Design and experiment of the sowing monitoring system for a wide-width rapeseed planter

Sowing is one of the most important processes in agricultural production. The efficiency and quality of sowing also dominate deeply the growth and yield of crops. Precision sowing in mechanized seeding aims to achieve a high yield of crops under reasonable planting density in recent years. It is highly urgent to accurately monitor the sowing amount, frequency, and work quality of planter for visualization and digitalization in precision sowing. A wide-width planter can be expected to significantly improve productivity during large-scale planting, especially for rapeseed production. However, most planters are difficult to adjust the working parameters suitable for agronomic requirements, because the real-time sowing frequency fails to be detected during the traditional sowing, particularly for the small seeds, such as rapeseed. Furthermore, the state-of-the-art precise sensors have increased the cost of the monitoring quite a lot. It is still lacking to deal with the real-time data during the sowing in the current system. In this study, a rapeseed sowing monitoring system was proposed for the wide-width planter (named 2BMF-12), thereby significantly improving the quality of planting scale and mechanized production. The quality of sowing was also evaluated during the rapeseed sowing. The 2BMF-12 planter was utilized to support up to 12 rows of sowing at the same time. The monitoring system included 12 seed-sensing devices and a detecting terminal. The seed-sensing devices were placed on the planter under the sowing pipe lines, in order to collect the information of seeds crossing the sensing area and generate electrical signals. The terminal received the electrical signals as external interrupts using multithreading I2C bus to expand ports, in order to receive more interrupts-channel at the same time. In this case, the terminal was designed to monitor 12 rows, where more rows were set under parallel connecting the chips, named PCF8574 on the I2C bus. Then, the sowing amount and frequency were measured for each row of the planter. The leakage and qualified index of sowing were thus counted, according to national standards. Local database was used to manage and store in the assigned table of the database once a second. MariaDB server was built in the sowing monitoring terminal for real-time storage and management parameters during the planters running, such as sowing amount, sowing frequency, the leakage, and qualified index. Bench and field tests showed that the seeding monitoring system was stable to real-time detect the status and leakage index during sowing. Specifically, the system error was 2.21 percentage points for the degree of leakage, while there was no severity leakage occurred in the planter. The accuracy rate of the monitoring system was not less than 96.5% for rapeseed sowing, suitable for the farm. This finding can provide for the quality evaluation of rape seeding operations in the future. It can be possible to combine with high-precision BeiDou positioning satellite and mobile internet, thereby supporting the sowing amount and leaking state maps generating for the sowing in the fields.