玉米精密播种粒距在线监测与漏播预警系统研究

Design of On-line Seed Spacing Monitoring and Miss Seeding Warning System for Maize Precision Planting

摘要:

针对玉米精密播种粒距偏差导致播量分布不均匀的问题，设计了玉米精密播种粒距在线监测与漏播预警系统。该系统主要由车载计算机、排种监测ECU及相关传感器组成，设计了上位机监测软件和基于移动平均粒距在线监测的下位机程序，通过监测玉米精密播种作业过程中的粒距及其误差，完成漏播预警。首先，设计并进行了排种计数监测精度试验，结果表明，在模拟车速3~12km/h范围内，以1km/h递增变化的10个车速下，系统对指夹式排种器和气吸式排种器的排种计数监测平均准确率分别为99.12%、99.71%，标准差分别为0.52%、0.44%，总体排种计数监测误差平均值小于1%。其次，基于高速摄像的播种粒距测量试验台进行了实验室环境下的粒距监测精度试验，采用指夹式排种器进行排种，目标粒距为25cm，在车速3~12km/h范围内，以1km/h为间隔的10个车速下，系统对粒距监测误差绝对值的平均值为2.34cm，标准差为2.56cm。针对试验结果存在较多的随机异常点问题，采用移动平均滤波对监测粒距进行分析，得出粒距监测误差绝对值的平均值为0.79cm，标准差为0.62cm，单车速下对应的粒距监测误差绝对值的平均值最大为1.69cm，标准差为0.23cm，经移动平均滤波处理后，粒距误差异常点明显减少，系统粒距监测误差小于2.00cm。最后，基于气吸式玉米精密播种机设计了试验样机，设置播种车速为5.49、8.49km/h，目标粒距为25cm，进行了田间播种粒距监测精度试验，分别采集350个连续的出苗粒距进行对比分析，结果表明，与出苗粒距移动平均值相比，系统粒距监测误差的平均值分别为1.84、2.22cm，标准差分别为1.61、2.13cm，粒距监测值曲线与出苗粒距移动平均值曲线的变化趋势基本相同。

Abstract:

Aiming to solve the problem of that the seed spacing error of corn precision planting caused the seed rate distribution nonuniformity, an online seed spacing monitoring and miss seeding warning system of maize precision planting was designed. The system was mainly composed of an onboard computer, a seeding monitoring electronic control unit (ECU) and related sensors. The corresponding host computer monitoring software and a slave computer program based on an online moving average seed spacing monitoring algorithm were designed for the purpose of achieving seed missing warning according to the seed spacing and its error monitoring on the process of maize precision planting. Firstly, the precise seed dropping monitoring was the premise of seed spacing monitoring, thus a seed count accuracy experiment was designed. The results showed that the 10 simulated travel speeds increased from 3km/h to 12km/h at an interval of 1km/h, the average seed count accuracy of a finger pick-up seed meter and an air suction seed meter was 99.12% and 99.71%, with standard deviation of 0.52% and 0.44%, respectively. The average error of the whole seed count monitoring was less than 1%. Secondly, under the condition of laboratory environment, a system seed spacing monitoring precision experiment based on a highspeed photogrammetry seed spacing testbed was designed. The target seed spacing was set to 25cm, the travel speeds increased from 3km/h to 12km/h at an interval of 1km/h, and the finger pick-up seed meter as the test object. The experimental results showed that the average of absolute seed spacing monitoring error was 2.34cm with standard deviation of 2.56cm. However, there were many seed spacing monitoring outliers. To solve the above problem, a moving average program was designed to analyze the monitoring seed spacings, the results showed that the average absolute of seed spacing monitoring error was 0.79cm with standard deviation of 0.62cm, the maximum of the average seed spacing monitoring error at any travel speed was 1.69cm with standard deviation of 0.23cm. The seed spacing outliers were significantly reduced after moving average processing, and the seed spacing monitoring error was less than 2.00cm. Finally, a test prototype was designed based on an air-suction maize precision planter, and a seed spacing monitoring performance experiment was done in field. In the test, the travel speeds was 5.49km/h and 8.49km/h，respectively, the target seed spacing was set to 25cm, and then 350 continuous emergence spacings were collected. The test results indicated that compared with the emergence moving average spacings, the average seed spacing monitoring error was 1.84cm and 2.22cm, with standard deviation of 1.61cm and 2.13cm, respectively, at the travel speed of 5.49km/h and 8.49km/h. The curve of seed spacing monitoring values were basically coincident with the curve of the emergence moving average spacings values. The online seed spacing monitoring was implemented, and the monitoring precision of the seed spacing uniformity was improved. The results can give a reference to achieve the precise online monitoring system of the maize precision planting uniformity.