As one of the most core technologies of unmanned driving system, GPS Global Positioning system plays a very important role in unmanned positioning. However, unmanned vehicles are driving in complex dynamic environments, especially in large cities, with GPs multipath reflecting problems. The GPS positioning information is easy to get a few meters of error. Such errors are likely to lead to traffic accidents for automobiles with a limited width at high speeds. Therefore, other sensors must be assisted to locate and enhance the positioning accuracy. In addition, due to the low frequency of GPS updating, it is difficult to give precise real-time positioning when the vehicle travels fast.
An inertial sensor (IMU) consisting of a gyroscope, accelerometer, and other sensors is the type of high frequency sensor for detecting acceleration and rotating motion. By processing the data of inertial sensors, we can get the information of the displacement and rotation of the vehicle in real time. However, the inertia sensor itself has the effect of deviation and noise. By using the sensor fusion technology based on Kalman filter, we can integrate the data of GPs and inertial sensors, each take the director to achieve better positioning effect.
It is important to note that because of the high reliability and safety requirements of unmanned aerial vehicles, therefore, the positioning of GPs and inertial sensors is not the only way to locate unmanned aerial vehicles, in reality, the LiDAR point clouds are used to match the high precise maps, and the method of locating the visual mileage calculation, so that various positioning methods can be corrected to achieve more precise effect.