| In the related technology research of autonomous mobile robots,navigation technology is its core technology and it is the prerequisite for other intelligent tasks.The navigation is the process that mobile robot estimating its own position by a sensor carried by itself or an external sensor.The traditional method based on Lidar has some shortcomings such as slow positioning,poor positioning accuracy and easy to converge in the wrong position when the global positioning.In order to overcome these shortcomings,this paper proposes the method that fusion of visual sensor and lidar sensor information to locate mobile robot.Experiments show that the proposed method has certain improvement in some environments compared to the traditional laser positioning method.The main contents are as follows:First of all,analyzing the overall requirements of the navigation system and define the performance indicators of the system.According to it,the entire program was designed that including structural design,hardware design,and software design.And the specific hardware module is selected and assembled according to the design scheme to build a complete hardware experimental platform.Secondly,Introduce the Bayesian localization method and the mathematic model of wheel odometry and laser sensor.Besides,proposed an algorithm that combining Bayesian localization method and scan-match method to improve the positioning accuracy.Experiments show that the proposed method has better accuracy.Finally,in order to fuse the data of the vision sensor into the Bayesian estimation model,establishing a Bayesian estimation model for the vision sensor,and gives the detailed establishment process of the model.The results show that the model meets the design requirements.Based on this model,an algorithm that combining vision sensor and lidar is proposed.The experiments show that the fusion algorithm is superior to the traditional algorithm in the global positioning speed and success rate in some environments. |
PDF Full Text Request