The annular nozzle is an important part of the dyeing machine as its impact and traction performance directly affect the quality and efficiency of dyeing. The annular gap forms the core of the nozzle. In this paper, the performance differences between six nozzles with different annular gap shapes were analyzed under various conditions using numerical simulation. The results indicate that the performance of the cone straight nozzle is more stable than other nozzle shapes under various conditions. To further enhance the performance of the nozzle, a multi-objective optimization of the cone straight nozzle structure was performed using a backpropagation neural network and a non-dominated sorting genetic algorithm, thereby obtaining the optimal combination of structural parameters. Finally, the optimization results were verified via experiments and numerical simulations. The results show that the pulling force of the optimized cone straight nozzle is more than 10% higher than that before optimization.