A Novel Method for Detecting Joint Angles Based on Inertial and Magnetic Sensors

The development of various robotic and biomedical devices has raised demands for accurate, cheap and quickresponsive joint angle sensors. To address the problem, a novel system together with a computation method for joint angle sensing based on inertial and magnetic sensors has been proposed in this paper. This method utilizes a uniform gravitational field and a uniform magnetic field in space. Low-pass filter is applied to reduce the interference of high-frequency noise. The data are also calibrated using ellipsoidal calibration model. Then the system synthesizes the two filtered and calibrated data to calculate the angle between the arms of a joint. Compared to other existing methods, this method keeps a good accuracy while greatly reduces the cost and complexity of the system. One example of the implementation of this method is given in this paper. And the results of both static and dynamic measurement experiments validate the feasibility of the method.