Search

Results:

Vol. 18 (2015 year), No. 1

Agarkov S. A., Pashentsev S. V.
Parametric identification of the Nomoto generalized model using the apparatus of variational calculus

A new approach to the identification of parameters of the Nomoto generalized vessel model has been proposed. The apparatus of classical calculus and the method of least squares have been used

(in Russian, стр.5, fig. 5, tables. 0, ref 6, Adobe PDF, Adobe PDF 0 Kb)

Vol. 26 (2023 year), No. 4, DOI: 10.21443/1560-9278-2023-26-4

Pashentsev S. V.
Neural networks as a tool for improving the mathematical model of ship motion

Using neural networks opens up great opportunities for studying mathematical models of ship motion. Correction by a network of identified parameters of the selected model should be as adequate as possible to the results of standard full-scale tests defined by the IMO Resolution N 137 of 2002. A mathematical model in displacements is considered, containing 16 parameters that determine the hydrodynamic forces acting on the ship's hull and steering gear, and is the source of a data set for training the network by randomly varying the parameters and subsequent computer testing. The standard maneuver is a steady-state circulation with fixation of the maneuvering elements: diameter, linear velocity, drift angle and angular velocity of rotation. Improving the quality of the model has consisted of changing its parameters and minimizing the mean square errors of the values of the maneuvering elements obtained during testing. For these purposes, a neural network with 16 inputs (model parameters) and four outputs (maneuvering elements for steady-state circulation) has been built. The data set for training the network was obtained using a program developed by the authors and intended for calculating parameters and conducting maneuver tests. A tanker with a displacement of 30,000 tons was chosen as a test object. Various options for network architecture and tools for working with it have been considered; the Statistica Neural Nets (SNN) software environment and the ANN package in the SciLab environment have been used. Comparative assessments of the results of working with these tools have been given.

(in Russian, стр.17, fig. 18, tables. 4, ref 12, AdobePDF, AdobePDF 0 Kb)

Vol. 28 (2025 year), No. 4, DOI: 10.21443/1560-9278-2025-28-4/1

Pashentsev S. V.
Zigzag maneuvering test and trained neural network as tools for adequate identification of the mathematical model of vessel motion

The neural network is applied for correction of the mathematical model of vessel motion. The data obtained during the model tests in the standard maneuver mode "zigzag 20/20" have been used for its training. The data set training the neural network has been obtained by means of random variations with normal distribution of the initially calculated parameters of the model. During the computer tests of the varied model, the measurable kinematic parameters for the characteristic moments of maneuvering have been recorded. These are the moments of the beginning of the rudder throwing from side to side and the moments of the subsequent maximum yawing of the vessel. For six such moments, seven parameters are saved: time, linear speed, angular rate of turn, course and coordinates of the vessel (42 input data for network training). In the Statistica Neural Nets (SNN) software environment, the network has been trained on the basis of 600 sets of such data using the IPS intelligent problem solver built into the SNN environment. The listed data are the network input, and the output ones are the parameters of the mathematical model. The network trained in this way allows for the given maneuvering characteristics, for example, determined by full-scale tests, to find a set of model parameters. If it is necessary to correct the model to meet the changed maneuvering requirements, using them as input to the already trained network, at the output we will obtain a set of model parameters adequate to these changed requirements. The most complex mathematical model in movements is considered, which is expanded to 19 parameters by additionally including two coefficients of added masses and the added moment of inertia of the vessel. All this makes it possible to obtain refined parameters of the mathematical model of the vessel's motion as output variables of the network. The analysis of the results allows us to draw a number of conclusions about the applicability of this approach and the degree of its effectiveness.

(in Russian, стр.14, fig. 17, tables. 1, ref 7, AdobePDF, AdobePDF 0 Kb)