Optimization of large scaled structures under effect of dynamic loads is a difficult task because of very time consuming analysis as a result of high dimension calculation data in every cycle of optimization process. In this book, the Genetic Algorithm method has been used for optimization of a Jacket platform. Dynamic nature of forces that effect on the Jacket platform results in a time consuming analysis and so, a lengthy optimization process. In such cases, approximate analysis is a good solution. Therefore, the wave-net method has been used for approximate analysis. Wave-net method has been modified by combining the Back-Propagation neural network and wavelet theory. Wave-nets have been trained by several training sets. Then, the trained network will be used many times in optimization process. All of the required programs for design optimization process have been prepared in MATLAB software environment. For exact analysis and generating the training sets the ANSYS software has been used. At the end, the capability of the method has been indicated by a numerical example.