-
A Multi-step Dynamics Modeling Framework For Autonomous Driving In Multiple Environments
Authors:
Jason Gibson,
Bogdan Vlahov,
David Fan,
Patrick Spieler,
Daniel Pastor,
Ali-akbar Agha-mohammadi,
Evangelos A. Theodorou
Abstract:
Modeling dynamics is often the first step to making a vehicle autonomous. While on-road autonomous vehicles have been extensively studied, off-road vehicles pose many challenging modeling problems. An off-road vehicle encounters highly complex and difficult-to-model terrain/vehicle interactions, as well as having complex vehicle dynamics of its own. These complexities can create challenges for eff…
▽ More
Modeling dynamics is often the first step to making a vehicle autonomous. While on-road autonomous vehicles have been extensively studied, off-road vehicles pose many challenging modeling problems. An off-road vehicle encounters highly complex and difficult-to-model terrain/vehicle interactions, as well as having complex vehicle dynamics of its own. These complexities can create challenges for effective high-speed control and planning. In this paper, we introduce a framework for multistep dynamics prediction that explicitly handles the accumulation of modeling error and remains scalable for sampling-based controllers. Our method uses a specially-initialized Long Short-Term Memory (LSTM) over a limited time horizon as the learned component in a hybrid model to predict the dynamics of a 4-person seating all-terrain vehicle (Polaris S4 1000 RZR) in two distinct environments. By only having the LSTM predict over a fixed time horizon, we negate the need for long term stability that is often a challenge when training recurrent neural networks. Our framework is flexible as it only requires odometry information for labels. Through extensive experimentation, we show that our method is able to predict millions of possible trajectories in real-time, with a time horizon of five seconds in challenging off road driving scenarios.
△ Less
Submitted 3 May, 2023;
originally announced May 2023.
-
Parametric Rectified Power Sigmoid Units: Learning Nonlinear Neural Transfer Analytical Forms
Authors:
Abdourrahmane Mahamane Atto,
Sylvie Galichet,
Dominique Pastor,
Nicolas Méger
Abstract:
The paper proposes representation functionals in a dual paradigm where learning jointly concerns both linear convolutional weights and parametric forms of nonlinear activation functions. The nonlinear forms proposed for performing the functional representation are associated with a new class of parametric neural transfer functions called rectified power sigmoid units. This class is constructed to…
▽ More
The paper proposes representation functionals in a dual paradigm where learning jointly concerns both linear convolutional weights and parametric forms of nonlinear activation functions. The nonlinear forms proposed for performing the functional representation are associated with a new class of parametric neural transfer functions called rectified power sigmoid units. This class is constructed to integrate both advantages of sigmoid and rectified linear unit functions, in addition with rejecting the drawbacks of these functions. Moreover, the analytic form of this new neural class involves scale, shift and shape parameters so as to obtain a wide range of activation shapes, including the standard rectified linear unit as a limit case. Parameters of this neural transfer class are considered as learnable for the sake of discovering the complex shapes that can contribute in solving machine learning issues. Performance achieved by the joint learning of convolutional and rectified power sigmoid learnable parameters are shown outstanding in both shallow and deep learning frameworks. This class opens new prospects with respect to machine learning in the sense that learnable parameters are not only attached to linear transformations, but also to suitable nonlinear operators.
△ Less
Submitted 5 February, 2021; v1 submitted 25 January, 2021;
originally announced January 2021.
-
Episodic Koopman Learning of Nonlinear Robot Dynamics with Application to Fast Multirotor Landing
Authors:
Carl Folkestad,
Daniel Pastor,
Joel W. Burdick
Abstract:
This paper presents a novel episodic method to learn a robot's nonlinear dynamics model and an increasingly optimal control sequence for a set of tasks. The method is based on the {\em Koopman operator} approach to nonlinear dynamical systems analysis, which models the flow of {\em observables} in a function space, rather than a flow in a state space. Practically, this method estimates a nonlinear…
▽ More
This paper presents a novel episodic method to learn a robot's nonlinear dynamics model and an increasingly optimal control sequence for a set of tasks. The method is based on the {\em Koopman operator} approach to nonlinear dynamical systems analysis, which models the flow of {\em observables} in a function space, rather than a flow in a state space. Practically, this method estimates a nonlinear diffeomorphism that lifts the dynamics to a higher dimensional space where they are linear. Efficient Model Predictive Control methods can then be applied to the lifted model. This approach allows for real time implementation in on-board hardware, with rigorous incorporation of both input and state constraints during learning. We demonstrate the method in a real-time implementation of fast multirotor landing, where the nonlinear ground effect is learned and used to improve landing speed and quality.
△ Less
Submitted 3 April, 2020;
originally announced April 2020.
-
Design and Autonomous Stabilization of a Ballistically Launched Multirotor
Authors:
Amanda Bouman,
Paul Nadan,
Matthew Anderson,
Daniel Pastor,
Jacob Izraelevitz,
Joel Burdick,
Brett Kennedy
Abstract:
Aircraft that can launch ballistically and convert to autonomous, free flying drones have applications in many areas such as emergency response, defense, and space exploration, where they can gather critical situational data using onboard sensors. This paper presents a ballistically launched, autonomously stabilizing multirotor prototype (SQUID, Streamlined Quick Unfolding Investigation Drone) wit…
▽ More
Aircraft that can launch ballistically and convert to autonomous, free flying drones have applications in many areas such as emergency response, defense, and space exploration, where they can gather critical situational data using onboard sensors. This paper presents a ballistically launched, autonomously stabilizing multirotor prototype (SQUID, Streamlined Quick Unfolding Investigation Drone) with an onboard sensor suite, autonomy pipeline, and passive aerodynamic stability. We demonstrate autonomous transition from passive to vision based, active stabilization, confirming the ability of the multirotor to autonomously stabilize after a ballistic launch in a GPS denied environment.
△ Less
Submitted 17 May, 2020; v1 submitted 22 November, 2019;
originally announced November 2019.
-
Extended Dynamic Mode Decomposition with Learned Koopman Eigenfunctions for Prediction and Control
Authors:
Carl Folkestad,
Daniel Pastor,
Igor Mezic,
Ryan Mohr,
Maria Fonoberova,
Joel Burdick
Abstract:
This paper presents a novel learning framework to construct Koopman eigenfunctions for unknown, nonlinear dynamics using data gathered from experiments. The learning framework can extract spectral information from the full nonlinear dynamics by learning the eigenvalues and eigenfunctions of the associated Koopman operator. We then exploit the learned Koopman eigenfunctions to learn a lifted linear…
▽ More
This paper presents a novel learning framework to construct Koopman eigenfunctions for unknown, nonlinear dynamics using data gathered from experiments. The learning framework can extract spectral information from the full nonlinear dynamics by learning the eigenvalues and eigenfunctions of the associated Koopman operator. We then exploit the learned Koopman eigenfunctions to learn a lifted linear state-space model. To the best of our knowledge, our method is the first to utilize Koopman eigenfunctions as lifting functions for EDMD-based methods. We demonstrate the performance of the framework in state prediction and closed loop trajectory tracking of a simulated cart pole system. Our method is able to significantly improve the controller performance while relying on linear control methods to do nonlinear control.
△ Less
Submitted 18 March, 2020; v1 submitted 20 November, 2019;
originally announced November 2019.
-
Design of a Ballistically-Launched Foldable Multirotor
Authors:
Daniel Pastor,
Jacob Izraelevitz,
Paul Nadan,
Amanda Bouman,
Joel Burdick,
Brett Kennedy
Abstract:
The operation of multirotors in crowded environments requires a highly reliable takeoff method, as failures during takeoff can damage more valuable assets nearby. The addition of a ballistic launch system imposes a deterministic path for the multirotor to prevent collisions with its environment, as well as increases the multirotor's range of operation and allows deployment from an unsteady platfor…
▽ More
The operation of multirotors in crowded environments requires a highly reliable takeoff method, as failures during takeoff can damage more valuable assets nearby. The addition of a ballistic launch system imposes a deterministic path for the multirotor to prevent collisions with its environment, as well as increases the multirotor's range of operation and allows deployment from an unsteady platform. In addition, outfitting planetary rovers or entry vehicles with such deployable multirotors has the potential to greatly extend the data collection capabilities of a mission. A proof-of-concept multirotor aircraft has been developed, capable of transitioning from a ballistic launch configuration to a fully controllable flight configuration in midair after launch. The transition is accomplished via passive unfolding of the multirotor arms, triggered by a nichrome burn wire release mechanism. The design is 3D printable, launches from a three-inch diameter barrel, and has sufficient thrust to carry a significant payload. The system has been fabricated and field tested from a moving vehicle up to 50mph to successfully demonstrate the feasibility of the concept and experimentally validate the design's aerodynamic stability and deployment reliability.
△ Less
Submitted 13 November, 2019;
originally announced November 2019.
-
Towards a characterization of the uncertainty curve for graphs
Authors:
Bastien Pasdeloup,
Vincent Gripon,
Grégoire Mercier,
Dominique Pastor
Abstract:
Signal processing on graphs is a recent research domain that aims at generalizing classical tools in signal processing, in order to analyze signals evolving on complex domains. Such domains are represented by graphs, for which one can compute a particular matrix, called the normalized Laplacian. It was shown that the eigenvalues of this Laplacian correspond to the frequencies of the Fourier domain…
▽ More
Signal processing on graphs is a recent research domain that aims at generalizing classical tools in signal processing, in order to analyze signals evolving on complex domains. Such domains are represented by graphs, for which one can compute a particular matrix, called the normalized Laplacian. It was shown that the eigenvalues of this Laplacian correspond to the frequencies of the Fourier domain in classical signal processing. Therefore, the frequency domain is not the same for every support graph. A consequence of this is that there is no non-trivial generalization of Heisenberg's uncertainty principle, that states that a signal cannot be fully localized both in the time domain and in the frequency domain. A way to generalize this principle, introduced by Agaskar and Lu, consists in determining a curve that represents a lower bound on the compromise between precision in the graph domain and precision in the spectral domain. The aim of this paper is to propose a characterization of the signals achieving this curve, for a larger class of graphs than the one studied by Agaskar and Lu.
△ Less
Submitted 27 April, 2016;
originally announced May 2016.
