ω1 ≥ 14.0Hz  (9)  

The objective function combines linearly the weight of fresh water tank with natural frequency of structure like Eq. (10). The objective is to get an economic and sound structure to reduce the weight of stiffener W and to increase the first natural frequency ω 1.  

 where, subscript t and 0 mean target and current values, respectively. α and β are weighting factors and set α = 0.5, β = 0.5 in this paper.  

4.3 Optimization results and discussion 

The optimum design was carried out to get an optimal size of stiffener and plate thickness on the fresh water tank to maintain its anti-vibration design. Table 4 shows the results of the design variables before and after optimization. It shows that the stringer S8 is increased by 72% and the others by 4.0-52%. This result indicates that the most reasonable modification method is to increase the stringer, which has an effect on decreasing the span of the vertical stiffeners. In this case, however, the plate thickness does not have any effect on the natural frequency of the structure. Table 5 shows the variation of natural frequency and weight of structure before and after optimization. According to the results, the first natural frequency increased by 163% from 8.6Hz to 14.02Hz, and the safety margin with twice passing frequency of the propeller correspondingly changed from -29.1% to 15.5%. Therefore, the structure is free from resonance. Moreover, the weights of stiffeners which are applied to the design variables also decreased in spite of the higher natural frequency. In summary, the local vibration problems which require avoidance of structure resonance through the movement of natural frequency without additional weight have been successfully solved by the proposed optimization method. Table 6 and Fig. 7 show the comparison of optimization results between GA and IEOA. The evaluation number means a total evaluation number of the objective function used in the optimization procedure, and is directly proportional to the total calculation time. According to the results, IEOA can give better solutions than GA on accuracy and convergent speed. These results lead us to draw the conclusion that the proposed new hybrid algorithm is a more powerful global optimization algorithm from the view of convergent speed and global search ability.

摘要

本文提出了一个集成的进化优化算法（IEOA）并结合遗传算法（GA），随机禁忌搜索方法（TS）和响应面法（RSM）。这个算法，为了提高收敛速度，采用响应面法和单纯形法，而收敛速度一直被认为是遗传算法的缺点。遗传算法虽然具有随机变化，但系统的多样性还是要通过使用禁忌表来保证。这种方法的效率一直采用传统的测试函数和与遗传算法比较来证明结果。这是一个证据表明，新建议的算法可以有效地找到全局最优的解决方案，通过应用它来最大限度地减少布置在船后方，设计用来避免共振的淡水舱的重量。结果表明，遗传算法的收敛速度，已经在初始阶段被利用响应面法改进了。一个最优化的解决方案旨在没有额外的实际目标函数下进行评价计算。最后，可以得出结论，从收敛速度和全局搜索能力来看，IEOA是一个非常有用的全局优化算法。

毕业论文关键词：进化优化算法；遗传算法；响应曲面法；禁忌搜索法；单纯形算法；淡水舱源:自*优尔~·论,文'网·www.youerw.com/

1． 介绍

许多动态分析的重点是要找到最大的响应点，并避免给定结构在所有激励力下的的共振。通常，这些特性提供设计极限的基本准则，从而来确定一个结构的动态特性及其重量。为了这个原因，减少响应和避免共振的最小重量一直是设计工程师主要关注的问题。