Research On Method To ASP And Albproblem About Panel Ship-block

Generally speaking, assembly work is one kind of time-consuming workand will occupy the most key production resources. Among them, it willconsume about45%man-hour and50%costs of the total hull assembly. So,the optimization of the assembly process can not only shorten the cycle ofhull construction, reduce the assembly cost, but also improve the quality ofassembly.The dissertationproposes the methodology to solve the ASP and ALBproblems about the panel ship-block. The achievements can help thedesigners to create the optimal production plan of the complex product. Theresearch work mainly includes the following contents:(1)Case-based reasoning (CBR) combined with constraint-basedreasoning method for assembly sequence planning. Hull structure is complexand contains many parts, which will bring great difficulties to the feasibleassembly sequence reasoning. In addition, there is always no strict geometricconstraint relationship among parts of the ship-block. So the number ofgeometric feasible sequences is large, it’s hard to evaluate and choose theoptimal one from all feasible sequences. This dissertationproposes thecase-based reasoning method. This method will give the assembly precedencerelations refers to the most similar case retrieved from the case base. And thenthe complete assembly sequence will be generated with the help of theconstraint-based reasoning. Finally, we use some specific examples to verifythe effectiveness of the proposed method.(2)Research on modeling and solving of the Batched-model assemblyline balancing problem. In the assembly process of the sub-block, theallassignmentsneed not only meet the processing time, occupied space constraints, but also meet the certain completion precedence relations. Inorder to achieve the optimization objectives of the productivity maximizationand workload balancingon the assembly line,the mathematical models forbatched assembly line balancing problem are established.And then a Memeticalgorithm is proposed to search the optimal assembly plan for sub-blocks. Theintelligent division for all tasks can be realized by the decoding process ofchromosomes. The operator of automatic sequence adjustment is designed toguarantee that the generated optimal solutions are according with theprecedence constraints. In addition, the local search operator is joined tospeed up the solving process of the algorithm. At last, the effectiveness of thealgorithm was verified using a practical assembly example of sub-blocks inshipbuilding.It can give some value references to solve the division orsequencing problems for a set of tasks with precedence relations.(3) Research on modeling and solving of the mixed-model assembly linebalancing problem. The objective of the mixed-model assembly line forpanel-block of hull is to find the optimal ordering of the tasks, which canachieve the shortest production cycle time, the least number of task tardinessand the workload balancing objectives. In order to obtain the Pareto optimalsolutions of the multi-objective optimization problem, the mathematicalmodels of the mixed-model assembly line balancing problem are constructedbased on the reasonable hypothesis. A modified multi-objective geneticalgorithm is proposed, in which the assignment of the individualfitnessstrategy and Pareto solutions updating mechanism are designed toobtain the set of non-dominated solutions satisfying optimization targetsquickly. Finally, the effectiveness of the proposed algorithm is verified using apractical assembly example of panel-blocks in shipbuilding. It can providesome references for obtaining the non-dominated solutions of themulti-objective combinatorial optimization problems.(4) Research on system integration method of product assemblysequence planning and assembly line balancing. We study the relationshipbetween sequence planning and assembly line balancing, and then construct the assembly optimization application system about panel-blocks of the hull.At last the system is validated using the real panel-blocks assembly data ofone shipbuilding in china.In summary, this dissertation proposes a methodology for combinationoptimization problem solving in assembly process of the panel-blocks, whichprovides decision supports for manufacturers in shorting the assembly timeand saving the production cost of the shipbuilding.