A Coal Output Scheduling Research Based On The Closed Silo Coal Terminals

The production scheduling problem in the terminals has always been the engineering problem that the terminals need to overcome,and it is also a hot spot in academic research.The coal terminals are the transit points in the process of coal transportation.Optimizing the production scheduling of the coal terminals plays an important role in improving the operating efficiency of the coal supply chain.Coal can easily lead to dust pollution during loading and unloading.Traditional open dumps are no longer in line with sustainable development strategy,so the terminals are trying to build siloes for storing coal which are more friendly to environment.However,scholars both at home and abroad mainly study the container terminals and the bulk cargo terminals,and do some research on silo design,coal spontaneous combustion and other problems caused by closed storage mode.There are less research on silo process and operation process from the perspective of management and decision.In real life,the coal terminals are manually operated on the basis of the previous experience.There are many machinery and equipment involved in the process of coal loading and unloading,and the operation rules are complex.The operation efficiency of the dock is often lower by the artificial scheduling model.In order to seek a more effective decision-making solution,this dissertation studies the silo handling system and coal production scheduling problem,Combines the coal output operation process,analyses the complexity of silos distribution,discusses the constraint relationship of terminal scheduling,taking minimizing the ship's total port time as an optimization goal based on the actual selection.Finally,two algorithms are used to verify the validity of the model.The main research results are as follows:(1)First,this dissertation analyses characteristics and advantages of this new type of silos,and studies the specific process of coal input and output in the case of using silos,and analyzes the essence of the scheduling problem.After making reasonable assumptions about flow of silos operations,it is find that silo scheduling problem can be transformed into a parallel machine scheduling problem with special constraints.Combined with the constraints of the actual operation of the terminals and the model of the parallel machine problem,a mathematical model of coordinated scheduling of silo lines,berths and ship loaders is established for the coal output operation.(2)The solution of this problem is an NP-hard problem.After further analyzing the constraints from the perspective of optimization,it is find that in this model constraints are global constraints and logical constraints,and the decision variables belong to the interval variables containing starting point and length and end point.Constraint programming has more advantageous in dealing with these complex constraints and representing this interval variable.Therefore,this dissertation borrows the modeling method provided by the CP system of ILOG CPLEX to build a constrained programming model based on interval variables.(3)In order to find other effective methods,this dissertation adopts the most mature genetic algorithm to solve this problem.According to the characteristics of the problem,we propose an effective encoding and decoding method.Based on the basic GA,the algorithm is improved from three aspects: initial population design,local search and adaptive crossover mutation probability.The experiment proves that the quality of the improved genetic algorithm has been improved.In order to compare the performance of constrained programming with improved GA,we randomly generated multiple sets of experimental datasets to test.It is found that whether the dataset is small or largescale,whether the objective value of the optimal solution or the running time of the algorithm,the result of constraint programming is better than genetic algorithm under the combination of different demand types,ship type proportions and ships' arrivals.(4)Taking the actual data of a large port as an example,an algorithm that is in accordance with the scale of an example is used to solve the problem.The experimental results show that algorithm scheduling result can meet the scheduling requirements of terminals and is better than manual scheduling.In order to meet the requirement of ship loading,this dissertation designs a train scheduling method based on ship scheduling.Taking the solution of the case as an example,using this method to determine schedule of train scheduling and draw Gantt chart,and then we can find that this method can satisfy terminal loading plan.