For the European maritime industry to retain an edge over its competitors, it must provide a cheap cost effective, efficient service to the customer. In providing such a service safety levels cannot be compromised and indeed should be enhanced. It is a false economy to employ poorly trained crews on ill-equipped ships as such operations carry increased risk of accidents, collisions and equipment failure, and frequently leads to loss of life and damage to the environment. The only realistic solution, as some industrial nations have realised, is to gain competitive edge by exploiting current and future technology to its full extent. This paper considers the possibility of a fully automatic intelligent integrated ship guidance system that is able to track, with a high degree of accuracy, a pre-defined state trajectory. Built within the system is the ability to make decisions when there is a potential hazard on its predetermined route, and take, automatically, with due regard to international regulations, corrective action. This paper reviews the current state of hardware and software and demonstrates, using highly accurate simulation models of surface vessels, the feasibility of such a system. INTRODUCTION Autopilot Systems Automatic guidance of ships has its origin near the beginning of this century following the invention of the gyrocompass. Traditionally, ship autopilots are single input (desired heading) and single output (actual heading) control systems using generally only a compass as a sensor. Most autopilots employ a proportional, integral and derivative (PID) control law as described by Bech [1]. However, vessel response is sensitive to controller parameter settings, which ideally should be adjusted to cope with changes in environmental conditions such as sea state, or the loading and trim conditions of the vessel. For this reason there have been moves to employ adaptive algorithms to implement selftuning autopilots. Originally these were based on the work of Astrom and Wittenmark [2]. Since the proposition of fuzzy set theory in 1965 by Transactions on the Built Environment vol 11, © 1995 WIT Press, www.witpress.com, ISSN 1743-3509 642 Marine Technology and Transportation Zadeh, there have been numerous applications, including the application of fuzzy sets to ship autopilot design by Sutton and To will [3]. The fundamental problem with conventional autopilots is that they are single dimensional, controlling one parameter only, whereas by its very nature a ship is a multidimensional system with many inputs and outputs which, for complete integrated ship guidance, need to be controlled simultaneously. By the use of multi-variable system theory it is possible to describe the complete motion of a ship and hence formulate an optimal control policy that can view the global problem, and so minimise the errors in the controlled variables according to some predefined order of priority. Optimal guidance systems have been designed and implemented [4] that can provide high accuracy performance for such parameters as position on earth's surface, course and forward velocity control. Navigation Systems Hyperbolic electronic positioning systems such as Omega, Decca and Loran C have been around for some time, but systems such as Decca, although reasonably accurate, do not have world-wide coverage. Satellite navigation has been with us for about twenty years. Although early systems such as Transit have world-wide coverage, continuous position fixes are not available. The recently introduced Global Positioning System (GPS) gives continuous 3-D world-wide navigation coverage, as does the Russian GLONASS satellite navigation system, due to be fully operational by 1995. GPS has an accuracy of 16 m for military users (P code) and 100 m for general users (C/A code). However, there is an increased use of differential GPS, where land-based receivers of known position, transmit errors to local users, thus providing a high level of accuracy. System Integration There have been a number of Ship of the Future projects in recent years [5][6][7j. One of the main conclusions to be drawn from these projects is the need for system integration. The concept of total ship system integration is shown in Figure 1 and requires that all shipboard instruments have a standard interface to a common data highway, communicating using a standard protocol. Such a system currently being developed is the Maritime Information Technology Standard (MITS) being developed by European maritime IT suppliers [8]. SHIP MOTION SIMULATION In order to demonstrate the effectiveness of a fully automatic guidance system, it is necessary to have an accurate dynamic mathematical model of a surface ship, that responds to control inputs (rudder, main engines) and environmental inputs (wind, waves and tide). A realistic six degree of freedom model has been developed to cope with such situations. Transactions on the Built Environment vol 11, © 1995 WIT Press, www.witpress.com, ISSN 1743-3509 Marine Technology and Transportation 643 fl CONTROLS AND NAVIGATION '.••.; : ',.'*•'''''}'''' SHIF EMENT
Read full abstract