The Method Of Analyzing The Multi-directional Irregular Wave Spectrum By The Genetic Algorithm

The wave which is usually described by the directional spectrum is a three-dimensional random process, and the height and period of the wave are distributed randomly. The actual wave is multi-directional irregular wave, and the wave energy is distributed with the wave frequency and direction. This can be described by the directional wave spectrum. It is important to calculate the wave action on the sea-shore buildings and sediment movement and analyze the vibration of sham structure reasonably, and it is also the fundamental information for the reason of the wave induced reaction of the ship,floating cylinder and other floating body. Therefore, people pay more and more attention on the research of the directional wave spectrum. However, the observation and analysis of the directional wave spectrum is much more complicated than the frequency spectrum. It needs to layout several measuring point(not less than 3) to measure the wave characteristics(such as waves, water particle velocity, acceleration, etc.) or the water wave surface of a sea area at the same instantaneous time(such as three-dimensional photography, etc.). The directional wave spectrum can be analyzed by the characteristics of wave properties and phase relationship of the wave properties. The cost of observation is very high. Therefore, limited by the economic and technical conditions, it is impossible to increase in the number of measured wave characteristics excessively. So people proposed different methods to get satisfied directional spectrum from the limited data.This paper investigates the method of the multi-directional irregular wave spectrum by the combination of numerical simulation and physical model test methods. Two kinds of analysis methods of multi-directional irregular wave spectrum, i.e., the method of direct solution(here called direct algorithm) and genetic algorithm(called GA)are established. The genetic algorithm is a bionic random optimization and searching method. It is also called simulation life evolution algorithm. In the process of analyzing the wave spectrum, the directional spreading is obtained by optimizing a fitness parameter which is a function of the difference between the simulated and theoretical cross spectra. By the numerical simulation of cross-spectrum and the wave time series, the validity of the analyzing method of the multi-directional irregular wave spectrum by the method of genetic algorithm is verified. The results show that the simulated directional spreading by the method of GA and input directional spreading is close to each other. The comparison of the results of GA and BDM method indicates that GA is better than BDM. Therefore, GA can be accurately applied to analyze the directional spectrum of the multi-directional irregular wave.