A diversity genetic algorithm(DGA) is designed to maintain the diversity of population during the evolutionary process. Through the applications of DGA and Monte Carlo simulation method into design optimization, it has ability to find the most fault-tolerant solution from all the optimal solutions. As the result of the numerical experimentation, it is intuitively verified that maintaining the diversity of the population plays an important role in the optimization of fault-tolerant, and the result of the method used into the structure optimization design of a spring is a strong proof that it is more effective than existing methods.