Research On Indoor Environment Perception And Map Building Of Mobile Robot Based On SLAM

The evolution of the times, and technology development, more and more fields have been put forward to the demand of robot. The related research of mobile robot has been concerned by all walks of life. If indoor mobile service robot wants to be real intelligence, it must be aware of the indoor environment for intelligent decision. And the robot’s perception of the indoor environment is accurate and complete or not, just also to determine the accuracy of the robot intelligent decision. Autonomous mobile robot is one of the most common intelligent robots, which solves the problem of "where I am", which is the prerequisite for the autonomous navigation of mobile robotSimultaneous Localization And Mapping, which is called of SLAM, is able to solve this problem, it can be better in the unknown indoor environment for perception and synchronization of the robot’s own position and attitude acquisition and environment map building. The main work of this paper includes these following aspects:1. For the mobile robot indoor environment perception and map building, and then introduced the thought and implementation of SLAM algorithm based on particle filter by theoretical analysis and mathematical formula. Through laboratory experiments on the real environment, and analysis the advantages and disadvantages of the environment in deraile. Provided a theoretical support to optimize scheme.2. SLAM algorithm based on particle filter, which uses a large number of particles to represent the probability distribution of the position pose, also the number of particles to determine localization algorithm accuracy and anti-interference ability in a certain extent, However, increase of the number of particles, the calculation of the algorithm will also increase, which leads to the positioning delay, because the SLAM algorithm is time sensitive, the positioning of the delay will result in the final positioning error. Based on this problem, this paper proposes a method of using the GPU parallel computing, based on RBPF-SLAM algorithm is improved to reduce the execution time and the positioning error. And finally through experiments demonstrate the effectiveness of this method.3. The map constructed by SLAM is still incomplete, because the map information does not contain information such as indoor environment temperature, humidity, object properties, etc. This will lead to the robot’s perception not deep enough of the environment, and it determines the robot lose the smart decision of autonomous navigation in the future. Based on this problem, this paper presents a new method of ontology sematic combined with SLAM algorithm, to get the environment data from Zigbee wireless sensor network. And convert the indoor environment to ontology format through ontology knowledge and ontology web language. At the same time, the robot is guided by the reasoning mechanism of the ontology, and this will provide technical support for the robot to realize intelligent decision in the future.