Navigation Using The Wheel Mounted MEMS-IMU

In civil positioning scenes such as car navigation and wheeled robots,the microelectromechanical system inertial measurement unit(MEMS-IMU)has been widely used and has become one of the essential navigation methods.In particular,in the autonomous vehicle navigation and positioning technology based on Dead-Reckoning,the inertial navigation system(INS)composed of MEMS-IMU plays an irreplaceable role as the main navigation system.However,MEMS-IMU has a problem of rapid accumulation of positioning errors due to its low precision and high temperature sensitivity,as well as the conventional installation method,which often has an inadvertent heading angle.In order to reduce the error accumulation to improve the navigation performance of the system,this paper designs and verifies a wheeled vehicle navigation scheme that installs the MEMS-IMU in the center of the vehicle wheel.First of all,a MEMS-IMU is fixed at the center of the wheel and rotated coaxially with the wheel.The motion characteristics of the periodic rotation are used to extract the motion velocity information,and the three-dimensional velocity observation information is constructed in combination with the motion constraint,thereby improving the positioning accuracy of the navigation system.The solution replaces the reliance on odometer sensors from the installation and algorithm design levels.The measured results based on the wheeled robot test platform show that the planar positioning error of this scheme is reduced by 58.53%(RMS)compared with the traditional MEMS-IMU navigation scheme,and by 29.67%(RMS)compared with the traditional MEMS-IMU and odometer integrated navigation scheme.Then,due to the particularity of a MEMS-IMU installation location(wheel center),the usage scenario has certain limitations and is only applicable to a flat environment without slope,such as an indoor environment.In order to make the wheel mounted MEMS-IMU navigation with more versatile,this paper further improves and perfects this scheme,and proposes to build a multi-system combined positioning scheme using multiple MEMS-IMUs.By installing multiple MEMS-IMUs in different positions,the positioning system can be applied to complex road surfaces,and the constraint algorithm is constructed by using the installation position relationship of multiple IMUs,and the positioning accuracy is further improved.The measured results based on the wheeled robot test platform show that the planar positioning error(RMS)of the multi-MEMS-IMU combined positioning scheme is reduced by 72.76% compared with the single MEMS-IMU navigation scheme.