Research On Mobile Robot Location Based On Multi - Sensor Information Fusion

The positioning of mobile robot is essential for mobile robot to complete other actions, at the same time, the localization of mobile robot is easily affected by environment and its own movement, thus the achievement of precise localization comes to an importance. The positioning system can accurately detect and estimate robot’s position based on the multi-sensor information fusion, the positioning results can be sent back to the positioning system at the same time, which leads to the greater application of mobile robots in industry.Based on the multi-sensor information fusion technology, the positioning system of mobile robot is researched and studied by the developments in the domestic and foreign advanced technology. The specific research contents are divided into four parts as follow:Firstly, according to the static positioning of mobile robot, the theory of trilateral algorithm and quadrilateral algorithm are introduced. The improved quadrilateral algorithm model is set up with the distance between the anchor node and mobile robot, which can greatly shorten the computing time, weaken the loss of information and improve the positioning accuracy.Secondly, according to the constant motion of mobile robot in the two-dimensional plane, the theory of Kalman filtering, extended Kalman filtering and unscented Kalman filtering are introduced. Taking the mobile robot’s driving force and ground friction as the reference of filtering algorithm, which makes the tracking trajectory correct and positioning precision improved.Thirdly, according to the non-constant motion of mobile robot in the two-dimensional plane, the two-wheeled differential drive mobile robot’s model is set up, the constant model and non-constant model of mobile robot are established apparently, then by the combination of filtering inputs and filtering outputs the Multi-Mode Kalman Filtering algorithm is proposed, which improves greatly the real-time tracking performance of positioning system.Fourthly, according to the constant motion of mobile robot in the non-Gaussian noise environment, combining with the model of flicker noise then the improved particle filtering is put forward based on the introduction of degradation factor, which avoids the failure of particle filtering algorithm and the problem of single sample, reduces the affect of non-Gaussian noise and improves the positioning accuracy greatly.