Abstract: Blade groups have been limited to the violent vibration of the machine in the sticky and heavy consolidated soil of the oil tea plantation forest. In this study, the blade groups were designed with layered cutting and throwing ditching in the oil tea forest, in order to obtain better ditching with low power consumption. The control group was then set as the single outer cutter and a positively mounted throw blade. The interaction between soil and cutting blade was analyzed under the ditching environment and fertilization agronomic requirements of oil tea forests with the sticky and heavy consolidated soil. The kinetic differential equations were established for the relative slip of soil particles along the surface of the throwing blade. The motion behavior of soil particles was then determined using the kinetic analysis. The eccentricity coefficient was obtained as the key influencing factor on the performance of cutting and throwing in the blade group. The sliding angle was optimized at the maximum working radius r1, setting angle, and the blade axis rotational speed. The single factor and quadratic orthogonal rotational tests were carried out with the power consumption and stability coefficient of ditch depth as the evaluation indexes. EDEM-RecurDyn simulation was coupled to clarify the influence of the eccentricity coefficient, the sliding angle at the maximum working radius r1, the setting angle, and the blade axis rotational speed on each index. A second-order polynomial response surface model (RSM) was then constructed for the power consumption and stability coefficient of ditch depth. The cyclic approximation was also optimized using the NSGA-III. The optimal combination of parameters was determined: eccentricity coefficient of 1.3, the sliding angle at r1 of 64.7°, setting angle of 55.1°, and blade axis rotational speed of 301 r/min. The torque fluctuation tests were conducted with/without layered cutting and throwing. The torque fluctuation was reduced by 21.53 % for the blade groups with the layered cut and throw ditching, compared with the control group. A systematic investigation was made on the soil disturbance by the optimal parameter blade group during a single-layered cutting and throwing of the soil. The continuously alternate layered operation of layered cutting and throwing ditching blade group was used to break up and disturb the consolidated soil. The relative errors of power consumption and stability coefficient of ditch depth between simulation and field test were 10.25 % and 4.55 %, respectively, under the optimal parameter blade group. The mechanical properties of the simulated soil model were basically consistent with the actual suitable for the movement of soil particles. The accuracy of the coupled model was verified in the power consumption, where the simulated values were less than the measured in the field test. The residual roots and gravel were attributed to the variation in the actual furrowing environment. The power consumption and stability coefficient of ditch depth were 25.96 kW and 88.31 %, respectively, for the field tests. The specific power consumption of the layered cutting and throwing ditching blade group was reduced by 8.31% (0.298 kW·h/m³) for the optimal combination of parameters. The stability coefficient of ditch depth fully met the technical requirements of national standards. The finding can provide theoretical support and reference for the design and optimization of ditching blade groups in oil tea forests.