Orthogonal experiments have mostly been used in the structural optimization of drip irrigation emitter flow channels. To further improve the efficiency of the optimal design, this study used a genetic algorithm to optimize the structure of the bionic pit flow channel. Based on the structural similarity and performance optimization of the torus-margo bordered pit structure, the constitutive equation of the flow channel unit was constructed. The selection, crossover and mutation operators were set by the genetic algorithm, and the objective function value was calculated. The design variables and known variables that met the requirements were put into the computational domain model, and the pit flow channel structure was simulated and optimized. The results showed that there were large low-velocity regions at the junctions and corners of the pit flow channel units at a working pressure of 50 kPa, and no complete low-velocity vortices were observed, indicating that the flow channels had good anti-clogging performance. The distribution of flow velocity on the same cross-section was quite different, which made the flow layers collide and mix, which intensified the loss of energy, indicating that it had a good energy dissipation effect. The multivariate linear regression analysis showed that the four variables of tooth stagger value (j), flow channel angle (θ), tooth spacing (l) and inner and outer boundary spacing (h) had a decreasing degree of influence on the flow index (x). The flow index (x) was negatively correlated with the tooth stagger value (j), flow channel angle (θ) and tooth spacing (l), and positively correlated with the inner and outer boundary spacing (h). The test results of physical samples showed that the average error between the simulation results and the real values was 3.4%, indicating that the accuracy was high, which can provide a basis for the structural optimization design of related pit drip irrigation emitters.