Design On Visual Odometry Experiment Bench And Research On Stepper Motor Subdivision Control

In the fields of robotics and computer vision, visual odometry is one of the research focuses and this technology has been successfully applied to autonomous mobile robots. By analyzing a series of related image sequences, visual odometry can ascertain the position and posture of robots and estimate their displacement. In comparison with the traditional method, visual odometry enhances the robot navigation accuracy and is better able to adpt to changing conditons. In order to test software, this paper first designed experiment bench to launch a complete set of equipments and then did some rearch on improving motor’s performance.First, this paper designed the experiment bench’s mechanical structure and relevant driving mode, setting up framework with sections in order to fix hardware equipments of visual odometry; Secondly, the paper choosed electric actuators as bench’s drive source. According to the demands of load movement, the concrete hybrid stepper motor was selected. In order to control the stepper motor, the paper established hardware system with the control core of MCU. In the design of software, the paper scheduled speed and location control’s design flow, especially placed more emphasis on the acceleration and deceleration process. Then, as a start point to improve stepper motor’s servo performance, this paper studied on the motor’s sine-wave subdivision drive method and put forward a certain hardware system with the above way. Finally, the sine-wave subdivision control system was simulated using the software of Simulink.The simulation results showed that the sine-wave subdivision drive method can significantly improve stepper motor’s performances in low frequency conditon with low cost, which is enough to meet the needs of general-purpose application.