Research On Human-computer Interaction Technology Based On Natural Language Of Mobile Robot

With the rapid development of science and technology,Robotics has become an international hot research field.Especially,the service robot technology which is closely linked to human life is a hot topic.In modern life,more and more human beings rely on the precision of the machine intelligent,in order to meet the needs of human services machinery,human-computer interaction technology came into being.The most natural and convenient interface is the human language,the human-computer interaction based on natural language makes the information exchange between human and robot more friendly and natural.Now there are still some limitations in robot intelligence,it is more difficult for robot in the recognition of unfamiliar environment,so combining the human experience with the robot intelligence together using natural language,can make the task completed better.The paper systematically develops the human-computer interaction interface based on the natural language on the basis of the crawler-type mobile robot in the laboratory,putting forward a way that combine the stem library with autonomous behavior,combining the mobile robot path planning technology,using the natural language to complete the robot autonomous behavior sequence analysis and path planning simulation experiments.Firstly,the whole frame of human-computer interaction system is designed.The human-computer interaction system is divided into three layers: the reaction layer,the deliberation layer and the human-machine interaction layer.The system structure and workflow of each layer are clarified,and the human-computer interaction framework is coupled with the natural language interface.The remote monitoring interface is designed as the human-computer interaction interface to complete the transmission of the panoramic image of the monitoring interface in the range of the remote network connection.Then,the study of natural language interaction technology is finished.Based on the autonomous behaviors involved in robots,a special knowledge base of word stem is established,and the attribute fields which should be covered in the stem library are designed.In this paper,the automatic Chinese word segmentation algorithm is studied,and the forward maximum matching algorithm is used to combine word segmentation knowledge with natural language.The logical reasoning rules are designed for different autonomous behavioral modules,so that the natural language can be processed to output the autonomous behavior sequence automatically and complete the task of natural language analysis.Secondly,a natural language semantic map is designed.This paper introduces the types of environment map that mobile robot involved in path planning,and designs a semantic grid map for natural language interaction based on knowledge base of stemming.Mark the obstacle target location on the map in the stem knowledge base,Defining natural language properties.The path planning method of mobile robot is studied,and the grid map-based genetic algorithm is adopted as the path planning method.At last,the experiment and analysis are finished.Combining the basic theoretical research in the previous chapters,the experiment of natural language semantic analysis is completed,that is,the experiment of converting natural language into the instruction sequence of robot autonomous behavior directly in interactive interface.Natural language path planning simulation experiments and motion control experiments are carried out,that is,natural language instructions are issued in the man-machine interface,and the robot path planning and robot movement are achieved directly after the analysis.