Design And Application Of A Localization And Mapping Module For Indoor Mobile Robots

In recent years,with the continuous progress of science and technology and the change of people's production and lifestyle,the development of indoor mobile robots has entered the fast lane.The application field of indoor mobile robots starts to move from the traditional industrial to the diversified commercial service,and from the small-scale trial stage to the mass production application.It is foreseeable that mobile robots will be widely used in all walks of life.Under these circumstances,in this paper,aiming at the problems existing in the function realization and internal design of the existing mobile robot,a low-cost and universal localization and mapping module is designed,which can independently realize simultaneous localization and mapping,and reduce the design difficulty and functional cost of indoor mobile robots.It has great practical significance and application value The main work of this paper is as follows:Firstly,the system design is carried out for the localization and mapping module.In terms of hardware,on the basis of ensuring the realization of the design function,the principle of low-cost and small volume is adopted to compare and select the chip and sensor required by the module.And according to the connection between them,,the hardware system frame and control circuit are designed.In terms of software,a distributed system framework is used to package processing and release the data of each part.Meanwhile,a mobile APP is configured to make the module convenient to useSecondly,the fundamental principles of the three sensors used in the module is briefly analyzed and the raw data collected by the sensors is pre-processed on the basis of theoretical analysis and practical needs.Aiming at the problem that some special objects/environments cannot be effectively identified by single lidar,a method of data fusion of ultrasound and lidar is proposed to improve the quality of mapping.This method uses a calibrated model to convert part of the ultrasonic radar's ranging data into a point cloud-like form,which is then fused with the lidar data for subsequent construction of environment maps.Moreover,Compared with several mainstream laser SLAM methods,The Cartographer,as it posesses the best mapping capacity,is selected as the core functional framework for the localization and mapping module,and the basic principles of algorithms and implemention steps are illustrated.Aiming at the problem that the low-cost lidar has a small measurement range and is prone to generate continuous blank data,an indoor scan matching method is proposed to merge the missing message.This method calculate the variation of the direction angle with the blank data set that produced by lidar in the indoor environment,which can complement the angle measurement value of the attitude sensor to enhance the speed and reliability of the scan matching algorithmFinally,the designed localization and mapping module is placed in the actual indoor environment for experiments..Through comparison and analysis,the experimental results show that:1)The software and hardware systems of the module runs normally,and the basic performance parameters have also reached the actual use requirements;2)In the multi-glass partition environment,the grid map integrated based on the above data fusion method is better.