An assessment of the error inherent in rotation vector algorithm utilized in strapdown inertial navigation system is presented, which shows that the accuracy cannot be improved boundlessly only by increasing the number of the used samples. After reaching the upper limit for accuracy, increasing sample size would only enlarge the algorithm error rather than improve the performance of rotation vector algorithm. After accuracy assessment, a new method to construct strapdown attitude algorithm is originated. Unlike previous algorithms adopting rotation vector, the new method utilizes the direction cosine matrix differential equation as its basis, and discretizes this differential equation with Taylor time series meanwhile extracts angular velocity and its derivatives by polynomial fitting. With this method, three direction cosine matrix attitude algorithms, taking 4, 5, 6 samples respectively, are deduced, and their performances are evaluated with coning motion and severe maneuver profile. The results show that the new algorithms based on direction cosine matrix differential equation can effectively handle the coning error and severe maneuver error.

Presently, many researches have been carried out on rough set based data reduction. However, this method encounters a problem when dealing with real-valued data and fuzzy information. By lucubrating the theory of fuzzy rough set and utilizing the definition of information entropy presented in literature [5], the information entropy model of fuzzy rough set has been constructed. Then the conditional information entropy of attributes is adopted to measure the significance of attributes. On this condition, a heuristic fuzzy-rough data reduction method based on information entropy (E-FRDR) has been put forward. Finally, the method is validated by an example that indicates the method is feasible.

Rapid advancements in autonomous driving technology have highlighted the challenges of ensuring vehicle safety and driving efficiency in complex dynamic traffic environments. Current approaches typically define potential risks as safety constraints for compliance and use them in trajectory planning. However, the risks predefined in these constraints are often fixed, reducing driving efficiency. To address this limitation, we proposed a dynamic risk-information-driven adaptive trajectory planning method for autonomous vehicles (AVs). This study dynamically adjusted safety constraints using risk assessment results to improve driving efficiency without compromising safety. Firstly, considering the influence of vehicle suspension characteristics on driving safety, collision, and instability risk assessment indices were designed using a three-way-coupled dynamic model to assess driving safety risks. Next, we used the safety risk assessment module to evaluate specific potential risks and adaptively adjusted the safety constraints for constraint-based adaptive trajectory planning. Furthermore, considering trajectory traversal constraints, trajectory selection and optimization were performed on pre-planned trajectories using the cost function to determine the optimal driving trajectory. Lane-changing trajectory planning experiments showed that the method adaptively adjusts safety constraints based on risk assessment results. Under the premise of ensuring driving safety, driving efficiency improved by 55.9% in the preset instability constraint scenario and 27.86% in the preset collision constraint scenario.

The random error of fiber optic gyros is a critical factor affecting their measurement accuracy. However, the statistical characteristics of these errors exhibit time-varying properties, which degrade model fidelity and consequently impair the performance of random error suppression algorithms. To address these issues, this study first proposes a recursive dynamic Allan variance calculation method that effectively mitigates the poor real-time performance and spectral leakage inherent in conventional dynamic Allan variance techniques. Subsequently, the recursive dynamic Allan variance is integrated with the process variance estimation of Kalman filtering to construct a dual-adaptive Kalman filter capable of autonomously switching and adjusting between model parameters and noise variance. Finally, both static and dynamic validation experiments were conducted to evaluate the proposed method. The experimental results demonstrate that, compared to existing algorithms, the proposed approach significantly enhances the suppression of angular random walk errors in fiber optic gyros.

Hyperspectral image (HSI) clustering has been a fundamental but challenging task with zero training labels. Currently, some deep graph clustering methods have been successfully explored for HSI due to their outstanding performance in effective spatial structural information encoding. Nevertheless, insufficient structural information utilization, poor feature presentation ability, and weak graph update capability limit their performance. Thus, in this paper, a homophily structure graph learning with an adaptive filter clustering method (AHSGC) for HSI is proposed. Specifically, homogeneous region generation is first developed for HSI processing and constructing the original graph. Afterward, an adaptive filter graph encoder is designed to adaptively capture the high and low frequency features on the graph for subsequence processing. Then, a graph embedding clustering self-training decoder is developed with KL Divergence, with which the pseudo-label is generated for network training. Meanwhile, homophily-enhanced structure learning is introduced to update the graph according to the clustering task, in which the orient correlation estimation is adopted to estimate the node connection, and graph edge sparsification is designed to adjust the edges in the graph dynamically. Finally, a joint network optimization is introduced to achieve network self-training and update the graph. The K-means is adopted to express the latent features. Extensive experiments and repeated comparative analysis have verified that our AHSGC contains high clustering accuracy, low computational complexity, and strong robustness. The code source will be available at https://github.com/DY-HYX.