The design of an unmanned ground vehicle and its trajectory-tracking control by using GPS

The autonomous mobile robots (AMR) are automated machinery which is able to self-motion and have the ability to navigate in their intended environment without the need for special guidance hardware or electromechanical devices. The autonomous mobile robots are becoming more common and have widely used in all areas of life (commercial, industrial, military, etc.). This study consists of the design and trajectory/path tracking control of autonomous mobile robot which is an unmanned ground vehicle (UGV). It has the ability to navigate and follow a pre-defined trajectory or path with the lowest of positional error. The motion of this vehicle to follow the pre-defined trajectory or path is controlled by a feedback control in the position and velocity level. This feedback control system compensates the errors between the real position of the vehicle and the reference trajectory or path. For this aim, the classical PD controller with some modifications is used to control the mobile robot and compared with another control method designed for this type of mobile robots cited in literature. At first, the unmanned ground vehicle is designed with a CAD program in detail with all the parts and systems. Then, according to this design, the chassis of the vehicle was manufactured and all necessary hardware and devices were assembled. The self-propelled unmanned ground vehicle designed in this study consists of four wheels which of two rear wheels are separately driven by two DC-Motors. This mobile robot is steered from the front wheels. Three DC-Motors are used in the mobile robot two for driving longitudinally and one for adjusting the direction of the vehicle. These DC with the redactors runs in the range of 0-24 volts and get the required voltages through the driver cards, which amplify the control voltages coming from the microcontroller. In this system, Arduino board is used as microcontroller because it is simple, cheaper, and easy to program. All communications between the computer and devices on the mobile robot are executed via the Arduino card with a USB connection. Generating a control voltage between 0 and 5 volts, Arduino can control the voltage of 0-24 V for each DC-Motor over DC-Motor drivers. The steering system is a type of rack and pinion and actuated by a DC-Motor. An encoder device is installed on the output of the steering DC-Motor to measure the angular velocity of the front wheels about the vertical axis, i.e. the value of the steering angle indirectly. This system involves a Global Positioning System (GPS) device, an inertial measurement unit (IMU), encoder and magnetometer as sensors to accurately measure the real position and orientation of the vehicle. All these measured data are received in real time and processed synchronously to give the corresponding real position. In this study, the unmanned ground vehicle was modeled using the only kinematic bicycle model to describe the vehicle's behaviors. The dynamics effects in the model were neglected because the mobile robot has the light weight and slow motion. The Bicycle model simplifies the four-wheel vehicle model by combining the rear wheels together and the front wheels together to take a shape of two-wheeled model. In this simplified kinematic model, the lateral dimension of the vehicle was eliminated due to no effect on the lateral stability. The control algorithm based on the kinematic model was proposed to control the position, speed and orientation of an unmanned ground vehicle to follow the desired reference routes. The proposed control system using a modified PD controller has proved highly effective in the process of controlling the vehicle to follow the desired trajectory or path. In order to prove the efficiency of the control system, the results obtained in simulations were compared with the sliding mode control (SMC) method used in control of mobile robots in literature. The results of the comparison show that the proposed system has a high performance with respect to the positional error, speed restriction and steering angle. In the theoretical and experimental studies, the trajectories or paths are defined as two dimensional splines or known geometric curves. For these curves to be created the control points are selected by the Google Earth program. The errors in the position and orientation of the mobile robot (Xe, Ye, e) are calculated by comparing the current vehicle position and the related reference curve. The real position will be determined with the GPS/IMU integration in real time applications. All real time tests are implemented via the user defined programs written on MATLAB software. All features of the mobile robot prototype related to the motion behavior of the mobile robot are figured for the different type for applications. The simulation results show the success of the unmanned ground vehicle in designing and control.