This study proposes a strategy algorithm based on the fuzzy inference method for the rapid optimization of unmanned underwater vehicle (UUV) hull design parameters.

The initial solutions generated by the genetic algorithm are first fuzzified during the fuzzification stage, then used as training samples, and the antecedent parameters of the fuzzy rules are obtained using an equal interval fuzzy partition strategy with the membership values of all calculated UUV solutions. Next, a least learning machine (LLM) is employed to solve the consequent parameters of the fuzzy rules. Based on the generated antecedent and consequent parameters, new UUV solutions are created and the evaluation membership values for speed and range are calculated. Finally, these new UUV solutions are tested against the constraint conditions to obtain optimized and compliant UUV design parameters.

The experimental results show that within 20 seconds, the intelligent fuzzy inference method can infer multiple UUV hull parameter schemes with a combined evaluation membership degree score for speed and range of over 170 points based on the initial UUV hull parameters generated by genetic algorithms.

The proposed method effectively enhances design efficiency and balances speed and range. The findings of this study can provide valuable references for the intelligent and rapid generation of UUV hull parameters.

Aiming at the problem that it is difficult to model user requirements with fuzzy characteristics in existing ship conceptual scheme design (SCSD), this study focuses on a rapid SCSD system based on a fuzzy inference system (FIS) that can quickly obtain ship performance values that meet the user requirements and generate the corresponding ship hull design parameter values.

First, the existing ship hull parameter data is collected and sorted as prior knowledge for hyper-parameter initialization in the process of user requirement modeling. Second, based on the fuzzy set theory, the user requirements are quantitatively mapped to the fuzzy space, and the sorted prior knowledge is used to infer the ship performance values that meet the user requirements in an interpretable way. Finally, the best ship hull design parameters are matched for each ship performance value that meets the user requirements, and through the "AND" principle in fuzzy set theory, the ship design parameters that meet all the user requirements for ship performance are further inferred and taken as the SCSD.

The experimental results show that the intelligent FIS can quickly infer multiple SCSDs with deviations within 3.5% according to the fuzzy user requirements within 30 seconds.

The intelligent FIS quantifies the fuzziness in user requirements, and after two stages of inference, the SCSD that meets the various user requirements is efficiently generated. The results of this study can provide useful references for rapid intelligent SCSD generation.