Passenger-cargo RoRo Ships Stowage Planning Decision With Two-dimensional Packing Constraints

In recent years,to serve the burgeoning tourist industry and business in domestic coastal waterways,the passenger-cargo RoRo(Roll-on/roll-off)transportation has emerged in large numbers.It integrates the characteristics of the traditional RoRo transportation and ferries transportation,and adapts to the increasing demand of passenger and cargo transport.Compared with other types of ships,passenger-cargo RoRo ships require fewer operators and shorter berthing time due to the loading and unloading mode of vehicles directly entering and leaving the cabin.As a core part of the passenger-cargo RoRo transportation,the stowage planning directly affects the utilization of cabin area,and then spreads to the transportation efficiency.At present,the research on RoRo ships stowage mainly focuses on traditional RoRo ships but rare studies on passenger-cargo RoRo ships.The differences in service objects,vehicle types and stowage processes between passenger-cargo RoRo ships and traditional RoRo ships make the existing stowage planning methods cannot be directly adapted to passenger-cargo RoRo ships stowage planning.It leads to the fact that the stowage planning still mainly relies on manual experience.The lack of computer decisionmaking on stowage makes it difficult to realize the port closed-loop management,which seriously restricts the improvement of port production efficiency.Therefore,based on the process and characteristics of the passenger-cargo RoRo ships stowage and the theory of two-dimensional packing and online decision-making,this research studies the decision-making model and algorithm that are in line with the reality to realize the mathematical absorption of the problem characteristics,to adapt the decision-making method to the passenger-cargo RoRo ships stowage process.It will help the passenger-cargo RoRo port realize the computer intelligent stowage planning decision.The main researches of this subject are as follows:(1)Research on characteristics of the passenger-cargo RoRo ships stowage planning.Based on the current status of passenger-cargo RoRo services,the characteristics of the passenger-cargo RoRo ships stowage planning are extracted including two-dimensional packing,two-phase,complex rotation,safety navigation,and online decision-making.Simultaneously,according to the progressive order of the characteristic of the problem,the research is divided into two types: static passengercargo RoRo ships stowage planning problem and dynamic passenger-cargo RoRo ships stowage planning problem.(2)Research on the static passenger-cargo RoRo ships stowage planning problem in offline environment.The mixed integer linear programming(MILP)model and the relaxed upper bound model(UBM)are constructed for the two-dimensional packing and other characteristics without considering the online decision characteristics of the problem.A framework of biased random key hybrid algorithm(BRKHA)is proposed to solve the problem.The core of the framework is the multi-phase heuristic(MH),which generates initial solutions through three stowage methods,and adjusts ship stability using the longitudinal and horizontal moment optimization strategy.At the same time,multiple swarm intelligence algorithms are combined to form three schemes:biased random key hybrid genetic algorithm(BRKHGA),biased random key hybrid particle swarm optimization(BRKHPSO),and biased random key hybrid whale optimization algorithm(BRKHWOA).Effectiveness and practicality of the algorithm and the proposed model are validated by computational experiments.(3)Research on the dynamic passenger-cargo RoRo ships stowage planning problem in online environment.Considering the online decision-making need that the vehicle information can not be fully known when loading,the characteristic constraints are reconstructed through mathematical discretization to realize the effective transmission of the phase state.Then a multi-objective discrete linear decision(MDLD)model is constructed and a multi-stage dynamic decision algorithm(MDDA)is proposed to solve the problem.The core of MDDA is hybrid genetic algorithm based on grey entropy parallel analysis(HGA-GEPA),which carries out the multi-objective parallel optimization of vehicle placement in each stage.In the case study,artificial loading heuristic(ALH)and hybrid genetic algorithm(HGA)are embedded in MDDA to compare the results and verify the effectiveness of the algorithm.