Moving State Estimation Of The Mini-Mobile Robot

With the fast development of technology, mini-mobile-robot has become more and more popular. Nowadays, mini-mobile-robot control system have been included automatic control, information processing, test-switch and other kinds of professional technology, and becoming one of the important sides of the mobile robot research. In engineering, the requirements such as cost, power consumption, volume, are raising quickly. Traditional strap-down inertial navigation system uses high-accuracy gyroscope and accelerometer to get the azimuth and the position of the mini-mobile robot. But the high-accuracy sensors need a high price and bring up the cost of the project, and the low-accuracy sensors’measurements can not reach the precision of the navigation system. In this thesis, adopt low-accuracy sensors and fuse the measurements after analysising the character of each sensor to improve the estimation accuracy of the mini-mobile-robot moving state.There are five parts in this thesis, which can be summarized as:(1) Introduced the development and research status in domestic and international of the mobile robot navigation and positioning system, and the popular methods of mobile robot localization. It laid the foundation for the next step of the thesis’s research.(2) Showed the hardware structure of the mini-mobile-robot and the sensors that adopted. It also analyzed the character of each sensor, and gave the mathematical models.(3) Chose the proper algorithm of multi-sensor fusion, and introduced the unscented Kalman filter in detail. Introduced the Federated Unscented Kalman Filter in this thesis.(4) Introduced the traditional strap-down inertial navigation system and the usual coordinate system. In this thesis, introduced a new kind of strap-down inertial navigation system– multi-sensor strap-down inertial navigation system, and introduce the character and the principle of it in detail.In the fifth part showed the simulation results of the Federated Unscented Kalman Filter used in azimuth estimation and multi-sensor strap-down inertial navigation system used in position estimation. And the results showed that in the multi-sensor strap-down inertial navigation system the low accuracy sensors can still estimate the mini-mobile-robot moving state well.