For Intelligent Cleaning Robot Path Planning Technology Research

As a special application of intelligent mobile robot, cleaning robots inherited a number of key techniques of mobile robots. Such as sensor technology, location technology, etc, while it has own characteristics, for example there's no need to understand three-dimensional environment but a full-coverage of the environment is needed. Cleaning robots can substitute for manual labors. So it is widely used in home, public places, schools and other complex environments, so there are broad market prospects of cleaning robot. This paper analyzes several popular cleaning robots, and provides several solutions for a complex environment, which was not solved in those cleaning robots.This paper is based on family environments. It shows a detection of feature points algorithm, which are provided as the reference points in SLAM. Robots could locating and mapping while it is working. And this paper provides a extend Kalman filter of the SLAM algorithm. And finally robots can decompose the full region into small pieces and link those areas and then sweep these small pieces. The methods are as follows:1. In this paper, data are transferred from sensors. Firstly the sensor data are separated, and then lines are fitting by these discrete data. Secondly, cross points of two lines are found which are seen as the corner points. At last, these points are features of SLAM.2. This paper implements a SLAM algorithm, which is base on an extend Kalman filter. Considering that robots are moving slowly, so we improve the algorithm into the SUT-based SLAM. We improve the accuracy of this algorithm by SUT, which constructs new statistical features by non-linear function.3. This paper implements a round-based algorithm to determine the edge of obstacles. Robots determine the environment through working along the wall, and construct the grid map of the environment. Similarly to the inside obstacles.4. This paper presents a region-segmentation algorithm based on the determination of the obstacle in unknown environment. Considering that the existing segmentation algorithms are based on given environments, in this paper, robots create the grid map by working along the wall, after that, finish segmentation algorithm in grid maps. If any obstacle was seen inside the room, find the position of it, and then do the segmentation algorithm in this region.5. After long-time working, robots could run off of the power, so we use A* algorithm for the path planning, the robots can return to charge.