The procedure also applies to the simulated e-puck.įor a differential drive robot, the position of the robot can be estimated by looking at the difference in the encoder values: Δ s r. Note that the calibration must be done for each robot and preferably on the surface on which it will be used later on. Tiny differences in wheel diameter will result in important errors after a few meters, if they are not properly taken into account. To get such results, it is crucial to calibrate the robot. Over short distances, however, odometry can provide extremely precise results. This error accumulates over time and therefore renders accurate tracking over large distances impossible. At each step (each time you take an encoder measurement), the position update will involve some error. The major drawback of the above procedure is error accumulation. In the following exercise we will learn odometry basics. By recording this information, the trajectory of the robot can be estimated. These encoders count the number of steps done by the motors (1000 steps for one rotation of the wheel). We saw in previous exercises that the e-puck robot has two DC motors and posses an encoder for each. Eventually, we will speak of unsupervised learning with particle swarm optimization (PSO) and the simultaneous localization and mapping problem (SLAM) will be investigated. The third one is about pattern recognition using artificial neural networks and supervised learning. The second is about path planning with NF1 and potential fields. The first exercise is about position estimation using odometry. Instead, we will focus on five hot topics to get an insight into today robotics. At this level, the subject becomes increasingly rich and complex, so we will not try to be exhaustive. It will guide you through a selection of advanced topics by means of exercises all based on the e-puck robot. This chapter requires advanced knowledge in computer science.
0 Comments
Leave a Reply. |