Multi-objective optimization design of envelope shape of a certain airship with deviations considered

To be practical in engineering, we must eliminate certain Pareto-set (possible) optimization solutions for violating constraint conditions when unavoidable deviations are considered. In the envelope shape design, we propose a method, which is a multi-objective decision-making method based on the SN (signal noise) concept, that can eliminate the Pareto-set solutions that are not practical. In the full paper, we explain in some detail the decision-making method and how the deviations are considered; in this abstract, we just add some pertinent remarks to listing the three topics of explanation. The first topic is multi-objective optimization model of the envelope shape of a certain airship. In this topic, in order to minimize the drag and structural weight of the airship, we optimize the airship envelope shape using the multi-objective particle swarm algorithm. The second topic is: multi-objective optimization design. In this topic, we select the parameters of the particle swarm algorithm and present the optimization results of a certain airship as shown in Fig. 2 in the full paper. The third topic is: the decision-making on selecting the best design. In this topic, we discuss the following three advantages of multi-objective decision-making method: (1) the decision-maker can easily decide preference information; (2) the method takes into consideration unavoidable deviations; (3) it is highly adaptable to any multi-objective decision-making problems. Also in this topic, we select the best design out of a large number of the Pareto-set (possible) solutions for airship envelope shape. From Fig. 4 in the full paper, we reach the conclusion that it is essential to adopt the multi-objective decision-making method to carry out multi-objective optimization of the envelope shape of a certain airship. We also find that, out of 38 Pareto-set solutions, 24 designs violate constraint conditions due to unavoidable deviations and if these 24 designs are not eliminated, they may lead to dangerous optimization results.