The Study On SLAM Algorithm Of Mobile Robot Based On WSN

In many occasions, human being can be substituted by mobile robots to implement some dangerous mission with the function of interacting with the environment actively. So it has caused wide public concern of research communities. However, because of shortage of information collection in disaster site, the mobile robot localization accurately and environment map building is an aporia limiting its application and development. As a new technology of information acquisition and processing, WSN can be used as a technical support. Therefore, the area combined between WSN and mobile robot has become the hot point of international research. Combined with WSN, localization and SLAM technologies of mobile robot is studied on this thesis.Focusing at the problem, which mobile robot using dead reckoning method can produce error accumulation to decrease position accuracy, an mobile robot localization algorithm based on information fusion is proposed on this thesis. By using EKF merging into robot dead reckoning and distance measurement information, accumulated error with robot dead reckoning is corrected accordingly through distance measurement information of WSN nodes. Simulation experiment is shown that the method proposed on thesis can weaken the influence of positioning and have a better accuracy than dead reckoning localization.Based on EKF fusion location of hybrid coordinate system, a modified algorithm is proposed to overcome the shortcomings of the naive method and it builds model of mobile robot motion model and measurement model under hybrid coordinate system. In this algorithm, Dynamic Weighted Average algorithm is used to update the measurement variance to adjust residual of system localization. With both location tracking and global positioning of mobile robot simulation experiments, the proposed method can overcome the shortcomings and limits of the naive method. What's more, it has higher position accuracy and stronger robustness.When locations of nodes are unknown, it can use Range-Only SLAM to solve the problem of mobile robot with SLAM. Based on traditional algorithm, EKFSLAM with hybrid coordinate system is proposed. And it can estimate movement uncertainty of mobile robot more accurately. The result of simulation experiment is shown that the proposed algorithm has higher accuracy for localization both mobile robot and nodes synchronously. The proposed method combined with laser radar to measure data can build environment maps correctly.In addition, a WSN-mobile robot system for positioning experiments is set up. It consists of UP-voyagerll mobile robot platform, WSN nodes and data acquisition and control software. Based on the test data collected by positioning experiment system, corresponding simulation experiments are carried out for the proposed algorithm. Combining with experiment results, it tests and verifies the reasonableness and effectiveness of proposed method.