The robot mobile ackermann is designed based on the ROS (robot operating system) and can realize multi-functions including human feature action recognition, SLAM map scanning and drawing, obstacle avoidance and remote control.robot mobile ackermann It can also support different wireless communication protocols including Wi-fi and Bluetooth.
The underlying architecture of the system is shown in Figure 1.robot mobile ackermann The control center receives input from heterogeneous sensors, which includes a camera, YOLOv4-tiny and simple online real-time tracking sensors for detection of pedestrians and obstacles. These sensors provide data to the navigation control module, which fuses the information and outputs linear and angular velocity commands for movement.
In addition, the system is equipped with a motor driver and a power-control unit for actuation of the front two wheels.robot mobile ackermann This allows the system to flexibly complete turns while avoiding collisions with static obstacles and pedestrians. The kinematic model of the Ackermann steering robot is shown in Figure 2.
The Ackermann steering robot features two independent steering axles with a differential drive. Each axle has an inner and outer wheel. The inner wheel has a smaller turning radius than the outer wheel, so it can rotate more quickly and change direction more easily. The inner wheel can also move a greater distance with the same amount of force as the outer wheel. In this way, the robot can turn more quickly than a differential-drive robot with the same number of degrees of freedom.
To avoid dynamic obstacles and reach the destination, the system combines global navigation with local navigation to establish an optimal path using a cost map. A Kalman filter is used to combine the sensor inputs and outputs in order to estimate the robot position. The result of this calculation is a trajectory that can be compared with a pre-selected ground truth trajectory. The trajectory is then converted into linear and angular velocities, and the Ackerman UMV is command to move.
A series of experiments were conducted to test the navigation performance of the Ackermann UMV. The first experiment involved navigation with static obstacles, while the second experiment involved navigation with moving dynamic obstacles. The results indicated that the Ackerman UMV was able to avoid the dynamic obstacles while maintaining a speed of 8.07 m and a travel time of 18.2 s. The average error of the entire motion path was 0.357 m.
In the future, it will be necessary to improve the accuracy of sensor fusion in order to reduce the error of the Ackermann UMV. In addition, the sensitivity of the system should be increased in order to respond to changing environmental conditions more quickly. In the meantime, the Ackermann UMV can be deployed in various applications in which safety and efficiency are important, such as logistics, public security, rescue, and agriculture. The Ackermann UMV can also be used for education and entertainment purposes. It can be a good choice for the development of educational robotics courses and for public demonstrations. It is easy to learn and operate, making it a popular choice for many users.
Tianjin Weide Aviation Technology Co., Ltd.