Mechatronics Implementation of Inverse Dynamics-Based Controller for an Off-Road UGV

Abstract

This paper presents a project developed at the University of Alabama at Birmingham (UAB) aimed to design, implement, and test an off-road Unmanned Ground Vehicle (UGV) with individually controlled four drive wheels that operate in stochastic terrain conditions. An all-wheel drive off-road UGV equipped with individual electric dc motors for each wheel offers tremendous potential to control the torque delivered to each individual wheel in order to maximize UGV slip efficiency by minimizing slip power losses. As previous studies showed, this can be achieved by maintaining all drive wheels slippages the same. Utilizing this approach, an analytical method to control angular velocities of all wheels was developed to provide the same slippages of the four wheels. This model-based method was implemented in an inverse dynamics-based control algorithm of the UGV to overcome stochastic terrain conditions and minimize wheel slip power losses and maintain a given velocity profile. In this paper, mechanical and electrical components and control algorithm of the UGV are described in order to achieve the objective. Optical encoders built-in each dc motor are used to measure the actual angular velocity of each wheel. A fifth wheel rotary encoder sensor is attached to the chassis to measure the distance travel and estimate the longitudinal velocity of the UGV. In addition, the UGV is equipped with four electric current sensors to measure the current draw from each dc motor at various load conditions. Four motor drivers are used to control the dc motors using National Instruments single-board RIO controller. Moreover, power system diagrams and controller pinout connections are presented in detail and thus explain how all these components are integrated in a mechatronic system. The inverse dynamics control algorithm is implemented in real-time to control each dc motors individually. The integrated mechatronics system is distinguished by its robustness to stochastic external disturbances as shown in the previous papers. It also shows a promising adaptability to disturbances in wheel load torques and changes in stochastic terrain properties. The proposed approach, modeling and hardware implementation opens up a new way to the optimization and control of both unmanned ground vehicle dynamics and vehicle energy efficiency by optimizing and controlling individual power distribution to the drive wheels.