IMU rotation of an inertial navigation system (INS) can bound the free propagation of the INS error introduced by the drifts of inertial sensors. The rotation scheme of the IMU will directly affect the accuracy, structure and costs of the system. A reasonable rotation scheme should remove most of the system errors arising from the drifts of the inertial sensors, and at the same time, should not introduce other additional errors. First, this paper discusses the design and analysis approach of the rotation scheme based on the error propagation equations of the INS. Then, a conventional 8-position rotation scheme is analyzed for the applications of the optical gyro INS, and its drawbacks are discussed in detail. Following these, an improved 8-position rotation scheme and a novel 16-position rotation scheme are proposed for the optical gyro INS, and their merits are also discussed. Simulation results have shown that the 16-position rotation scheme, which can compensate not only the drifts but also the scale factor errors and the misalignment errors of the inertial sensors, is the best rotation scheme and can be used as a practical solution to compensate the drifts of the inertial sensors in the rotational INS.

http://iopscience.iop.org/article/10.1088/0957-0233/23/2/025102/pdf

Error compensation of an optical gyro INS by multi-axis rotation

In this paper, the in-motion coarse alignment (IMCA) for odometer-aided strap-down inertial navigation system (SINS) is investigated with the main focus on compensating for the dynamic errors of gyroscope induced by severe maneuvering. A new Kalman-filtering-based IMCA method for an odometer-aided SINS is presented. A novel closed-loop approach to estimating the attitude matrix from the current body frame to the initial body frame is proposed, in which the attitude error between the closed-loop calculation and the true attitude matrix is first estimated, and then, the estimated attitude matrix is obtained by refining the closed-loop calculation with the estimated attitude error. A linear state-space model for the attitude error is derived, and then, a Kalman filter is employed to track the attitude error. Experimental results illustrate that the proposed closed-loop approach can estimate the attitude matrix from the current body frame to the initial body frame better than the existing open-loop approach, which results in improved alignment accuracy as compared with the existing optimization-based alignment method for the odometer-aided SINS when the vehicle maneuvers severely.

Kalman-Filtering-Based In-Motion Coarse Alignment for Odometer-Aided SINS

In general, the strap-down inertial navigation system (SINS)/Doppler velocity log (DVL)-integrated navigation method can provide continuous and accurate navigation information for autonomous underwater vehicles (AUV). This SINS/DVL fusion is the loosely integrated method, in which DVL may contain large error or does not work when some beam measurements are inaccurate or outages for complex underwater environment. To solve these problems, in this article, a novel tightly integrated navigation method composed of an SINS, a DVL, and a pressure sensor (PS) is proposed, in which beam measurements are used without transforming them to 3-D velocity. The simulation and vehicle test show that the proposed method can significantly outperform the traditional loosely integrated method in providing estimation continuously with higher accuracy when DVL data are inaccurate or unavailable for a complex environment. Compared with loosely integrated method, the position accuracy of the proposed method has improved by 32.5%.

A Novel SINS/DVL Tightly Integrated Navigation Method for Complex Environment

Vehicle localization is essential for intelligent and autonomous vehicles. To improve the accuracy of vehicle stand-alone localization in highly dynamic driving conditions during GNSS (Global Navigation Satellites Systems) outages, this paper proposes a vehicle localization system based on vehicle chassis sensors considering vehicle lateral velocity. Firstly, a GNSS/On-board sensors fusion localization framework is established, which could estimate vehicle states such as attitude, velocity, and position. Secondly, when the vehicle has a large lateral motion, nonholonomic constraint in the lateral direction loses fidelity. Instead of using nonholonomic constraint, we propose a vehicle dynamics/kinematics fusion lateral velocity estimation algorithm, which combines the advantage of vehicle dynamic model in low dynamic driving conditions and the advantage of kinematic model in highly dynamic driving conditions. Thirdly, vehicle longitudinal velocity estimated by WSS (Wheel Speed Sensor) and lateral velocity estimated by proposed method are as measurements for the localization system. All information is fused by an adaptive Kalman filter. Finally, vehicle experiments in U-turn maneuver and left-turn maneuver at a traffic intersection are conducted to verify the proposed method. Four different methods are compared in the experiments, and the results show that the estimated position accuracy of our method is below half a meter during a 5s GNSS outage and could keep a sub-meter-level during a 20s GNSS outage while the vehicle has a relatively large lateral motion.

Improved Vehicle Localization Using On-Board Sensors and Vehicle Lateral Velocity

In-motion alignment of strapdown inertial navigation systems (SINS) without any geodetic-frame observations is one of the toughest challenges for autonomous vehicles. Considering the characteristics of SINS, this paper presents a dual-model-based in-motion alignment method for the odometer-aided SINS. Two inertial navigation calculation loops are established for an inertial measurement unit (IMU), one of which executes the in-motion gyrocompass horizontal alignment algorithm to decompose the body velocity measured by odometer to navigation frame and attenuate the disturbance. And the other is the attitude determination-based alignment loop, where the vector observation-based SINS alignment is executed. The contributions of the work presented here are twofold. First, the dual-model initial alignment (DMIA) algorithm for SINS is proposed, which introduces the idea of constructing multiple calculation loops for an IMU to maximize the advantages of SINS. Second, depending on the body-frame speed-aided attitude determination and navigation-frame speed-aided attitude determination, the body-frame velocity observation with disturbance is accurately decomposed to navigation frame with the noise attenuated by gyrocompass horizontal alignment. The experimental results show that the proposed DMIA algorithm can achieve a rapid and accurate in-motion alignment.

A Novel Autonomous Initial Alignment Method for Strapdown Inertial Navigation System

A type of fixed integral interval sliding method for optimization-based alignment is proposed in this paper. The method is more reasonable and suitable than those in the references due to the appropriate treatment of the noise in the outputs of the Inertial Measurement Unit (IMU). The undesired sensor noise of sensors includes both random noise and bias. Our optimal alignment is based on an integrated form where the bias will be integrated during the process of alignment. In this respect, the length of the integrated data is a key factor in determining current attitude due to the ambivalent effect of integrating the sensor noise. If the length is too great the precision deteriorates due to the integration of bias. If the length is too small the precision also deteriorates because of the randomness of the noise. The validity of the method is verified by simulation and measured data.

An Improved Optimal Method For Initial Alignment

The task of strapdown inertial navigation system (SINS) initial alignment is to calculate the attitude transformation matrix from body frame to navigation frame. In this paper, such attitude transformation matrix is divided into two parts through introducing the initial inertially fixed navigation frame as inertial frame. The attitude changes of the navigation frame corresponding to the defined inertial frame can be exactly calculated with known velocity and position provided by GNSS. The attitude from body frame to the defined inertial frame is estimated based on the SINS mechanization in inertial frame. The attitude, velocity and position in inertial frame are formulated together as element of the group of double direct spatial this http URL is proven that the group state model in inertial frame satisfies a particular "group affine" property and the corresponding error model satisfies a "log linear" autonomous differential equation on the Lie algebra. Based on such striking property, the attitude from body frame to the defined inertial frame can be estimated based on the linear error model with even extreme large misalignments. Two different error state vectors are extracted based on right and left matrix multiplications and the detailed linear state space models are derived based on the right and left errors for SINS mechanization in inertial frame. With the derived linear state space models, the explicit initial alignment procedures have been presented. Extensive simulation and field tests indicate that the initial alignment based on the left error model can perform quite well within a wide range of initial attitude errors, although the used filter is still a type of linear Kalman filter. This method is promising in practical products abandoning the traditional coarse alignment stage.

Strapdown Inertial Navigation System Initial Alignment based on Group of Double Direct Spatial Isometries

This paper concerns the application of Huber-based robust unscented Kalman filter (HRUKF) in nonlinear system with non-Gaussian measurement noise. The tuning factor $$\gamma $$
 is key factor in determining the form of Huber cost function. Traditionally, $$\gamma $$
 is mainly determined by experience and/or experiments. It is hard to acquire optimal parameter or achieve an optimal filtering. To solve this problem, the influence of tuning factor $$\gamma $$
 on the performance of HRUKF is analyzed, and then, an adaptive strategy based on projection statistics algorithm for this parameter is proposed to improve filtering performance under the conditions that the measurement noise is contaminated by heavier tails and/or outliers. Simulation results for the problem of Reentry Vehicle Tracking demonstrate the superiority of the proposed method over the traditional ones.

Huber-Based Adaptive Unscented Kalman Filter with Non-Gaussian Measurement Noise

In underwater navigation systems, Global Navigation Satellite System (GNSS) information cannot be used for navigation. The mainstream method of autonomous underwater vehicles (AUV) underwater navigation system is Doppler Velocity Log (DVL) aided strapdown inertial navigation system (SINS). However, because the DVL is an instrument based on Doppler frequency shift to measure velocity, it is easily affected by the external environment. In a complex underwater environment, DVL output is easily polluted by outliers or even interrupted. In this paper, A new integrated navigation algorithm based on deep learning model is proposed to deal with DVL malfunctions. First, use RKF based on Mahalanobis distance algorithm to eliminate outliers, and then train the Nonlinear AutoRegressive with eXogenous input (NARX) model when DVL is available. When DVL is interrupted, use the NARX model to predict the output of DVL and continue integrated navigation. The proposed NARX-RKF scheme’s effectiveness verification was performed on the data set collected by the SINS/DVL ship-mounted experimental system. For comparison, different methods are also compared in the experiment. Experimental results show that NARX-RKF can effectively predict the output of DVL and is significantly better than other methods.

/pdf/an-underwater-integrated-navigation-algorithm-to-deal-with-1np5yhqpnb.pdf

Jiangning Xu

Papers

A Novel Autonomous Initial Alignment Method for Strapdown Inertial Navigation System

An Improved Optimal Method For Initial Alignment

Strapdown Inertial Navigation System Initial Alignment based on Group of Double Direct Spatial Isometries

Huber-Based Adaptive Unscented Kalman Filter with Non-Gaussian Measurement Noise

An Underwater Integrated Navigation Algorithm to Deal With DVL Malfunctions Based on Deep Learning