The Research On Autonomous Localization And Mapping Technology Of Mobile Robot Based On Multi Sensor

Today the mobile robot has been integrated into our daily lives, it plays an increasingly important role in the help-age disabled, home cleaning, medical rehabilitation, multimedia entertainment and other fields. The mobile robot usually work at home or office and other indoor environments, such structure is relatively complex environment, you need to realize autonomous robot to locate and build maps for robot services basis.In this paper, the status of basic research in related fields of study abroad robot navigation above to multi-sensor data fusion technology as the core, autonomous robot localization and map building and indoor environmental issues such as dynamically updated research work. For the complexity of the indoor environment, this paper presents fusion technology based on odometer, laser sensors, RFID, multi-sensor data inertial module raster maps autonomous indoor environment construction method, which can prevent the existence of a single sensor can not be overcome poor fault tolerance drawbacks. For each multi-sensor data, this paper presents data BAYES algorithm based on evidence theory and D_S fusion algorithm, which is the optimal algorithm for data fusion. This paper first realized RFID tags based on particle filter positioning, and designed to match the location and parameter matching algorithm.In order to achieve autonomous mobile robot localization, take advantage of the location information is distributed in the indoor environment provides a static RFID tags, we design a robot localization method based on multi-sensor data fusion. In order to determine the label with respect to the direction of the robot, we use fuzzy inference method for probabilistic model of the antenna is processed.In order to achieve a more accurate positioning of the robot, we design a special node correction method of multi-sensor data fusion for odometer through BAYES and D_S fusion algorithms and theoretical evidence for the reliability of the sensor data obtained were evaluated to be different credibility weights, multiple sensor data fusion. Finally, experiments validate the use of weights correction algorithm, the optimal positioning results.After the barrier around the mobile robot localization information based on the information returned by the sensor and laser scanning points obtained by linear fitting of different disorders constitute an enclosed area. Is then converted to the corresponding grid shows the local area surrounding accessibility. Meanwhile, the robot through the different line segments to fit the local grid map to merge into the global grid map in a special node. Articles based on the location information to complete the semi-dynamic global grid map of the semi-dynamic objects and semantic information nodes updated database updates. Since this semantic map update is only possible to move the semi-dynamic items updated, thus greatly improving the grid map constructed after the update efficiency.