Robert Sutton

Robust Adaptive Control of an Uninhabited Surface Vehicle

A robust adaptive autopilot for uninhabited surface vehicles (USV) based on a model predictive controller (MPC) is presented in this paper. The novel autopilot is capable of handling sudden changes in system dynamics. In real life situations, very often a sudden change in dynamics results in missions being aborted and the uninhabited vehicles have to be rescued before they cause damage to other marine craft in the vicinity. This problem has been suitably dealt with by this innovative design. The MPC adopts an online adaptive nature by utilising three algorithms, individually: gradient descent, least squares and weighted least squares (WLS). Even with random initialisation, significant improvements over the other algorithmic approach were achieved by WLS by maintaining the intermittent continuous values of system parameters and periodically reinitialising the covariance matrix. Also, a time frame of 25 seconds appears to be the optimum to reinitialise the parameters in simulation studies. This novel approach enables the autopilot to cope well with significant changes in the system dynamics and empowers USVs to accomplish their desired missions.

Non-linear control algorithms for an unmanned surface vehicle

Although intrinsically marine craft are known to exhibit non-linear dynamic characteristics, modern marine autopilot system designs continue to be developed based on both linear and non-linear control approaches. This article evaluates two novel non-linear autopilot designs based on non-linear local control network and non-linear model predictive control approaches to establish their effectiveness in terms of control activity expenditure, power consumption and mission duration length under similar operating conditions. From practical point of view, autopilot with less energy consumption would in reality provide the battery-powered vehicle with longer mission duration. The autopilot systems are used to control the non-linear yaw dynamics of an unmanned surface vehicle named Springer. The yaw dynamics of the vehicle being modelled using a multi-layer perceptron-type neural network. Simulation results showed that the autopilot based on local control network method performed better for Springer. Furthermore, on the whole, the local control network methodology can be regarded as a plausible paradigm for marine control system design.

An Autopilot Based on a Local Control Network Design for an Unmanned Surface Vehicle

Over recent years, a number of marine autopilots designed using linear techniques have underperformed owing to their inability to cope with nonlinear vessel dynamics. To this end, a new design framework for the development of nonlinear autopilots is proposed. Local Control Networks (LCNs) can be used in the design of nonlinear control systems. In this paper, a LCN approach is taken in the design of a nonlinear autopilot for controlling the nonlinear yaw dynamics of an unmanned surface vehicle known as Springer. It is considered the approach is the first of its kind to be used in marine control systems design. Simulation results are presented and the performance of the nonlinear autopilot is compared with that of an existing Springer Linear Quadratic Gaussian (LQG) autopilot using standard system performance criteria. From the results it can be concluded the LCN autopilot out-performed that based on LQG techniques in terms of the selected criteria. Also it provided more energy saving control strategies and would thereby increase operational duration times for the vehicle during real-time missions.

Adaptive navigation systems for an unmanned surface vehicle

This paper reports the design of two potential navigation systems for use in an unmanned surface vehicle (USV) named Springer. The approaches adopted are based on fuzzy multisensor data fusion (MSDF) and multiple model adaptive estimation (MMAE) algorithms with adaptive capabilities. A general description of the Springer USV is given along with details of its navigation sensor suite. Of particular interest are the three different types of electronic compass used to supply heading information. Using a system identification technique, state space models of the compasses are derived for use in a simulation study to compare the navigation systems. From the results presented, it is concluded the fuzzy MSDF algorithm is better than the MMAE methodology in terms of heading (yaw error) accuracy and robustness.

Interval Kalman filtering in navigation system design for an uninhabited surface vehicle

This paper reports on the potential application of interval Kalman filtering techniques in the design of a navigation system for an uninhabited surface vehicle named Springer. The interval Kalman filter (IKF) is investigated for this task since it has had limited exposure for such usage. A state-space model of the Springer steering dynamics is used to provide a framework for the application of the Kalman filter (KF) and IKF algorithms for estimating the heading angle of the vessel under erroneous modelling assumptions. Simulations reveal several characteristics of the IKF, which are then discussed, and a review of the work undertaken to date presented and explained in the light of these characteristics, with suggestions on potential future improvements.

Application of artificial neural networks to weighted interval Kalman filtering

The interval Kalman filter is a variant of the traditional Kalman filter for systems with bounded parametric uncertainty. For such systems, modelled in terms of intervals, the interval Kalman filter provides estimates of the system state also in the form of intervals, guaranteed to contain the Kalman filter estimates of all point-valued systems contained in the interval model. However, for practical purposes, a single, point-valued estimate of the system state is often required. This point value can be seen as a weighted average of the interval bounds provided by the interval Kalman filter. This article proposes a methodology based on the application of artificial neural networks by which an adequate weight can be computed at each time step, whereby the weighted average of the interval bounds approximates the optimal estimate or estimate which would be obtained using a Kalman filter if no parametric uncertainty was present in the system model, even when this is not the case. The practical applicability and robustness of the method are demonstrated through its application to the navigation of an uninhabited surface vehicle.

Neural network fault diagnosis of a trolling motor based on feature reduction techniques for an unmanned surface vehicle

This article presents a novel approach to the diagnosis of unbalanced faults in a trolling motor under stationary operating conditions. The trolling motor being typically of that used as the propulsion system for an unmanned surface vehicle, the diagnosis approach is based on the use of discrete wavelet transforms as a feature extraction tool and a time-delayed neural network for fault classification. The time-delayed neural network classifies between healthy and faulty conditions of the trolling motor by analysing the stator current and vibration. To overcome feature redundancy, which affects diagnosis accuracy, several feature reduction methods have been tested, and the orthogonal fuzzy neighbourhood discriminant analysis approach is found to be the most effective method. Four faulty conditions were analysed under laboratory conditions, where one of the blades causing damage to the trolling motor is cut into 10%, 25%, half and then into full to simulate the effects of propeller blades being damaged partly or fully. The results obtained from the real-time simulation demonstrate the effectiveness and reliability of the proposed methodology in classifying the different faults faster and accurately.

A Robust Navigation Technique for Integration in the Guidance and Control of an Uninhabited Surface Vehicle

This paper proposes the novel use of a weighted Interval Kalman Filter (wIKF) in a robust navigational approach for integration with the guidance and control systems of an uninhabited surface vehicle named Springer. The waypoint tracking capability of this technique is compared with that of one that uses a conventional Kalman Filter (KF) navigational design, when the model of the sensing equipment used by the filter is incorrect. In this case, the KF fails to predict correctly the vehicle’s heading, which consequently impacts negatively on the performance of its integrated navigation, guidance and control (NGC). However, the use of a wIKF technique that is immune to this kind of erroneous modelling endows the integrated NGC system with better accuracy and efficiency in completing a mission.

Subsea cable tracking in an uncertain environment using particle filters

Localization of subsea cables is a demanding and challenging task. Among the few methods reported in the literature, magnetic field detection is the most promising one, as the cable does not require to be seen visually. Magnetic noise and a quick attenuation of the magnetic field propagating in sea water often make available methods unreliable. The authors propose a novel method of using particle filters for estimating the position of a subsea cable in a highly uncertain environment. The method was tested on data collected from a buried cable in the Baltic Sea, Denmark and shown to have a close approximation to the true location of the subsea cable. The method can be used to localize a subsea cable in an offshore noisy and uncertain environment and provides an inexpensive alternative to the use of a diver or a remotely operated platform.

Local Model and Controller Network Design for a Single-Link Flexible Manipulator

This paper describes a new genetic learning approach to the construction of a local model network (LMN) and design of a local controller network (LCN) with application to a single-link flexible manipulator. A highly nonlinear flexible manipulator system is modelled using an LMN comprising Autoregressive–moving-average model with exogenous inputs (ARMAX) type local models (LMs) whereas linear Proportional-integral-derivative (PID) type local controllers (LCs) are used to design an LCN. In addition to allowing the simultaneous optimisation of the number of LMs and LCs, model parameters and interpolation function parameters, the approach provides a flexible framework for targeting transparency and generalisation. Simulation results confirm the excellent nonlinear modelling properties of an LM network and illustrate the potential benefits of the proposed LM control scheme.