The Grey Forecasting Model Based On Ant Colony Algorithm And Its Application In Prediction Of Ship Motion

The short-term prediction of ship motion has significant meaning to the safety of ship movement and it will enhance the safety factor of capturing planes. Because of the influence of ocean waves, winds and other interferers, ships have complex movement in six freedoms which have strong random and nonlinear characters. So, it brings many difficulties to predict the ship motions.In the paper, the characteristics of pitch motion are studied for the short time prediction of ship motion. Based on the learning the Grey System, we introduce the Ant Colony Algorithm into the prediction of ship motion. The paper establishes a real time prediction model for nonlinear system, that is the Ant Colony Algorithm GM(1,1) model(ACGM). And the model is applied to the short-term prediction for ship motion for the first time. The main works in this thesis are as follows:1. According to the character of pitching data of ship motion, the paper chooses the Functional transformation GM (1, 1) model, because of the traditional GM (1, 1) model is fit for the monotonous incremental sequence.2. We study some measures to optimize the basic grey forecasting model when we construct model to predict some data. Adding a constant parameter w to each element of the functional-transformed series and then build grey system model with the new series, this method will increase the precision of the model.3. It is considered that the parameterαin background of GM (1, 1) model will influence the precision too. So the paper applies the Ant Colony Algorithm in the parameters chosen of GM (1, 1) and establishes the Ant Colony GM (1, 1) Model. And the Ant Colony Algorithm performs better than other methods in choosing the best w andα, we get a useful prediction model above it.4. We do lots of simulation experiments via software Matlab, and determine the parameters w and a in the ACGM (1, 1).By the analysis of simulated results aimed at the prediction of pitch motion from quantities of experiments through the ACGM (1, 1) model, it is known that the prediction model can increase the prediction precision and prolong the prediction time, which implies the prediction model present in the paper is reasonable and feasible.