Abstract: Abstract: With the increasing demand for straw resource utilization under the rice-wheat rotation system in the middle and lower reaches of the Yangtze River, less tillage or no-tillage (conservation tillage) on rice stubble fields has become a new trend. A large amount of residual surface straw can easily cause the entanglement and blockage of the soil-contacting parts for the no-till planter, which leads to a sharp increase in power consumption and a decrease in passability of the machine. To predict and evaluate the performance of the soil-contacting components for no-tillage planters, it's necessary to establish the interaction model of the implement-soil coupling system. Traditional models regard the straw scattered on the soil surface as a rigid body and fail to simulate the rupture and deformation characteristics of the straw under rotary tillage operating conditions, which reduces the prediction accuracy of the model. In order to overcome the shortcomings of the previous models, an improved flexible model of the straw was proposed and the corresponding discrete element model of the tool-straw-soil coupling system was constructed considering the effect of scattered straw on the soil surface was constructed in this work. The soil and straw model parameters were calibrated by repose angle and compression experiments, respectively. Then, the correctness of the improved straw model was verified by soil bin test. At the same time, in order to solve the winding and obstructing problems caused by the operation of the standard rotary blade, a novel structure scheme of trapezoidal straight blade is designed. On this basis, straw displacement, shaft torque, stubble crushing and burying performance of both standard and trapezoidal straight blades under rotary tillage conditions are analyzed. Simulation results show that the existence of straw flexibility reduces the displacement of the straw, and plays a buffering role in the interaction between the straw and the cutter. The standard blade has a larger soil disturbance ability and causes the straw to move a longer distance, while the trapezoidal straight blade has a stronger stubble smashing ability after breaking the soil as a result of sharp blade edge. Stubble crushing and burying performance of the trapezoidal straight blade is better than that of the standard blade and the groove section shapes of both standard and trapezoidal straight blades under rotary tillage conditions are trapezoidal and triangular, respectively. The groove depth of the standard blade is 57 mm and that of the trapezoidal straight blade is 52 mm. The bending structure of standard blade causes a large amount of soil to be scattered in the seed ditch, especially when the depth of furrow is large, which will lead to the phenomenon of over-burying and affect the development of seeds. After the trapezoidal straight blade operation, the shape of the groove is more regular, and the residual soil is less. Triangular impact peak exits in the torque of the cutter shaft. The torque of the cutter shaft under the operation of the trapezoidal straight blade is greater than that of standard blade, while the soil breaking performance of the standard blade is better than that of the trapezoidal straight blade. In addition, taking the bending angle, flank angle and blade height as design variables, the influence of the variance and the interaction of design variables on the regression model was analyzed using the response surface method and the simplified regression equation of the objective function (rotary tillage torque and soil breakage rate) was determined. Based on the NSGA-II algorithm, the structure of the trapezoidal straight blade is optimized, and the optimal structural parameters are determined as follows: bending angle (150°), flank angle (60°), and blade height (88 mm). Through the comparative analysis of field experiments, it can be seen that the average torque and peak torque of cutter shaft were reduced by 11.70% and 6.28%, respectively, and the soil breaking rate is increased from 82.53% to 89.22%, which verifies the effectiveness of the optimization design of the tool structure.