Multi-degree-of-freedom Attitude Sensing System Of Ball Joint Based On Magnetic Measurement

The posture detection of the ball joint is the key to accurately control the motion posture of the ball joint.At present,the posture detection of the ball joint is mainly divided into two methods: contact measurement and non-contact measurement.Contact measurement often requires other measurement devices such as inertial measurement units,guide rails,and encoders to introduce new inertial and frictional forces.Non-contact measurement mainly has multiple solutions based on vision,optics,inductance,and magnetic fields.The posture measurement of ball joints based on vision,optics,and inductance often requires a lot of extra workload and computing resources,complex auxiliary structures and high installation requirements,and cannot be widely promoted in actual production applications.This paper proposes a ball joint multi-degree-of-freedom attitude sensing system based on magnetic measurement,which can realize up to three-degree-of-freedom ball joint attitude detection and is easy to install,low cost and practical.The sensing system uses machine learning,inverse solution and fitting based on Dipole model to establish a function mapping model between the magnetic field change and the Euler angle of the ball joint.The optimal layout and accuracy requirements of the sensors are obtained through numerical simulation,and the physical model of the sensing system is designed and compared with the inertial measurement unit to verify that the system has real-time,high-precision and antiinterference while avoiding zero bias and cumulative errors Features,can be applied to large-scale production and application.