Research On In-motion Alignment Technology Of Low-cost MEMS-INS/GNSS

Compared with the expensive and high performance inertial navigation systems(INSs),MEMS-INSs are considered as the first choice in the field of car navigation,precision agriculture and other commercial civil application for its low cost,small volume,light weight and other advantages.However,the low quality MEMS gyro can't realize the initial alignment on the static base by detecting the earth rotation.And in the low dynamic application scenarios(the velocity is about lm/s),the commonly used attitude determination method using GNSS velocity is also less reliable.Aiming at this practical problem encountered in the precision agriculture project,this paper designed,optimized and validated two sets of in-motion initial alignment schemes for the low cost MEMS-INS/GNSS integrated navigation system on the ground vehicle.For the first,the optimization-based alignment(OBA)algorithm is based on the least squares theory,attitude determination method using GNSS velocity depends on the velocity vector observations.combining their advantages,an improved OBA algorithm based on velocity vector observations was derived.By the means of an intuitive global observability analysis,the OBA algorithm was proved to require variable acceleration maneuvers,while the improved OBA algorithm only requires the carrier to have the speed.Simulation and experiments results showed that if not at a quick turn the heading alignment error of the improved OBA algorithm can be reduced to within 4 deg in 60s under low dynamic conditions with about 1m/s velocity.Secondly,the large heading uncertainty model was used to design an initial alignment algorithm.Ignoring some small variables,both the large heading uncertainty error propagation model for the strapdown inertial navigation system(SINS)and the GNSS position and velocity observation equations were deduced.In addition,the nonholonomic constraints(NHC)observation equation was also deduced.By the means global observability analysis,NHC was proved capable of improving the heading alignment accuracy in low dynamic conditions.Simulation and experiments results showed that heading alignment error with NHC can be reduced to below 2 deg in 80s under low dynamic conditions with about lm/s velocity.