Research On Improved RatSLAM Model On The Mobile Robot With Multi-sensor Fusion

Combined with cognitive science,mathematics,computer science,information theory,control theory and bionics,Intelligent robots attract domestic and foreign scholars to study in all fields.Research of mobile robot is an important part of the intelligent robot research,and the key technology to realize positioning and navigation is Simultaneous Localization And Mapping(SLAM).Milford,an Australian academic,proposed a bionic navigation model called Rat SLAM.The mechanisms of spatial information processing for the hippocampal structure of rodents is simulated in this model,but the matching effect of this model remains to be further improved under the circumstance of the angle of light changes.The Rat SLAM model proposed still belongs to visual navigation,and interference of moving obstacles seriously affect the visual odometer,and then resulting in a larger trajectory deviation.In addition,with the increase of navigation time of the mobile robot,the problem of high computation cost of the traditional closed-loop detection algorithm is also emerging.Relying on the traditional model of bionic navigation Rat SLAM and the visual SLAM technology,a closed-loop detection algorithm based on real-time key frames matching is proposed in this paper,this algorithm better estimates the closed loop assumption in the future by storing different signature under the same location,which improves the matching rate of complex scene under the circumstance of the angle of light changes.In the meantime,the improved closed-loop detection algorithm improves the real-time performance of the traditional closed-loop detection algorithm by referring to the mechanism of human brain memory.Fuse this algorithm into the Rat SLAM model and do experiment respectively on the firing rate of the activity packet in the pose cells network at the closed-loop detection,the visual template from the local view cells,the matching effect of the experience node,and experience map by the qualitative approach.Then analyzing the data quantitatively by the accuracy rate,the recall rate and F1 value.Experiences show that compared with the traditional closed-loop detection,the improved closed-loop detection algorithm has stronger robustness under the circumstance of the angle of light changes and real-time performance.This paper also borrows from the spatial navigation strategy of multi-sensor information fusion,fuse the accelerometer,magnetometer,gyroscope with Rat SLAM model to establish a new type of dead reckoning model,which avoids wrong information about speed and angle extracted from visual odometer and the experience map with a larger trajectory deviation,enables the mobile robot to adapt to long-term navigation tasks under complex scenarios.Build platform on Voyager II mobile robot to collect experimental data and transmit data to upper computer to carry on MATLAB simulation experiment and compare the effect of the speed of the two model,the angular speed of the two model and the experience map of the two model respectively.As a conclusion,Rat SLAM model with multi-sensor fusion has smaller error,better stability and stronger adaptability.