Search

Graph convolution subspace clustering has been widely used in the field of hyperspectral image (HSI) unsupervised classification due to its ability to aggregate neighborhood information. However, existing methods focus on using graph convolution techniques to design feature extraction functions, ignoring the mutual optimization of the graph convolution operator and the self-expression coefficient matrix, leading to suboptimal clustering results. In addition, these methods directly construct graphs on raw data, which may be easily affected by noises and then degrade the clustering performance, as the constructed topology is not credible for the training procedure. To address these issues, we propose a novel method called spatial-spectral adaptive graph convolutional subspace clustering (S2AGCSC). We employ the reconstruction coefficient matrix to devise a graph convolutional operator with adjacency matrix, which collaboratively computes both the feature representations and coefficient matrix, and the graph-convolutional operator is updated iteratively and adaptively during training. In addition, we harness a combination of spectral and spatial features to introduce additional view information to help learn more robust features and generate more refined superpixels. Experimental validation on three HSI datasets confirms the efficacy of S2AGCSC.

In the context of Industry 4.0 and based on the demand of digital logistics construction of industrial enterprises, this paper integrates the concept of digital twin into Refined Logistics Supply Chain construction. Considering the constraints of multi-distribution center, heterogeneous vehicle performance, distribution cost, quasi-shipment certificate and humanized management, Refined Logistics Supply Chain System (RLSCS) and cross-regional scheduling optimization model of logistics vehicles with multi-distribution center were established. The designed model can minimize the transportation cost, reduce the transportation time, and improve the vehicle load rate. An adaptive elite honey badger target algorithm based on cubic mapping mechanism (IHBA), is designed to solve the model. Further, performance evaluation by optimizing test functions, the convergence performance of IHBA algorithm was demonstrated. Finally, the simulation experiment is carried out according to the actual business data and it is compared to eight other optimistic algorithms. The experimental results show that the proposed algorithm is more effective and more robust, and the related models and algorithms can provide research basis for industrial products digital supply chain system.

The requirement for motion control of robotic arms in industrial settings is a dynamic field. This study examines the principles and derivation of the kinematics of the robotic arm based on the D-H parameter model. Additionally, the introduction of the seventh joint is proposed as a faster solution for solving the inverse kinematics of the robotic arm. Swarm intelligence optimization for path planning is currently advancing, and our proposed improved algorithm, the Golden Eagle search algorithm, enhances the traditional Golden Eagle search algorithm Jining by integrating a stochastic gradient descent strategy and Cauchy mutation strategy. We compare our IGEO algorithm with various other algorithms, and the findings demonstrate that the robotic arm can adeptly circumnavigate obstacles while walking seamlessly through environments with multiple obstacles. The IGEO algorithm is adept at navigating paths obstructed by multiple obstacles. It improves the accuracy by 15.38% as compared to the conventional algorithm and also improves it a lot as compared to other optimisation algorithms by up to 29.88%. It provides a solution to the path planning problem of robotic arms with excellent robustness and accuracy in finding the shortest collision-free path.

We propose a new adaptive modeling algorithm, nonlinear multidegree-of-freedom (NLMDF). The underlying idea is to assimilate the adaptive modeling process of iteration for the optimal solution to a particle's movement on the mean-square error (MSE) surface geometry, which is imagined to be located in the gravitational field. The new algorithm is derived by incorporating a viscous damping force in the particle movement differential equation. A new variable step size least mean square (LMS) algorithm--new variable step LMS (NVLMS)--is derived from the NLMDF algorithm by choosing specific value for the damping parameter in the NLMDF recursion. The simulation results show a faster rate of convergence for the new algorithms as compared with the damping force LMS (DFLMS) algorithm and the LMS algorithm with a white or colored noise input signal. [ABSTRACT FROM AUTHOR]