Robust Optimization Of Marine Debris Collection Routing For Hybrid Energy Vessels

Since the early 1900 s,it has been widely recognised that marine debris has developed into a serious environmental problem and that it is one of the most widespread pollution problems affecting the environment and human beings.Pollution of the marine environment can damage biological resources and marine life,endanger human health and other harmful effects.There are few studies on the optimal recycling of marine debris.How can marine debris be effectively recycled and how to develop a reasonable vessel pathway is the focus of this research paper.Marine debris recovery route is different from land-based debris recovery.The location of marine debris in the ocean is constantly changing due to the constant flow of seawater and changes in the wind field.We can first obtain the initial location of the marine debris point by means of satellite detection images,and then we simulate the drifting trajectory of marine debris by means of an oil spill model combined with an ocean dynamics model.When the input data is more accurate,the solution can be found to be more feasible.Combined with GNOME’s diagnostic model to simulate the marine environment,a time window is set for each litter point.In order to be able to efficiently collect rubbish at sea while reducing atmospheric pollution from the vessel during its journey,this paper uses a new hybrid LNG vehicle as a carrier.Considering the actual complex environment at sea,the weight of marine debris cannot be accurately obtained.In this paper,the weight of the rubbish is considered as an uncertain perturbation parameter and robust optimisation models with single and dual objectives are established to investigate the collection routes of marine rubbish using logistics networks.(1)In this paper,a routing robust optimisation model with the objective of minimising the total cost is developed based on the uncertainty of the weight of debris at sea,the fuel capacity constraint and the time window constraint at the debris point.Equivalent transformation of robust optimization models into deterministic robust peer-to-peer models by the Bertisimas robust optimization method.We also design a hybrid heuristic algorithm combining genetic algorithm with mileage saving,neighbourhood search and simulated annealing algorithms to solve the optimisation model.The model and algorithm are tested for accuracy and validity analysis by using an LNG hybrid energy vessel as a carrier to collect floating debris in the Yangtze River estuary region.The extent to which vessels with different fuel-to-gas ratios affect the operational results is analysed,showing that the greater the diesel share of the new hybrid LNG vessel,the smaller the collection time will be,while the carbon emissions will increase.By analysing the effect of different waste weight fluctuations and the effect of different robust control parameters,it is clear that robust optimisation is a method that pursues a stable objective and solves the problem with a certain degree of conservatism.The total cost obtained by the improved algorithm is compared with the results of the traditional genetic algorithm and the simulated annealing algorithm solution,which are reduced by 8.00% and 10.70% respectively.(2)Based on the established single-objective robust route optimisation model,the selection of the vessel’s speed at discrete speeds is considered.A dual-objective route robust optimisation model is developed,which aims at minimising the total travel time and carbon emissions,and also considers the time window constraint for marine debris,the load capacity constraint and the fuel capacity constraint and cost budget constraint of the vessel.Based on the problem characteristics,the H-MOEA/D algorithm is designed to optimise the route and speed of the dual-objective model,and the proposed model and algorithm are tested with a case study of marine debris in the Yangtze River inlet.The results show that the established dualobjective optimisation model and the designed H-MOEA/D algorithm can effectively find the Pareto optimal solution set of the marine refuse collection path,and comparing the designed HMOEA/D algorithm with the NSGA-Ⅱ algorithm,the total travel time can be saved by 23.48%and the carbon emission can be saved by 17.08%,and the results of the H-MOEA/D algorithm always outperform the NSGA-II algorithm.