Portfolio Of Mobile Robot Navigation System Research And Design

Research in intelligent robotics, navigation is a very important issue. In a sense, the navigation is a core technology of intelligent robots. Intelligent robot navigation objective of the study is that no one intervention, a destination for the robot to move and complete a specific task to carry out specific operations. This is the real robot fully autonomous mobile intelligent and critical technologies. There are many intelligent robot navigation methods, such as inertial navigation, visual navigation, the navigation sensor data, GPS navigation and satellite navigation and so on. These navigation methods were applied to a variety of different environments, with respective characteristics.In the inertial navigation system, gyro sensors and accelerometer drift error will increase over time. Resulting navigation accuracy can not meet the actual needs. Visual navigation system due to hardware limitations, its navigation speed is very difficult to meet the requirements of the actual navigation. In this paper, combined with information processing technology, we applied the combination of inertial and visual navigation methods of navigation. On inertial/visual navigation system for the exploration of the issues we discussed and put forward concrete navigation algorithm. First, the paper introduces the inertial navigation and visual navigation works. We studied the navigation algorithm, analyzes its own characteristics. Combining two methods of navigation, the paper discusses the combination of navigation mode. Paper conduct the delay calibrated of the navigation, solve the problem of time synchronization. Usually we through the Kalman filter to deal with the error and noise of the navigate date. Features for navigation data, bring forward an improved filtering method. Using Unscented Kalman filtering, processing the navigation data, the results show that compared with the traditional Kalman filter more accurate. Finally, the error of navigation simulation data obtained higher accuracy than a single way to navigate the results.