Browsing by Author "Galvão Wall, David"

Now showing 1 - 7 of 7
  • Thumbnail Image
    ItemOpen Access
    A graph-theory-based C-space path planner for mobile robotic manipulators in close-proximity environments
    (2016-08-10) Galvão Wall, David; Economou, John T.
    In this thesis a novel guidance method for a 3-degree-of-freedom robotic manipulator arm in 3 dimensions for Improvised Explosive Device (IED) disposal has been developed. The work carried out in this thesis combines existing methods to develop a technique that delivers advantages taken from several other guidance techniques. These features are necessary for the IED disposal application. The work carried out in this thesis includes kinematic and dynamic modelling of robotic manipulators, T-space to C-space conversion, and path generation using Graph Theory to produce a guidance technique which can plan a safe path through a complex unknown environment. The method improves upon advantages given by other techniques in that it produces a suitable path in 3-dimensions in close-proximity environments in real time with no a priori knowledge of the environment, a necessary precursor to the application of this technique to IED disposal missions. To solve the problem of path planning, the thesis derives the kinematics and dynamics of a robotic arm in order to convert the Euclidean coordinates of measured environment data into C-space. Each dimension in C-space is one control input of the arm. The Euclidean start and end locations of the manipulator end effector are translated into C-space. A three-dimensional path is generated between them using Dijkstra’s Algorithm. The technique allows for a single path to be generated to guide the entire arm through the environment, rather than multiple paths to guide each component through the environment. The robotic arm parameters are modelled as a quasi-linear parameter varying system. As such it requires gain scheduling control, thus allowing compensation of the non-linearities in the system. A Genetic Algorithm is applied to tune a set of PID controllers for the dynamic model of the manipulator arm so that the generated path can then be followed using a conventional path-following algorithm. The technique proposed in this thesis is validated using numerical simulations in order to determine its advantages and limitations.
  • Thumbnail Image
    ItemOpen Access
    Hypersurface normalised gain-scheduled controller for a non-linear 6-DOF fast jet
    (Elsevier, 2020-07-25) Hamilton, Jordan; Galvão Wall, David; Saddington, Alistair J.; Economou, John T.
    This paper describes a novel approach for improving the dynamic response of a bank-to-turn autopilot for a non-linear six degree-of-freedom (6-DoF) aircraft model. The autopilot consists of a series of gain-scheduled (GS) proportional, integral and derivative (PID) controllers that govern the aircraft's angular velocities for roll, pitch and yaw. The controller gains have been optimised for localised trim points and applied continuously to the controllers using linear interpolation to form a hypersurface. Our novel solution has been achieved by implementing a set of scheduled gains for near-zero reference signals and integrating this with a set of gains that are normalised to the reference signal. The proposed approach has been compared to conventional gain scheduling techniques using a series of step input simulated manoeuvres, applied individually to the roll and pitch controllers. The results show improved rise and fall times, steady state errors, as well as reduced controller effort
  • Thumbnail Image
    ItemOpen Access
    H∞/LQR optimal control for a supersonic air-breathing missile of asymmetric configuration
    (Elsevier, 2019-11-25) Vincent, Raymond Vin; Economou, John T.; Galvão Wall, David; Cleminson, John
    Robust control is challenging to achieve for air-breathing missiles operating in a high Mach number regime, such as at high supersonic speeds (M > 3). The challenge arises because of strong couplings, significant non-linearities and large uncertainties in the aerodynamics and propulsion system. The feasibility of achieving robust control in such applications is strongly linked to the development of an appropriate control design structure. The purpose of this paper is to illustrate that in order to stabilise a highly unstable airframe and achieve the required performance, a hybrid of two control schemes may be used to achieve best results. A state feedback linear quadratic regulator is used to stabilise the plant and a forward path H∞ optimal controller is used to achieve the required performance and robustness. We also highlight the complementary attributes of the two control schemes that together can generate a more robust controller; LQR is used since it can achieve good gain and phase margins, whereas, the H∞ control method is better equipped to deal with uncertainties.
  • Thumbnail Image
    ItemOpen Access
    Indirect tuning of a complementary orientation filter using velocity data and a genetic algorithm
    (Taylor and Francis, 2024-04-23) Maton, Dariusz; Economou, John T.; Galvão Wall, David; Khan, Irfan; Cooper, Robert; Ward, David; Trythall, Simon
    In this paper, the accuracy of inertial sensor orientation relative to the level frame is improved through optimal tuning of a complementary filter by a genetic algorithm. While constant filter gains have been used elsewhere, these may introduce errors under dynamic motions when gyroscopes should be trusted more than accelerometers. Optimal gains are prescribed by a Mamdani fuzzy rule base whose membership functions are found using a genetic algorithm and experimental data. Furthermore, model fitness is not based directly on orientation but the error between estimated and ground truth velocities. This paper has three interrelated novel elements. The main novelty is the indirect tuning method, which is simple, low-cost and requires a single camera and inertial sensor. The method is shown to increase tracking accuracy compared with popular baseline filters. Secondary novel elements are the bespoke genetic algorithm and the time agnostic velocity error metric. The contributions from this work can help improve the localization accuracy of assets and human personnel. This research has a direct impact in command and control by improving situational awareness and the ability to direct assets to safe locations using safer routes. This results in increasing safety in applications such as firefighting and battlespace.
  • Thumbnail Image
    ItemOpen Access
    Intelligent based terrain preview controller for a 3-axle vehicle
    (AVEC 16, 2016-09-16) Economou, John T.; Purdy, David J.; Galvão Wall, David; Diskett, D.; Simner, D.
    The paper presents a six-wheel half longitudinal model and the design of a dual level control architecture. The first (top) level is designed using a Sugeno fuzzy inference feedforward architecture with and without preview. The second level of controllers are locally managing each wheel for each axle. As the vehicle is moving forward the front wheels and suspension units will have less time to respond when compared to the middle and rear units, hence a preview sensor is used to compensate. The paper shows that the local active suspensions together with the Sugeno Fuzzy, (locally optimised using subtractive clustering), Feedforward control strategy is more effective and this architecture has resulted in reducing the sprung mass vertical acceleration and pitch accelerations.
  • Thumbnail Image
    ItemOpen Access
    Quasi-real-time confined environment path generation for mobile robotic manipulator arms
    (SAGE, 2018-01-10) Galvão Wall, David; Economou, John T.; Knowles, Kevin
    Path generation for mobile robotic manipulator arms is challenging in dynamic environments because high-speed calculations are required to deal with fast-moving obstacles. A novel path-planning algorithm has been developed which solves in quasi-real time the problem of path generation in confined environments for interconnected multi-body systems, specifically a robotic manipulator arm with three links. The work presented in this article builds upon the previous work by reformulating the technique to increase the speed at which the algorithm is able to calculate a safe path. The complexity of the task space has increased substantially compared to previous work, and the algorithm has been reformulated to speed up the calculation in order to maintain or even improve its ability to plan a safe path in real time. The method is now able to calculate a safe path through environments significantly more quickly than the previous method, and the results presented in this article expand the complexity of the environment by a large amount and test the ability of the reformulated algorithm to still operate in real time, which the method achieves. It was found that the reformulated method reduces the calculation time for path generation exponentially when used to plan safe paths through test environments involving different numbers of obstacles. The new algorithm thus has the potential to facilitate path planning in challenging dynamic environments, such as those used in sensitive manufacturing and maintenance tasks as well as bomb disposal and similar applications.
  • Thumbnail Image
    ItemOpen Access
    Subtractive clustering Takagi-Sugeno position tracking for humans by low-cost inertial sensors and velocity classification
    (Sage, 2023-04-12) Maton, Dariusz; Economou, John T.; Galvão Wall, David; Ward, David; Trythall, Simon
    In this work, open-loop position tracking using low-cost inertial measurement units is aided by Takagi-Sugeno velocity classification using the subtractive clustering algorithm to help generate the fuzzy rule base. Using the grid search approach, a suitable window of classified velocity vectors was obtained and then integrated to generate trajectory segments. Using publicly available experimental data, the reconstruction accuracy of the method is compared against four competitive pedestrian tracking algorithms. The comparison on selected test data, has demonstrated more competitive relative and absolute trajectory error metrics. The proposed method in this paper is also verified on an independent experimental data set. Unlike the methods which use deep learning, the proposed method has shown to be transparent (fuzzy rule base). Lastly, a sensitivity analysis of the velocity classification models to perturbations from the training orientation at test time is investigated, to guide developers of such data-driven algorithms on the granularity required in an ensemble modelling approach. The accuracy and transparency of the approach may positively influence applications requiring low-cost inertial position tracking such as augmented reality headsets for emergency responders.