Weeds have posed a serious impact on rice growth and yield in paddy fields. In this study, an inter-row weeding device was designed in the paddy field, including a pressing-weeds floating plate and a weeding roller. The weeding procedure was that the floating plate pushed over the weeds in the inter-row area, and then the weeding roller pressed the weeds into the soil. The probability that the weeding roller pressed the weeds into the soil was improved by 9.98% after the floating plate was pushed over the weeds. Furthermore, the pressing-weeds floating plate was in a closed box structure, which gradually pushed over the weeds, reduced the forward resistance and moved the rice for less damage to the stems and leaves. The weeding roller was composed of weeding discs and weeding cutter teeth. The weeds were pressed into the soil to isolate them from sunlight and air, in order to achieve the purpose of weeding. The two ends of the weeding roller were equipped with the chamfer, in order to reduce the damage to rice roots. The weeding width was expanded by 3%, compared with the regular weeding roller. Some parameters were determined, including the expanded weeding width, chamfer angle, soil depth, root crown root angle and weeding width of the weeding roller. The simulation test was implemented to clarify the interaction between the weeding device and soil. The quadratic orthogonal rotation combination test was carried out to analyze the effect of the weeding depth and weeding speed on the amount of heaped soil on the floating plate, the amount of soil disturbance, and the force of the floating plate and weeding roller in the horizontal and vertical directions. The interaction relationship between the factors and test indexes was determined by variance significance and response surface method (RSM). As such, the optimal weeding parameters were determined: the weeding depth was 35 mm, and the weeding speed was 0.8 m/s. The field experiment was conducted to analyze the effects of different weeding modes on the plant height, yield, and roots of rice. Four treatments were set: non-weeding, chemical, artificial and mechanical weeding using this inter-row weeding device. Among them, three widths of weeding roller 150, 175, 200 mm and two types with/without chamfer were selected in the weeding rollers. A systematic investigation was made to determine the effect of different types of weeding rollers on the weeding rate, rice roots, rice growth and yield. The experimental results showed that the weeding rate of the device reached 87.51%, which fully met the requirements in the paddy field. The weeding device was also used to turn the soil and increase the permeability of the soil. The weeding rate increased gradually with the increase in the width of the weeding roller. There was an increase in the weeding rate of 7.3 percentage points by expanding the inter-row weeding area. There were also significant effects among different weeding modes on the plant height and yield of rice（P＜0.05）. Mechanical weeding can be expected to promote the growth of rice. The yield can also reach or even exceed the levels of chemical and artificial weeding. The weeding roller with different parameters has a significant impact on the weeding rate, plant height, yield and roots of rice. The weeding roller with chamfer at both ends when increasing the width of the weeding roller can effectively avoid or reduce damage to the roots of rice, for better growth of rice, yield and weeding performance（P＜0.05）. The inter-row weeding device can provide a strong reference to the design of weeding machinery and equipment in paddy fields.

To improve the GNSS receiver’s accuracy, continuity, and stability in measuring the height of agricultural implements, this study proposed a variable-parameter Kalman filter (VPKF) algorithm based on GNSS and accelerometer to estimate the height of the implements optimally. The VPKF was verified, and its accuracy was evaluated by parallel rail platform and field tests. From the parallel rail test results, when the GNSS receiver was in real-time kinematic (RTK) positioning and the time delay of differential correction data (TDDCD) was less than or equal to 4 s, the root mean square error (RMSE) of the VPKF estimation was 9.82 mm. The RMSE of the GNSS measurement was 18.85 mm. When the GNSS receiver lost differential correction data within 28 s, the absolute error of VPKF was less than 30 mm, and the RMSE was 16.93 mm. The field test results showed that when the GNSS receiver was in RTK positioning and the TDDCD was less than or equal to 4 s, the RMSE of VPKF estimation was 13.43 mm, and the GNSS measurement was 14.56 mm. When the GNSS receiver lost differential correction data within 28 s, the RMSE of the VPKF estimate was 15.22 mm. These results show that VPKF can optimally estimate implement height with better accuracy. Overall, the VPKF can obtain a more accurate, continuous, and stable height of the implement, and increase the application scenarios of the GNSS receiver to measure the implement height.