Yunlong Wang

Study on semi-finished Ship structural components assembly sequence optimization

Ship is a huge system, assembled by the millions of components, and ship building process is a more complicated systematic project. Assembly is an important part during the building process because assembly sequence has a direct impact on the shipbuilding cycle, assembly quality and construction cost. Therefore it is significant to study the ship assembly sequence optimization for reducing shipbuilding cycle and improving the quality. According to the characteristics of real ship building process and the requirements of assembly sequence optimization, Ship structural components disassembly interference matrix was also generated based on component interference conditions. And then the model of assembly sequence optimization is build based on ant colony algorithm(ACO). Finally, using CATIA three-dimensional modeling software, taking double bottoms section of a bulk carrier as an example, based on the ACO, a reasonable assembly sequence was generated with the ship structural components optimizing programming.

A novel two-stage optimization method for beam–plate structure design

In this article, a novel optimization method for beam–plate structure is proposed based on two optimization levels, including the topology optimization and the section optimization. First, the optimal topology of structure with fixed material and dimensions of design domain is obtained using bidirectional evolutionary structure optimization method, and then, the section optimization is made to produce final optimal solution based on response surface model. In order to deal with beam–plate structures, the traditional bidirectional evolutionary structure optimization method is improved using cubic box as the unit cell instead of solid unit to construct periodic lattice structure. To handle irregular structural element layout produced in bidirectional evolutionary structure optimization method, section conversion process is employed to smooth the edges of sections; in this way, the manufacturability of optimal solution will be significantly improved. To determine the section parameters, response surface method is used to perform the section optimization, and the structure mass is further reduced. One example of cantilever beam structure design is provided and discussed. The applied two-stage optimization procedure resulted in 86 iteration steps to obtain the optimal topology, and a further decrease in 2.65% mass on the basis of optimal topology, which shows that the proposed method has advantages on convergence rate and computational cost.

Notice of Retraction Model study on the principal dimensions of the revolving floating crane based on SVM

During the design of the revolving floating crane, mathematical model for the principal dimensions is not available at present. In the paper, on the basis of collecting and classifying the cranes principal dimensions, the mathematical model for the principal dimensions of the revolving floating crane is developed using the SVM,in which the relationship between the lifting capacity and main dimensions has been established respectively. The model is helpful for mastering the essential variation rules on the cranes principal dimensions and can be used for technical and economic demonstration during its design. Practical result shows that the method of SVM possesses better regression ability than BP neural networks, which provides a new method to forecast the cranes principal dimensions.