5. MASTERING CHARACTERISTICS OF RADAR SIMULATOR TRAINING FOR BEGINNERS
In the experiment using on-board radar described in the pre-chapter, the experiment for mate of 3 persons is also carried out. They also have the sufficient boarding experience on anyway. Therefore, it is possible to regard their experimental results as desired level of beginners training. Then, their experimental results were set at desired level, and the learning curve should be obtained.
At the first, the approximation which shows time required to the ship position decision is obtained on the basis of the result of the radar simulator experiment. Time required to the ship position decision can be shown by following approximations.
f(t) = Ae-at + C
f(t) : Approximate value
A : Constant indicated value
a : Constant indicated learning process
C : Desired level of the training
t : Training process(times)
Radar simulator training was newly carried out in order to verify the validity of the approximation in Toba National College of Maritime Technology. Examinees are students without boarding experience, and 6 times per 1 person are performed. The own ship position is fixed in Ise Bay in the experiment. Each examinee chooses three land objects on the basis of the radar image, and distance and azimuth of the selection object are measured. Using only distance data of three objects, the ship position was decided by the description. Though the training sea area is more different than the experiment carried out in the last time Tokyo Univ. of Fisheries, and other condition was all similar.
Fig.12-a shows time required to the ship's position decision. Transverse axis is the training frequency, and vertical axis is the training hours (seconds). Average measured value of six examinees is shown in the black point. The graph shows the learning process of beginners, and the experimental results for the mate has been set at desired level. It is proven that the approximate curve corresponds to the measured value. It was confirmed again that the radar simulator training contributed to the improvement of degree of rapidity. The training frequency necessary for the improvement in degree of rapidity was obtained by utilizing the learning curve, and the validity was confirmed in present verification experiments.
Fig.12-b shows the mean value of the positioning error. Vertical axis is distance difference from true position, and average measured value is shown in the black point as well as Fig.12-a. At set desired level, performance ability of radar simulator system such as range resolution has been considered. The effect of the briefing is observed resistant for the improvement of degree of accuracy, when it has already been described. Therefore, the learning curve is being obtained only using the data in the case of the briefing. Though measured value is plotted with the learning curve, so the improvement of the accuracy is not remarkable. There is a difference between learning curve and measured value, and the validity of learning curve could not be confirmed from present verification experiment.
It is necessary to skillfully deal with the cocked hat in order to accurately require the position of the ship. These skills are also included for the desired level (data of the mate), and this cannot sufficiently supplement it only in the training by the simulator. It seems to improve more the accuracy, if the simulator training is carried out, after these skills are acquired.
In comparison with both figures, the degree of convergence in the data of Fig. 12-a is remarkable. The performance of radar simulator system such as range resolution and bearing resolution contributes to the improvement in the accuracy. And, constraints by using range and effect by the pulse duration are also matters to be considered . Therefore, the performance of radar and simulator system should be improved of faithfully reproducing the image of on-board radar, in order to improve degree of accuracy.
Fig. 12-a Learning process of Degree of Rapidity (Total times)
Fig. 12-b Learning process of Degree of Accuracy (Positioning Error)
The technique needed for the radar positioning was systematically classified from viewpoint of degree of rapidity and accuracy. Then, the results of the simulator training could be quantitatively and qualitatively arranged. The results of getting from radar simulator experiment and on-board radar experiment are summarized in the following.
1) By accurately carrying out the briefing, the improvement in the accuracy can be attempted.
2) Degree of rapidity is improved by the repetition of the training.
3) The simulator training is more predominant for the improvement in the accuracy.
4) The fidelity of the simulator image to the on-board radar image is required, when the improvement in degree of accuracy is attempted.
5) The necessary training frequency by utilizing the learning curve, for the improvement of degree of rapidity could be proposed.
6) The improvement of the performance of radar and simulator system is necessary in order to improve degree of accuracy.
7) On the accuracy, it seems to be also a cause on not only performance limitations of the system but also skill difference which can not correspond in simulator training learning curve measured value difference.
This time, the method for quantitatively and qualitatively evaluating the training result was described referring to the positioning training. And, it was proven that the mastering characteristics of beginners without boarding experience could be arranged by using the approximation obtained by the verification experiment. On the basis of present results, the research will be developed in future.
 T.Kataoka, Y.Arai. T.Kakihara, Y.Miyamoto and M.Takita, "Collision Avoidance raining using Rader/ARPA Simulator with Two Own-ships", Proceedings of INSLC 12th, Vallejo, California USA, 2003
 T.Kataoka, Y.Arai, T.Kakihara, Y.Miyamoto and M.Takita, "Evaluation of Collision Avoidance Training using Radar/ARPA Simulator", Journal of Japan Institute of Navigation, No.106, ISSN 0388-7405 pp. 13-19, 2002
 T.Kataoka, Y.Arai, T.Kakihara, Y.Miyamoto and M.Takita, "Basic Study on Collision Avoidance Training using Radar/ARPA Simulator for Begineers", Japan/Korea Workshop on Marine Simulator and Simulation Research, pp. 177-184, Japan, 2002
 T.Kataoka, Y.Arai, T.Kakihara, Y.Miyamoto and M.Takita, "Systematic Validation on Radar Simulator Training", Japan/Korea Workshop on Marine Simulation, pp.9-20, Korea,2001
 T.Kataoka, Y.Arai, T.Kakihara, Y.Miyamoto and M.Takita, "A Study on the Systematic Validation on the Radar Simulator Training", Proceedings of INSLC 11th, Kalmar, Sweden, pp. 39-45, 2000
 Research Committee of Ship-Handling Simulator in Japan Institute of Navigation. "Education and Training Using of Radar/ARPA Simulator", Text of Marine Simulator Symposium, Tokyo, Japan, 2000
 T.Kataoka, et al 4. "Systematic Evaluation of Radar Simulator Training", Journal of Japan Institute of Navigation, No. 102, ISSN 0388-7405 pp. 271-277, 2000
 H.Kobayashi, T.Kataoka et al 3 "Evaluation Methods for Ship Handling Training Using Ship Maneuvering Simulator", Journal of Japan Institute of Navigation. No. 98, ISSN 0388-7405 pp. 161-169, 1998
 K.Inoue and A.Ohno "Quantitatively Evaluation Method of the Achievement Process of Simulator Training", Journal of Japan Institute of Navigation, No. 98, ISSN 0388-7405 pp. 171-180, 1998
 Research Committee of Ship-Handling Simulator in Japan Institute of Navigation, "Education and Training Using of Radar/ARPA Simulator", Text of Marine Simulator Seminar, Tokyo, Japan, 1997
 H.Kobayashi, K.Inoue, Y.Arai, T.Fujii, M.Endo, Y.Matsuura, M. Endo and Y.Sakaguchi, "Development of Shiphandling Techniques into Elemental Techniques", Journal of Japan Institute of Navigation, No. 96, ISSN 0388-7405 pp. 119-125, 1997
 K.Tsuji, H.Iwasaki et al 6 "The Study about the Effect on Student's Training of Ship Maneuvering Simulator", Journal of Japan Institute of Navigation, No. 96, ISSN 0388-7405 pp. 111-118, 1997
 Y.Arai & T.Fujii, "The Utilities and Possibilities on the Ship Maneuvering Simulator Training", Navigation 127, Japan Institute of Navigation, ISSN 0919-9985 pp. 28-37, 1997
 H.Kobayashi, Y.Sakaguchi, C.Hattori, A Study on the Training Methods of Ship Handling", Journal of Japan Institute of Navigation, No. 93, ISSN 0388-7405 pp.109-116, 1995
Takashi KATAOKA, graduated from the Faculty of Navigation, Kobe University of the Mercantile Marine in 1983. From 1992 to 1996 lecture, from 1996 to present associate professor in charge of maritime laws and simulator training at Navigation Department in Toba National College of Maritime Technology. He is the member of Ship Handling Simulator Research Committee of Japan Institute of Navigation.