Research On Environmental Cognition Technology Of Mobile Robot In Unknown Environment

The technology of cognize the unknown environment by autonomous mobile robot, which is the premise of the autonomous mobile robot to achieve a variety of tasks. The environmental cognition can solve the problem of unknown environment into known environment, also we can convenient to consider the issue from the view of the global map. Autonomous mobile robot in unknown environments cognitive research has important theoretical significance and research value.The autonomous mobile robot environment cognitive technology mainly includes two aspects:exploring and mapping of the unknown environment. Unknown environment Exploration of unknown environment is mainly through the sensor of the mobile robot,which is a process of constantly looking for the next best point (Next Best View). The process of exploration requires considering the problem of path planning, selecting the best path to explore the environment; Build the environment map is using the specific environmental map to express cognitive results, usually has a close relationship with the autonomous mobile robot positioning technology.On the basis of the theory of frontier, we comprehensive consider the travel loss, the pose of the mobile robot and environmental ergodicity, then explain the exploration technologies of mobile robot in unknown environments. For multi-candidate detection point in the exploration process, we use of Multi-Criteria Decision Making method to determine the optimal candidate detection point. The build environment map is accompanied by exploration at the same time, the mobile robot to explore an area and build map in this region immediately. We use the topology map to express the environment information, and the map is building by the Growing Neural Gas Network. The growth neural gas network has the features of growth and topology. Exploration and topological map building test in Matlab simulation environment, and achieved satisfactory results, an effective solution to the traditional methods of computational complexity and inefficiencies.