Parameter estimates are obtained to assess the validity of the model in both the time domain and frequency domain. Measured behavior in all evaluated conditions can be captured with the commonly used quasi-linear operator model for compensatory tracking, extended with two viewpoints of the previewed target. The derived model provides new insights into how human operators use preview information in tracking tasks.