Abstract: Over years' development, the crude-oil ocean shipping in China has evolved to a mode of collectivize alliance operation. In this mode, all refineries affiliated to an alliance assemble the small patches of purchased oil products rationally in a large oil tanker for transportation and distribution, so that the ocean shipping expenses can be reduced significantly. The optimization of crude-oil ocean shipping plan is a NP-hard problem involving multidimensional variables (e.g. space, time, ship modes, and oil types), which leads to a great challenge for the establishment and solution of transportation optimization model. The crude-oil ocean shipping plan can be optimized using the modern self-heuristic algorithm. While the algorithm is implemented, the encoding/decoding rules shall be improved so as to satisfy the constraint factors and reduce the use of penalty function. As for large-scale data, it is suggested to apply the decomposition optimization method to obtain local optimization schemes, and then integrate them together into a global optimization plan. Besides, it is recommended to explore how to adopt the parallel computation and even the "cloud computing" to optimize the searching time efficiency of transportation plans.