arXiv subjects

Ankit Goel

Publications and source records attributed to Ankit Goel.

49 records · Page 3Linked to original sources

arXiv2023

Experimental Flight Testing of an Adaptive Autopilot with Parameter Drift Mitigation

This paper modifies an adaptive multicopter autopilot to mitigate instabilities caused by adaptive parameter drift and presents simulation and experimental results to validate the modified autopilot. The modified adaptive controller is obtained by including a static nonlinearity in the adaptive loop, updated by the retrospective cost adaptive control algorithm. It is shown in simulation and physical test experiments that the adaptive autopilot with proposed modifications can continually improve the fixed-gain autopilot as well as prevent the drift of the adaptive parameters, thus improving the robustness of the adaptive autopilot.

eess.SY↗

arXiv2022

Euler's Equation via Lagrangian Dynamics with Generalized Coordinates

Euler's equation relates the change in angular momentum of a rigid body to the applied torque. This paper fills a gap in the literature by using Lagrangian dynamics to derive Euler's equation in terms of generalized coordinates. This is done by parameterizing the angular velocity vector in terms of 3-2-1 and 3-1-3 Euler angles as well as Euler parameters, that is, unit quaternions.

math.DS↗

arXiv2022

A Tutorial on Neural Networks and Gradient-free Training

This paper presents a compact, matrix-based representation of neural networks in a self-contained tutorial fashion. Specifically, we develop neural networks as a composition of several vector-valued functions. Although neural networks are well-understood pictorially in terms of interconnected neurons, neural networks are mathematical nonlinear functions constructed by composing several vector-valued functions. Using basic results from linear algebra, we represent a neural network as an alternating sequence of linear maps and scalar nonlinear functions, also known as activation functions. The training of neural networks requires the minimization of a cost function, which in turn requires the computation of a gradient. Using basic multivariable calculus results, the cost gradient is also shown to be a function composed of a sequence of linear maps and nonlinear functions. In addition to the analytical gradient computation, we consider two gradient-free training methods and compare the three training methods in terms of convergence rate and prediction accuracy.

eess.SY↗

arXiv2022

Experimental Flight Testing of a Fault-Tolerant Adaptive Autopilot for Fixed-Wing Aircraft

This paper presents an adaptive autopilot for fixed-wing aircraft and compares its performance with a fixed-gain autopilot. The adaptive autopilot is constructed by augmenting the autopilot architecture with adaptive control laws that are updated using retrospective cost adaptive control. In order to investigate the performance of the adaptive autopilot, the default gains of the fixed-gain autopilot are scaled to degrade its performance. This scenario provides a venue for determining the ability of the adaptive autopilot to compensate for the degraded fixed-gain autopilot. Next, the performance of the adaptive autopilot is examined under failure conditions by simulating a scenario where one of the control surfaces is assumed to be stuck at an unknown angle. The adaptive autopilot is also tested in physical flight experiments under degraded-nominal conditions, and the resulting performance improvement is examined.

eess.SY↗

arXiv2022

On the Accuracy of the One-step UKF and the Two-step UKF

The most accurate version of the unscented Kalman filter (UKF) involves the construction of two ensembles. To reduce computational cost, however, UKF is often implemented without the second ensemble. This simplification comes at a price, however, since, for linear systems, the one-step variation of the two-step UKF does not specialize to the classical Kalman filter, with an associated loss of accuracy. This paper remedies this drawback by developing a modified one-step UKF that recovers the classical Kalman filter for linear systems. Numerical examples show that the modified one-step UKF also recovers the accuracy of the two-step UKF in nonlinear systems with linear outputs.

eess.SY↗

arXiv2022

Retrospective Cost Parameter Estimation with Application to Space Weather Modeling

This chapter reviews standard parameter-estimation techniques and presents a novel gradient-, ensemble-, adjoint-free data-driven parameter estimation technique in the DDDAS framework. This technique, called retrospective cost parameter estimation (RCPE), is motivated by large-scale complex estimation models characterized by high-dimensional nonlinear dynamics, nonlinear parameterizations, and representational models. RCPE is illustrated by estimating unknown parameters in three examples. In the first example, salient features of RCPE are investigated by considering parameter estimation problem in a low-order nonlinear system. In the second example, RCPE is used to estimate the convective coefficient and the viscosity in the generalized Burgers equation by using a scalar measurement. In the final example, RCPE is used to estimate thermal conductivity coefficients that relate temporal temperature variation with the vertical gradient of the temperature in the atmosphere.

eess.SY↗

arXiv2021

An Adaptive Digital Autopilot for Fixed-Wing Aircraft with Actuator Faults

This paper develops an adaptive digital autopilot for a fixed-wing aircraft and compares its performance with a fixed-gain autopilot. The adaptive digital autopilot is constructed by augmenting the autopilot architecture implemented in PX4 flight stack with adaptive digital control laws that are updated using the retrospective cost adaptive control algorithm. In order to investigate the performance of the adaptive digital autopilot, the default gains of the fixed-gain autopilot are scaled down to degrade its performance. This scenario provides a venue for determining the ability of the adaptive digital autopilot to compensate for the detuned fixed-gain autopilot. Next, the performance of the adaptive autopilot is examined under failure conditions by simulating a scenario where one of the control surfaces is assumed to be stuck at an unknown angular position. The adaptive digital autopilot is tested in simulation, and the resulting performance improvements are examined.

eess.SY↗

arXiv2021

An Adaptive PID Autotuner for Multicopters with Experimental Results

This paper develops an adaptive PID autotuner for multicopters, and presents simulation and experimental results. The autotuner consists of adaptive digital control laws based on retrospective cost adaptive control implemented in the PX4 flight stack. A learning trajectory is used to optimize the autopilot during a single flight. The autotuned autopilot is then compared with the default PX4 autopilot by flying a test trajectory constructed using the second-order Hilbert curve. In order to investigate the sensitivity of the autotuner to the quadcopter dynamics, the mass of the quadcopter is varied, and the performance of the autotuned and default autopilot is compared. It is observed that the autotuned autopilot outperforms the default autopilot.

eess.SY↗

arXiv2021

Two Modifications of the Unscented Kalman Filter that Specialize to the Kalman Filter for Linear Systems

Although the unscented Kalman filter (UKF) is applicable to nonlinear systems, it turns out that, for linear systems, UKF does not specialize to the classical Kalman filter. This situation suggests that it may be advantageous to modify UKF in such a way that, for linear systems, the Kalman filter is recovered. The ultimate goal is thus to develop modifications of UKF that specialize to the Kalman filter for linear systems and have improved accuracy for nonlinear systems. With this motivation, this paper presents two modifications of UKF that specialize to the Kalman filter for linear systems. The first modification (EUKF-A) requires the Jacobian of the dynamics map, whereas the second modification (EUKF-C) requires the Jacobian of the measurement map. For various nonlinear examples, the accuracy of EUKF-A and EUKF-C is compared to the accuracy of UKF.

eess.SY↗

arXiv2020

Experimental Implementation of an Adaptive Digital Autopilot

This paper develops an adaptive digital autopilot for quadcopters and presents experimental results. The adaptive digital autopilot is constructed by augmenting the PX4 autopilot control system architecture with adaptive digital control laws based on retrospective cost adaptive control (RCAC). In order to investigate the performance of the adaptive digital autopilot, the default gains of the fixed-gain autopilot are scaled by a small factor, which severely degrades its performance. This scenario thus provides a venue for determining the ability of the adaptive digital autopilot to compensate for the detuned fixed-gain autopilot. The adaptive digital autopilot is tested in simulation and physical flight tests, and the resulting performance improvements are examined.

eess.SY↗

arXiv2020

Adaptive Digital PID Control of a Quadcopter with Unknown Dynamics

This paper develops an adaptive autopilot for quadcopters with unknown dynamics. To do this, the PX4 autopilot architecture is modified so that the feedback and feedforward controllers are replaced by adaptive control laws based on retrospective cost adaptive control (RCAC). The present paper provides a numerical investigation of the performance of the adaptive autopilot on a quadcopter with unknown dynamics. In order to reflect the absence of prior modeling information, all of the adaptive digital controllers are initialized at zero gains. In addition, moment-of-inertia of the quadcopter is varied to test the robustness of the adaptive autopilot. In all test cases, the vehicle is commanded to follow a given trajectory, and the resulting performance is examined.

eess.SY↗

arXiv2020

Recursive Least Squares with Variable-Direction Forgetting -- Compensating for the loss of persistency

Learning depends on the ability to acquire and assimilate new information. This ability depends---somewhat counterintuitively---on the ability to forget. In particular, effective forgetting requires the ability to recognize and utilize new information to order to update a system model. This article is a tutorial on forgetting within the context of recursive least squares (RLS). To do this, RLS is first presented in its classical form, which employs uniform-direction forgetting. Next, examples are given to motivate the need for variable-direction forgetting, especially in cases where the excitation is not persistent. Some of these results are well known, whereas others complement the prior literature. The goal is to provide a self-contained tutorial of the main ideas and techniques for students and researchers whose research may benefit from variable-direction forgetting.

math.OC↗

arXiv2020

Convergence and Consistency of Recursive Least Squares with Variable-Rate Forgetting

A recursive least squares algorithm with variable rate forgetting (VRF) is derived by minimizing a quadratic cost function.Under persistent excitation and boundedness of the forgetting factor, the minimizer given by VRF is shown to converge to the true parameters. In addition, under persistent excitation and with noisy measurements, where the noise is uncorrelated with the regressor, conditions are given under which the minimizer given by VRF is a consistent estimator of the true parameters.The results are illustrated by a numerical example involving abruptly changing parameters.

math.OC↗