Research On Sonar-Based Mobile Robot Localization Algorithm

Robot is playing an increasingly important role in people's life. Mobile robot is doing a good job in the application of robotics. The localization problem is a key problem in mobile robotics. It is estimating a robot's pose relative to its environment from sensor data. That is one of the pre-requistion to practical use of mobile robots. This thesis makes a summary on different research aspects of mobile robot and compares the different development level of foreign and domestic. Some uncertainty factors during the mobile robot localization are given .We can classify the localization methods into the map-based localization and the not-map-based one in terms of the known environment data or not, the former we choose. Aim for the experiment flat, this theiss introduces several aspects which must be focused on during the localization, for example, the coordinate transformations, map-building, model establishment and so on. The Monte Carlo (MCL) localization algorithm and Uniform MCL come after those contents to resolve the dynamic and computation questions. MCL is a recursive Bayes filter that estimates the posterior distribution of robot poses conditioned on sensor data. The key idea of MCL is to represent the belief by a set of samples, drawn according to the posterior distribution. When the robot gets odometer data or takes an action, the movement model updates; when the robot gets sensor data, the sensor (perception) model updates. MCL has an advantage of low computational cost. Uniform Monte Carlo Localization uses uniform distribution as probability of local position in MCL. Even though the distribution is simple, the UMCL works better. This algorithm is more robust and has lower computational cost. Thanks for the use of the Bayes rule , the algorithm also fuses the sensor datas. We experiment with the simulator on Pioneer II robot and the experiment's results verify that the above strategy and algorithm is correct and effective. In the end this paper summarizes about full text and looks ahead several mobile robots localization research aspects.