AN IMPROVED COMPLEMENTARY FILTER ALGORITHM IN THE APPLICATION OF THE MOBILE ROBOT ATTITUDE ESTIMATION

For the embedded mobile robot attitude estimation system, the convergence speed of the general algorithm is too slow to get the stable data when starting. This paper proposes an improved complementary filter algorithm based on low-cost microeElectro-mechanical system (MEMS) inertial sensors. According to the output of inertial sensors and the motional characteristics of the mobile robot, we establish a time-correlated parameter equation. The algorithm can adjust the filter parameters in real time and choose the cut-off frequency adaptively. Therefore, convergence speed is increased and the accuracy of the attitude estimation is improved, which makes the mobile robot start smoothly and stably. The experiment results show that the convergence time of the attitude estimation is about 97.4 ms that is 49% lower than that of the general complementary filter.