Zhihong Deng

A new sampling method in particle filter based on Pearson correlation coefficient

Abstract Particle filters have been proven to be very effective for nonlinear/non-Gaussian systems. However, the great disadvantage of a particle filter is its particle degeneracy and sample impoverishment. An improved particle filter based on Pearson correlation coefficient (PPC) is proposed to reduce the disadvantage. The PPC is adopted to determine whether the particles are close to the true states. By resampling the particles in the prediction step, the new PF performs better than generic PF. Finally, some simulations are carried out to illustrate the effectiveness of the proposed filter.

A Self-Calibration Method for Nonorthogonal Angles Between Gimbals of Rotational Inertial Navigation System

Navigation accuracy of an inertial navigation system can be significantly enhanced by rotating inertial measurement unit with gimbals. Therefore, nonorthogonal angles of gimbals, which are coupled into the navigation error during rotation, should be calibrated and compensated effectively. In this paper, the relationship model of nonorthogonal angles and navigation error is established. Then, the calibration scheme and observation equation during gimbals rotation is proposed. Proved by a piecewise constant system method, all of the error parameters are observable and can be estimated by an extended Kalman filter. Experimental results show that compared with the traditional method, the proposed method can substantially reduce velocity error on static base. Moreover, the position accuracy of long-term navigation under moving base is also significantly increased.

Estimation of Information Sharing Error by Dynamic Deformation Between Inertial Navigation Systems

Dynamic deformation of a vehicle and relative position of inertial navigation systems will cause large errors during information sharing. Such errors should be estimated and compensated effectively, or it will cause large errors to navigation information output. Therefore, the model between deformation angle and dynamic lever arm is established to verify that the influence is not coaxial but decussate. Then, estimation and compensation methods for deformation angle and dynamic lever arm are proposed with real-time closed-loop correction of the lever-arm length. Simulation results demonstrate that, for the estimation of misalignment angle, the proposed method has a faster convergence than the traditional method. Moreover, the deformation angle and dynamic lever-arm estimation also takes less time with a higher accuracy than the traditional method.

A Particle Filter-Based Matching Algorithm With Gravity Sample Vector for Underwater Gravity Aided Navigation

Gravity matching algorithm is a key technique of gravity aided navigation for underwater vehicles. The reliability of traditional single point matching algorithm can be easily affected by environmental disturbance, which results in mismatching and decrease of navigation accuracy. Therefore, a particle filter (PF)-based matching algorithm with gravity sample vector is proposed. The correlation between adjacent sample points of inertial navigation system is considered in the vector matching algorithm in order to solve the mismatching problem. The current sampling point matching result is rectified by the vectors composed by the selected sampling points and matching point. The amount of selected sampling points is determined by the gravity field distribution and the real-time performance of the algorithm. A PF-based on Bayesian estimation is introduced in the proposed method to overcome the divergence disadvantage of the traditional point matching algorithm in some matching areas with obvious gravity variation. Simulation results prove that compared with the traditional methods, the proposed method is robust to the changes of gravity anomaly in the matching areas, with more accurate and reliable matching results.

A motion-based integer ambiguity resolution method for attitude determination using the global positioning system (GPS)

Loss of the satellite signal and noise disturbance will cause cycle slips to occur in the carrier phase observation of the attitude determination system using the global positioning system (GPS), especially in the dynamic situation. Therefore, in order to reject the error by cycle slips, the integer ambiguity should be re-computed. A motion model-based Kalman predictor is used for the ambiguity re-computation in dynamic applications. This method utilizes the correct observation of the last step to predict the current ambiguities. With the baseline length as a constraint to reject invalid values, we can solve the current integer ambiguity and the attitude angles, by substituting the obtained ambiguities into the constrained LAMBDA method. Experimental results demonstrate that the proposed method is more efficient in the dynamic situation, which takes less time to obtain new fixed ambiguities with a higher mean success rate.

An adaptive Kalman filter estimating process noise covariance

Abstract In this paper, a new adaptive Kalman filter algorithm is proposed to cope with the unknown a priori covariance matrix of process noise for the linear discrete-time systems. The process noise covariance matrix is estimated by the proposed algorithm based on the measurement sequence. Accordingly, we construct a new measurement sequence to sequentially estimate process covariance matrix in terms of the relationship between the measurement and process noise sequence. Then the stability of the proposed algorithm is analyzed. The algorithm shows a simple recursive form and great performance enhancement of application. Finally, the navigation simulation results are presented to illustrate the validity and practicality of the proposed algorithm.

Analysis and Calibration of the Nonorthogonal Angle in Dual-Axis Rotational INS

Navigation errors of inertial navigation system can be effectively restrained by rotating inertial measurement units (IMU), enhancing accuracy in long-endurance navigation. However, nonorthogonal angles between rotating axes are inevitable during manufacturing. Coupled into the navigation error, nonorthogonal angles can lead to navigation accuracy, especially attitude accuracy decline during rotation. Therefore, the calibration and compensation of nonorthogonal angles will enhance the systemic reliability. In this study, we first established nonorthogonal angles model, then estimated the effect on sensitive errors of IMU using the proposed model. Based on the relationship between nonorthogonal angles and IMU attitude, an improved calibration algorithm of estimating nonorthogonal angles is presented. The results of theoretical analysis demonstrated that this method can effectively calibrate nonorthogonal angles in a simple and fast way, which is desirable for practical requirements. Simulation and comparison analysis indicated that the calibration of nonorthogonal angles is effective and practical. Finally, the feasibility of calibration and compensation is also verified by the improvement of navigation accuracy in the experiment.

An Acoustic Communication Time Delays Compensation Approach for Master–Slave AUV Cooperative Navigation

Time delay is a major problem in the acoustic communication technology. Considering such a background, a new dynamic model is proposed for an improved error estimation algorithm in this paper. And error propagation equations are constructed for an inertial navigation system/Doppler velocity log integrated navigation system. The time delay problem is reconsidered for acoustic communication with an ultra-short baseline acoustic positioning system that uses a multi-autonomous underwater vehicle (AUV) cooperative navigation process in master-slave mode. The characteristics of time delays in acoustic communication are considered and the time delay is converted into a measurement bias within an observation equation for the slave AUV platform. Under the framework of a standard Kalman filter, an improved error estimation algorithm is presented to address the problem that occurs when random measurement bias exists in the dynamic linear system model. Based on the Monte Carlo method, the simulation results illustrate that compared with traditional methods, the error estimation algorithm proposed in this paper can effectively decrease positioning errors caused by time delays in acoustic communication.

The Gravity Matching Area Selection Criteria for Underwater Gravity-Aided Navigation Application Based on the Comprehensive Characteristic Parameter

Matching area selection is a key problem of the gravity matching navigation method, especially for underwater application. The gravitational profile varies when the vehicle gets into the matching area in different directions. Therefore, the matching navigation efficiency is also distinct due to the directionality of matching area. In this study, based on the fractal geometry theory, a characteristic parameter-isotropic coefficient, which can measure the directionality of gravity matching area-is proposed by analyzing the spectrum character of gravity anomaly sequence on 3-D surface. Instead of single parameter, the comprehensive characteristic parameter is built by combining traditional gravity parameters and isotropic coefficient through an analytic hierarchy process for matching area selection. Compared with the traditional method, simulation results prove that the matching area selected by the improved comprehensive characteristic parameter method not only has larger coverage, but also has better directivity.

An Improved TERCOM-Based Algorithm for Gravity-Aided Navigation

Gravity-aided inertial navigation is a leading issue in the application of autonomous underwater vehicle. An improved gravity matching algorithm, based on the principle of the terrain contour matching (TERCOM) algorithm, is proposed. The matching algorithm applies the shortest path algorithm to increase update frequency. In addition, the positioning error can be limited due to the novel correlation analysis method. Compared with existing algorithms, the improved TERCOM algorithm has better real-time performance, positioning accuracy, and reduced calculation burden. The reliability and the accuracy of the algorithm are verified via simulation tests.