scholarly journals Submarine Autopilot Performance Optimization with System Identification

Author(s):  
F Belanger ◽  
D Millan ◽  
X Cyril

Computer simulation models play a vital role in the assessment of a ship’s autopilot design. A well-tuned autopilot will contribute to reducing rudder activity, thereby minimizing wear on the actuation plant and also generally reducing fuel consumption. The equations that describe the ship motion dynamics contain a large number of hydrodynamic coefficients that must be calculated as accurately as possible to justify the use of a simulation model and its relevance to predicting the ship manoeuvring characteristics. Proper prediction of the ship performance is an essential pre-requisite in the process of tuning the autopilot. The hydrodynamic coefficients can be calculated by using theoretical methods or by carrying out experiments on the actual ship or on a scaled model of the ship. System Identification (SI) is an experiment-based approach and in this paper the authors present an algorithm that can estimate the coefficients with great accuracy. These coefficients can classically be obtained in a towing tank using a captive model, and with a planar motion mechanism and a rotating arm. Generally, these systems are costly and entail expensive trials programs, and SI methods have been developed in an effort to obviate some of those problems and limitations. They typically process ship manoeuvring data obtained from a free-running scaled model or full-scale trials. While similar to a surface ship, the motion dynamics of a submarine introduce additional challenges for SI methods. This is because the submarine manoeuvres in “three dimensions”, which adds complexity and more hydrodynamic coefficients to the equations. The standard submarine simulation model, also referred to as the Gertler and Hagen equations, incorporates over 120 coefficients. To calculated these coefficients, the SI algorithm uses a Square-Root Unscented Kalman filter (SR-UKF). One of its appealing features is that it calculates all the coefficients by processing data from a single submarine manoeuvre that has a repeating sinusoidal pattern in both depth and course. The manoeuvre can be performed in a towing tank by a free-running scaled model of the submarine, or it can be performed at sea on the full-scale submarine as part of the sea trials schedule.

Author(s):  
Sergey Gavrilin ◽  
Sverre Steen

Increasingly complex marine operations dictate higher need for precise and reliable modelling. For the last decades several different approaches to modelling of ship manoeuvring were developed, including scaled model testing, numerical and empirical methods. Increasingly, time-domain simulation models for ship manoeuvring are developed and used for training and planning of marine operations. Especially when using the simulation models to plan operations, it is essential that the simulation model is properly validated. There is a need for better and more standardized validation methods for such simulation models. A key issue is the uncertainty of the data used in the validation. Typically, the validation will be against full scale trials results. In the study reported in this paper it was found by inspection of repeated tests results that even under relatively calm environmental conditions outcomes of sea trials can be highly uncertain. However, it is very expensive to investigate uncertainty of each type of trial experimentally. Therefore it can be very beneficial to estimate it by means of a simulation model. This paper presents results of analysis of full scale turning circle trials of research vessel “Gunnerus”. Turning circle trials with 20° and 35° rudder angle executed both to starboard and port sides are analyzed. Experimental uncertainty analysis is performed. Effectiveness of IMO correction procedure is discussed. Also paper describes a method for determining uncertainty of trial results due to environmental effects by means of simulation model and compares them with experimental uncertainty. The method is based on Monte-Carlo techniques.


Author(s):  
Caio Swan de Freitas ◽  
Vinicius L. Vileti ◽  
Paulo de Tarso T. Esperança ◽  
Sergio H. Sphaier

Investigating the hydrodynamics of a ship manoeuvring is a continuing concern for researchers. In recent years, with the growth of the operations complexity, understanding the hydrodynamics of a ship manoeuvring has become a central issue for supply vessels operating in close proximity of oil platforms. One of the main obstacles in understanding this problem is the difficulty of reproducing the manoeuvring on real ships with acceptable measurements, uncertainties and environmental control, not to mention the cost involved. A natural approach to address the issue is to run model tests, where it is possible to control a great number of variables. In the desire to develop new methods to evaluate hydrodynamic coefficients and to improve the understanding of the phenomena, this paper proposes different types of free-running tests to be conducted in an experimental ocean basin. An Anchor Handling Tug Supply Vessel (AHTS) scaled model was used to perform the classic turning circle and the novel turning eight, a substitute to the zigzag that fits in limited manoeuvring facilities. Optimizations of mathematical manoeuvring models were applied to estimate the hydrodynamic coefficients with a new proposed metric of manoeuvres comparison. Simulations results were compared with the experimental measurements of the model during the tests. The experimental tests were performed at LabOceano, the Ocean Technology Laboratory of the Federal University of Rio de Janeiro.


2018 ◽  
Vol Vol 160 (A3) ◽  
Author(s):  
Haitong Xu ◽  
M A Hinostroza ◽  
C Guedes Soares

Free-running model tests have been carried out based on a scaled chemical tanker ship model, having a guidance, control and navigation system developed and implemented in LabVIEW. In order to make the modelling more flexible and physically more realistic, a modified version of Abkowitz model was introduced. During the identification process, the model’s structure is fixed and its parameters have been obtained using system identification. A global optimization algorithm has been used to search the optimum values and minimize the loss functions. In order to reduce the effect of noise in the variables, different loss functions considering the empirical errors and generalization performance have been defined and implemented in the system identification program. The hydrodynamic coefficients have been identified based on the manoeuvring test data of free-running ship model. Validations of the system identification algorithm were also carried out and the comparisons with experiments demonstrated the effectiveness of the proposed system identification method.


2020 ◽  
Vol 8 (5) ◽  
pp. 378
Author(s):  
Tetsuhiro Yuura ◽  
Hirotada Hashimoto ◽  
Akihiko Matsuda

Free-running model tests were conducted using a scaled model of a large cruise ship with a damaged compartment, to investigate the effects of damage opening and floodwater on the manoeuvring performance in calm water and regular and irregular head waves. Drifting tests in regular beam waves were also performed. The experimental results indicated that the course-keeping ability in waves and turning ability became worse in the damaged condition. However, the target ship retained its manoeuvrability for safe return to the port, on its own, even in a damaged state. As it is time- and cost-consuming to conduct a free-running model experiment, a captive model test was also carried out to develop a system-based simulation model for evaluating the manoeuvrability of large cruise ships after damage.


2016 ◽  
Vol 2016 ◽  
pp. 1-10 ◽  
Author(s):  
Weilin Luo

To improve the feasibility of system identification in the prediction of ship manoeuvrability, several measures are presented to deal with the parameter identifiability in the parametric modeling of ship manoeuvring motion based on system identification. Drift of nonlinear hydrodynamic coefficients is explained from the point of view of regression analysis. To diminish the multicollinearity in a complicated manoeuvring model, difference method and additional signal method are employed to reconstruct the samples. Moreover, the structure of manoeuvring model is simplified based on correlation analysis. Manoeuvring simulation is performed to demonstrate the validity of the measures proposed.


Author(s):  
Haitong Xu ◽  
M A Hinostroza ◽  
C Guedes Soares

Free-running model tests have been carried out based on a scaled chemical tanker ship model, having a guidance, control and navigation system developed and implemented in LabVIEW. In order to make the modelling more flexible and physically more realistic, a modified version of Abkowitz model was introduced. During the identification process, the model’s structure is fixed and its parameters have been obtained using system identification. A global optimization algorithm has been used to search the optimum values and minimize the loss functions. In order to reduce the effect of noise in the variables, different loss functions considering the empirical errors and generalization performance have been defined and implemented in the system identification program. The hydrodynamic coefficients have been identified based on the manoeuvring test data of free-running ship model. Validations of the system identification algorithm were also carried out and the comparisons with experiments demonstrated the effectiveness of the proposed system identification method.


2011 ◽  
Vol 115 (1170) ◽  
pp. 453-470 ◽  
Author(s):  
L. Lu ◽  
G. D. Padfield ◽  
M. White ◽  
P. Perfect

AbstractHigh fidelity modelling and simulation are prerequisites for ensuring confidence in decision making during aircraft design and development, including performance and handling qualities, control law developments, aircraft dynamic loads analysis, and the creation of a realistic simulation environment. The techniques of system identification provide a systematic framework for ‘enhancing’ a physics–based simulation model derived from first principles and aircraft design data. In this paper we adopt a frequency domain approach for model enhancement and fidelity improvement of a baseline FLIGHTLAB Bell 412 helicopter model developed at the University of Liverpool. Predictability tests are based on responses to multi–step control inputs. The techniques have been used to generate one, three, and six degree-of-freedom linear models, and their derivatives and predictability are compared to evaluate and augment the fidelity of the FLIGHTLAB model. The enhancement process thus involves augmenting the simulation model based on the identified parameters. The results are reported within the context of the rotorcraft simulation fidelity project, Lifting Standards, involving collaboration with the Flight Research Laboratory (NRC, Ottawa), supported with flight testing on the ASRA research helicopter.


Sign in / Sign up

Export Citation Format

Share Document