The control of warp tension is a key part of the intelligent control of rapier loom, which determines the quality of the product. In order to further improve the control performance, this paper proposes a PID control scheme optimized by genetic algorithm based on PID control. By analyzing the dynamic characteristics of warp feed and winder system of loom, the warp tension model was established and the fitness evaluation mechanism of genetic algorithm was optimized, which effectively solved the problem of lack of stability of traditional control scheme. Then, simulation experiments are carried out to compare and analyze the performance indexes between the optimized control scheme and the traditional scheme, and the results show the outstanding performance of the optimized control scheme. Finally, a platform is built on VSI22 loom to verify the simulation results.