Abstract:
The articulated frame steering vehicles (ASV) are widely implemented in agriculture, mining, construction and forestry sectors due to their high maneuverability. The ASVs, however, are known to possess lower dynamic stability and yield high magnitude of whole-body vibration, which are reported to be harmful to the operators. Automatic driving system is thus necessary for the ASVs to exclude the human driver from detrimental operations, especially for the agricultural ASV. In order to enable the automation of ASV, path tracking strategies are essential to maintain the normal work of the vehicles. As the ASV dynamics significantly are different from the conventional vehicles with front wheel steering, the path tracking controller derived for conventional vehicles is considered not to be applicable for the ASVs. Moreover, large variations of the vehicle load and the off-road excisions challenge the robustness of path tracking algorithms. In this paper, a path tracking strategy is proposed for the ASVs on the basis of sliding mode control (SMC). The kinematic model of the ASV is derived neglecting the vehicle dynamics. Three measurable errors are defined to indicate the deviation of real path from reference path, i.e. lateral displacement error, orientation error and curvature error. These errors serve as the inputs in order to synthesize the SMC. The exponential reaching law is selected in order to increase the reaching speed and reduce chattering. The sign function of exponential reaching law is replaced by a continuous function to further suppress the chattering. Lyapunov function is then utilized in order to assess the system stability. The system transition performances in terms of response time, setting time and overshooting are tuned via pole placement method. The differential transformation method is implemented to determine the poles, in order to obtain the transition performances while preserving the system stability. Ackermann's formula is used to improve traditional pole placement algorithm and further design the control law. The open loop eigen-polynomial of the system is thus not requisite. Furthermore, the hardware-in-the-loop (HIL) simulation is conducted to evaluate real-time performance of the proposed control law. The HIL platform is set up on the basis of national instruments PXI-8110 and cRIO-9024 as well as a host computer. The real-time ASV kinematic model established in the MapleSim platform is downloaded into the PXI-8110 as the simulation plant, and the SMC path tracking algorithm compiled by Simulink is embedded to the cRIO-9024 as the real electronic control unit. The host computer couples the real-time vehicle model and the path tracking algorithm via the LabVIEW platform and displays the simulation status as the upper monitor. The path tracking algorithm then controls the vehicle to follow a circle path in real time. The results suggest that the simulated vehicle path is smooth and almost identical to the reference path. The 3 kinds of errors achieve steady state in 10 s. The proposed SMC controller is then demonstrated to be robust even neglecting the vehicle dynamics. The results also suggest that the SMC with Ackermann's formula can fulfill the prescribed request of the dynamic and the steady performance. The real-time performance of the path-tracking is even better than the off-line simulation. The simulation duration 60 s is equal to the calculating duration in the HIL simulation that means the time is synchronous. Compared with the real vehicle test, the HIL simulation is economical and efficient. This research can provide a reference for the design of agricultural ASV automatic driving system.