The Design And Implementation Of Navigation System For Family Service Robot Based On Multi-sensor Fusion

The application field of robot is constantly expanding.It was used to improve production efficiency in industries in the past and now it begins to create better life for people in service industry.Based on the need of domestic service,the unique advantages of family service robot are more and more obvious.It has not been applied in real life so far,because there are some practical problems.One of the problems is that it is unable to achieve autonomous navigation in complex and unknown family environment,so it is extremely important and urgent to research its navigation system.This paper focus on the navigation system of the small penguin shaped family service robot in family environment.Since the complex and unknown family environment,the author designs the navigation system to realize the function of autonomous navigation and obstacle avoidance in the family environment that the obstacles are irregular and distributed randomly in different spatial place.The main contents of this paper are as follows:1.According to the functional requirements of family service robot,designing the autonomous navigation system of home service robot with multi-sensor fusion,including the design of system hardware and the development of software in the robot operating system platform.In addition,the contact type automatic charging mode is designed so that the home service robot can charge automatically when the power is insufficient.2.Constructing the grid map of family environment by using the technology of simultaneous localization and mapping of lidar.And the positioning accuracy is improved by studying the particle filter algorithm,so as to build an accurate map.In the path planning,using A* algorithm and dynamic window algorithm,we can search the optimal feasible route from the initial position to the target position in the constructed map,and replan the route to the target position when the robots encounter obstacles in the process of walking to the target position.3.In view of the obstacles distributed in the above-mentioned family space environment,the multi-sensor information fusion real-time perception detection such as lidar,ultrasonic sensor and depth camera is adopted in the process of walking towards the target position,which lays the foundation for realizing the autonomous obstacle avoidance and partial route planning,which is also the focus of this paper.4.Combined with the developed navigation system,the navigation system function is programmed according to the system operation flow.According to the function requirements of the project,a series of experimental tests are carried out on the navigation system of home service robot.After continuous debugging and improvement,the experimental results show that the home service robot can realize the function of non-collision safe walking and automatic charging in the complex and unknown home space environment.