播种单体对地下压力测量方法和数学模型

对播种下压力实时测量是精准控制的基础,为了提高现有测量方法的通用性、准确性和稳定性,该研究在对播种下压力和播种深度关系模型分析基础上,采用轴销传感器播种下压力测量方法,进行了传感器力学分析和设计选型方法研究,并针对不同播深设定下单一测量模型误差大的问题,建立了融合播种深度因素的播种下压力测量修正模型,模型决定系数(R^2)为0.991 6,均方根误差(root mean square error,RMSE)为28.88 N,验证试验表明,不同播深设定下,模型预测误差绝对值最大为44.13 N,最大相对预测误差为3.28%,提高了播种下压力测量模型通用性和准确性。播种下压力田间动态变化分析试验结果表明,在4~8km/h车速下,播种下压力振荡主频幅值随车速增加而减小,且免耕处理下主频幅值和功率谱密度(powerspectraldensity,PSD)峰值均大于旋耕处理。不同车速和耕作方式下,播种下压力振荡主频变化较小,主要集中在0~1 Hz,为后续信号稳定输出的滤波处理提供依据。该研究结果可为播种下压力的精准控制奠定基础。

Measurement method and mathematical model for the seeding downforce of planter row unit

Appropriate and consistent seeding downforce can reduce the vibration of row units and improve the stability of sowing depth. Studies have shown that the existing real-time measurement methods are not suitable for precise control of downforce due to the poor universality of methods, inadequate stability and low accuracy of output. To solve these problems and promote the development and application of downforce control technology, a universal measurement method was adopted by replacing the pin shaft of the limit block with the axle pin sensor. The mechanical analysis and design selection methods of the axle pin sensor were studied, which could provide a reference for the design of pin sensor with similar structure in the future. Based on the analysis of the motion of a gauge wheel, a relationship model between the seeding downforce and the sowing depth was established, which pointed out that the angle of limit shank was an important factor affecting the accuracy of downforce measurement. Then a sowing depth measurement device based on the angle of limit shank was designed and the corresponding depth measuring model was established to reduce the measurement error of single variable model with different sowing depth settings. Equipped with the sowing depth measurement device and the axle pin sensor, an indoor test platform was built and a modeling experiment with six kinds of sowing depth and seven levels of downforce was conducted. By using the polynomial approximation fitting method with Matlab, the sensor data was analyzed and a bivariate correction downforce measurement model was established with determination coefficient (R2) of 0.991 6 and root mean square error (RMSE) of 28.88 N. To accurately evaluate the predictive performance of the model, a validation test with another three sowing depth settings and six downforce values was designed and carried out. The results showed that the maximum absolute value of prediction error was 44.13 N and the maximum relative prediction error was 3.28% with different sowing depth settings, which indicated that the downforce measurement model had good universality and accuracy. Furthermore, to analyze the frequency composition of dynamic change of seeding downforce caused by collision and impact during seeding operation, a field experiment of two-factor split plot was carried out with tillage mode and speed as experimental factors, and the data was collected by an electronic control unit (ECU) with the sample frequency of 200 Hz. Spectrum analysis of the data by discrete Fourier transform (DFT) showed that the time-domain variation of downforce was sharper at higher planting speed and a larger margin of downforce occurred in no-tillage field, which led to the increase of high-frequency components. Besides that, the main frequency amplitude of downforce vibration decreased with the increase of planting speed, whose maximum value was at 4 km/h, corresponding to 219.1 N and 161.4 N for the no-tillage field and the rotary tillage field respectively. The results of power spectral density (PSD) analysis of downforce signal showed that the main frequency amplitude and peak value of PSD in the no-tillage field were larger than that in the rotary tillage field. Moreover, the vibration frequency was less affected by tillage mode and planting speed, mainly concentrated in 0-1 Hz, which could provide a reference for low-pass filtering of signals in downforce control. The study can lay a foundation for precise control of seeding downforce.