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.

This study aims to improve the cooperative operation efficiency of multiple unmanned intelligent harvesters and grain trucks. A control system of multi-machine cooperative harvesting was proposed in the intelligent agricultural machinery. Two types of rice harvesters and one-grain truck were also taken as the research objects. An improved model of a continuous-time Markov chain was established for the finite state processes in the collaborative harvesting operation, according to the constraints of control decisions. The unloading conditions of harvesters varied greatly over time. A grain truck only served one harvester at a time, while the service time depended on the total amount of grain harvested. The minimum total time of collaborative harvesting was highly required to configure the harvesting sequence of each harvester. Particularly, there was the tradeoff between harvesting first or unloading first in a collaborative grain unloading task. The time that the grain truck stayed in a certain state was independent of that in the system. The process of transferring from one state to another was also irrelated with the previous and subsequent state. The unloading sequence and time of each harvester were predicted and dynamically updated to reduce the non-operation time using the improved model. The simulation results show that the control system of multi-machine cooperative harvesting effectively reduced the non-operation time for the high working efficiency. The completion time of grain truck, harvesters 1 and 2 was reduced by 10.71%, 10.25%, and 17.28%, respectively, compared with the full warehouse calling for grain unloading. The average completion time of agricultural machinery was reduced by 13.58% for the cooperative harvesting task. The waiting time was also greatly reduced in the grain unloading area. The field experiments show that the control system of multiple intelligent agricultural machineries was realized in the cooperative autonomous operation of two rice harvesters and one-grain truck. In scenario 1, there was a similar time for harvesters 1 and 2 to call the grain truck. The non-operation time for the two harvesters was 3.45 and 6.95 min, respectively. Among them, the non-operation time was gradually reduced for harvester 2 to wait in the field. The non-operation time of harvesters 1 and 2 decreased by 71.25% and 42%, respectively, whereas, the harvest efficiency increased by 6.65% and 5.22%, respectively, compared with the full-call grain unloading. In scenario 2, the intelligent agricultural machinery was able to independently enter the grain unloading area. The average non-operation and the completion time of the two harvesters were 54.67 and 59.66 min, respectively. The non-operation time of harvesters 1 and 2 decreased by 77.64% (1.9 min) and 37.09% (5.85 min), respectively, whereas, the harvest efficiency increased by 12.07% and 5.78%, respectively. The autonomous operation of harvesting/unloading transportation was realized to reduce the non-operation time of harvester for the high efficiency. Two scenarios can be expected to allocate the other ones in the vast majority of fields. The findings can provide support to the intelligent harvesting collaborative operation in unmanned farms. The improved model can also be further optimized for the scenarios when a circle of harvesting operation cannot be completed in a larger field, due to the limited granary capacity of the harvester.