Research On Virtual Intelligent Housekeeper System And Related SLAM Technology Based On Six Wheel Car

In recent years,whether Alpha Go has successively defeated the world champion of Chinese and Japanese go,or the world's first "AI synthesis anchor" released by Xinhua news agency and Sogou has successfully completed a 100 second news broadcast,or Baidu Apollo unmanned car has made a high-profile appearance in CCTV Spring Festival Gala,all of which are evidence of the breakthrough of AI in recent years.The wave of industrial revolution is rapidly expanding to various industries and permeates all aspects of our lives.As an important application field of artificial intelligence,autopilot still has many shortcomings,and it still needs a long research and development process to make the humanoid robot housekeeper enter every household.The laboratory has tried to replace the humanoid robot with the UAV as the carrier,focusing on the collection,communication and processing of information,to build an open light intelligent Butler system.However,the experiments show that the UAV flying in the indoor limited space has the disadvantages of high noise,difficult navigation and difficult to ensure the safety.Based on the past experience,this paper develops an indoor intelligent housekeeper system with virtual reality technology.This paper is mainly responsible for the intelligent car positioning and map making,path planning,obstacle avoidance,as well as the development of environmental monitoring and other functions.The main research work and achievements are as follows.Based on the six wheel intelligent car,the hardware system of intelligent housekeeper in laboratory is developed.The system includes laser radar,camera,notebook computer,driving circuit of intelligent car,temperature and humidity environment sensing module and electrical control module.In order to overcome the shortage of the low chassis of the intelligent car,a lifting camera platform is also designed and installed.This paper studies the SLAM technology based on lidar.Through particle swarm optimization algorithm,in the presence of noise and occlusion and other adverse conditions,the accurate matching and splicing of lidar ranging data of two adjacent acquisition points are realized,and the map splicing of the whole laboratory is further made.With the help of the whole laboratory map,the position and direction of the intelligent car at any point are also measured.In this paper,two kinds of car location technology based on computer vision are discussed,that is,the car location based on checkerboard and the car location based on feature matching.In particular,a simple and fast algorithm is proposed to locate the camera according to the three corners of the chessboard,and a detailed derivation of the relevant formula is given.In the method of car location based on feature matching,the ORB algorithm with high real-time and accuracy is used to extract feature points.At the same time,the Lowe's algorithm and Ransac algorithm are combined to eliminate the mismatches and screen the true matching points,so as to control the car to move towards the right direction,so that the car can return to the starting position exactly after one week of cruising.Therefore,when the ground is not flat,the orb algorithm based on It is difficult to make laboratory map with single line lidar.Under the VS2017 platform,with the help of OPEN CV,the paper developed a set of Laboratory Intelligent Butler software system based on intelligent car,which realized the basic functions of intelligent car such as indoor positioning,map making,path planning and obstacle avoidance,as well as the development of additional functions such as network communication,remote control command and environmental monitoring.At the same time,the camera calibration is realized by MATLAB.With the help of other members of the research group,the article identification and management,VR modeling of Laboratory Intelligent housekeeper,and interactive display and management in VR environment are further realized.Experiments show that the intelligent housekeeper system based on the light intelligent car developed in this paper works stably,especially the key technologies such as indoor positioning and automatic obstacle avoidance control of the intelligent car in this paper meet the preset requirements,and basically meet the requirements of the intelligent housekeeper system in the laboratory.The system is an open system.With the addition of more software and hardware modules,its functions will be more and more perfect and its intelligence will be higher and higher.