| As the importance of polar marine environments for the sustainable economic development and national security is increasingly emerging,the exploitation and detection for polar marine resource has a decisive position.However,since the anomalies of magnetic fields and multipath effect in the polar region,the conventional navigation approaches,such as radio navigation,geomagnetic navigation and satellite navigation,fail to work in the polar region.Strapdown inertial navigation system(SINS)has the advantages of high autonomy and relatively independent,so it has been widely used in the polar navigation of various military and civilian large-sized ships.For the affiliated aircrafts and workboats of large-sized ships,their initial alignment is usually realized by transfer alignment(TA)to improve the alignment speed and accuracy.At present,researches for the polar TA just get started,so it still exists lots of problems and technical difficulties.In order to achieve the rapid and accurate result of polar TA,a rapid TA approach based on the grid frame for the affiliated aircrafts and workboats is proposed in this paper.Due to the meridian convergence in the polar region,navigation approaches of master and slave SINSs based on the geographic frame would suffer the problems of calculation overflow and sharply increased errors.And the TA approaches based on the geographic frame also suffer the same problems,which would decrease the accuracy of polar TA.The grid navigation approach can avoid the problems caused by meridian convergence,so it can be used for polar TA.Since the measurement cannot be direct observed,existing grid TA approach has low observable degrees and requires high mobilities.Thus,the existing approach is not applied for transfer alignment of the low-mobility large-sized ship.Moreover,the existing approach increases the complexity of measurement matrix and does not have a definite compensation method for interference errors.Therefore,the existing grid TA approach should be modified in this paper.Meanwhile,interference errors of TA should be analyzed to supply references for the modified gird TA approach.Based on the rapid TA approach of velocity plus attitude matching method,the completely observable measurement misalignment angles and velocity errors of master and slave SINSs are chosen as the states and measurements of TA.By considering the effects of flexural deflection on lever-arm,the model of flexural lever-arm is established.In virtue of the white noise principle,the noise compensation method is used to compensate the flexural deflection.Based on the modified grid TA errors equations and interference errors compensation method,the modified TA filter models are designed.Due to the noise compensation method of flexural deflection would decrease the accuracy of TA,an improved Sage-Husa adaptive Kalman filter is proposed.To supply references for the design of mobilities,the observability of modified grid TA approach is analyzed by the singular value decomposition(SVD)analysis method.The modified grid TA approach proposed in this paper can effective achieve the estimation of misalignment angles and the compensation of interference errors.Meanwhile,the modified approach has a higher observability and better alignment result under the linear motion with constant acceleration.In the application of TA,the azimuth misalignment angle of can become large.In the case of large azimuth misalignment angle,the modified grid TA approach based on the assumption of small angles would become inapplicable.By considering the system as nonlinear,the grid nonlinear error equations are derived.Based on the global observability analysis method,the observability of grid TA nonlinear approach is analyzed.Combined the modified compensation method of TA interference errors,the nonlinear compensation models of grid TA is designed.According to the problems that the strength of process noise is oversize and the statistical characteristic is uncertain,an adaptive unscented Kalman filter is proposed.The nonlinear approach of grid TA can be applied for polar TA in the case of large azimuth misalignment angle.Meanwhile,no matter the ship is under any kind of motion,the nonlinear approach of grid TA has a good alignment accuracy.Due to the multipath effect in the polar region,the accuracy of global navigation satellite system(GNSS)is decreased and even be diverging.Since the master SINS cannot well calibrate its errors by GNSS,accumulative errors of the master SINS would appear.In the case of master SINS has accumulative errors,an external navigation aided grid TA approach is proposed to ensure the performance of TA.By the principle analysis of Doppler velocity log and star sensor,the errors models are established.Based on the velocity and attitude matrix measured by Doppler velocity log and star sensor,the high-accuracy velocity and attitude measurements are designed.Combined the nonlinear models of grid TA and adaptive Kalman filter,the nonlinear grid TA approach with the aid of acoustic and astronomic navigation is proposed in this paper,which can ensure the performance of polar TA in the case of master SINS has accumulative errors.In the case of small angles and large azimuth misalignment angle,the grid TA approaches are modified and improved in this paper.According to the case of master SINS has accumulative errors,the grid TA approach with the aid of acoustic and astronomic navigation is researched in this paper.Results of simulation and experiment indicate that the modified grid TA approach can achieve a high-accuracy(the estimation errors of three-axis are within 0.3?)alignment result within 15 seconds.In the case of large azimuth misalignment angle,the nonlinear approach of grid TA can achieve a high-accuracy(the estimation errors of three-axis are within 3.3?)alignment result within 40 seconds.In the case of master SINS has accumulative errors,the nonlinear grid TA approach with the aid of acoustic and astronomic navigation can achieve a high-accuracy(the estimation errors of three-axis are within 6?)alignment result within 40 seconds. |
PDF Full Text Request