Regarding the published researches on bicycles, they fail to design fatigue characteristics of the bicycle. In addition, dynamics and fatigue characteristics are not further improved by using advanced optimization algorithms. Aiming at these questions, this paper tries to optimize dynamics and fatigue characteristics of the bicycle through combining finite element model with advanced algorithms. The advanced algorithm applies ideas of cellular automation (CA) to Particle Swarm Optimization (PSO), and then a hybrid CA-PSO algorithm is proposed. Moreover, the finite element model is also validated by experimental test. Computational results show that: the maximum stress of bicycles is mainly distributed on the frame, especially on joints of different round pipes at different moments mainly because a dead corner is at the joint, and the dead corner can easily cause stress concentration. Under alternating forces, the stress concentration at joints will cause fatigue damage. Therefore, the service life of this position will be the shortest. As a result, the dynamics and fatigue characteristics of the joint position are taken as the optimized objective. In order to verify the optimized effectiveness of the proposed CA-PSO algorithm in the paper, the widely used PSO algorithm and PSO-GA algorithm are also used to optimize the bicycle. When the traditional PSO algorithm is used to optimize the bicycle, the root-mean-square value and maximum difference of vibration accelerations are decreased by 11.9 % and 14.3 %. When the PSO-GA algorithm is used to optimize the bicycle, the root-mean-square value and maximum difference of vibration accelerations are decreased by 20.3 % and 12.9 %. When the proposed CA-PSO algorithm is used to optimize the bicycle, the root-mean-square value and maximum difference of vibration accelerations are decreased by 27.1 % and 18.6 %. Compared with other two kinds of PSO algorithms, optimized effects of vibration accelerations are very obvious. In addition, the fatigue life of the original structure is 5 years, while the fatigue life of the optimized bicycle is 7 years. Therefore, the fatigue life is improved obviously.