A Novel Attitude Algorithm for 12 Accelerometer Based GFINS Using Hermite Interpolation

A novel attitude algorithm for attitude matrix computing is proposed in order to alleviate reconstruction errors of discrete angular rates in a 12 accelerometer based Gyroscope-Free Strap down Inertial Navigation System(GFINS). Different from usual attitude algorithms which use only angular rates, the new algorithm combines angular rate with angular acceleration. 3-order Hermit interpolation method(H-method) substitutes linear interpolation(L-method) to reconstruct the discrete angular rates. With the reconstructed angular rates, attitude matrix can be determined. Basic steps of new algorithm was analyzed, and simulation experiments based on typical coning motion were done. Results show that the precision of new algorithm improves by more than 45%.