LNG Hybrid Diesel Power Vessels Routing Optimization For Marine Debris Collection Under Uncertain Environment

The solid waste that people discharge into the marine environment is called marine debris.The huge amount of marine debris,spread over four oceans,poses a great danger to human beings and the entire ecosystem,so cleaning up marine debris has become an urgent task.In order to reduce environmental pollution while collecting marine debris,this paper uses LNG hybrid diesel power vessels to study the vessel routing optimization of marine debris collection under uncertain environment.Marine debris drifts with seawater and its location information is constantly changing.According to the characteristics of the problem,the initial position of marine debris is firstly obtained,data on the flow field,wind field and hydrological elements of the sea area are collected,the marine environment is simulated using the GNOME diagnostic model,the trajectory of marine debris movement is predicted,and the debris position and time window are determined.Then,the logistics network method is used to optimize the marine debris collection path of LNG hybrid vessels under uncertain environment.A stochastic chance constrained model is developed with the objective of minimizing total cost for the debris collection vessel routing optimization from LNG hybrid vessels under uncertainty environment.Weight of debris is introduced into the model as a random variable,a collection time window constraint,a vessel capacity constraint,a tank capacity constraint,a carbon emission constraint,and an LNG storage tank capacity constraint is added to the LNG hybrid vessel.The distribution function transformation method was used to transform.PSO+ALNS hybrid algorithm is designed by using ALNS neighborhood search.The algorithm selects the removal and insertion operators according to the adaptive criteria,and updates the particle velocity and position with nonlinear decreasing inertia weight.Taking the mixed ship with LNG diesel ratio of 7:3 as the carrier to collect garbage in the Yellow Sea area of Qingdao port in China as an example,the effectiveness of the model and algorithm are analyzed,and the sensitivity is analyzed from different angles.The results show that choosing the right time to collect garbage is very important.Choosing the fifth day to collect garbage can save 15630.45 yuan and reduce 22.44%;Save 34.42 hours and reduce 24.38%.The oil gas ratio of LNG hybrid ship is an important factor affecting the total cost,total time and carbon emission of ship recovery.By analyzing the ratio of natural gas to diesel,it is found that the total cost and carbon emission of ship collection with 4:6 are the smallest,which can save 10406.02 yuan,reduce 18.77%,5.66 tons and 24.12%.Through the analysis of ship capacity,it is found that comprehensively considering the weight of garbage and time window,selecting the appropriate ship type can improve the use efficiency of ship capacity and realize the maximum utilization of resources.In terms of algorithm,when the scale of the example is 60,PSO+ALNS algorithm saves 17374.37 yuan,14.91%,94.93 seconds and30.67% compared with the target value of PSO algorithm.PSO+ALNS algorithm shows better performance.On the basis of the above,the discrete speed variable is introduced to further consider the relationship between segmented speed,operation time and carbon emission.This section establishes a double objective stochastic chance constrained programming model with the goal of minimizing the total time and carbon emission,and considers the constraints such as time window,LNG storage tank capacity and cost budget.The random chance constraint of waste weight is treated by distribution function transformation method.For the double objective problem of adding speed optimization,a PHALNS algorithm is designed based on AMOSA algorithm.The algorithm uses AMOSA algorithm to update the non-dominated solution set,ALNS algorithm performs neighborhood search operation,designs pheromone guided insertion operator,and optimizes the speed factor by introducing priority.Taking the East China Sea area and Shanghai port as an example.The results show that choosing the right time to collect marine garbage will achieve the optimal target value.Through the analysis of waste distribution function,it is found that the distribution function will affect the total time and carbon emission of waste collection.When the weight obeys the exponential distribution,the total time and carbon emission are the largest,with a total time increase of 11.28%,carbon emission increase of 5.43% and total cost increase of 8.96%.The analysis of velocity difference shows that with the increase of velocity difference,the larger the variation range of total collection time and total carbon emission,the wider the distribution range of solution set.When using 25-30-35 speed,more total collection time can be saved at the expense of less carbon emission.The analysis of ship oil-gas ratio shows that compared with the example ship,using the 4:6 ship type can save 15.29% time,increase the total carbon emission by 65.09% and increase the total cost by 41.49%,indicating that the ship with a larger proportion of natural gas shows strong advantages in total cost and carbon emission.In terms of algorithm efficiency,comparing the data of different scales with NSGAⅡ algorithm,it is found that the average total time saved by using PHALNS algorithm is 13.55%,the average total carbon emission saved is 18.52%,and the average running time saved is 20.25%.It shows that PHALNS algorithm shows good performance in solving efficiency and optimal solution set.