This paper presents the comprehensive hardware architecture of an autonomous vehicle, focusing on the integration of an onboard computer, the Nvidia Jetson Nano, multiple microcontrollers, and sensors including a camera and radar. The microcontrollers, specifically Arduino Nano boards, are tasked with individually controlling the brake, steering, and throttle systems, forming a drive-by-wire system. The onboard computer sends instructions to the microcontrollers, offloading some processing tasks and facilitating easier debugging.
Each microcontroller manages its respective actuators through corresponding motor drivers. Communication between the onboard computer and the microcontrollers is facilitated by the CAN protocol, ensuring robust and reliable data exchange. The steering system employs a DC geared motor with an encoder, a steering rod controlled by a spur gear mechanism, and a potentiometer attached by a belt pulley mechanism for feedback, ensuring precise control and monitoring of steering actions.
For the hydraulic brake system, a linear actuator is used to actuate the master cylinder, generating the necessary brake pressure. A pressure sensor attached to the brake line provides feedback, ensuring effective and responsive braking. The throttle system processes the required RPM value from the onboard computer into PWM signals via the microcontroller. These signals control the throttle line, which is connected to a PMSM motor, thus manipulating its RPM and controlling the vehicle's speed. To achieve precise and stable control, fine-tuned PID controllers are deployed on the respective microcontrollers, implementing closed-loop control for each drive-by-wire system.
The vehicle employs radar and camera as the main sensors. A dedicated CAN bus is utilized for radar, which also shares data with the IMU via an Arduino. The camera interfaces directly with the onboard computer through a CSI port using a CSI cable.
This modular control architecture not only offloads processing from the onboard computer but also enhances the overall reliability and maintainability of the vehicle. Integrated feedback mechanisms in each control system provide precise and responsive operation.
In summary, this paper outlines the fundamental hardware architecture and control mechanisms of the autonomous vehicle, emphasizing the synergy between microcontrollers and the onboard computer. The described system architecture provides a solid foundation for developing advanced autonomous driving capabilities.