Simultaneous Localization And Mapping Based On Line Feature For AUV

Autonomous Underwater Vehicle (AUV), which can navigate autonomously in complex unknown undersea environment, has been widely applied in ocean exploration, submarine rescue, resource development and other fields. And AUV has broad development prospects. The navigation technology, which is the key issue for AUV, has been one of the hot spots that researchers are working on recently. Simultaneous localization and mapping (SLAM) algorithm is the basic to solve the AUV autonomous navigation problem. It means that AUV navigates from an unknown location, detects the environment by self-carrying sensor, and builds the environment map. Meanwhile AUV uses the map and sensor data to calculate its pose, and achieves self locating.In underwater, the environment is very complex, thus it puts forward special requirements on the sensors. Point features, with describing limitations, are not qualified to describe the environment well. So this paper uses line feature descriptive method for AUV SLAM algorithm and carries out further research. This paper first introduces the research development and prospects of AUV both in the domestic and international, summarizes the simultaneous localization and mapping (SLAM) algorithm and makes a comparison for several SLAM algorithms. Then, this work takes the EKF for example and introduces SLAM algorithm in detail, which contains the basic principle and the concrete implementation process. Also, this work elaborates a feature extraction method based on improved Hough transform to extract line features from sonar data. With a given confidence level, this algorithm, what's more, can represent the uncertainties of the extracted features with a Gaussian distribution, which makes the results of extraction can be directly added to the SLAM in the future. Based on these theories, this paper researches a line-feature-based EKF SLAM algorithm. The concrete method of fusing the line features into the EKF SLAM is elaborated, and real-world experimental data are also used to verify the algorithm. The experimental results confirm that the algorithm is effective. Finally, this paper summarizes the full text, and puts forward the improvement areas in future.The research work shows that it is feasible to apply the line-feature-based simultaneous localization and mapping method in the navigation of AUV.