Automatic Guidance of Agricultural Wide-Span Implement Carrier (WSIC)

Abstract

The Wide-Span Implement Carrier (WSIC) is a versatile agricultural platform for the support and operation of different agricultural field equipment. It has been used in cranberry production since the 1990s. The main components of the WSIC include a long truss and two low-profile tractors that provide support to the truss at each end. WSIC operations require the motions of the two tractors to be well synchronized. Currently, this process is controlled by human operators on the tractors and the guidance accuracy depends on their experience and on-site judgement. The objective of this research is to develop an automatic guidance system to replace the manual guidance and control process of the WSIC, so that the demand of operators can be reduced and the operation accuracy can be improved. The original contributions of this work are the development and experimental validation of the two groups of control algorithms designed for the WISC, one for its mobile operating mode and one for its stationary operating mode. The mobile mode was modelled as a synchronous trajectory tracking problem and a master-slave cooperative method was designed. The stationary mode was modelled as a parallel point-to-point tracking problem and a second master-slave cooperative method was designed. For the mobile mode, the master vehicle performs a path following task by controlling its steering angle, and the slave vehicle performs a path following task and a synchronous tracking task by controlling both the steering angle and the velocity. During the operation, the motion states of the master including its position, orientation, and velocity are sent to the slave in real time. For the stationary mode, both vehicles perform a point tracking task and repeat the task in a periodic control sequence, which is executed on the master. The slave follows the commands sent from the master to perform point tracking processes. The designed control algorithms were first verified through several simulations. To test the control models and algorithms in practical applications, a scaled-down experimental WSIC platform using two heavy-duty mobile robots and an original automatic guidance system adapted to WSIC operations were developed. The hardware and software of the guidance system were designed and developed. The hardware included a dual-rover Real-Time Kinematic Global Positioning System, inertial measurement units, XBee wireless communication modules, and a group of control processors. The control structure of the software was developed at two levels: upper level for guidance algorithms calculation and lower level for velocity and steering angle control. Validation experiments were conducted using the robotic platform under restricted conditions including flat and firm test grounds and reasonably slow operation velocities. For each operating mode, two series of experiments were performed. For the mobile mode, the first series tested the path following control for a single robot, and the second series tested the cooperative synchronous tracking control. For the stationary mode, the first series verified the designed velocity and steering angle control laws for the point tracking task, and the second series tested the sequential point-to-point tracking control. Experimental results showed that the developed guidance system performed with satisfactory accuracy. For the mobile mode, the master and slave robots could automatically follow the reference paths with lateral errors less than 0.1 m and orientation errors less than 5°, and the motions of the two robots were well synchronized with offset errors less than 0.1 m. For the stationary mode, the robots performed the forward point tracking tasks with lateral errors less than 0.16 m and orientation errors less than 11° , and the root mean square and maximum of the offset errors were 0.04 m and 0.1 m respectively. The time differences when the robots stopped at their target points were less than 2 s. This motion synchronization could permit high efficiency for autonomous WSIC operations.