Abstract: Abstract: The jacking mechanism of ZGM-7 automatic tray-seedling grafting machine developed by our research group has failure problems in the process of jacking seedling by needle. For solving those failure problems, the software EDEM (enhanced discrete/distinct element method) is used in the research of the process of jacking seedling by needle in this paper. The Edinburgh elastic-plastic cohesion model (ECM) is chosen as particle contact model and a variety of complex particle models with various material properties are established to simulate the realistic seedling substrate in the software EDEM. Furthermore, the relationship between the mechanism and the seedling substrate is studied in simulation. It takes the maximum jacking height of the bottom of seedling substrate as index parameter and analyzes the principles how the different factors, including needle diameter, needle length and jacking speed, influence the process of jacking seedling by needle. The simulation and experimental tests are designed and performed by response surface methodology. The actual experimental result is consistent with the simulation result. The both results show that needle diameter and jacking speed have a significant influence on the maximum jacking height of the bottom of seedling substrate, and the interaction between needle diameter and jacking speed is significant. The maximum jacking height of the bottom of seedling substrate increases monotonically and then slowly decreases monotonically with the increasing of needle diameter in the test level. The maximum jacking height of the bottom of seedling substrate increases monotonically with the increasing of jacking speed in the test level. The maximum jacking height of the bottom of seedling substrate increases with the increasing of jacking speed, when the needle diameter is large. The maximum jacking height of the bottom of seedling substrate decreases with the increasing of the needle diameter, when the jacking speed is small. The maximum jacking height of the bottom of seedling substrate gradually increases and then decreases with the increasing of the diameter of steel needle, when the jacking speed is large. The error between simulation and experiment result is only 0.7%-7.2%. Among those errors, larger errors only exist in phenomenon of dropping and crooking seedling substrate. The realistic seedling substrate has a root system, which prevents the insertion of needle. Thus, the seedling substrate is easier to be dropped and crooked early. However, the error when vertically jacking seedling substrate is less than 2.1% in simulation and experiment. The successful maximum jacking height of the bottom of seedling substrate is 144 mm, which is proposed as the optimization target. The software Design-Expert is used to optimize the simulation results within test level. The optimal parameters of the simulation result are needle diameter of 2.28 mm, needle length of 12.28 mm and jacking speed of 0.09 mm/s. The simulation optimization parameters are applied to the actual mechanism. It's found that the effect of simulation optimization is significant and the success rate of jacking seedlings is 95.3%. This study greatly improves the operation effect of jacking mechanism of ZGM-7 automatic plug-seedling grafting machine, and it provides reference for the similar problems like jacking up the discrete matrix by needle. Meantime, the reliability of simulation results has been proved in this paper. So, the experimental results can be replaced by simulation results, which will reduce actual test workload significantly and shorten the equipment development cycle obviously. The method in this paper achieves a good simulation effect, which can be applied to other similar complicated environments.