Takeo Koyama

Automatic Berthing Using the Expert System

Abstract The automatic berthing by the expert system is presented in this paper. The system is made on the desktop simulator on the Sun-3 workstation combined with the OPS83 expert system to deal with the final approach and berthing. The rules for the berthing were obtained through the simulation by the authors. Fairly good berthing had been achieved by the present expert system.

H∞ control design to include nonlinearities. First report : Saturation of the rudder

Abstract: In designing a control system for large vessels, there are two kinds of nonlinearities which must be considered. One is the nonlinearity in equations of motion and the other is the saturation in control devices such as engine output or rudders. These nonlinearities can strongly affect the performance of the control system. A new method to deal with the nonlinearity of saturation has been proposed in this paper. The saturation is considered as a variation of gain in the system. A control system with gain variation can be designed by H∞ control theory. As an example, a course-keeping control system for a course-unstable ship has been presented. The results show that the method proposed is effective.

Development of a hull form definition tool with a related knowledge-based advisory system

The development of a parametric hull form design system utilizing combined knowledge-based and objectoriented methodologies, approaches, and techniques is the major purpose of this research. One objective was the development of a powerful, intuitive, and parameter-based hull definition system with a related hull definition advisory system that would provide useful and necessary advice to the user. Object and knowledge modeling activities were conducted during the analysis and design stage to facilitate the implementation of a hull design program and advisory system. The hull design system includes the generation of a geometric model from a process of aggregating simple user-selected shapes representing different boundaries. Design involves the capability to vary simple parameters such as entrance angles, length-breadth ratios, and radii to define local and general surface geometries. Hydrostatic calculations are included for analysis; results are referred to an advisory system for evaluation and the determination of the acceptability of parameters and analysis results. The advisory system includes a knowledge bases consisting of hull form databases, results of the statistical analysis of data, design parameter constraints, and expert knowledge acquired from designers, reference materials, and technical reports. Use of the design tool with the advisory system results in the rapid generation of hull forms and associated design evaluations. Flexibility in decision making results from the separation of the design system from the advisory system. The independence of the two systems allows for a wide range of user experience with experts using the design tool with minimal reference to the advisory system. On the other hand, the expertise and knowledge stored in the advisory system is fully available to inexperienced designers and users.

An Unified Longitudinal Control System for Hydrofoil Based on Wave Prediction

Abstract Hydrofoil longitudinal control has been made by mode selection between platforming and contouring. In addition, the height control has been made by wave observation by human operators. To exclude this human operation, the unified automatic control system(Unified ACS) consists of optimal servo system and prediction of waves was proposed. First, the wave elevation and significant wave height were estimated by using sensor signals and Kaiman filter, and the wave elevation was predicted by Kaiman filter. And also optimal servo system(OSS) was designed to contour waves. Finally, the Unified ACS was achieved by using the predicted wave elevation and using modification of the reference input according to estimated significant wave height in ACS. The validity of the Unified ACS was confirmed through simulation.

Speed Performance and Longitudinal Control of SES

Abstract The Surface Effect Ship (SES) is a promising concept and it is adopted for the Techno-Super Liner (TSL) Project to achieve high speed with big payload. The SES has the inherent problem in the ride quality, moreover, the SES cannot keep high speed performance in rough seas since the unfavorable discharge from the pressure chamber causes increase in draft and resistance. In this paper, the pitch and heave control system to improve ride quality and speed performance at once is discussed. The simulator including control systems was build on the MATRIXx, a control system design software, and conventional PD, Pseudo Bang-Bang and Fuzzy controller have been examined in the simulator. The estimate of the resistance is one of the key issues in this paper and it was given by the summation of the static resistance components The 100 meter-length SES was defined based on the TSL SES to exemplify the system. It is revealed that the Fuzzy controller can provide good result in the ride quality and speed performance up to the wave of three meters in height.

A Study on the Design System Based on the Ship Database

The database of the ship constructed in the past has already been generated in the particular shipyard. However, the application of the database to the design of the new ship is still very difficult since the methodology to extract and to use the knowledge from the database has not been concluded.This paper proposes the methodology to use the database for the design of the new ship based on the multivariate analysis. The database containing more than three hundred ships were established in this paper. First, the category and its related parameters were defined by the user. For example, the category “Hull Form” was treated using the parameters Cb, L/B, L/d, B/D, D/d and so forth. The whole database was partitioned into clusters and relationship between parameters were established in each cluster. Parameters were determined by the formula obtained here. As some parameters were fixed previously, one must make sure that the set of parameters was in the same cluster. If the ship was assured in the same cluster, one can go into the next category such as “Principal Particulars” and so forth. The process was applied repeatedly to give the final set of design parameters.The prototype system was built on the Macintosh Quadra 700 computer. Designs of a high speed liner and a car ferry by this system showed good accordance with existing ships.