The use of topology optimization to assist in the completion of engineering design is a new method in recent years. The paper adopts the genetic bi-directional evolutionary structural optimization (GBESO) discrete model to obtain the optimal reinforcement topology based on the uniform stress distribution, which provides a new idea for solving the engineering reinforcement design problems of reinforced concrete deep beams and other two-dimensional components. With reference to the solution of the optimization algorithm and the empirical design method of deeply flexural members recommended by the Chinese code, a group of reinforced concrete simple-supported deep beam reinforcement designs have been completed and static comparison experiments were carried out. Results show that the component designed by GBESO has lower steel bars consumption, higher ultimate bearing capacity, stronger crack development capacity, and better energy dissipation capacity during the failure process. It proves that the reinforcement design method of the oblique section reinforced with oblique steel bars is more in line with the force mechanism and characteristics of the members. Ability of the GBESO discrete model in the design of deep beam reinforcement has been verified to a certain extent.