Abstract: The optical intelligent dry coal separator is a dry coal separation equipment based on advanced imaging and artificial intelligence technology, mainly composed of material system, recognition system, and execution system[1]. In the environment of container operation, a large amount of dust, coal ash, and coal slurry can easily cause the nozzle of the dry coal separator to clog, resulting in a significant decrease in the striking force and seriously affecting the separation effect[2]. In response to this problem, this paper proposes an optimization method for the nozzle, which can ensure that the nozzle has sufficient striking force during separation and can be used normally in harsh working environments. After practical testing, the optimized nozzlesignificantly improved the striking force, improved the material separation recognition rate, and brought significant economic benefits.