Objective To optimize the structure parameters of cell wheel notoginseng precision seeding apparatus and improve seeding performance.

Method Notoginseng seeds were chosen as materials. Number, diameter and depth of type hole, along with cell wheel rotational speed of seed metering device were selected as four influence factors to carry out quadratic orthogonal rotation combination test. Seeding qualified index, repeat sowing index and leakage sowing index were tested by the self designed soil bin test bed. The test results were analyzed by response surface methodology, and quadratic regression models between various factors and test indexes were established. The influence rules of various factors, interaction terms and quadratic terms on the seeding qualified index were studied. Test factors were optimized and verified by actual experiment.

Result The order of factors affecting the seeding qualified index was diameter of type hole > depth of type hole > number of type hole > rotational speed of cell wheel. The seeding qualified index was above 90, the repeat sowing index and leakage sowing index were below 5 when the depth of type hole was 5.4 mm, the rotational speed of cell wheel was 35 r·min^-1, the number of type hole was 12-14 and the diameter of type hole was 7.7-8.2 mm. Factors were verified after optimization, and the experimental results were generally in agreement with the optimization results.

Conclusion The performance of this seeding apparatus can be optimized by response surface methodology. This research provides a theoretical basis for designing notoginseng seeder and mechanized sowing of notoginseng seeds.