| Presently, wave information can be acquired through several ways. These include wave spectra, remote sensing techniques, wave buoys, and ship borne wave measurement system .Our country have achieved great improvements in wave prediction, but it is still combined with some limitations for special area and wave state, and can not satisfy the requirements for real-time or nearly real-time wave information in seakeeping (rials, ship operation and control. Many studies have been carried out to establish ship-borne wave measuring system to identify wave states in navigatioaFor high speed ship in navigation, it is not easy to use ship borne wave sensors to get wave information. Our study concentrates on developing an adaptive wave states predicting system which only use ship motion time series measured in navigation and with the assumption that the ship motion transfer functions are unknown in advance ,so the wave predicting system would not be disturbed by the change of the transfer functions caused by many factors ,and the comparing between estimated and original ship motions frequency response enables us to gain the insight about ship motion response and wave input, and the cumulating of wave spectrum information in navigation will provide useful evaluating data for ship design. By converting ship motions (pitch, roll) into 'two axis wave meter' output equivalency , This paper also presents a simple way to extract wave direction information from ship motions ,it costs least effort of data collecting and calculating.Without knowing wave-excitation, based on the adaptive data processing of the ship motions time series, the wave predicting system provide wave information in a simple, low cost way, it is a reliable basis for future engineering project. Here, the main work of this study was described as follow:1. Generating wave-excitationIn irregular sea, wave excitation of ship motions can be described as a composite of harmonic waves possessing different frequencies, amplitude, random phases, and going in different direction. In order to concentrate on the main task of extracting wave statistics from ship motions, this paper did not consider the factor of wave energy spread, wave data was simulated as long crest wave.2. Simulating of ship motionsShip motion is a narrow band, stationary, stochastic process. The numeric studies ofthe wave prediction system based on the simulation of wave excitation and ship motions response. It is assumed that under slight wave input, ship motion response is liner, so ship motion in irregular sea can be decomposed into a sum of responses in regular sea. Ship motions are the outputs of ship dynamic system which is characterized by the transfer functions under wave excitation. Using 2-D strip method ,ship transfer functions in regular sea were calculated from 0?to 165?, at 15?interval, thus got the transfer functions in 12 directions . According to the concept of linear system input and output, using long crest irregular wave as input, the ship motions response were calculated through the transfer functions.3. Adaptive estimation of ship motion frequency response in real timeThe estimation of ship frequency response is the essential step for the wave prediction system under study, it affects the accuracy of the wave prediction system directly. The change of wave amplitude, wave period, wave direction, ship speed, and load states of ship causes great diversity of encounter states of ship in navigation, so the features of measured ship motions time series vary with the different encounter states, such as encounter frequency, skewness, kurtosis, SNR and so on .In order to establishing a wave identifying system that is capable of extracting wave information in such ever changing situation, after simulation comparing with many other system identifying methods in computation cost, adaptability for the encounter situations ,stability ,and so on , an adaptive optimization method is established .The ship motion transfer functions are optimized in an iterative circle by minimizing...
|
PDF Full Text Request