Research On Ship Route Optimization Based On Fuel Consumption Prediction

In recent years,with the rise of artificial intelligence,big data and so on,intelligence ships have become the frontier and trend of development.As the key technology of intelligent ships,ship route optimization is beneficial to improve the safety and economy of ship navigation,and is a great significance to promote the green and intelligent development of ships.However,there is still a lack of effective mining and analysis of energy efficiency data in the current research on route optimization,which makes it impossible to accurately build fuel consumption prediction models to achieve efficient route optimization.Therefore,in order to improve the superiority of ship route optimization results,this article will study the ship fuel consumption prediction model and route optimization methods.Considering that the working conditions are complex and affected by many factors,and the current working condition identification using clustering is mostly based on the research of working condition data points,which makes it difficult to describe the working conditions effectively,in this paper a method for condition identification based on deep temporal clustering is studied and a measure of similarity of ship condition sequence is proposed.Based on this similarity measure and Kmeans clustering,a work condition recognition model based on depth automatic encoder is constructed,which can extract and cluster the deep feature of work condition sequence at the same time.The results show that the accuracy of ship navigation condition division can be improved by combining condition similarity measurement and deep clustering network.Considering that the characteristic importance of different parameters to ship fuel consumption at the same time and the time importance of the same parameters to ship fuel consumption under time-varying conditions are different,while the existing black box model for fuel consumption modeling lacks the consideration of the importance of various factors affecting fuel consumption,this paper introduces the characteristic attention mechanism and time attention mechanism,and a ship fuel consumption prediction model based on dual attention mechanism is built.Combined with the deep temporal clustering model,the fuel consumption prediction model under different working conditions is constructed,and the overall method of fuel consumption prediction is given.The results show that the fuel consumption prediction model based on the dual attention mechanism of deep temporal clustering can achieve an efficient prediction of fuel consumption.Considering the space-time complexity of many factors such as wind,wave and current,and the requirement of safety and cost for ship route optimization,a ship route optimization method based on improved ant colony algorithm is proposed.Firstly,a bi-objective optimization function is constructed to minimize fuel consumption and risk.Secondly,the grid method is used to model the real marine environment;Finally,for ship navigation application scenarios,the basic ant colony algorithm is improved from heuristic information,state transfer rules and pheromone updating respectively to solve ship route optimization problems.The results show that this method can realize the effective route optimization and give the ship operators a reliable navigation method.Combining the content of this article and the requirement of shipping companies,this article also designs and develops the ship energy efficiency intelligent management system,which provides technical support for shipping companies to carry out energy efficiency management.