مسعود زارع زاده بغدادآباد

In the present work, the optimization of a surface piercing propeller was investigated using a multi-objective genetic algorithm and kriging interpolation methods as a meta-model. Based on previous simulation studies and experimental tests performed, suitable propeller performance data sets with different geometries were provided. First, a suitable method was sought to estimate the performance of the propeller based on its geometrical parameters. Since the goal is to optimize the propeller section and considering various resources as well as the ability of the kriging algorithm to provide suitable estimates at unknown points, this method was used. Finally, the kriging algorithm is integrated with the multi-objective genetic optimization method for optimization purposes. In this work, the optimization method itself is innovative and new. After each optimization, the result is compared with data obtained from numerical simulations, and if there is a missing match, it is added to the original data field. The addition of data is thus done intelligently to ensure that the number of simulations is optimized, and excessive simulations and experimental tests are avoided. The algorithm and the fitting model (kriging) converge to the optimal value simultaneously. To ensure the effectiveness of the above algorithm, a single-objective optimization was performed first. Then the main optimization problem was performed, and the results were presented and eva‎luated. If the necessary match is achieved, the output of the optimization algorithm converges to the numerical solution; otherwise, convergence continues until it is achieved. The optimized propeller was manufactured using a five-axis CNC machine and experimentally tested in the Hydrotech Center's water tunnel. A comparison of the experimental results of the basic propeller and the optimized propeller shows a 20% improvement in the thrust coefficient and a 10.5% improvement in the torque coefficient. The results show that the optimization algorithm has reasonable performance and can be applied to similar problems.

پروانه نيمه مغروق , كريجينگ , الگوريتم ژنتيك چندهدفه , بهينه سازي

surface piercing propeller , kriging , NSGA-II , optimization

masoud zarezadeh

Dr. reza madoliat