The application of advanced control theory and intelligent optimization algorithm in actual production is an effective way to further improve the coiling temperature control accuracy of hot strip. A genetic algorithm for parameter intelligent design is proposed in this paper based on the design of two new genetic operations, eugenic and atavistic. And it is applied in the coiling temperature prediction system based on Genetic Neural Network. Using the historical data of the hot rolling operation department of a steel mill, the model was trained and in-line tested and the simulation results of MFC (Microsoft Base Class Library) show that the coiling temperature prediction system has high precision and fast convergence speed. It can not only improve the quality of hot rolled strip products, but also provide strong technical support for the development of high-end and high value-added strip products.