Research On Embedded Coverage Algorithm And Auto-Recharging Algorithm For Family Cleaning Robot

With the rapid progress of science and technology and the development of Robotics, service robots come into human's life. Cleaning Robots have received more and more attention as the combinations of the technology of mobile robot and cleaner. Cleaning robot can clean room automatically and release human's burden. As one of particular applications of the intelligent robots, Cleaning Robot specifically embodies many key technologies and has the typical role of example from the point of technology.In this dissertation, we have made one embedded family cleaning robot which is a collaborate project developed by HIT(Harbin Institute of Technology university) between HIT and LongChang company, Guangdong province as the instance. We make a specific and roundly introduction about the specific development of embedded software for the Cleaning Robot.Two important algorithms have been proposed, the Internal Spiral Coverage algorithm based on grid-map and Auto-docking and recharging algorithm. And we take round and deep discussion and research the two algorithms.The detection range of the sensors has great influence on the planning ability of the covering robot. It demands more planning ability with shorter detection-range sensors. The coverage for the robot with contact sensors has been studied in the dissertation, and the Internal Spiral Coverage (ISC) algorithm that is based on grids is proposed in the paper. The ISC algorithm realizes simply and has less repeat coverage which guarantees complete coverage. There are two stages in ISC: in the first stage the robot explores along the boundary of the environment, and in the second stage the robot plans the coverage path on line. By repeat covering some grids and setting the GATE grids it not only guarantees the complete coverage, but also reduces the coverage repetition, thus improves the efficiency. There are no complex mathematical calculations in the planning which can meet the requirement of real-time and is fit for the embedded system. And it doesn't search for the path in the whole space which overcomes the shortcomings of the grid-based representation that the real-time property will get sharp decline as the space enlarges.