We present a new trajectory generation algorithm for autonomous guidance and control of unmanned vehicles from a given starting point to a given target location. We build and update incomplete a priori maps of the operating environment in real time using onboard sensors and compute level sets on the map reflecting the minimal cost of traversal from the current vehicle location to the goal. We convert the trajectory generation problem into a finite-time-horizon optimal control problem using the computed level sets as terminal costs in a receding horizon framework and transform it into a simpler nonlinear programming problem by discretization of the candidate control and state histories. We ensure feasibility of the generated trajectories by constraining the solution of the optimization problem using a simplified vehicle model. We provide strong performance guarantees by checking for stability of the algorithm through the test of matching conditions at the end of each iteration. The algorithm thus explicitly incorporates the vehicle dynamics and constraints and generates trajectories realizable by the vehicle in the field. Successful preliminary field demonstrations and complete simulation results for a marine unmanned surface vehicle demonstrate the efficacy of the proposed approach for fast operations in poorly characterized riverine environments. / Master of Science

On-the-fly laser drilling provides a highly productive method for producing hole clusters (pre-defined groups of holes to be laser drilled) on freeform surfaced parts, such as gas turbine combustion chambers. Although the process is capable of achieving high throughputs, current machine tool controllers are not equipped with appropriate trajectory functions that can take full advantage of the achievable laser drilling speeds. While the problem of contour following has received previous attention in time-optimal trajectory generation literature, on-the-fly laser drilling presents different technological requirements, needing a different kind of trajectory optimization solution, which has not been studied prior to this thesis. The duration between consecutive hole locations, which corresponds to the laser pulsing period, has to be kept constant, ideally throughout the part program. However, the toolpath between the holes is not fixed and can be optimized to enable the shortest possible segment duration. To preserve the dynamic beam positioning accuracy and avoid inducing excessive vibrations on the laser optics, the axis velocity, acceleration, and jerk profiles need to be limited. Furthermore, to ensure that hole elongation does not violate the given part tolerances, the orthogonal component of part velocity relative to the laser beam needs to be capped. All of these requirements have been fulfilled in the trajectory optimization algorithm developed in this thesis. The hole locations are provided as pre-programmed sequences by the Computer Aided Design/Manufacturing software (CAD/CAM). A time-optimized trajectory for each sequence is planned through a series of time-scaling and unconstrained optimization operations, which guarantees a feasible solution. The initial guess for this algorithm is obtained by minimizing the integral square of the fourth time derivative (i.e. ‘snap’). The optimized trajectories for each cluster are then joined together or looped onto themselves (for repeated laser shots) using a time-optimized looping/stitching (optimized/smooth toolpath to repeat/loop a cluster or connect/stitch between consecutive clusters) algorithm. This algorithm also minimizes the integral square of jerk in the faster axes. The effectiveness of the overall solution has been demonstrated in simulations and preliminary experimental results for on-the-fly laser drilling of a hole pattern for a gas turbine combustion chamber panel. It is shown that the developed algorithm improves the cycle time for a single pass by at least 6% (from kinematic analysis of the motion duration), and more importantly reduces the integral square of jerk by 56%, which would enable the process speed to be pushed up further.

On-the-fly laser drilling provides a highly productive method for producing hole clusters (pre-defined groups of holes to be laser drilled) on freeform surfaced parts, such as gas turbine combustion chambers. Although the process is capable of achieving high throughputs, current machine tool controllers are not equipped with appropriate trajectory functions that can take full advantage of the achievable laser drilling speeds. While the problem of contour following has received previous attention in time-optimal trajectory generation literature, on-the-fly laser drilling presents different technological requirements, needing a different kind of trajectory optimization solution, which has not been studied prior to this thesis. The duration between consecutive hole locations, which corresponds to the laser pulsing period, has to be kept constant, ideally throughout the part program. However, the toolpath between the holes is not fixed and can be optimized to enable the shortest possible segment duration. To preserve the dynamic beam positioning accuracy and avoid inducing excessive vibrations on the laser optics, the axis velocity, acceleration, and jerk profiles need to be limited. Furthermore, to ensure that hole elongation does not violate the given part tolerances, the orthogonal component of part velocity relative to the laser beam needs to be capped. All of these requirements have been fulfilled in the trajectory optimization algorithm developed in this thesis. The hole locations are provided as pre-programmed sequences by the Computer Aided Design/Manufacturing software (CAD/CAM). A time-optimized trajectory for each sequence is planned through a series of time-scaling and unconstrained optimization operations, which guarantees a feasible solution. The initial guess for this algorithm is obtained by minimizing the integral square of the fourth time derivative (i.e. ‘snap’). The optimized trajectories for each cluster are then joined together or looped onto themselves (for repeated laser shots) using a time-optimized looping/stitching (optimized/smooth toolpath to repeat/loop a cluster or connect/stitch between consecutive clusters) algorithm. This algorithm also minimizes the integral square of jerk in the faster axes. The effectiveness of the overall solution has been demonstrated in simulations and preliminary experimental results for on-the-fly laser drilling of a hole pattern for a gas turbine combustion chamber panel. It is shown that the developed algorithm improves the cycle time for a single pass by at least 6% (from kinematic analysis of the motion duration), and more importantly reduces the integral square of jerk by 56%, which would enable the process speed to be pushed up further.

La marche est un mécanisme complexe impliquant l’élaboration de trajectoires dans des milieux divers et variés et la réalisation des mouvements segmentaires qui permettent de les parcourir. Elle est alors dépendante de l’environnement, des obstacles et autres individus qui le peuplent mais également des capacités physiques du corps humain. De part cette complexité, l’étude de la marche est généralement compartimentée en deux niveaux : la génération de trajectoires locomotrices d’une part et les mouvements des membres d’autre part. Ce travail vise à proposer un processus complet d’analyse de la marche, en s’intéressant au lien unissant les trajectoires locomotrices aux mouvements segmentaires. Dans un premier temps nous nous intéressons à la génération de trajectoires. Plus particulièrement, nous nous focalisons sur une situation de croisement entre deux piétons et sur les stratégies mises en place pour éviter la collision. Puis, nous portons notre attention sur la manière dont les endroits de pose de pieds sont influencés par la trajectoire. Cette analyse nous conduit à proposer un modèle de génération d’empreintes de pas. Enfin, nous nous intéressons à la générations des trajectoires articulaires menant au mouvement de marche. Par l’utilisation de la méthode de linéarisation locale nous proposons une nouvelle méthodologie pour la simulation de la marche à partir d’une entrée unique : la prochaine empreinte pas / Walking is a complex mecanism involving trajectories generation in various environments and motion generation in order to follow the path. Then, it is dependent on environment, obstacles and peoples moving around but also on body capabilities. This complexity lead scientits to split walking analysis in two levels : trajectory generation in one hand, and motion generation in the othe hand. This work aim to provide a global walking analysis processus by linking trajectorires and motion generation. First, we explore walking trajectories throw a particular situation : pedestrians crossing. Here we take interest in trajectories and speed adaptations. Then, we sink for the link between trajectory and heelstrike. It lead us to develop a model for heelstrike generation based on trajectory. Finally, we take interest in walking motion simulation. By the use of local linearization, we provide a new methodology for joints joints angles generation

The inspection of power lines and associated hardware is vital to ensuring the reliability of the transmission and distribution network. The repetitive nature of the inspection tasks present a unique opportunity for the introduction of robotic platforms, which offer the ability to perform more systematic and detailed inspection than traditional methods. This lends itself to improved asset management automation, cost-effectiveness and safety for the operating crew. This dissertation presents the development of a prototype industrial brachiating robot. The robot is mechanically simple and capable of dynamically negotiating obstacles by brachiating. This is an improvement over current robotic platforms, which employ slow, high power static schemes for obstacle negotiation. Mathematical models of the robot were derived to understand the underlying dynamics of the system. These models were then used in the generation of optimal trajectories, using nonlinear optimisation techniques, for brachiating past line hardware. A physical robot was designed and manufactured to validate the brachiation manoeuvre. The robot was designed following classic mechanical design principles, with emphasis on functional design and robustness. System identification was used to capture the plant uncertainty and a feedback controller was designed to track the reference trajectory allowing for energy optimal brachiation swings. Finally, the robot was tested, starting with sub-system testing and ending with testing of a brachiation manoeuvre proving the prospective viability of the robot in an industrial environment.

Cette thèse s'intéresse à la réalisation autonome de plans de vol cinématographiques par un quadrotor équipé d'une caméra. Ces plans de vol consistent en une série de points de passage à rejoindre successivement, en adoptant diverses méthodes de prise de vue et en respectant des références de vitesses ainsi que des couloirs de vols. Une étude approfondie de la dynamique du quadrotor est tout d'abord proposée et utilisée pour construire un modèle linéarisé du drone autour de l'équilibre de vol stationnaire. L'analyse de ce modèle linéaire permet de mettre en évidence l'impact de l'inertie des rotors du drone dans sa dynamique, notamment l'apparition d'un comportement à non minimum de phase en roulis ou tangage, lorsque les moteurs sont inclinés. Dans un second temps, deux algorithmes de génération de trajectoires lisses, faisables et adaptées à la cinématographie sont proposés. La faisabilité de la trajectoire est garantie par le respect de contraintes sur ses dérivées temporelle, adaptées pour la cinématographie et obtenue grâce à l'étude du modèle non linéaire du drone. Le premier repose sur une optimisation bi-niveaux d'une trajectoire polynomiale par morceaux, dans le but de trouver la plus rapide des trajectoires à minimum de jerk permettant d'accomplir la mission. Le second algorithme consiste en la génération de trajectoires B-spline non-uniformes à durée minimale. Pour les deux solutions, une étude de l’initialisation du problème d'optimisation est présentée, de même qu'une analyse de leurs avantages et limitations. Pour ce faire, elles sont notamment confrontées à des simulations et vols extérieurs. Enfin, une loi de commande prédictive est proposée pour asservir les mouvements de la caméra embarquée de manière douce et précise. / This thesis focuses on the autonomous performance of cinematographic flight plans by camera equipped quadrotors. These flight plans consist in a series of waypoints to join while adopting various camera behaviors, along with speed references and flight corridors. First, an in depth study of the nonlinear dynamics of the drone is proposed, which is then used to derive a linear model of the system around the hovering equilibrium. An analysis of this linear model allows us to emphasize the impact of the inertia of the propellers when the latter are tilted, such as the apparition of a nonminimum phase behavior of the pitch or roll dynamics. Then, two algorithms are proposed to generate smooth and feasible cinematographic trajectories. The feasibility of the trajectory is ensured by constraints on its time derivatives, suited for cinematography and obtained with the use of the nonlinear model of the drone. The first algorithm proposed in this work is based on a bi-level optimization of a piecewise polynomial trajectory, in order to find the fastest feasible minimum jerk trajectory to perform the flight plan. The second algorithm consists in the generation of feasible, minimum time, non-uniform B-spline trajectories. For both solutions, a study of the initilization of the optimization problem is proposed, as well as a discussion about their advantages and limitations. To this aim, they are notably confronted to simulations and outdoor flight experiments. Finally, a predictive control law is proposed to smoothly and accurately control the onboard camera.

abstract: Control algorithm development for quadrotor is usually based solely on rigid body dynamics neglecting aerodynamics. Recent work has demonstrated that such a model is suited only when operating at or near hover conditions and low-speed flight. When operating in confined spaces or during aggressive maneuvers destabilizing forces and moments are induced due to aerodynamic effects. Studies indicate that blade flapping, induced drag, and propeller drag influence forward flight performance while other effects like vortex ring state, ground effect affect vertical flight performance. In this thesis, an offboard data-driven approach is used to derive models for parasitic (bare-airframe) drag and propeller drag. Moreover, thrust and torque coefficients are identified from static bench tests. Among the two, parasitic drag is compensated for in the position controller module in the PX4 firmware. 2-D circular, straight line, and minimum snap rectangular trajectories with corridor constraints are tested exploiting differential flatness property wherein altitude and yaw angle are constant. Flight tests are conducted at ASU Drone Studio and results of tracking performance with default controller and with drag compensated position controller are presented. Root mean squared tracking error in individual axes is used as a metric to evaluate the model performance. Results indicate that, for circular trajectory, the root mean squared error in the x-axis has reduced by 44.54% and in the y-axis by 39.47%. Compensation in turn degrades the tracking in both axis by a maximum under 12% when compared to the default controller for rectangular trajectory case. The x-axis tracking error for the straight-line case has improved by 44.96% with almost no observable change in the y-axis. / Dissertation/Thesis / Real-time Flight Test of Circular Trajectories / Masters Thesis Aerospace Engineering 2020

This thesis deals with the general problem of robot motion planning and control. It proposes the hypothesis that it should bepossible to create a generic software framework capable of dealing with all robot motion planning and control problems, independent of the robot being used, the task being solved, the workspace obstacles or the algorithms employed. The thesis work then consisted of identifying the requirements and creating a design and implementation of such a framework. This report motivates and documents the entire process. The framework developed was tested on two different robot arms under varying conditions. The testing method and results are also presented.The thesis concludes that the proposed hypothesis is indeed valid.

This dissertation addresses motion planning, modeling, and feedback control for autonomous vehicle systems. A hierarchical approach for motion planning and control of nonlinear systems operating in obstacle environments is presented. To reduce computation time during the motion planning process, dynamically feasible trajectories are generated in real-time through concatenation of pre-specified motion primitives. The motion planning task is posed as a search over a directed graph, and the applicability of informed graph search techniques is investigated. Specifically, a locally greedy algorithm with effective backtracking ability is developed and compared to weighted A* search. The greedy algorithm shows an advantage with respect to solution cost and computation time when larger motion primitive libraries that do not operate on a regular state lattice are utilized. Linearization of the nonlinear system equations about the motion primitive library results in a hybrid linear time-varying model, and an optimal control algorithm using the L2-induced norm as the performance measure is applied to ensure that the system tracks the desired trajectory. The ability of the resulting controller to closely track the trajectory obtained from the motion planner, despite various disturbances and uncertainties, is demonstrated through simulation. Additionally, an approach for obtaining dynamically feasible reference trajectories and feedback controllers for a small unmanned aerial vehicle (UAV) based on an aerodynamic model derived from flight tests is presented. The modeling approach utilizes the two step method (TSM) with stepwise multiple regression to determine relevant explanatory terms for the aerodynamic models. Dynamically feasible trajectories are then obtained through the solution of an optimal control problem using pseudospectral optimal control software. Discrete-time feedback controllers are then obtained to regulate the vehicle along the desired reference trajectory. Simulations in a realistic operational environment as well as flight testing with the feedback controller demonstrate the capabilities of the approach. The TSM is also applied for system identification of an aircraft using motion capture data. In this application, time domain system identification techniques are used to identify both linear and nonlinear aerodynamic models of large-amplitude pitching motions driven by control surface deflections. The resulting models are assessed based on both their predictive capabilities as well as simulation results. / Ph. D.

No description available.